78000 towards better expenditure quality guatemala public expenditure review May 2013 C E N T R A L A M E R I C A CO U N T R Y M A N A G E M E N T U N I T | P O V E R T Y R E D U C T I O N A N D E CO N O M I C M A N A G E M E N T | L AT I N A M E R I C A A N D T H E C A R I B B E A N R E G I O N CURRENCY AND EXCHANGE RATE (As of May 11, 2012) CURRENCY UNIT = GUATEMALAN QUETZALES US$1.00 = Q.7.78 FISCAL YEAR January 1 – December 31 Vice President: Hasan A. Tuluy Country Director: C. Felipe Jaramillo Sector Director: Rodrigo A. Chaves Sector Manager: Auguste Tano Kouame Sector Leader: Oscar Calvo-Gonzalez Task Team Leader: Jasmin Chakeri ABBREVIATIONS AND ACRONYMS BAC Budget Availability Certificate MiFaPro Conditional Cash Transfer Program (Mi Familia Progresa) CACIF Coordinating Committee of Agricultural, MTEF Medium-Term Expenditure Framework Commercial, Industrial and Financial Associations (Comité Coordinador de Asociaciones Agrícolas, Comerciales, Industriales y Financieras) CAS Security Advisory Council MINUGUA United Nations Verification Mission for (Consejo Asesor de Seguridad) Guatemala CGC General Auditor’s Office MOF Ministry of Finance (Contraloría General de Cuentas) CICIG International Commission against Impunity in MOI Ministry of the Interior Guatemala (Comisión Internacional contra la Impunidad en Guatemala) COVIAL Road Conservation Executing Agency NGOs Non-governmental Organizations (Unidad Ejecutora de Conservación Vial) CREMA Contracts for Rehabilitation and Maintenance OECD Organization for Economic Cooperation and (Contrato de Rehabilitación y Mantenimiento) Development DGC General Directorate of Roads (Dirección General PEFA Public Expenditure and Financial Accountability de Caminos) DR-CAFTA Dominican Republic-Central America Free Trade PIP Public Investment Program Agreement with the United States EC European Commission PFM Public Financial Management ENCOVI Living Standards Measurement Survey PNC National Civil Policy (Policía Nacional Civil) (Encuesta Nacional de Condiciones de Vida) FMIS Financial Management Information System POA Annual Operating Program (Programa Anual Operativo) FONAPAZ National Peace Fund SAT Superintendence of Tax Administration (Fondo Nacional para la Paz) (Superintendencia de Administración Tributaria) FSS Social Solidarity Fund SIAF Integrated financial management system (Fondo Social de Solidaridad) (Sistema Integrado de Administración Financiera) GDP Gross Domestic Product SICOIN Integrated Accounting System (Sistema de Contabilidad Integrado) Guatecompras Guatemalan state procurement system SIGES Management System (Sistema de Información y Gestión) IDB Inter-American Development Bank SMEs Small and medium size enterprises LAC Latin America and the Caribbean SNIP National Public Investment System (Sistema Nacional de Inversión Pública) LAPOP Latin American Public Opinion Project STA Single Treasury Account MDG Millennium Development Goal UNDP United Nations Development Program MICIVI Ministry of Communications, Infrastructure VAT Value-Added Tax and Housing (Ministerio de Comunicaciones, Infraestructura y Vivienda) Table of Contents » Executive Summary A. Guatemala’s Development Challenges B. Overview of Public Finances C. The Quality of Public Expenditure in Selected Sectors 6 7 7 9 D. Citizen Security and Public Expenditures 11 E. Budget Process, Public Financial Management and Procurement 13 Chapter 1: Guatemala’s Development Challenges Ahead 16 A. Introduction 17 B. Historical and Political Context 18 C. Economic Development Challenges 19 Chapter 2: Overview of Public Finances 26 A. Introduction 27 B. Structure and performance of revenues 29 C. Structure and performance of expenditures 35 D. Fiscal sustainability 40 E. Conclusions and recommendations 43 Chapter 3: The Quality of Public Expenditure in Selected Sectors 44 A. Introduction 45 B. Quality of Public Expenditure in the Education Sector 46 C. Quality of Public Expenditure in the Health Sector 50 D. Quality of Public Expenditure in the Road Transport Sector 55 E. Conclusions and recommendations 60 Chapter 4: Citizen Security and Public Expenditures 64 A. Introduction 65 B. Trends in Crime and Violence 66 C. Drivers 69 D. The Institutional Response 71 E. Financing of Security and Justice 73 F. Assessing the Outcomes 77 G. Conclusions and recommendations 83 Chapter 5: Budget Process, Public Financial Management and Procurement 85 A. Introduction 86 B. Budget Process and Public Financial Management 88 C. Public Procurement 100 D. Conclusions and recommendations 107 List of references 110 Annexes 115 Annex 1: Results of OLS Regression 115 Annex 2: Data Envelopment Analysis (DEA) 116 Annex 3: Impact of Recent Natural Disasters on Road Sector 117 Annex 4: Public expenditures on citizen security 119 Map 121 List of Figures Figure 1.1: Governance indicators for Guatemala (percentile rank 0-100) 24 Figure 2.1: Central government tax revenue by type of tax (% of GDP) 31 Figure 2.2: Central government expenditure – functional classification (% of GDP) 39 Figure 2.3: Social expenditure in Central America 41 Figure 2.4: Municipal spending by functional classification (2009) 41 Figure 2.5: Municipal spending by economic classification (2009) 41 Figure 2.6: Municipal and central government per capita social expenditure and poverty, by department (2009) 41 Figure 2.7: Projected public debt service 42 Figure 3.1: Central government spending on education, by function (% of GDP) 46 Figure 3.2: Distribution of education expenditure by department (average 2008-09) 47 Figure 3.3: Relationship between completion rates (grade 9), adult literacy, spending on education per pupil (bubble size) and malnutrition rates (color) in selected municipalities (2010) 49 Figure 3.4: Relationship between test scores of graduates, adult literacy, spending on education per pupil (bubble size) and malnutrition rates (color) in selected municipalities (2010) 49 Figure 3.5: Distribution of central government health expenditure and doctors by department (average 2008-09) 53 Figure 3.6: Relationship between vaccination coverage, poverty, per capita spending on health (bubble size) and malnutrition rates (color) by department 54 Figure 3.7: Relationship between infant mortality, poverty, per capita spending on health (bubble size) and malnutrition rates (color) by department 54 Figure 3.8: Central and municipal expenditure in road sector by department (in quetzales per capita, 2008-09) 59 Figure 4.1: Central American homicide rates (2000-2011) 67 Figure 4.2: National Government expenditure on security and justice (2006-2010) 73 Figure 4.3: Security budget in Central American countries(2006-2010) 74 Figure 4.4: National Government expenditure on security, by institution (2006-2010)* 74 Figure 4.5: National Government expenditure on justice, by institution (2006-2010)* 74 Figure 4.6: External funding for citizen security (2005- 2011) 74 Figure 4.7: Spending on prevention and crime control (2006-2010) 76 Figure 4.8: Impunity rate for all crimes 79 Figure 4.9: Unreported crimes in Guatemala 80 Figure 4.10: Trust in the National Police (2010) 81 Figure 4.11: Police force per 100,000 habitants 81 Figure 5.1: CGC audit opinion regarding budgeted revenues and expenditures of decentralized and autonomous entities (2006-10) 95 Figure 5.2: Central government fideicomisos: resources allocated by primary purpose 95 Figure 5.3: Autonomous and decentralized fideicomisos: resources allocated by primary purpose (2004-10) 95 Figure 5.4: Performance of the Guatemalan public procurement system 102 Figure 5.5: Performance of the Guatemalan public procurement system 102 Figure 5.6: Institutional procurement practices – amounts contracted by method 105 Figure 5.7: Budgetary performance of MICIVI procurement (Q. million) 106 Figure 5.8: MICIVI contracting practices - contracts amendments 106 List of Tables Table 1.1: Central America: key social indicators* 20 Table 1.2: Key macroeconomic indicators 23 Table 2.1: Central America central government revenues (% of GDP) 28 Table 2.2: Revenue impact of tax reforms and reversals by the Constitutional Court 31 Table 2.3: Tax effort in Latin America 33 Table 2.4: Tax expenditures by type of tax (current Q. millions)^ 34 Table 2.5: Progress towards economic, fiscal and social objectives of the Peace Accords 37 Table 2.6: Rigidity classification of the 2011 budget (as of September 2011) 38 Table 2.7: Central government expenditure – economic classification (% of GDP) 38 Table 3.1: Central government spending on health* 52 Table 3.2: Condition of the road network in selected Latin American countries 57 Table 3.3: Public expenditure on road transport in Central American countries* 57 Table 3.4: Central government expenditure in the road transport sector (Q. millions) 59 Table 3.5: MICIVI’s investment and maintenance expenditures under Prog. 11 (Q. millions) 60 Table 3.6: Expected impact of alternative financing options 62 Table 4.1: Budget allocated to security and justice functions (Q. million, 2011) 76 Table 4.2: Modifications and execution of security budget (Q. million, 2010) 78 Table 4.3: Effectiveness in the resolution of cases filed in first instance, according to judicial categories, (2005 & 2009) 82 Table 5.1: Deviations of actual from originally approved budget (2005-2010) 92 Table 5.2: Expenditures of Central Government fideicomisos (2004-2010) (Q. millions) 97 Table 5.3: Expenditures executed by NGOs and other agreements 97 Table 5.4: Fideicomiso audits performed by the CGC (2007-2010) 99 Table 5.5: Procurement by method (% of total) 10 List of Maps Map 3.1: 9th grade completion rates vs. graduate test scores at municipal level (2010) 49 Map 3.2: Technical efficiency in education across Guatemalan municipalities (2009) 50 Map 3.3: Technical efficiency in Guatemalan municipalities 51 Map 3.4: Per capita expenditure on health services by department (average for 2008-09) 53 Map 3.5: Vaccination and infant mortality rates by department 54 List of Boxes Box 2.1: A short history of fiscal reforms post-Civil War 30 Box 2.2: Key provisions of the 2012 tax law 34 Box 2.3: The government’s reform program for fiscal transparency 36 Box 5.1: PFM strengths and weaknesses identified by the Guatemala PEFA (2009) 87 Acknowledgements This report was prepared by a team led by Jasmin Chakeri (LCSPE) under the overall super- vision and guidance of C. Felipe Jaramillo (Country Director, LCC2C), Rodrigo A. Chaves (Sector Director, LCSPR), Auguste Tano Kouame (Sector Manager, LCSPE), Oscar Calvo- Gonzalez (Lead Economist and Sector Leader, LCSPR) and Oscar Avalle (Country Manager, LCCGT). The peer reviewers were Emily Sinnott (ECSH4), Ulrich Lachler (EASPR), Rashmi Shankar (ECSP1) and Bernard Harborne (AFTCS). The core team included David Gould (SARCE), Miguel Eduardo Sanchez Martin, Luisa Fe- lino, Mateo Clavijo, Patricia Chacon Holt (all LCSPE), Rodrigo Serrano Berthet (LCSSO), Joao Veiga Malta (LCSPT), Massimo Mastruzzi (WBIOG), Ana Silvia Aguilera (consultant), Jose Eduardo Gutierrez Ossio (LCSPS), Carlos Perez-Brito (consultant) and Christian Schus- ter (consultant). Inputs and background papers were also received from Ricardo Barrientos (Instituto Centroamericano de Estudios Fiscales, ICEFI), Juan Belikow (consultant), Simone Bunse (consultant), Maynor Cabrera (ICEFI), Valentina Calderon (consultant), Verena Fritz (PRMPS), Katherine Grau (LCCGT), Hernan Pfluecker (consultant), Hector Rosada (consul- tant), Alys Willman (SDV) and Melissa Zumaeta (consultant). The team would like to thank Roland Clarke (LCSPS), Steven Brushett (LCSTR), Phil Keefer (DECMG), Alberto Leyton (LCRVP), Fernando Paredes (LCCGT), Mariano Rayo (consultant) and David Yuravlivker (consultant) for their valuable guidance and comments during the preparation of this report. Executive Summary A. GUATEMALA’S DEVELOPMENT system. Guatemala also continues to lag behind its CHALLENGES peers in terms of poverty incidence, inequality and hu- man development outcomes, especially in indigenous ES.1. Guatemala has made tremendous strides in areas. Concerns over rising crime and violence are moving toward democracy and building institutions also becoming more and more prominent. after a long civil war. The Peace Accords identified an ambitious agenda for reconciliation and reincorpora- B. OVERVIEW OF PUBLIC FINANCES tion of excluded groups, human development, sustain- able economic development and modernization of the ES.5. In the decade and a half since the Peace Ac- state and improved governance. Since the signing of cords, both tax revenue and expenditure have in- the Accords in 1996, progress has been made in re- creased. Guatemala has been able to gradually raise ducing poverty (especially in urban areas), improving a its tax intake thanks to incremental reforms: the tax number of human development outcomes, and main- to GDP ratio increased from 8.8 percent in 1995 to 12.1 taining a stable macroeconomic environment. percent in 2007 and 11.0 percent in 2011. Notably, the rise in revenues has taken place despite the absence ES.2. Economic growth has been weak despite pru- of a major overhaul of the tax system and despite a dent macroeconomic management. Since 1990, real gradual decline in import tariffs resulting from trade GDP growth per capita in Guatemala has averaged liberalization and regional integration. The increase in about 1.2 percent, about three-quarters of a percent- revenues was accompanied by an expansion in spend- age point less than the rate in the rest of Latin Ameri- ing to meet the commitments of the Peace Accords: ca and the Caribbean and significantly less than other total central government expenditure increased from middle-income countries. On the other hand, volatility 10.3 percent to 14.6 percent in the same period, with has been much lower than elsewhere in Latin Ameri- the social sectors accounting for a significant share of ca; this relative stability can be attributed to prudent this increase. macroeconomic policies that have kept inflation and public debt at manageable levels, while avoiding fiscal ES.6. While the increase in expenditure has helped imbalances that plagued much of the region. improve key human development indicators, several of the Peace Accord targets in the social sectors ES.3. Advances in structural reforms have been have not been achieved. As found in previous Pub- made. Guatemala has made progress in improving its lic Expenditure Reviews, Guatemala has been able to investment climate, ranking among the top reformers make significant improvements in most social indica- worldwide in Doing Business indicators in 2006 and tors since 1996, albeit not at the pace foreseen in the 2007. In addition, in 2006 Guatemala implemented Peace Accords. While the targets for literacy, gross the Dominican Republic-Central America Free Trade primary enrollment, and vaccination coverage for po- Agreement with the United States (DR-CAFTA) that lio were met or exceeded, those for infant and mater- helped secure a more open trading environment be- nal mortality have not yet been achieved. This raises tween Central America and the United States. Foreign questions about the quality of public expenditures, direct investment (FDI) increased nearly threefold especially in the social sectors, including issues relat- from 2003 to 2010. Financial sector legal and regula- ed to geographic targeting and technical efficiency. It tory reforms implemented since 2000 have improved also emphasizes the importance of improved monitor- banks’ solvency and resilience to market volatility. ing and evaluation mechanisms that can feed back into Public Expenditure Review Improvements have also been made with respect to policy design. transparency, especially in the area of public expen- diture through the introduction and expansion of the ES.7. Despite the progress made since the Peace integrated financial management system. Accords, Guatemala is still among the countries in the region with the lowest tax revenues, and the ES.4. Still, challenges remain. Key institutional ob- lowest per capita spending on the social sectors. At stacles to effective policy making remain, however, around 11 percent of GDP, Guatemala’s tax-to-GDP ra- including the instability in the country’s political party tio is well below the Central American average of 13.3 7 percent, and far lower than the average for all of Latin sustainable growth, the allocation of expenditures by America (19.2 percent).1 This low level of revenues is function seems to be broadly in line with the govern- reflected in the limited public expenditures, with Gua- ment’s development objectives: over 50 percent of temala occupying the bottom position in most interna- total spending goes to the social sectors, for instance. tional comparisons. However, when measured as a share of GDP, public expenditures in most priority areas (including educa- ES.8. Unable to build consensus and political sup- tion, health, transport and citizen security) lag behind port for comprehensive tax reforms, successive other countries in the region. governments passed a series of piecemeal mea- sures aimed at improving collections. These include ES.12. The regional distribution of public expendi- changes both in tax policy (which were frequently ture shows large disparities. Central government ex- reversed through legal challenges) and improve- penditure allocations tend to disproportionally bene- ments in tax administration. These measures did fit geographic areas that are not the neediest in terms not, however, bring about a major shift in the struc- of poverty or social indicators. While municipal gov- ture of Guatemala’s tax system, which relies heav- ernments do have limited own revenue assignments, ily on indirect taxes, especially the VAT, which ac- their budgets tend to be small and cannot make up counts for about half of total tax revenues. Direct for these disparities, especially in the social sectors. taxes – especially the income tax – remain hampered However, in sectors such as transport, municipal in- by small tax bases, numerous exemptions, ram- vestments do play an important role. pant evasion and widespread economic informality. ES.13. Conservative fiscal management has led to ES.9. However, the tax reforms approved in early relatively low public debt levels. However, debt lev- 2012 represent an important milestone. While the els have been rising. Moreover, Guatemala’s optimal direct revenue impact of the reforms is expected to public debt level may lie below that of other countries be modest (up to 1.3 percent of GDP by 2015), the of a similar income level, given its low tax collections. reforms are an important step towards improving in- come taxation as well as further strengthening tax and Recommendations customs administration, which will reinforce the coun- try’s tax system in the long run. ES.14. Measures aimed at increasing tax collections and improving the quality of expenditure are critical ES.10. The recent tax reforms are to be accom- in order to enable Guatemala to meet the targets of panied by a set of measures to improve public ex- the Peace Accords, enhance equity and improve so- penditure management. In recent years, the debate cial outcomes. The approval of the recent tax reforms, about tax reforms has been closely tied to the need and the government’s commitment towards strength- for improvements regarding expenditure efficiency ening transparency, are important steps in the right and transparency. A key argument used by opponents direction. However, in light of the low estimated rates of the tax reform was that an increase in tax revenue of tax effort, significant tax expenditures and rampant would only contribute to inefficient spending, corrup- evasion, institutional strengthening of the tax adminis- tion and – by extension – mediocre outcomes in public tration will be needed to ensure that the reforms can service delivery. In March 2012, the Executive submit- be implemented effectively. In addition, it will be im- ted a proposal to Congress, which includes reforms to portant to secure the approval of the key components Public Expenditure Review the civil service, public trust funds, procurement, and of the transparency package submitted to Congress. sanctions against illicit enrichments by public officials. ES.15. Better information is needed to improve ex- ES.11. Expenditure composition seems to be less penditure quality and adequately assess the region- of a problem than the overall small expenditure en- al distribution of public spending. Given the existing velope. While capital expenditures (at around 4 per- deficiencies in the integrated financial management cent of GDP) are considered insufficient to support system (SIAF) and the integrated accounting system 8 1. OCDE (2011). (SICOIN), it is difficult to obtain comprehensive infor- tion. Most of this spending takes place at the central mation on resource allocations by municipality or de- level, with municipalities accounting for less than 11 partment. Availability of sectoral outcome indicators percent. Incidence analysis suggests that education at the sub-national level is even more limited. In order expenditure is overall poverty neutral. While primary to allocate scarce resources in the most effective way, education expenditure is found to be progressive (i.e. better information is required and needs to be linked disproportionately benefiting the poor), spending on to budget preparation. secondary and tertiary education is regressive. ES.16. Finally, investment in infrastructure should ES.19. Preliminary analysis suggests that there is be boosted. The new government is committed to no clear relationship between the level of spending linking its medium-term debt management strategy and student achievement at the municipal level. For more closely to its investment plans, which would help example, the two municipalities of San Sebastián and utilize financing in a more efficient way. However, giv- Tectitán in the Department of Huehuetenango show en the limited public resources available, Guatemala’s very different 9th grade completion rates (12.4 per- investment needs cannot be met by the government cent and 35.5 percent, respectively) even though both alone; private sector participation will be important record very low adult literacy rates and high malnutri- (see Chapter 3). tion rates, and even though the municipality of San Se- bastián spends more per pupil than Tectitán. A similar C. THE QUALITY OF PUBLIC EXPENDITURE IN relationship can be found when looking at test scores SELECTED SECTORS of graduates as the main output indicator. The analy- sis also suggests that high performance in completion ES.17. Guatemala remains characterized by low de- rates is not necessarily correlated with high quality, velopment outcomes, considering its income level. as measured by higher test scores. In addition, higher In the areas of education, but especially in health and graduation rates and higher test scores are not neces- nutrition, Guatemala performs significantly worse than sarily correlated. When technical efficiency scores are other countries in Latin America. Its infrastructure de- mapped against overall spending and indices of need velopment is also lagging behind, and is negatively im- (extreme poverty rates) at the municipal level, two in- pacted by the natural disasters that affect the country teresting lessons emerge: first, several municipalities every year. While there are a multitude of factors that appear to be inefficient whether measured against affect these outcomes, both overall expenditure levels need or spending, or a combination of both. Second, as well as expenditure quality can play an important there appears to be regional clustering among effi- role. While overall spending levels may be a binding cient and inefficient municipalities. constraint, efficiency improvements would be benefi- cial in getting “more bang for the buck�. In particular, ES.20. The variation in educational outcomes does there is great variation in the efficiency2 of health and not appear to be primarily driven by public educa- education spending at the sub-national level. In the tion spending. Demographics and other socio-eco- transport sector, there is also ample room for improv- nomic indicators may be responsible for determining ing the quality of public expenditure. such large variations in performance. This is confirmed by regression analysis, which shows that education Education spending is not statistically significant with respect to explaining student achievement, as opposed to the ma- Public Expenditure Review ES.18. Education spending, especially at the primary jority of demographic and socio-economic indicators. level, has increased in recent years, in line with gov- ernment priorities. Since 2004, Guatemala has boost- Health ed its education spending from 2.6 percent of GDP to 3.2 percent, reflecting a strong focus on strength- ES.21. Despite a noticeable increase in public expen- ening social expenditure by the previous administra- diture on health and social assistance from 1.2 to 1.6 2. We define efficiency as follows: 1) allocative efficiency: the extent to which public expenditure allocations meet the government’s priorities and strategic goals; 2) technical efficiency: the relationship between inputs (public spending and other inputs) and outputs. 9 percent of GDP between 2008 and 2010, spending Road Transport levels remain low by international and regional stan- dards. Expenditures on health services in particular ES.24. Guatemala’s road network has deteriorated expanded between 2004 and 2008, and then sharply in part due to consecutive natural disasters. An es- increased by 25 percent in nominal terms in 2009. On timated 65-70 percent of the paved network is in bad average, the central government accounts for 96 per- condition. In the last years the score of the Guatema- cent of total public health spending. However, there is lan road network has deteriorated based on interna- a significant amount of geographic variation: municipal tional standards and in comparison with Latin Ameri- shares in total per capita public health expenditure can and Caribbean (LAC) countries. range from 1.1 percent in the department of Retalhuleu to 18.8 percent in Suchitepéquez. ES.25. Guatemala budgets about 1.4 percent of GDP per year for the road sector, which is rela- ES.22. Health expenditure and health outcomes do tively low. While Guatemala invests the second high- not appear to be strongly linked. Immunization rates est amount in US dollar terms (US$495.5 million as do not show a clear-cut relationship either with pub- of 2010) in Central America, as a share of GDP its lic spending on health or with poverty levels across investment (1.2 percent) is lower than that of Panama departments. Only infant mortality and under five (1.9 percent) and Nicaragua (1.8 percent), but above mortality indicators show a negative (but weak) rela- Honduras (0.9 percent) and El Salvador (0.3 percent). tionship with public spending on health. The percent- A recent IMF study estimates that Guatemala could age of children under care seems to be much more raise growth by over 1 percentage point by increas- related to mortality rates than variations in public ex- ing its investment-to-GDP ratio to the average for penditure per se. Results from the household survey Mexico, Peru and Chile.4 confirm that access to health care services remain a problem, especially affecting pregnant women in the ES.26. Financial needs in the sector are generally poorest quintile living in rural areas. Thus, efforts much higher than budgetary allocations, and there could be aimed at increasing access to health facilities is no contingency funding to cover the costs asso- and doctors in the less-favored areas. ciated with natural disasters. For example, DGC’s estimation of required financial resources for 2012 is ES.23. Weak health outcomes are concentrated in around four times the executed budget of 2010 (Q.8.2 the west and northwest of Guatemala. On the one billion), including signed contracts for reconstruction hand, Guatemala, Sacatepéquez, and Zacapa have works after Hurricane Agatha. Given Guatemala’s vul- higher spending on health, smaller poverty rates, low- nerability to natural disasters, investment needs in the er malnutrition rates and lower infant mortality rates; transport sector are driven to a significant extent by on the other hand, most of the departments in the damages caused by hurricanes, earthquakes and vol- west and northwest of the country exhibit high mor- canic eruptions. tality rates, clearly correlated with high malnutrition and an extremely high poverty incidence. The depart- ES.27. Budget overruns are common in road proj- ments presenting high infant mortality rates are also ects. Within DGC and COVIAL, it is general practice characterized by a low percentage of births attended that contracts are extended in terms of time and cost by doctors, even when the number of doctors in those increases. Overruns should be more the exception departments is not significantly lower than in the rest than the rule. The review undertaken by CoST in the Public Expenditure Review of the country. This suggests barriers in the access to baseline study found that among the 16 projects re- public health services among the poor that may be viewed in 8 agencies, the most significant variations of embedded in their own socioeconomic conditions. works contracts were within DGC and COVIAL, with This observation goes in line with other recent papers average overruns of 123 percent in time and 103 per- which also highlight access limitations for the poor in cent in cost. Guatemala.3 3. See, for instance, Bowser and Mahal (2011). 10 4. Swiston & Barrot (2011). Recommendations on education, health, and road decisions among programs and subprograms and transport among departments and municipalities in the country. ES.28. Better targeting of social spending is needed. ES.30. Greater results orientation of the budget The analysis suggests that public education and health could help ensure that resource allocations are spending does not benefit regions with the greatest aligned with sectoral priorities. It appears as though need. While some education programs actually reach inequities in resource allocations currently reinforce poor areas such as the school meal program, some historical trends and existing endowments. Regional others such as text books are concentrated in urban allocations of the health budget, for instance, have and semi-urban areas. Spending allocation is also car- been based on the number of beds and health ser- ried out without considering student performance vice entities in each location. As a result, the poor- and educational outcomes, such as completions rates est regions – which tend to have fewer health facili- in 6th and 9th grades. Similarly, in the health sector, ties – receive less funds, regardless of demographic there is a high concentration of public expenditure or epidemiologic needs.5 The recent progress to pilot in the departments of Guatemala, Sacatepéquez and results-based budgeting in the Ministry of Health is Santa Rosa. In the case of the department of Guate- promising and represents a step in the right direction. mala, this has to do to some extent with the fact that the country’s two largest hospitals are located there. It ES.31. There are a number of options that could be may also reflect the fact that health remains a central- analyzed in more detail to improve the condition of ized sector and some of its expenditures continue to the road network that could also reduce the fiscal be registered in the department of Guatemala (due to pressure for the sector. These options include (i) in- the location of the Ministry of Health and Public As- cluding private participation for the financing of new sistance), even though the recipient or beneficiary de- infrastructure (or significant work improvements) partment may be a different one. On the other hand, and charging user tolls; and (ii) combining rehabilita- except for Quetzaltenango, less money per capita tion and maintenance activities under performance- is allocated to departments in the southwest of the based multi-year contracts. In this context, the ap- country. The lack of clear prioritization in the alloca- proval of the Law on Public-Private Partnerships in tion of resources in the social sectors does not allow 2010 is an important step in the right direction, even for a maximization of the benefits for the most vulner- if its implementation will require the approval of de- able parts of the population. tailed implementing regulations, as well as extensive capacity building. ES.29. Monitoring and evaluation systems should be strengthened to facilitate deeper analysis of the D. CITIZEN SECURITY AND PUBLIC determinants of sectoral outcomes. A key finding EXPENDITURES from the analysis is that there exist large disparities in spending performance in education across the ter- ES.32. Crime and violence have become the most im- ritory. Disparities in the technical efficiency of public portant destabilizing factor for political, social, and education spending exist even after controlling for a economic life in Guatemala. Estimates suggest that variety of socio-economic factors, suggesting that bet- violence cost the country an estimated US$2.4 billion, ter data and more analysis would be needed to delve or 7.3 percent of GDP in 2005.6 This figure represents into the determinants of student achievement and more than double the damage caused by Hurricane Public Expenditure Review their linkages to education spending at the municipal Stan in the same year, and more than double the com- level. In the transport sector, the lack of outcome and bined budget for the Ministries of Agriculture, Health output information at the municipal level represents a and Education for 2006. The costs of crime and vio- challenge for decision makers to adequately measure lence to the business sector represent a serious drag the impact of public investment in the sector. System- on growth and development: the 2011 Global Com- atic information is needed to allow making informed petitiveness Report ranked Guatemala as the worst 5. World Bank (2012). 6. Acevedo (2008). 11 performer, out of 142 countries, in terms of the costs vulnerable groups, has produced the explosive situa- that crime imposes on businesses.7 However, these tion that the country finds itself in. economic costs pale in comparison to the immeasur- able human costs of Guatemala’s violence. In a 2010 ES.36. A review of government policies, budgets, report on Central America, the World Bank estimates outcomes and outputs suggests the need to rethink that the health costs, measured in Disability-Adjusted the approach the government has been following Life Years (DALYs), are around 1.43 percent of GDP, so far. Despite progress in some aspects or areas, the second highest in the region after El Salvador. overall, the government’s response has been frag- mented, discontinuous, and unbalanced. Confront- ES.33. Crime and violence in Guatemala is a multi- ing the challenge of criminal violence in Guatemala faceted problem that calls for a comprehensive requires a comprehensive approach that addresses policy response given its multiple forms and pres- the multiple dimensions of the problem through pre- ence across territories, demographic groups, and vention and law enforcement initiatives, coordinated income levels. In addition to being the country with by a clear authority within government, and with the the 7th highest homicide rate in the world, Guatemala capacity to oversee implementation of policies in a is afflicted by many other forms of crime and violence, consistent manner. from domestic violence and femicides to extortions, kidnappings, armed robbery, and all sorts of ille- Recommendations on citizen security gal trafficking. While each type of event tends to be clustered in specific geographic areas of the country, ES.37. There is a need to boost resources along when taken as a whole it is difficult to find places that the security and justice value chain, especially for: are free of any form of violence. Similarly, while male preventive programs and community policing; law youth tend to be particularly affected by homicides, enforcement’s manpower and equipment; policing women and other age groups are also affected by oth- operations; improved access to justice; improved er forms of violence. Finally, while victimization tends prosecutors office capabilities; and penitentiary and to increase with income level, poor people who can- rehabilitation systems. Major rethinking is required to not afford private security often have their neighbor- deal with the problem given the Guatemalan scarcity hoods controlled by criminal gangs, and are subject to of resources. everyday forms of violence. ES.38. Better coordination of prevention programs ES.34. Compared to its Central American neighbors, is also needed. The Ministry of Interior (MOI) should Guatemala’s security and justice expenditure is the develop a system to monitor the execution of all second highest in terms of overall budget share, but programs related to crime and violence prevention, the lowest in terms of GDP. El Salvador leads on both including those implemented by other sectoral min- metrics, spending more than 20 percent of its budget istries. The government should create an inter-minis- on security compared to Guatemala, Nicaragua and terial coordination mechanism to review this informa- Honduras that spent between 10 and 15 percent. El tion and identify areas for performance improvement. Salvador spent about 3.5 percent of its GDP compared A useful exercise for this group would be to assess to 2.3 percent in Guatemala and 2.8 in Honduras.8 whether the main risks related to crime and violence have a matching program or set of interventions to ad- ES.35. Governance institutions are being over- dress them and the extent to which these programs Public Expenditure Review whelmed by the ability of criminal groups to infil- are deployed in an integrated fashion in the hot spot trate them. Weak institutions have not been able to communities that concentrate these risks. The govern- cope with external pressures brought by increased ment should also consider whether (i) it is desirable to drug trafficking and this, combined with deep inequali- give formal authority for coordinating interventions to ty and lack of opportunities for youth at risk and other one of the ministries, for instance, authorizing it to be 7. World Economic Forum (2011). 8. There is a slight discrepancy between UNDP’s and our own calculations of the share of national budget that corresponds to security. This 12 is probably due to the inclusion of some different programs in either process. As a share of GDP, however, the numbers coincide. present in the decision making process and even re- security sector. Indicators could include: share of pub- source allocation of other ministries’ programs when lic expenditures per pillar; coverage in terms of num- they have an important citizen security function; and ber of personnel per capita assigned to each of the (ii) decide whether this role should be given to the pillars, whether police, public prosecutors, public de- Ministry of Interior. fendants, etc; managerial capabilities; recidivism rate; and citizen confidence in public institutions. ES.39. Guatemala should invest in a second genera- tion crime and violence information system. This E. BUDGET PROCESS, PUBLIC FINANCIAL system should (a) capture the range of crime that MANAGEMENT AND PROCUREMENT goes unreported, (b) validate the data that is gener- ated, (c) produce key descriptive information about Budget Process and Public Financial the characteristics of the crime, and (d) feed all this Management data into a geo-referenced, user-friendly system that is useful for pro-active, results-oriented decision mak- ES.41. Guatemala’s budget process is characterized ing on citizen security. Violence observatories are an by the co-existence of good practices and vulnera- institutional “technology� that has been proven use- bilities. It is supported by detailed norms and regula- ful to address this challenge. They usually combine a tions, an Access to Public Information Law and a good national level unit that works to coordinate with, and practice e-government system, the Integrated Finan- improve national government information systems cial Administration and Governmental Audit System with, a network of municipal observatories in hot spot (SIAF-SAG). Recent improvements to the SIAF, which municipalities that validate and use the data for local supports the budget cycle, mean that the main PFM citizen security strategies. Key descriptive informa- systems are now internally coherent, sophisticated tion needed for decision making include: type of inci- and powerful in recording and integrating financial dent (homicide, robbery, assault, car theft, rape, etc); management information for the entire public sec- date, time, day of the week, address of where incident tor. The SIAF provides timely, accurate, and compre- takes place, neighborhood; victim and perpetrator in- hensive financial and, increasingly, performance data formation (name, age, sex, occupation, literacy level, to officials and the public alike, and is a fundamental home address); type of weapon and context of event building block for the extension and consolidation of (interpersonal violence, domestic violence, organized modern results informed budgeting within the context crime, gang related, confrontations with law enforce- of multi-year framework. ment); and relationship of perpetrator to victim. The objective information captured by government sys- ES.42. Currently, annual budgets are formulated tems should be complemented with regular standard- based on past allocations. Traditional programs are ized surveys that produce subjective and situational not reassessed and revised, and an inertial budget information on victimization, perception of insecurity, execution has become common practice. Improved violent attitudes, coping behaviors to reduce victim- budget preparation would reduce the need to modify ization, and urban and social characteristics of hot allocations during the year and also enhance the cred- spot communities. ibility of the PFM systems. In addition, more realistic income projections would facilitate budget and cash ES.40. Guatemala should also strengthen its infor- management. mation system on the performance of security and Public Expenditure Review justice institutions. To assess the performance of ES.43. An initiative to improve the budget formula- citizen security institutions, government should build tion by revising the program structure of three pi- a core set of indicators for each of the pillars of the lot ministries (Health, Education, and Interior) has value chain (prevention, law enforcement, criminal yielded promising results. The progress achieved in investigation, criminal justice, penitentiary, rehabili- the Ministry of Health shows that using a results-ori- tation). These indicators should serve to benchmark ented approach is possible and useful. This approach progress and ensure greater dialogue and coordina- links policy objectives with the allocation of available tion across the multiple actors involved in the citizen resources, and provides a medium-term horizon for 13 achieving strategic goals. Further, better estimations the budget. Like fideicomisos, NGOs are not subject of unit costs would help monitor and evaluate the to regular budgetary oversight and controls. quality of expenditures by program and assess wheth- er intended results have been achieved. The program ES.46. The accumulation of payment arrears (“float- revision also includes investment projects to improve ing debt�) is a long-standing problem. During budget infrastructure. Thus, whether a ministry executes its execution sectors register commitments and accruals capital budget by itself or through a fideicomiso or when the payment order is processed. However, if NGO, this will make the public investment program there are not enough available funds, sectors may not more consistent with the sector’s desired results and register the invoice in the SICOIN system, resulting strategic goals. in an arrear not registered in the SICOIN, which be- comes floating debt. COVIAL generates floating debt ES.44. Reallocations within the budget during execu- in a different way, by the repeated extension of con- tion are highly discretionary and significant in scope. tracts associated with infrastructure projects without While constrained by budget rigidities such as legisla- the medium-term budget credits. The system does not tive earmarks and demands by public sector unions, alert public officials of the creation of an illegal accru- the executive may regain control over budget execu- al, and the credibility of the PFM system is seriously tion through intra-institutional and inter-institutional undermined. A complete inventory of floating debt is budget transfers. In view of the country’s exposure not available.9 to natural disasters, imperfections in budget planning and potential fluctuations in its (limited) revenues re- Recommendations on public financial management quiring spending adjustments, this executive power is well warranted. However, budget transfers lend them- ES.47. In order to deter the generation of floating selves to political bargaining, and make it difficult to debt, it is essential to enforce the law and sanction follow through on the policy objectives supposedly public officials who do not comply with it. The exist- represented by the budget. ing CGC Law must be amended to include sanctions that are substantial and credible. That would also ES.45. The use of trust funds and non-governmental make public servants, NGOs’ officials, and managers organizations (NGOs) has led to the circumvention of of fideicomisos more accountable. Also, due to the the Organic Budget and Procurement Laws, among fact that Guatemala has been confronting natural others, and raises transparency concerns. While disasters regularly, it is critical to better regulate the public trust funds (fideicomisos) are similar to private use of public funds and their accountability in those ones in that the Government (as fideicomitente) trans- situations. In particular, it would be important to make fers funds and rights to the fiduciario, there are a num- fideicomisos subject to public procurement rules. ber of important differences. First, public expenditure is generally required to follow public financial manage- ES.48. The Government should continue its efforts ment and procurement procedures; however, these to build consensus around a public fideicomisos Law. are not applied to public fideicomisos because they Most of the regulations that might be included in the are regulated by private sector law. Similarly, monitor- draft law are stated in the 2011 and 2012 budget laws. ing and accountability mechanisms (including audits) However, there is no guarantee that Congress will that would generally apply to public spending are not include such regulations in future years. A new Law applied equally to public fideicomisos. As a result, would provide a permanent legal basis governing the Public Expenditure Review there is limited information on the execution of public management of fideicomisos. spending through fideicomisos, compared to regular public spending carried out through the SICOIN sys- ES.49. Based on the experience with the pilot pro- tem. Public contracting of NGOs is another commonly gram in results-based budgeting, it is highly rec- used vehicle to execute public resources outside of ommendable that the MOF drafts a new PFM leg- 9. The Government has contracted an international audit firm to conduct a comprehensive audit of floating debt; however, the audit could 14 only be completed for a portion of Central Government ministries. islation to improve the PFM systems performance, that shifts the objective of the procurement system drawing from the lessons learned from the pilots’ from compliance with processes to compliance with experience. An improved medium-term expenditure outcomes. It is important to note that shifting the fo- framework (MTEF) will support the revision of pro- cus to outcomes does not imply that compliance with grams as well as all budget cycle processes. procedures is not important; rather, it means that both outcomes and procedures will become measur- Procurement able results. ES.50. There is a need to improve planning, consoli- ES.54. By redesigning the operational model for dation, competition and contract management in framework agreements, the government could le- order to enhance the efficiency and transparency verage its e-procurement platform, Guatecompras, of Guatemala’s public procurement system. Whilst to implement whole-of-government framework Guatemala was a pioneer in the introduction of frame- agreements for common use goods and services. work agreements in the Latin American region, it nev- This would eliminate the use of direct contracting and ertheless failed to capitalize on the potential benefits exceptions for the procurement of common use goods generated by the use of this tool. Design flaws in the and services, and thereby result in increased efficien- planning and operational model contributed to this. cy, reduced prices and increased transparency. This The decline of framework agreements is reflected in approach could also be applied to the procurement of the decreasing use of this tool by public sector enti- pharmaceuticals. ties from a maximum of nearly 7 percent of contracts in 2009 two less than 0.04 percent in 2011. ES.55. A review of the significant cost overruns, es- pecially in infrastructure, should be undertaken to ES.51. There has been a growing tendency to either identify their causes and terminate any contracts procure by exception or limited competitive meth- which should have been re-bid rather than extend- od rather than through open tendering processes. ed. Also, there is a need to implement stricter proce- The use of exceptions is particularly prevalent in infra- dures in the contract management system to ensure structure, whilst direct procurement tends to be as- that contracts cannot simply be extended or incur sig- sociated with goods and services. Likewise, there is a nificant overruns without the proper public scrutiny. significant reduction in the use of open contracts; this Another practice that should be avoided is the enter- reduction is accompanied by an increase in requested ing of contracts into the system after they have been exceptions for the procurement of items outside the complete (rather than when they are signed) only in open contract. order to make payments. ES.52. In the area of infrastructure, one of the key ES.56. Another key area for reform is the develop- deficiencies identified is the apparent lack of ac- ment of a cadre of procurement specialists whose countability in contract management. It appears qualifications are founded on a competencies based that, whilst the government captures the entire con- system that is linked to career advancement as well tract management data related to infrastructure proj- as responsibility. The adoption of such a certification ects, it does not use this data as a management tool. system would ensure that public officials tasked with The review identified a large number of contracts with carrying out procurement activities are trained and significant overruns, some of which were many times thus knowledgeable on the application of a wide range Public Expenditure Review the originally contracted amount. of procurement methodologies. This initiative would also serve to create specialized units that would in Recommendations on procurement turn have the skills and knowledge to address sector specific issues such as the procurement of pharmaceu- ES.53. In order for Guatemala to have a procure- ticals, setting up and administering framework agree- ment system that delivers the government’s invest- ments, as well as complex infrastructure projects. ment program as well as the goods and services de- manded by society, substantial reform is required 15 Chapter 1 Guatemala’s Development Challenges Ahead 1 » A. Introduction 1.1 While Guatemala has made tremendous strides in moving toward democracy and building institutions after a prolonged civil war, much remains to be done in generating the foundations for strong broad-based economic growth. Since the signing of the Peace Accords in 1996, progress has been made in reducing poverty (especially in urban areas) and maintaining a stable macroeconomic environment. However, the country still ranks as one of the lowest in Latin America and the Caribbean (LAC) in terms of basic social indicators. Indeed, largely due to its historical legacy of conflict and exclusion, over half the population is below the poverty line and measures of inequality are higher than the regional average. According to the 2010 UNDP Human Development Report, Guatemala ranks among the bottom three positions in LAC, along with Nicaragua and Haiti, in the overall human devel- opment index, including infant mortality, maternal mortality, adult literacy, and net second- ary enrollment. 1.2 As Guatemala enters its fifth presidential cycle since the 1996 Peace Accords, the newly elected administration faces tremendous economic challenges. Efforts during the previous administration to directly address poverty and improve education and health out- comes, particularly in the indigenous and rural areas, were intense, with tentative signs of success. However, these efforts, particularly the introduction and rapid expansion of the government’s conditional cash transfer program (Mi Familia Progresa, MiFaPro), also gen- erated an added fiscal burden and much political controversy regarding transparency and social benefits.10 Better monitoring and evaluation of all development programs is needed to provide the foundation for broad political support. 1.3 This introductory chapter reviews Guatemala’s development challenges. The idea is to place Guatemala in its own historical context and determine how the economy has evolved over the last fifteen years since the Peace Accords, in what areas of development the country made progress, where the country lagged, and what are the key challenges for future development. Subsequent chapters will delve more deeply into these issues examin- Public Expenditure Review ing the efficiency of public expenditures, spending on citizen security, and public financial management and procurement. 10. While the direct cost of the conditional cash transfers is a modest 0.4 percent of GDP, the additional supply of education and health services have led to a larger increase in social spending as a share of GDP. 17 B. HISTORICAL AND POLITICAL CONTEXT Some econometric analyses done by Bank staff indi- cate that had the war not occurred, per capita GDP 1.4 To understand why Guatemala has taken the would have been about 40 percent higher and pover- development path that it has, it is important to look ty 12 percent lower by 2000.11 Social indicators lagged at the fundamental roots that may be underlying even further behind growth, with Guatemala remain- its development process. Extractive colonial institu- ing among the worst performers in the region on most tions and a radical liberal reform period in the 1800s education and mortality measures. contributed to weak institutions and property rights, particularly with respect to land and labor, which led 1.6 The signing of the Peace Accords in 1996, fol- to an exclusionary pattern of development. In the late lowing a protracted negotiations process, marked 1800s and the first half of the 20th century, with the a turning point in Guatemala’s development path. spread of coffee cultivation, the practice of expropri- Beyond formally ending the armed conflict and ad- ating land and requiring forced labor from indigenous dressing related political issues, the Peace Accords communities accelerated. In turn, the pattern of large identified an ambitious agenda for reconciliation and plantations owned by small elite and manned by reincorporation of excluded groups, human develop- subsistence-wage indigenous workers provided little ment, sustainable economic development and mod- incentive for land owners or the workers themselves ernization of the state and improved governance. to invest in human capital. Thus, investments in edu- Three cross-cutting themes were emphasized through cation for indigenous groups lagged far behind other the Accords: (i) the rights of indigenous communities; countries in the region which relied on a small-holder (ii) commitments regarding the rights and position of model for developing their coffee sector (e.g., Costa women, and (iii) a strengthening of social participa- Rica and Colombia). The situation was exacerbated tion. Among the various accords concluded within the by discriminatory policies which reserved education framework of the peace process, the Social and Agrar- rights for “citizens�, a status not fully accorded to the ian Issues Accord (May 1996) established the overall indigenous or women until 1945. development agenda, with stronger social orientation and a general goal of closing the huge gap between 1.5 By 1960, when Guatemala’s civil war broke out, the rich and the poor. The Accord covered a range of the country was lagging well behind comparable issues including growth, tax revenues, public spend- countries on virtually all key social indicators. It was ing, education and health (each with monitorable tar- even faring worse than many much poorer neighbors gets).12 It also emphasized broader citizen participation (in GDP per capita terms) such as Nicaragua and Hon- in development projects through the establishment of duras. Guatemala’s civil war was one of the longest and decentralized development councils. With respect to bloodiest in Latin America. It is estimated that about land, the Accord focused on market-assisted land dis- 2 percent of the population (over 200,000 people) tribution, creation of a Land Fund and improved land were killed or disappeared, over a million (close to 10 administration. percent) were displaced and about 600 villages and communities completely destroyed. Indigenous com- 1.7 However, Guatemala’s institutional environ- munities bore the brunt of the violence, suffering the ment continues to complicate the cooperation of po- vast majority of casualties and damage or property litical actors over time and thus the inter-temporal loss. While it is impossible to quantify the full range bargains13 required to agree upon and sustain struc- of social and economic impact of the civil war, the His- tural reforms. Four key features particularly stand Public Expenditure Review torical Clarification Commission estimated that dur- out: (i) mechanisms for the aggregation of individual ing the 1980s alone, the costs of the war amounted to and group interests into public policies which multi- 15 months of production or 121 percent of 1990 GDP. ply the number of actors impacting the policy-making 11. López (2001). 12. See Chapter 2 for more information on these targets. 13. Inter-temporal reform bargaining refers to the idea that reforms frequently require the exchange of current actions or resources – such as legislative votes for a reform – for promises of future actions or resources – such as benefits accruing from a reform or concessions in other 18 policy areas (see Inter-American Development Bank 2008). process; (ii) informal rather than well-institutionalized nature of parties and party membership; survey data arenas for political exchange; (iii) exceptionally short indicates that partisan allies in Guatemala hold as political time horizons; and (iv) lack of credible en- many different opinions as they share on key socio- forcement mechanisms for political and policy agree- economic and political issues.18 Citizen identification ments. These four features in turn are a function of an with parties is impaired – with Guatemala ranking third inchoate party system as well as a judiciary and bu- from the bottom in Latin America– and the legislative reaucracy lacking capacity and autonomy. branch enjoys extraordinarily low levels of legitimacy – citizens rank political parties as the least legitimate of 1.8 The inchoate political party system plays an im- all Guatemalan public institutions.19 portant role in the weak institutional environment. Driven by the concurrence of a resource-rich private 1.10 Political parties tend to be electoral vehicles sector elite and the limited ability of citizens to act rather than parties proper.20 Parties are linked to collectively, Guatemala’s political party system is char- a popular leader and (frequently) founder; once the acterized by “high electoral volatility, severe party in- leader leaves office, the party disintegrates or is taken stability, weak links to society, ideological vagueness, over by factions. Electoral rules prohibiting presiden- limited territorial presence, lack of legitimacy and tial re-elections reinforce this effect and the concomi- opaque financing mechanisms�.14 These weaknesses tant brief political time horizons. As such, parties are can be observed in multiple ways. Guatemala’s party unable to mediate between society and the state. system undergoes average election-to-election shifts Rather than fulfilling programmatic parties’ functions in votes of close to 50 percent – compared to a Cen- such as representation and interest aggregation, par- tral American average of 21 percent.15 No governing ties are vehicles for traditional forms of representation party has ever secured re-election, political parties in which access to the state is a function of resources. survive on average 7.6 years and more than 50 have With underprivileged sectors of society deprived of been declared void since the advent of democratiza- institutionalized political parties as a key means to tion.16 As a result, the majority of elected legislators overcome collective action problems21, access to the have no prior parliamentary experience. This com- state is monopolized by those who can afford it. pounds the weakness in technical capacity among members of Congress and limits the number of law C. ECONOMIC DEVELOPMENT CHALLENGES initiatives emerging from Congress. Poverty and Inequality 1.9 Guatemala’s party system is also the most frag- mented in the region. This stems in part from a prac- 1.11 Largely due to Guatemala’s historical legacy, tice of ‘cross-voting’ by citizens; reflecting low levels poverty and inequality are high and social indicators of trust in parties, the parliamentary vote increas- (health, nutrition, education) are low compared to ingly differs from the presidential vote. As a result, all countries of similar income. According to the 2010 presidents since 2004 have headed minority govern- UNDP Human Development Report, Guatemala ranks ments.17 Party fragmentation is, however, also a func- among the bottom three positions in Latin America, tion of high levels of party defection. More than 25 along with Nicaragua and Haiti, in the overall human percent of members of Congress switched parties development index, including infant mortality, mater- each year over the last decade on average; in 2004 nal mortality, adult literacy, and net secondary enroll- and 2007, this share even exceeded 40 percent. Par- ment. Guatemala also ranks low on social indicators Public Expenditure Review ty defection in turn is related to the non-ideological within Central America, according to the World Devel- 14. ASIES (2009). 15. Achard and Gonzalez (2004). 16. Insight (2011); Menocal and Calvaruso (2008). 17. Cabrera (2011). 18. García Díez (2001). 19. LAPOP (2011). 20. Sánchez (2008). 21. See Keefer (2011). 19 Table 1.1: Central America: key social indicators* Average GTM SLV HND NIC CRI PAN CA LAC LMICs Population Growth 2,5 0,5 2 1,4 1,3 1,6 1,6 1,1 1,5 Education Literacy rate, adult total (% of people ages 15 and above) 76 84 84 78 96 94 85 n.a. 71 School enrollment, primary (% net) 97 93 96 92 90 98 94 94 85 School enrollment, secondary (% net) 40 55 20 46 46 69 46 73 50 Health Life expectancy at birth, total (years) 71 72 73 73 79 76 74 74 65 Maternal mortality ratio (modeled estimate, per 100,000 live births) 110 110 110 100 44 71 91 85 300 Mortality rate, infant (per 1,000 live births) 25 14 20 23 9 17 18 19 52 Malnutrition prevalence, weight for age (% of children under 5) 19 6 9 4 5 6 8 4 24 *Data for the most recent year available. Source: World Development Indicators and national sources. opment Indicators (Table 1.1). Though social indicators dicators in rural areas has been slower than progress remain low, some improvements were made in recent country-wide. years, particularly in the overall poverty rate. 1.13 Indigenous people suffer from the highest pov- 1.12 There were important differences in poverty erty rates and have the lowest social indicators. Ac- outcomes by regions and socio-economic groups cording to the 2011 Living Standards Measurement within Guatemala in recent years. Overall poverty Survey (ENCOVI), the incidence of poverty for indig- fell from 56 to 51 percent between 2000 and 2006, enous households is around 56 percent, significantly or 0.85 points per year, while extreme poverty re- higher than the non-indigenous rate of 44 percent. mained nearly unchanged at slightly above 15 percent. The gap in health and nutrition is even more worri- Much of the decrease in moderate poverty was due some. Chronic malnutrition (percent of children less to economic growth and the increase in wage remit- than 5 years of age with stunted growth) is about 65.9 tances, particularly from the United States, which in percent for indigenous children compared to 36.2 2006 rose to over 12 percent of GDP. However, as a percent for non-indigenous children.22 There are also result of the global economic crisis, moderate poverty differences among various indigenous groups. For ex- crept up again to 54 percent by 2011, while extreme ample, the Garífunas and Sakapultekos have on aver- poverty fell to 13 percent. Regionally, extreme poverty age about six years of schooling, while the Chuj and is most prominent in the North and Northeast, with the Ch’orti’ have less than two years. rates of over 20 percent; while poverty rates tends to be lowest in the Metropolitan and Central regions. 1.14 Indigenous women face particular constraints. Public Expenditure Review Regional differences in labor productivity, a function The 62 percent illiteracy rate for indigenous women of infrastructure, education, and other factors, as well is extremely high compared to 24 percent for non-in- as climate conditions help explain much of these dif- digenous women, and 43 percent for indigenous men. ferences in poverty outcomes. Poverty and inequality There is also evidence that indigenous women are remain concentrated in rural areas and among indig- disadvantaged in the labor market; their labor mar- enous groups. The pace of improvements in social in- ket participation rate is almost half that of men (i.e., 20 22. UNICEF and ICEFI (2011). 35 percent for indigenous women versus 67 percent could be met. However, reaching maternal mortality for indigenous men). When indigenous women do par- goals and reducing the spread of HIV would remain ticipate in the labor force their salaries tend to be 53 a challenge. Under a more conservative scenario that percent less than their indigenous male counterparts, assumes an average economic growth rate of three even accounting for educational background and job percent and policies costing 1.1 percent of GDP, only activity. the poverty goal could be achieved. The Bank’s 2005 Country Economic Memorandum (CEM) indicates 1.15 Though much remains to be done to improve that relying exclusively on growth to achieve the MDG social indicators for all groups, Guatemala has of cutting poverty by half by 2015 would require an shown important gains. Recent advances, even if average annual per capita growth rate of 2.4 percent, small, suggest that the Government has been able to a goal that Guatemala has infrequently achieved.23 improve outcomes. For example, primary enrollment rates rose from 79.9 to 95 percent and infant mortal- Growth and Stability ity rates declined from 39 to 33 percent during the same period. The 2008 Poverty Assessment bench- 1.18 Over the past two decades, economic growth marked these rates of change to historical rates of has been relatively stable but weak. Since 1990, real change for all countries globally over the last 25 years, GDP growth per capita in Guatemala has averaged and Guatemala’s performance compares favorably. about 1.2 percent, about three-quarters of a percent- Many improvements in Guatemala were better than age point less than the rest of Latin America and the 80 percent of all the advances recorded. The global Caribbean and significantly less than other middle- economic crisis, however, may have placed some of income countries. Despite weak growth, volatility was these advances at risk. less than half the regional average and experienced less of a decline compared to the rest of Latin Amer- 1.16 Poverty reduction programs are constrained by ica during the recent crisis.24 Much of Guatemala’s budget rigidities and low social spending as percent- relative stability can be attributed to prudent macro- age of GDP. About 88 percent of the budget is ear- economic policies that have kept inflation and public marked by law. Social spending, although increasing in debt manageable, while avoiding fiscal imbalances recent years—particularly due to the MiFaPro Program that plagued much of the region. that is administered by the Ministry of Education—as a share of GDP is relatively low (8 percent in 2010) 1.19 Sound macroeconomic management has al- compared to the rest of Latin America (see Chapter 3 lowed Guatemala to ensure macroeconomic stabil- for more detail). ity, although fiscal balances have shown some strains as a result of the impact of the global economic 1.17 An assessment of progress on the Millennium crisis. Despite Guatemala’s susceptibility to natural Development Goals (MDGs), undertaken jointly with disasters (e.g. Hurricane Stan in 2005 and volcanic the Government, the IDB and the United Nations, activity and multiple episodes of flooding in 2010), fi- shows tentative progress on meeting several goals nancial market volatility (e.g., collapse and resolution except for extreme poverty and reducing the spread of two commercial banks in 2006-2007), and terms of of HIV/AIDS. The assessment concluded that to reach trade shocks (e.g., the coffee crisis in 2003-2004), the most of the 2015 goals Guatemala needs to grow fast- economy grew 4.5 percent on average from 2004 to er and in a more inclusive and sustainable way. Under 2007, reaching 6.3 percent in 2007, its highest growth Public Expenditure Review the best economic growth scenario, which implies an in three decades. While the global financial crisis average growth rate of 5 percent combined with ad- in 2008-09 and the adverse external environment ditional social and redistribution policies costing an slowed growth substantially, falling to 0.5 percent estimated 0.6 percent of GDP, the goals on poverty, in 2009, it never turned negative and rebounded to child malnutrition, literacy, gender and infant mortality 2.8 percent in 2010. Much of the credit to weather- 23. World Bank (2005b). 24. Volatility measured by the coefficient of variation of annual real GDP growth. 21 ing the external environment during the crisis can be States. Despite some retrenchment during the global attributed to prudent macroeconomic policy prior to crisis, foreign direct investment (FDI) increased nearly the crisis and reforms to fortify the financial sector. threefold from 2003 to 2010. Still, challenges remain, Although the country is still relatively well positioned and the 2008 Guatemala Investment Climate Assess- on the fiscal front, the 2009 economic downturn and ment highlighted corruption, energy supplies and increased spending have begun to put some strain crime as the greatest obstacles faced by business.27 on government accounts (Table 1.2; see Chapter 2 for more details). 1.22 DR-CAFTA and other free trade initiatives ap- pear to have helped to boost trade, but extending 1.20 To accelerate broad-based growth, Guate- these benefits to the poor will require a broad- mala will need to improve the investment climate based development agenda. Since DR-CAFTA came for business, increase educational attainment and into effect, exports and imports have grown more make progress on governance, transparency and rapidly than in previous years, despite the downturn security. A challenge in meeting these goals will be in 2009, but it is still too early to determine whether to expand fiscal space to support priority public in- these gains can be attributed solely to the free trade vestment and social spending. Since the 1996 Peace agreement or to other favorable external demand fac- Accords, Guatemala’s target has been to raise tax tors. Global experience suggests that the extra growth revenues to 13.2 percent of GDP.25 Despite significant and poverty reduction that Guatemala extracts from improvements in tax administration, including the its free trade initiatives will largely depend on the effi- passage of Anti-Tax Evasion Laws26 (I and II in 2006 ciency and competitiveness of its economy. Competi- and 2012, respectively), further progress has been tiveness reflects many factors, including labor force modest and political opposition and unfavorable con- education and flexibility, reliability of energy sources, stitutional court rulings on taxes have kept tax collec- good transportation linkages within the country, port tions as a share of GDP relatively flat and below the facilities, and the capacity of the legal and judicial sys- Peace Accords target. Guatemala’s light tax burden, tem to enforce contracts and facilitate business activ- which at 11.0 percent in 2011 is among the lowest in ity, among others. the region, does not provide sufficient fiscal space to address social needs.   1.23 Financial sector legal and regulatory reforms implemented since 2000 have improved banks’ sol- Investment Climate, Trade and Finance vency and resilience to market volatility, but lack of access to finance remains a major issue for the fi- 1.21 Guatemala has made progress in improving its nancial sector. Domestic credit to the private sector investment climate although reforms have slowed. increased from around 20 percent of GDP in 2000 to According to the Bank’s 2011 Doing Business Report, 28.1 percent in 2006, but fell to 23.4 in 2010. Despite while Guatemala still ranks low in business climate in- the modest improvement since 2000, these figures re- dicators (101th out of 183 countries) it has improved main low compared to the Central American average since 2006 when it ranked 128th out of 155 countries. (i.e., 47.5 percent in 2010) and interest rate spreads In 2006 and 2007, Guatemala was among the top re- are also relatively high. The problem is especially seri- formers worldwide in Doing Business indicators, and ous for smaller firms, whose access to credit is even perceptions of the business environment improved more restricted. Currently, a wide range of mostly substantially. In addition, in 2006 Guatemala imple- non-regulated financial entities service clients that Public Expenditure Review mented the Dominican Republic-Central America cannot access the banking system. In this context, a Free Trade Agreement with the United States (DR- stronger regulatory and supervisory framework would CAFTA) that helped secure a more open trading en- help reduce risk and deepen the development of the vironment between Central America and the United financial system. 25. 12 percent using GDP base year 1958. 26. One major result of the Anti-Evasion Law I was the significant reduction of VAT tax evasion, estimated to have fallen from 34.4 percent in 2004 to 26.1 percent in 2007. Source: SAT. 22 27. The second Investment Climate Assessment for Guatemala was completed in 2008 based on 2007 survey data and allows for a compari- son with 2004 data. The survey interviewed more than 500 firms from the manufacturing, services and construction sectors. Table 1.2: Key macroeconomic indicators 2003 2004 2005 2006 2007 2008 2009 2010 2011 National accounts and prices Real GDP (% change) 2.5 3.2 3.3 5.4 6.3 3.3 0.5 2.9 3.9 Real GDP per capita (% change) 0.0 0.6 0.7 2.8 3.7 0.8 -1.9 0.4 1.3 Inflation rate (eop CPI, % change) 5.9 9.2 8.6 5.8 8.7 9.4 -0.3 5.4 6.2 Real effective exchange rate (% change)1 -1.5 2.0 7.7 3.0 -0.4 7.5 -3.0 1.1 3.6 Central government finance (% of GDP) Total revenues 12.5 12.3 12.0 12.7 12.8 12.0 11.1 11.3 11.8 o/w tax revenues 11.7 11.5 11.2 11.9 12.1 11.3 10.3 10.5 11.0 Total expenditures 15.1 13.4 13.7 14.7 14.3 13.6 14.2 14.5 14.7 Current expenditures 10.1 9.2 9.1 9.4 9.5 9.2 10.1 10.4 10.6 o/w wages and salaries 3.9 3.6 3.3 3.3 3.1 3.1 3.6 3.8 3.9 Capital expenditures 5.1 4.3 4.6 5.3 4.8 4.5 4.1 4.1 4.0 Overall balance -2.6 -1.1 -1.7 -1.9 -1.4 -1.6 -3.1 -3.3 -2.8 Primary balance -1.3 0.3 -0.3 -0.6 0.0 -0.3 -1.7 -1.8 -1.3 Total public sector debt 21.2 22.2 21.1 21.9 21.6 20.1 23.0 24.5 24.3 External 14.9 15.5 13.2 13.1 12.4 11.2 13.1 13.5 12.0 Domestic 6.3 6.7 7.9 8.8 9.2 8.9 9.9 11.0 12.3 Balance of payments (% of GDP) Current account balance -4.7 -4.9 -4.6 -5.0 -5.2 -4.3 0.0 -1.5 -3.1 Trade balance -13.5 -15.2 -15.4 -16.1 -16.1 -14.2 -8.9 -10.3 -10.6 Exports of goods 20.7 21.3 20.1 20.1 20.5 20.0 19.3 20.6 22.4 Imports of goods 34.2 36.5 35.5 36.2 36.6 34.3 28.2 31.0 33.0 Remittances 9.7 10.9 11.2 12.2 12.3 11.3 10.5 10.1 9.4 Foreign direct investment 1.0 1.1 1.7 1.8 2.1 1.9 1.5 1.9 2.1 Monetary sector Credit to private sector (annual % change) 9.0 12.2 21.4 29.2 26.0 11.0 1.1 5.7 14.1 Lending interest rate (annual average, %) 15.0 13.8 13.0 12.8 12.8 13.4 13.8 13.3 13.5 Nominal exchange rate (average, Q./US$) 7.9 7.9 7.6 7.6 7.7 7.6 8.2 8.0 7.8 Memorandum items GDP at current prices (Q. billions) 174.0 190.4 207.7 229.8 261.8 295.9 308.0 331.9 365.8 GDP at current prices (Q. billions) 21.9 24.0 27.2 30.2 34.1 39.1 37.7 41.4 47.0 Public Expenditure Review *Preliminary. Source: Central Bank of Guatemala, Ministry of Finance and IMF. 23 Figure 1.1: Governance indicators for Guatemala (percentile rank 0-100) 70 60 50 40 30 20 10 0 Voice and Political Stability/ Government Regulatory Rule Control of Accountability Absence of Violence Effectiveness Quality of Law Corruption Latin America (unw. average) Central America (unw. average) Guatemala Source: World Bank governance indicators (2010). Governance and Transparency 1.25 Guatemala made progress in fiscal transpar- ency, but gains will need to be consolidated. The 1.24 Despite efforts in recent years to improve implementation of the integrated financial manage- governance and strengthen institutions, public ment system (Sistema Integrado de Administración perceptions and measures of governance remain Financiera, SIAF), which was built upon a good norma- low. Among World Bank governance indicators, regu- tive framework and a modern technological platform, latory quality has shown some improvement, but contributed to efficiency in financial transactions and most indicators have not changed much in recent the production of timely, transparent information on years and the rule of law remains one of the most most government finances. Adherence to the system’s important challenges as it falls significantly below av- norms and usage, however, will have to be strength- erages for Latin America and countries with similar ened if financial progress on management is to be income (Figure 1.1). Increasing crime and violence and maintained as there are only weak mechanisms pe- a growing sense of impunity has contributed to the nalizing transactions that are executed outside the decline in governance indicators. In an effort to fight system. Under recent fiscal pressures some ministries impunity in the country, in 2007 Congress approved have executed contracts without funding, leading to the creation of the International Commission against unapproved and unrecorded payment arrears. Impunity (Comisión Internacional Contra la Impuni- dad en Guatemala, CICIG). The commission’s efforts 1.26 Guatemala’s transparency and anti-corruption resulted in important progress towards improving agenda is one the Government’s main priorities. governance. This was done working in close col- There have been efforts to improve transparency of laboration with the judicial, legislative and executive procurement, such as the introduction of the e-pro- branches, and the UN High Commissioner for Human curement platform (Guatecompras), and a new access Rights. Nonetheless, recent gains remain at risk and to information law was implemented in April 2009. It the government will need to consolidate advances is now more difficult to hide misallocation of resources Public Expenditure Review with institutional reform, anti-corruption measures, from public scrutiny.28 Indeed, many more reports of vetting mechanisms for government officials as well poor government procurement are coming to light as boarder political participation. precisely because of greater transparency. Guatemala 27. La segunda Evaluación del Clima de Inversiones para Guatemala fue completada en 2008 basada en una encuesta realizada con datos de 2007 y permite una comparación con la información de 2004. La encuesta entrevistó a más de 500 firmas de los sectores de manufac- tura, servicios y construcción. 24 28. For instance, detailed budget data for the central and municipal governments is available through the web-based SICOIN portal. was selected by the World Bank as a C-GAC (Country Governance and Anti-Corruption) pilot country, and the World Bank and other International Organizations have been providing support to this agenda. Main C- GAC actions include: (i) activities to reduce the risk of corruption and enhance transparency in the con- ditional cash transfer (CCT) program; (ii) assessment of the “Internal Audit Function� core responsibilities of the executive branch; (iii) support to create the Vice Minister of Finance for Transparency; and (iv) implementation of SEPA, the Sistema de Ejecución de Planes de Adquisiciones. Civil society organizations and the media are also tapping into publicly available information for monitoring and social auditing of pub- lic management. 1.27 Violence has risen in recent years, with height- ened impact on citizens and businesses. Crime and violence are now recognized as serious economic and social problems with severe economic and social costs. High crime rates directly affect human welfare in the short run, and exert long-run effects on eco- nomic growth and social development. Today, crime and violence in Guatemala are driven not only by per- sistent inequality and lack of opportunities, but also by external sources of stress (such as increased drug trafficking) and by the growing state capture by crimi- nal groups. Crime and violence drives away private in- vestment, diverts resources from productive activities to crime prevention, discourages investment in human capital, and disrupts the functioning of public and pri- vate institutions. According to Guatemala’s 2006 En- terprise Survey, security costs and direct losses due to crime averaged some 3.9 percent of firm sales, com- pared to 2.8 percent for the LAC region as a whole.29 Not only does this have an adverse impact on invest- ment and growth, but it also generates low public con- fidence in the criminal justice system. Public Expenditure Review 29. World Bank (2008). 25 Chapter 2 Overview of Public Finances 2 » A. Introduction 2.1 In Guatemala, there has long been a preference for a small government, which lev- ies a minimum of taxes on its citizens and, in return, provides a limited amount of public services. On the one hand, this meant that the presence of the state, especially in more remote areas, was minimal, and marginalized and poorer groups had virtually no access to public services. Even today, this long-standing lack of access is reflected in low human de- velopment outcomes, especially among indigenous groups. On the other hand, the spending restraint imposed by this preference for a small government enabled successive administra- tions to pursue conservative fiscal policies, resulting in low fiscal deficits and modest levels of public debt. In a region that has traditionally struggled with recurring fiscal crises, this relative fiscal stability has been a strong feature of economic management in Guatemala. 2.2 However, the civil war brought issues of marginalization and widespread poverty to the forefront. In fact, the commitments made in the 1996 Peace Accords called for a stron- ger role of the state in neglected areas and increased expenditure, especially in the social sectors. Among other targets, the Peace Accords foresaw an increase in spending on educa- tion and health, housing and justice, to be financed by a gradual increase in tax revenues. 2.3 In the decade and a half since the Peace Accords, both tax revenue and expenditure have increased, albeit not at the pace foreseen in the agreements. Even in the absence of a major overhaul of its tax system and despite a gradual decline in import tariffs resulting from trade liberalization and regional integration, Guatemala was able to gradually raise its tax intake thanks to incremental tax policy measures and improvements in tax admin- istration: the tax to GDP ratio increased from 8.8 percent in 1995 to 11 percent in 2011. This increase was accompanied by an expansion in spending to meet the commitments of the Peace Accords: total central government expenditure increased from 10.3 percent to 14.6 percent in the same period, with the social sectors accounting for a significant share of this increase. Public Expenditure Review 27 2.4 Despite this shift, Guatemala is still among the countries in the region with the lowest tax revenues, and the lowest per capita spending on the social sectors. At around 11 percent of GDP, Guatemala’s tax-to-GDP ratio is well below the Central American average of 13.3 percent (Table 2.1), and far lower than the average for all of Latin America (19.2 percent).30 While other countries with low tax-to-GDP ratios, including Mexico and Panama, have sig- nificant non-tax revenue sources, Guatemala does not. As a result, total central government revenues have been below 13 percent of GDP in the past decade. This low level of resources is reflected in the limited public expenditures, with Guatemala occupying the bottom posi- tion in most international comparisons. 2.5 This chapter provides an overview of the underlying trends in Guatemala’s fiscal ac- counts. Section B provides an analysis of public revenues, with a focus on the tax system and the reforms recently approved by Congress. Section C presents a general overview of public expenditures, and Section D touches briefly on issues related to the sustainability of Guatemala’s public debt. Table 2.1: Central America central government revenues (% of GDP) 2001 2002 2003 2004 2005 2006 2007 2008 2009 Costa Rica Total Revenues 13.4 13.3 13.9 13.6 13.9 14.2 15.5 15.9 14.1 Tax Revenues 13.2 13.2 13.3 13.3 13.6 14.0 15.2 15.6 13.8 El Salvador Total Revenues 11.9 12.5 13.2 13.3 13.4 14.3 14.6 14.7 13.5 Tax Revenues 10.5 11.1 11.5 11.5 12.4 13.3 13.4 13.1 12.4 Honduras Total Revenues 16.7 16.4 16.5 17.2 17.6 18.1 19.1 19.7 17.4 Tax Revenues 13.6 13.3 13.7 14.5 14.5 15.2 16.3 16.0 14.4 Guatemala Total Revenues 12.4 12.8 12.5 12.3 12.0 12.7 12.8 12.0 11.1 Tax Revenues 10.8 11.9 11.7 11.5 11.2 11.9 12.1 11.3 10.3 Nicaragua Total Revenues 16.2 17.6 19.7 20.5 21.4 23.0 23.7 22.3 21.5 Tax Revenues 12.7 13.5 15.2 15.8 16.7 17.7 18.4 18.0 17.7 Panama Total Revenues 17.7 16.8 15.4 14.4 15.2 18.6 19.2 19.7 18.3 Tax Revenues 8.8 8.6 8.7 8.5 8.7 10.3 10.6 10.5 10.9 Sources: National authorities for Guatemala and CEPAL statistics for other countries. Public Expenditure Review 28 30. OCDE (2011). B. STRUCTURE AND PERFORMANCE OF rence of three factors: (i) a cycle of temporary taxes, REVENUES (ii) amplifications of tax incentives and exemptions, and (iii) improvements in tax administration and Guatemala’s tax system combating tax evasion. The cyclicality of temporary reforms stems from the repeated introduction of fiscal 2.6 Apart from the low tax intake, two other fea- reforms in the early years of a presidential mandate, tures of Guatemala’s tax regime have historically often as a response to budgetary and fiscal crises. stood out: (i) its reliance on indirect taxes; and (ii) These ad hoc reforms tend to introduce temporary the incentives it provides for tax avoidance. First, taxes which reach their time limit or are rolled back the value-added tax (VAT) generates close to 50 at the end of a presidential mandate. In addition, tax percent of total tax revenue. The intake from direct policy reforms are frequently challenged in the consti- taxes – between 24.5 and 30.6 percent of total tax tutional court and subsequently reversed (Table 2.2).35 revenues in the last decade – is among the lowest in Incoming presidents thus face fiscal difficulties and the the region.31 Second, the tax regime is particularistic. need to introduce short-term fixes. This cyclical move- Different rates are applied to income derived from dif- ment of temporary taxes has been paralleled by sec- ferent sources; tax planning to minimize obligations is tor-specific laws amplifying exceptions and weaken- thus widespread. This is complemented by numerous ing progressivity. However, over the last two decades tax exemptions. Finally, corruption scandals in tax col- reforms were also paralleled by a major transforma- lection and public expenditures have contributed to tion of the tax administration. This transformation has a culture of tax avoidance, which further complicates been a key driver behind the incremental (but not enforcement of the formal tax regime. necessarily steady) expansion of the country’s tax in- take – without, however, allowing for more structural 2.7 Guatemala’s tax system has retained these change of the country’s small-government paradigm.36 fundamental characteristics through several reform waves. As early as in the 1950s, tax reforms were ad- Revenue structure vocated, in particular a progressive income tax.32 From the 1960s onwards, a tax reform commission allowed 2.9 Despite the lack of comprehensive reforms, for the deliberation of reforms between governments the structure of Guatemala’s tax revenues has and private sector organizations, including the influ- changed significantly in the past 15 years. The liberal- ential Coordinating Committee of Agricultural, Com- ization of trade through regional integration and, more mercial, Industrial and Financial Associations (Comité recently, the DR-CAFTA agreement with the United Coordinador de Asociaciones Agrícolas, Comerciales, States, has resulted in a significant loss in revenues Industriales y Financieras, CACIF). Since the advent from import tariffs. In 1995, tariff revenue accounted of democratization, more than 130 fiscal reform laws for 24 percent of the total tax intake, falling to 8 per- have been passed (see Box 2.1 for an overview of the cent in 2007 (when the DR-CAFTA agreement came most important reform initiatives). Nonetheless, tax into effect) and to 6 percent in 2011. As a share of GDP, collection has only seen a moderate improvement, ris- tariff revenue fell from 2.1 percent of GDP in 1995 to ing from 8.2 percent of GDP in 1992 to 11 percent in 2011. 0.7 percent in 2011. This development was accompa- nied by a gradual strengthening in the administration 2.8 The series of piecemeal reforms and modest of domestic taxes, as well as incremental reforms in tax revenue expansion has been driven by the concur- policy, which especially strengthened VAT collections. Public Expenditure Review 31. Barreix, Bes and Roca (2009). 32. ICEFI (2007). 33. Schneider (en impresión) 34. Schneider (en impresión) 35. Vague language and strong property rights protection in the 1985 Constitution renders virtually any fiscal reform vulnerable to legal chal- lenge. In particular, the Constitution prohibits double and confiscatory taxation and limits fiscal authority. The vote of three (out of 5) judges on the Constitutional Court is sufficient to invalidate a reform. Between 1988 and 2009, 187 constitutional appeals were brought against fiscal reforms, in particular related to income taxes (22 percent) and beverage taxes (22 percent). Almost half of these appeals were successful. See ICEFI (2007b) and Mendoza (2010). 36. Schneider (in press). 29 Box 2.1: A short history of fiscal reforms post-Civil War Concluding twelve years of negotiations, the signing of the 1996 Peace Accords placed fiscal reform at the forefront of development priori- ties to tackle. While progress in raising tax revenues was made in the 1990s, including through the creation of a Superintendence of Tax Administration (SAT) in 1998, advances fell short of the established target. In response, the Commission for the Accompaniment of the Peace Accords formed a Preparatory Commission tasked with the preparation of a Fiscal Pact– in essence a political strategy to impulse the fulfillment of the fiscal elements of the Peace Accords. Its preparation was accompanied in 1999 by broad-based consultations – at which more than 120 organizations presented 47 written proposals – and capacity building of civil society in fiscal issues throughout the territory, with a view to reducing the knowledge asymmetry traditionally characterizing fiscal policy. This process led to the juxtaposition of CACIF and the Collective of Social Organizations (COS) at the core of the negotiations (ASIES 2005). In May 2000, the Fiscal Pact was signed by presidents of the three branches of government and over 100 organizations, including private sector associations, unions, indigenous and peasant organizations, religious groups and universities. Next to short-term objectives, the Pact laid out medium- and long-term principles for not only tax reforms but public finance more generally, including reforms of public expenditures, tax administration, debt management, transparency and decentralization (Fuentes and Cabrera 2006). This was followed in June 2000 by a political accord between CACIF and the COS about a specific fiscal reform proposal to be approved by Congress. In addition to public expenditure targets, the accord included increases to the VAT (from 10 to 12 percent) and maximum personal income taxes (from 25 to 31 percent), the introduction of new taxes such as an alcohol tax, revisions of tax exemptions and incentives, and measures to strengthen tax administration and combat tax evasion (Fuentes and Cabrera 2005). Rather than seizing this window of opportunity to approve a fiscal reform at little political cost, however, the recently elected governing party Frente Republicano Guatemalteco (FRG) of President Portillo chose to reject the Fiscal Pact on the grounds that it had not participated in its negotiations. Instead, it used its legislative majority to pass, without much consultation, in a piecemeal fashion modifications to the VAT, income and business taxes, among others. In this manner, the Portillo Government polarized the consensual climate created by the Pact negotiations. In conjunction with perceptions of corruption and authoritarianism, this led to a widespread backlash against the Government’s fiscal reforms, including through business strikes, media campaigns against the FRG and, in part co-opted by the private sector, civil society protests. Private sector appeals to the Constitutional Court subsequently led to the cancelation of several of the approved taxes. The Fiscal Pact as a political strategy was subsequently picked up by President Berger (2004-08), albeit in a diluted manner. A Technical Fiscal Pact Commission was entrusted with developing a reform to respond to an imminent fiscal crisis, since a judicial repeal of tax reforms passed under the FRG had cut revenues by US$256 million. Despite civil society opposition to the Commission – based on distrust of the strong government-private sector ties – these ties coupled with the crisis allowed for passage of a temporary income tax (IETAAP). The Berger administration also successfully passed an anti-evasion law in 2006 which strengthened the SAT, in particular regarding VAT collection. While attracting little attention, the reforms helped reduce VAT evasion and contributed – jointly with an economic upturn – to an increase in tax revenues to over 12 percent in 2007 – a first in Guatemala’s history (Schneider in press). Source: Schuster (2011). The VAT share in total tax revenues increased from 36 tions account for 85 percent of income tax receipts. percent in 1995 to 48 percent in 2011, replacing some This is partly due to the fact that the tax base for the of the revenue that was previously generated by im- personal income tax is relatively small and consists port tariffs (Figure 2.1). However, VAT revenues remain primarily of employees in the formal sector, with little heavily reliant on imports, which account for roughly income from self-employed professionals and other Public Expenditure Review 60 percent of total VAT collections. sources. While this problem is endemic in Latin Amer- ica – estimates suggest that only 10 percent of the 2.10 All the while, revenue from direct taxes in- economically active population in the region is cap- creased only slightly, with the share of income taxes tured by the personal income tax37 – it is particularly rising from 20 to 25 percent. Income tax collections pronounced in Guatemala. In the last decade, collec- rely heavily on firms rather than individuals: corpora- tions from the personal income tax increased from 0.2 30 37. Gomez Sabaini & Jimenez (2011). Table 2.2: Revenue impact of tax reforms and reversals by the Constitutional Court (2002-07, Q. million) 2002 2003 2004 2005 2006 2007 Reduction in import tariffs 93 -174 -73 1,027 -143 -562 Changes in income taxes 927 -98 -516 263 290 -167 Changes in VAT 914 -19 -9 56 238 1,625 Changes in other taxes 192 -356 -431 -1,396 TOTAL 2,126 -647 -1,029 -50 385 896 % of GDP 1.3% -0.4% -0.5% 0.0% 0.2% 0.3% Total tax collections (% of GDP) 11.9% 11.7% 11.5% 11.2% 11.9% 12.1% Source: Author’s calculations based on SAT data. Figure 2.1: Central government tax revenue by type of tax (% of GDP) 12 10 8 6 4 2 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 VAT Income taxes Other indirect taxes Other direct taxes Import tariffs Source: SAT. percent of GDP to 0.4 percent, while those from the formula: 35 percent in equal parts to all municipali- corporate income tax rose from 1.4 to 2.4 percent. ties; 30 percent proportional to each municipality’s population; 25 percent proportional to each munici- 2.11 In addition to the taxes levied by the central pality’s per capita own source revenue; and 10 percent government (a part of which is shared with sub-na- proportional to the number of villages and hamlets in tional governments), municipalities collect a limited each municipality.38 Transfers from the central govern- amount of revenue from their own taxes. Guatema- ment account for approximately two-thirds of total la is the most decentralized of all Central American municipal revenues.39 In addition, municipalities raise Public Expenditure Review countries with respect to legally mandated revenue a limited amount of own source revenue from the real transfers to sub-national governments. According to estate tax (Impuesto Único sobre Inmuebles, IUSI), the Constitution, 10 percent of general revenues must which accounts for an average of 75 percent of total be transferred to municipalities. These resources are municipal tax revenue, as well as a number of user allocated by a commission according to the following fees and charges. 38. According to Law 22/2002 (Revised Municipal Code). 39. The exception is Guatemala City, which generates almost three-quarters of its revenues from own sources. 31 2.12 The heavy reliance on indirect taxes, and a 2.15 Tax expenditures are particularly high in the generally weak taxpayer culture, raise concerns case of income taxes. This is partly due to the way over equity and efficiency of the tax system. The the SAT calculates tax expenditure, which includes next section will consider in more detail (i) to what the deduction of the first Q.36,000 of workers’ sala- extent the state captures the tax intake that would ries.42 If this deduction was not included, tax expendi- be expected given the country’s characteristics; and tures would be approximately 2 percentage points of (i) to what extent the Guatemalan tax system is re- GDP lower. Other tax expenditures for the personal gressive. income tax include the VAT credit that is counted against the income tax bill43, as well as other deduc- Efficiency of Guatemala’s tax system tions for pension contributions, donations etc. In the case of the corporate income tax, the exemptions 2.13 One measure of the efficiency of a country’s enjoyed by maquilas represented the single most im- tax system is tax effort. Tax effort is generally de- portant tax expenditure (estimated at 0.25 percent of fined as the ratio between actual collections and tax GDP in 2007). capacity, where tax capacity is the maximum tax rev- enue obtainable based on a country’s economic, so- 2.16 But even with the current structure of the cial, institutional and demographic characteristics.40 A tax system, collections could be increased through study conducted by Pessino and Fenochietto (2010) a curb on tax evasion. Guatemala has one of the high- finds a significant and positive relationship between est evasion rates in Latin America. Estimates suggest tax collections and the level of development, trade that VAT evasion is as high as 37.5 percent in Guatema- and education. Most Latin American countries includ- la45, while the rate reaches 63.7 percent in the case of ed in the study perform far below their tax capacity; income taxes. While in most Latin American countries among these countries, Guatemala shows the lowest evasion is more rampant in the case of the corporate tax effort of all (Table 2.3). This suggests that Gua- income tax, in Guatemala it appears that it is more temala could significantly increase its tax revenues pervasive in the personal income tax (69.9 percent vs. with an adequate mix of tax policy and administration 62.8 percent).46 reforms. In fact, the findings suggest that if tax effort was increased to the Latin American average, Guate- Equity of Guatemala’s tax system mala could raise an additional 6 percent of GDP in tax revenue. 2.17 In general, indirect taxes tend to be less pro- gressive than direct taxes. This is because the most 2.14 One of the reasons behind the low tax col- important indirect tax, the VAT, is linked to consump- lections is that tax expenditures are very high. The tion and thus captures a wider tax base, including the amount of foregone revenue due to exemptions and poorer segments of the population. The most impor- other tax incentives are expected to reach 8.1 percent tant direct tax, the income tax, on the other hand, has of GDP in 2012, according to SAT estimates included a much smaller base, consisting primarily of salaried in the annual budget annex. In fact, tax expenditures employees in the formal sector. Similarly, the VAT is have hovered around 8 percent in the past 6 years, a flat rate tax applied to most products, while income representing up to three-quarters of total tax collec- taxes tend to be designed on a progressive scale (i.e. tions (Table 2.4). This compares to much lower levels richer taxpayers pay a larger share of their income). in most other Latin American countries, ranging from The strong reliance on the VAT would thus suggest Public Expenditure Review 2.2 percent in Argentina to 5.4 percent in Mexico.41 that the Guatemalan tax system is regressive. 40. Gomez Sabaini and Jimenez (2011). 41. Data for other countries is for 2007. Villela et al. (2010). 42. There is some debate over whether this standard deduction should be counted as a tax expenditure, or should be considered part of the structure of the tax brackets. 43. However, this tax credit was reduced by the new tax law approved in February 2012. 44. Cabrera (2010). 45. This estimate (for 2006) is considerably higher than estimates by Guatemalan authorities (26.1 percent for 2007). 32 46. Gomez Sabaini and Jimenez (2011). Table 2.3: Tax effort in Latin Americaa Tax revenue (a) Tax capacity (b) Tax effort (a/b) Year Actual tax revenue Truncated Truncated normal Truncated Truncated normal (% of GDP)* model heterogeneous model heterogeneous model^ model^ Brazil 2006 34.2 34.8 36.0 98.4 95.1 Jamaica 2005 32.4 33.9 35.2 95.8 92.2 Colombia 2003 19.6 26.6 28.3 73.6 69.2 Costa Rica 2006 22.2 32.6 33.4 67.9 66.3 Trinidad and Tobago 2005 29.3 44.5 44.1 65.8 66.3 Bolivia 2006 26.6 39.2 42.5 67.8 62.5 Argentina 2006 27.4 42.9 44.6 63.9 61.4 LAC average 20.6 32.7 34.4 62.8 59.5 Paraguay 2006 15.3 24.0 25.8 63.7 59.3 Honduras 2006 17.9 27.4 30.3 65.3 59.0 Nicaragua 2004 21.5 31.9 37.8 67.3 56.9 Peru 2005 15.3 26.9 28.7 56.9 53.4 El Salvador 2006 15.3 28.1 30.0 54.3 50.8 Mexico 2006 19.9 39.4 40.5 50.5 49.2 Panama 2001 14.3 30.6 30.4 46.8 47.1 Dominican Republic 2005 14.2 29.3 31.3 48.5 45.3 Venezuela 2004 16.2 36.3 36.2 44.6 44.7 Guatemala 2005 10.7 27.6 29.9 38.7 35.8 *Includes social security contributions. ^Model shows impact of decreased corruption and inflation. Source: Pessino and Fenochietto (2011) 2.18 As for specific taxes, the evidence is some- more progressive, its distributional impact would be what mixed. Income taxes are generally found to be severely limited. For instance, estimates suggest that proportional.47 The personal income tax is slightly pro- only 3.3 percent of salaried workers currently have a gressive, given that only the highest income quintile positive income tax bill49, which means that the scope has a positive tax bill48; while the incidence of the cor- for income redistribution through the tax system is porate income tax ranges from proportional to mildly small. Any redistribution would thus have to come regressive. The VAT, on the other hand, is regressive from the expenditure side. when measured relative to income, but mildly pro- gressive when consumption is considered – probably The 2012 tax reforms – a break with history? due to the fact that certain products (especially food Public Expenditure Review items) that represent an important share of the poor’s 2.20 The new government pushed through an im- food basket, are exempt. portant package of tax reform within its first month in office. This package includes two key laws: (i) the 2.19 However, given the small overall size of the Anti-Evasion Law II (Law No. 4/2012), which introduces tax intake, it is likely that even if the tax system was a number of measures strengthening tax administra- 47. See, for example, Cabrera (2010) and Cubero & Hollar (2010). 48. I.e. they are above the minimum income threshold and their VAT deductions are smaller than the total tax liability. 49. ICEFI 2012. This percentage is likely to increase with the recently approved tax reforms. 33 Table 2.4: Tax expenditures by type of tax (current Q. millions)^ 2006 2007 2008 2009 2010* 2011* 2012* Direct taxes 14,565.0 16,477.4 18,611.8 18,845.2 20,126.3 21,537.9 23,366.7 Income taxes 13,742.7 15,739.3 17,737.8 17,931.6 19,158.2 20,510.0 22,199.1 Financial product taxes 40.7 94.8 147.1 238.0 252.8 269.1 433.1 Solidarity tax 781.6 643.3 726.9 675.6 715.3 758.8 734.5 Indirect taxes 3,858.9 4,631.2 5,087.7 5,406.6 5,902.3 6,462.1 8,174.1 VAT 3,414.8 3,989.8 4,510.2 4,800.7 5,212.0 5,672.7 7,477.0 Import tariffs 400.7 595.5 541.3 564.9 644.4 738.0 639.3 Fuel taxes 36.5 39.0 30.3 34.5 38.7 43.4 46.1 Tobacco taxes 1.4 1.3 0.7 1.2 1.5 1.8 1.7 Beverage distribution tax 0.1 0.1 0.1 0.1 0.1 0.1 0.5 Vehicle taxes 4.5 4.7 4.7 5.1 5.5 6.0 9.0 Cement distribution tax 0.9 0.8 0.4 0.1 0.1 0.1 0.5 Total tax expenditures (Q. millions) 18,423.9 21,108.6 23,699.5 24,251.8 26,028.6 28,000.0 31,540.8 Total tax expenditures (% of GDP) 8.0% 8.1% 8.0% 7.9% 7.8% 7.7% 8.1% Total tax expenditures 67.6% 66.9% 71.0% 76.2% 74.9% 68.9% 72.3% (% of total tax collections) Notes: ^The definition of tax expenditures used by the government may differ from internationally accepted definitions. *Indicates projection. 2012 projections do not take into account the recently approved tax reforms. Source: SAT. Box 2.2: Key provisions of the 2012 tax law Corporate income tax: • Rate increase for the simplified regime: from single general rate of 5 percent on gross income, to two marginal rates on gross income: 5 percent up to Q.30,000 per month, and 7 percent for income over Q.30,000 per month; • Rate reduction in the net profits regime, from 31 to 25 percent; • Tightening of documentation requirements for deductible expenses in the net profits regime; • Inclusion of transfer pricing. Personal income tax: • Taxation of dividends at 5 percent in the capital income regime; • Reduction in marginal tax rate from 15-31 to 5-7 percent on net income; • Elimination of VAT credit; • Deduction raised from Q.36,000 to Q.48,000 per year, expandable to Q.60,000 per year if VAT invoices are presented; • Reduction in the list of deductible expenses Vehicle registration tax: Public Expenditure Review • Replacement of import tariff with a domestic tax on the first registration of vehicles Customs • Domestic legislation required to apply regional customs regulations; • Strengthening customs authority’s capabilities to fight contraband and customs fraud Sales of real estate • Elimination of cascade effect of VAT for second and successive sales, replaced by stamp tax with lower base. Other taxes • Tax rate doubled for circulation of vehicles 34 Source: ICEFI (2012). tion; and (ii) the Law updating Guatemala’s tax system livery. As a result, the original GPDF reform proposal (Ley de Actualizacion Tributaria, Law No. 10/2012), included both measures on the revenue as well as on which makes important changes to the tax rates, tax the expenditure side. These included, among others: bases and other aspects of the existing system. Both a ban on the financing of current expenditure with laws are fundamentally based on the proposals devel- debt; strengthening of regulations on fiscal transpar- oped by a non-partisan group of economists in 2008 ency and accountability; justification of budgetary (Grupo Promotor del Dialogo Fiscal, GPDF), repre- transfers between ministries that take place through- senting broad-based consensus. out the budget year; enforcement of sanctions against the creation of commitments without budgetary re- 2.21 The Anti-Evasion Law II builds on the 2006 law sources; reforms of the procurement law; extend the by further strengthening the role of SAT. In particu- compulsory use of the Public Financial Information lar, the agency’s mandate to control and audit the pay- System (SIAF) to decentralized entities of the public ments of VAT, personal income taxes and other taxes sector. While the tax tax reform proposal was broadly was strengthened, and the Tax Code was modified to adopted in the two laws approved in February 2012, modernize administrative procedures. The law also the new government is planning to include some of includes measures designed to reduce the informal the proposed expenditure measures in a separate set economy and incorporate small VAT taxpayers into of laws on fiscal transparency as part of its “Fiscal Pact the system. Finally, it reinstates the tax on tobacco for Change� (see Box 2.3). consumption. 2.24 Concerns over public spending are espe- 2.22 The 2012 tax law represents the first compre- cially prominent because of the perceived lack of hensive tax policy reform approved in many years. progress since the Peace Accords. The 1996 Peace Most notably, it includes a new legal basis for the in- Accords, which ended a protracted civil war, repre- come tax, introduces a new tax on vehicle registration sented an important milestone in that they presented and strengthens customs administration (Box 2.2).50 a consensus on the country’s key development pri- While this represents an important step in the right orities and included clearly defined targets for tax direction, the revenue impact of the law is limited: es- revenues, public expenditure in priority sectors and timates suggest that taken t ogether, the Anti-Evasion human development outcomes. While the spending and Tax Laws together would raise an additional 1-1.3 targets have generally been met, Guatemala is lagging percent of GDP in tax collections. Given the nature far behind several of the outcome targets, especially of the reforms, whether the revenue impact can be regarding infant and maternal mortality (Table 2.5). achieved or not also depends crucially on the ability of SAT to improve its controls. Finally, the new law – just like 2.25 A key problem is that the high degree of ri- many previous reform attempts – is vulnerable to legal gidity in the budget leaves little room for changes challenges that could lead to a partial or full reversal. in budget priorities. In the 2011 budget, for instance, 83 percent of expenditures may be considered rigid. C. STRUCTURE AND PERFORMANCE OF As illustrated in Table 2.6, these include constitution- EXPENDITURES ally-mandated transfers to autonomous and decen- tralized entities (10 percent of the total budget), the 2.23 In recent years, the debate about tax reforms earmarked VAT funding social programs (7 percent has been closely tied to the need for improvements excluding decentralized entities), remunerations and Public Expenditure Review regarding expenditure efficiency and transparency. pensions (26 percent) and debt service (16 percent). A key argument used by opponents of the tax reform As a corollary of these rigidities and the country’s lim- was that an increase in tax revenue would only con- ited fiscal envelope, budget priorities have remained tribute to inefficient spending, corruption and – by largely unaltered over the last fifteen years, with only extension – mediocre outcomes in public service de- defense (from 10 percent to roughly 3 percent of pub- 50. The income tax reforms are expected to increase the number of effective taxpayers (i.e. those with a positive tax bill) from 55,000 to 300,000, while 90 percent of salaried workers in the lower income ranges will continue to pay zero income tax. 35 Box 2.3: The government’s reform program for fiscal transparency In the weeks leading up to the approval of President Pérez Molina’s tax reforms, opposition parties and private sector representatives made clear that fiscal reform will remain incomplete without a comprehensive legal package bolstering transparency. Beyond the recent creation of the Secretariat for Control and Transparency on February 9, 2012, at least eight transparency laws are now waiting to be presented before Congress. Of these, five initiatives should be highlighted. • Reform to the Budget Law. The aim of this reform is to gain greater flexibility in executing the annual budget by making the rules that govern its implementation more transparent. Current practice involves passing a series of norms and rules every year that regulate the new budget, making the execution of public resources cumbersome. This reform would eliminate this annual practice. • Reform to the Law Governing Trust Funds. This reform seeks to make Trust Funds that manage public resources more transparent by holding them responsible for complying with the norms set forth in Law of Public Contracts and accountable before the Auditor General of Public Accounts. Currently, Trust Funds are not legally bound to either. • Law of Public Contracts. This reform aims to increase transparency in state contracts by updating the bidding process, making it more accessible to the public. Additionally, it would reevaluate the contractual period of contracts in forces. The later of these is tar- geted at combating the floating debt. • Public Service Reform. This reform would strengthen the establishment of a career in public service through creating legal certainty in labor relations between the State and its employees. Such actions will involve guarantying the application of norms that govern the management of public personnel. • Law against Illicit Enrichment. Currently, the sanction for such crimes consists of paying a fine. If passed, this law would increase sanctions for public officials suspected of illicit enrichment, allowing them to be prosecuted under the penal code. It should be men- tioned that this legislation received favorable judgment by Congress in 2008, yet has not been made into law. Source: Bank staff. lic spending) and education (from 15 percent to 19 per- to specific companies, NGOs or other institutions; cent) seeing substantial variation (Table 2.6).51 (ii) modifications to budget transparency norms; (iii) restrictions to reduce executive discretion for bud- 2.26 The Executive’s control over spending allo- get execution; and (iv) assignment of quotas to loans cations and budget transparency is also somewhat still awaiting congressional approval. Politically most limited by the negotiations during Congressional salient, the Commission in practice defines the Geo- approval. While there is some assurance that the graphic List of Public Works, an annex to the bud- Executive’s budget proposal reflects line ministries’ get which specifies the geographic location of public priorities52, once the draft law reaches Congress it works – a key source of social legitimacy in their con- becomes the subject of intense bargaining. Much of stituencies. Once the budget law leaves the Commis- the parliamentary power over the budget rests with sion, further modifications are made by the plenary, the 21 members of the Public Finance and Monetary although these tend to be smaller in scope. Partly as Commission; once the budget law has been approved a result of these Congressional modifications, the by the Commission, the Plenary on its own can only Executive makes extensive use of inter- and intra- alter up to 30 percent of the articles.53 Discussions institutional budgetary transfers. The combination Public Expenditure Review in the Commission as the “epicenter of political ne- of these practices represents an important obstacle gotiations�54 tend to center around four types of to efficient, predictable and transparent expenditure modifications: (i) budget allocations for public works management.55 51. Schneider and España-Najera (2011). 52. See Chapter 4 for more information. 53. A higher share can be modified if at least 11 members of the Commission sign the modifications. 54. Barrientos (2011, p.9). 36 55. Chapter 4 has a more detailed analysis of the budget process. Economic classification of central government percent of GDP. Current transfers, especially those expenditure to public trust funds (fideicomisos), have increased from 1.4 to 2.4 percent of GDP since 1995.56 Capital 2.27 After the significant expansion in public ex- expenditure has experienced larger fluctuations, from penditure following the Peace Accords, spending 3 percent of GDP in 1995 to 5.8 percent in 1999 and 4 has stabilized around 14 percent of GDP in the past percent in 2011. 10 years. Current expenditures have accounted for roughly two-thirds of total spending, with wages and 2.28 Guatemala’s wage bill is relatively small. Gua- salaries representing the biggest share. However, the temala’s civil service comprises 254,000 workers, of wage bill has accounted for a relatively small share which the education sector accounts for 70 percent.57 of the increase in total spending, remaining below 4 Consequently, over half of all salaries are paid in the Table 2.5: Progress towards economic, fiscal and social objectives of the Peace Accords Objective 1995 2000 2011 Baseline Target 1 Actual 1 Actual2 Economic and fiscal objectives GDP growth rate (%) 4.9 6.0 4.1 3.7 Tax revenue (% of GDP) 7.9 13.2 10.6 11.0 Public expenditure Health and soc. assistance (% of GDP) 0.9 1.5 1.5 2.0 Education (% of GDP) 1.7 2.7 2.7 3.2 Judicial sector/Public Ministry (% of GDP) 0.3 0.5 0.6 0.5 Military (% of GDP, ceiling) 1.0 0.8 0.9 0.3 Low-income housing (% of tax revenue) 0.2 0.2 0.0 0.1 Internal security (% of GDP) 0.4 0.6 0.8 0.9 Social objectives Literacy rate (%) 64.2 70.0 68.9 76.0 Primary education coverage Gross enrollment 84.0 100.0 125.0 n.a. Net enrollment 69.0 100.0 72.0 97.0 Infant mortality (per 1,000 live births) 46.0 20.0 39.0 24.8 Maternal mortality (per 100,000 births) 140.0 48.5 110.0 110.0 Vaccination coverage Public Expenditure Review Polio 80.0 85.0 88.0 91.7 Measles 83.0 95.0 86.0 93.0 Notes: 1\ Using GDP w/base year 2001. 2\ Preliminary for 2011 or latest year available. GDP growth rates for 2000 and 2011 are averages for 1995-2000 and 1995-2011, respectively. Sources: Ministry of Finance; World Bank (2005a); Central Bank of Guatemala; INE-ENCOVI 2006 & 2011; INE – Encuestas Nacionales de Salud Materno Infantil; World Development Indicators. 56. See Chapter 4 for a discussion of fideicomisos and their role in public expenditure management. 57. As of 2010, according to data from the National Civil Service Office (Oficina Nacional de Servicio Civil, ONSEC). 37 Table 2.6: Rigidity classification of the 2011 budget (as of September 2011) Expenditure Type Quetzales (million) Share of total budget A. Rigid Expenditures 47,625.70 86% Remunerations 12,692.00 23% Pensions and other obligations 3,334.50 6% Public debt service 7,582.00 14% Peace VAT (excluding departmental councils and municipalities) 2,998.40 5% Institutional contributions 5,834.80 10% Specific usages (excluding departmental councils and municipalities) 5,913.10 11% Constitutional transfers (excl. departmental councils and municipalities) 2,581.10 5% Central government transfers to departmental councils 1,917.10 3% Central government transfers to municipalities 4,772.70 9% B. Flexible Current Expenditures 1,185.50 2% C. Physical Infrastructure 6,805.90 12% Total 55,617.10 100% Source: Ministry of Finance, Barrientos (2011). Table 2.7: Central government expenditure – economic classification (% of GDP) 2003 2004 2005 2006 2007 2008 2009 2010 2011 Total Expenditure 15.1 13.4 13.7 14.7 14.3 13.6 14.2 14.6 14.7 Current expenditure 10.1 9.2 9.1 9.4 9.5 9.2 10.1 10.4 10.6 Wages and salaries 3.9 3.6 3.3 3.3 3.1 3.1 3.6 3.8 3.9 Goods and services 1.5 1.1 1.1 1.3 1.3 1.8 1.9 1.9 2.0 Interest 1.3 1.4 1.4 1.4 1.5 1.4 1.4 1.5 1.5 Current transfers 2.5 2.3 2.5 2.6 2.7 2.1 2.3 2.4 2.4 Others current expenditure 0.8 0.8 0.8 0.8 0.8 0.8 0.9 0.9 0.9 Capital expenditure 5.1 4.3 4.6 5.3 4.8 4.5 4.1 4.1 4.0 Fixed capital formation 1.1 1.1 1.2 1.6 1.8 1.9 1.8 1.8 1.9 Capital transfers 3.9 3.1 3.4 3.6 3.0 2.5 2.3 2.3 2.2 Other capital expenditure 0.0 0.1 0.0 0.0 0.0 0.1 0.0 0.0 0.0 Source: Staff calculations based on data from Ministry of Public Finance and Central Bank. Public Expenditure Review education sector (54 percent). Internal security (14 GDP Guatemala has by far the smallest wage bill in percent) and health and social protection (13 percent) Central America. together account for another quarter of the wage bill. In international comparison, Guatemala’s wage bill – 2.29 Transfers represent the second largest cate- which accounts for 27 percent of total expenditures gory of current expenditure. This spending category 38 in 2011 – is not excessively large. In fact, as a share of includes the MiFaPro conditional cash transfer pro- gram, which accounts for approximately 15 percent Functional classification of central government of the total. It also includes transfers to other parts expenditure of the public sector, including the judicial sector, the public university and municipalities. 2.32 Social spending has expanded significantly since 1995. Expenditure in the social sectors59 has in- 2.30 Central government capital expenditure is creased from 4.4 percent of GDP in 1995 to 7.6 percent concentrated in the transport sector. The vast bulk in 2011. As a share of total expenditure, it rose from 43 of fixed capital formation at the central level is allo- percent in 1995 to 52 percent in 2011, representing one cated to rural and urban development and transport, of the highest shares in the region. The increase re- which on average together account for two-thirds flects both the c ommitments made under the Peace of the total. All other sectors account for 5 percent Accords in 1996, as well as a strong focus on social sec- or less each of the total. In addition to direct invest- tors under the previous administration, including the ment by the central government, 54 percent of capi- introduction and expansion of the conditional cash tal spending is channeled through transfers, including transfer program Mi Familia Progresa.60 The increase to regional development councils, municipalities and in social spending is likely to have contributed to the other private and public entities, for which a detailed achievement of several of the targets agreed to under sectoral breakdown is not available. the Peace Accords. 2.31 Research suggests that Guatemala’s historical 2.33 However, measured against the size of its public investment levels are not sufficient to sup- economy, Guatemala has the lowest social expendi- port sustained growth. As a share of GDP and on a ture levels in Central America (Figure 2.3). As a share per capita basis, Guatemala’s infrastructure invest- of GDP, Guatemala ranks lowest among the six coun- ment is the second lowest in Central America, just tries; when measured in per capita terms, it ranks just above El Salvador and Nicaragua, respectively. A re- above Nicaragua but below all other Central American cent IMF study estimates that Guatemala could raise countries, spending roughly one-tenth of the amount growth by over 1 percentage point by increasing its Costa Rica spends per person. It is thus not surprising investment-to-GDP ratio to the average for Mexico, that Guatemala also lags behind in human develop- Peru and Chile.58 ment outcomes. Figure 2.2: Central government expenditure – functional classification (% of GDP) 16% 14% 12% 10% 8% 6% 4% 2% Public Expenditure Review 0% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Social services Transportation Defense and public order Other economic services Public debt General public services Source: Staff calculations based on data from Ministry of Public Finance and Central Bank. 58. Swiston & Barrot (2011). 59. “Social spending� is defined as all expenditures under the budget code 300. 60. MiFaPro was introduced in 2008 and its coverage expanded from 281,000 households in 2008 to 873,000 households in 2011. 39 2.34 Capital expenditure is concentrated in two This clearly shows that transport is a key priority for sectors. The bulk of public investment goes into trans- municipal governments.63 port and urban and rural development, both of which have accounted for 33 percent of total capital spend- 2.37 There is great divergence in social expendi- ing since 2004. The remainder has been allocated to tures across sub-national governments. Combined agriculture (5.5 percent), health and social assistance public spending in the social sectors ranges from (3.5 percent), environment (3.1 percent) and other sec- Q.774 per capita in the department of Alta Verapaz tors. Despite the high concentration of investment in to Q.2,193 in the department of Guatemala. There is the transport sector, spending is not sufficient to meet no clear relationship between needs (measured by actual spending needs (including new construction, the incidence of poverty) and per capita spending; maintenance and disaster reconstruction), as demon- in fact, it is striking that at 78.2 percent, Alta Verapaz strated by the deterioration in the quality of Guate- has the highest poverty incidence of all departments mala’s road network in recent years.61 but the lowest per capita social expenditure, com- pared to Guatemala, which has the lowest poverty in- Geographic distribution of expenditures: cidence (18.6 percent) and highest per capita expen- municipalities62 diture. This lack of an obvious association of needs indicators with spending allocations also holds for 2.35 Municipalities are responsible for a limited individual components of social spending. Historical range of public services: water and sanitation, mar- trends, as well as other socio-economic factors such kets, street cleaning, waste management, municipal as migration, may explain some of the variation (see police, municipal roads, and local environmental and Chapter 3 for a more in-depth analysis of health and recreational services. Municipalities are also involved education expenditure). In addition, the municipality in the management of pre-primary and primary edu- of Guatemala is home to two of the country’s largest cation and in the execution of preventive health ser- hospitals, as well as the University of San Carlos, all vices, but are not responsible for the design or full of which absorb large budgets. Finally, shortcomings financing of these services. Even though the financing in the way expenditures are registered in the SICOIN of health and education do not formally fall under the may explain the apparent discrepancies: for instance, municipalities’ span of control, many municipalities some ministries, especially those that remain heavily hire contract teachers and health workers to supple- centralized, record expenditures in the municipality ment the deconcentrated staff from the central gov- where the executing agency is located (often Guate- ernment ministries. mala City), not in the municipality that benefits from the spending. This means that central government 2.36 In line with their primary expenditure respon- expenditures for the department of Guatemala are sibilities, sub-national fiscal data suggests that mu- likely to be lower than the analysis of the SICOIN nicipalities allocate a larger share of their budgets data suggests. to capital investments than the central government. A breakdown by economic classification reveals that D. FISCAL SUSTAINABILITY municipalities spend 63 percent of their budgets on capital expenditure, compared to 27 percent at the 2.38 Guatemala’s public debt levels are among the central government level. In 2009, 20 percent of mu- lowest in Central America. Traditionally conservative nicipal capital spending went to the transport sector, macroeconomic management resulted in debt levels Public Expenditure Review 11 percent to water and sanitation, 5 percent to educa- of around 20 percent of GDP before the 2008 cri- tion and 4 percent to urban and rural development, sis. Given the increase in the fiscal deficit since the with the remainder accounted for by other sectors. 2008 global economic crisis, public debt increased to 61. See Chapter 3 for a more in-depth discussion for issues related to transport spending. 62. Based on the data available through the SIAF-Muni. Municipal budget data is only available beginning in 2008, and not at the same level of detail as central government expenditure. 63. One reason for the prioritization of transport by municipalities is the earmarking of transfers from the fuel tax for investments in roads 40 and highways. Figure 2.3: Social expenditure in Central America (a) As a share of GDP (%) (a) US$ per capita 20 1000 18 900 16 800 14 700 12 600 10 500 8 400 6 300 4 200 2 100 0 0 Costa Rica LAC average Nicaragua Honduras El Salvador Panama Guatemala Costa Rica LAC average Panama El Salvador Honduras Guatemala Nicaragua Note: Data shown is for latest year available. Source: ECLAC. Figure 2.4: Municipal spending Figure 2.5: Municipal spending by functional classification (2009) by economic classification (2009) Debt service Current spending 3% 37% General Transport 20% public services 31% Water and Economic sanitation 11% services Education 5% Defense and 33% public order Urban and rural 1% development 4% Other 23% Social services Capital spending 32% 63% Source: BOOST database using SICOIN data. Source: BOOST database using SICOIN data. Figure 2.6: Municipal and central government per capita social expenditure and poverty, by department (2009) 2500 90 80 2000 70 60 1500 50 1000 40 30 Public Expenditure Review 500 20 10 0 0 Guatemala El Progreso Zacapa Sacate Pequez Baja Verapaz Jutiapa Chiquimula Santa Rosa Peten Retalhuleu Solola Quetzaltenango Jalapa Suchitepalque San marcos Izabal Escuintla Quiche Huehuetenango Chimaltenango Totonicapan Altaverapaz Mun. Centr. Poverty (right axis) Note: Social expenditure is de�ned as all items in budget category 300. Excludes central government spending classi�ed as “multi-departmental� or “multi-regional�. Source: ENCOVI 2011 (poverty) and BOOST database using SICOIN data. 41 US$11.4 billion (24.3 percent of GPD) in 2011. Domestic to be lower than that for the other three countries. debt accounts for 12.3 percent, while the remainder is As a result, a jump from its current non-investment made up of loans from multilateral and bilateral credi- grade to “borderline� investment grade would require tors, as well as Eurobonds held by private investors (is- a reduction in the country’s public debt to below 15 sued in 2003 and 2004). The share of domestic debt percent of GDP.65 has increased in recent years (from 44 to 50 percent of total public debt) as the government has increas- 2.40 Debt service will be substantial in the next 8 ingly tapped local markets for financing, primarily in years. While the repayment schedule for long-term local currency. As a result, the share of public debt de- multilateral loans is fairly smooth, the shorter matu- nominated in Quetzales has increased from 12 percent rity of domestic bonds (10 years on average) issued in in 2000 to 45 percent in 2011. In addition, 63 percent recent years, in addition to the large principal repay- of the government’s debt is at fixed interest rates, of ments on Eurobonds falling due in the next 8 years, between 1 and 9 percent. means that debt service will average Q.10-12 billion an- nually between 2012 and 2016, compared to Q.7 billion 2.39 Given the country’s low tax-to-GDP ratio, it in recent years and equivalent to between 17 and 22 may be prudent to maintain relatively conservative percent of projected total revenues. debt levels. Despite relatively low public debt as a share of GDP, central government debt was equiva- 2.41 A comprehensive debt management strategy lent to 203 percent of total revenue in 2011, compared would help keep debt levels at sustainable levels to 211 percent in Costa Rica (with a debt-to-GDP ra- and ensure consistency with the government’s me- tio of 41 percent) and 185 percent in Honduras (with dium-term fiscal policy. A strategy could also help to a debt-to-GDP ratio of 33 percent). Recent research achieve a more favorable debt service schedule. As also suggests that Guatemala’s “debt intolerance�64 part of such a strategy, the government would need threshold is lower than of its middle-income Central to constantly evaluate the costs and benefits of dif- American neighbors (Costa Rica, Panama and El Salva- ferent types of debt, including domestic and interna- dor). In other words, the debt-to-GDP ratio required tional bonds and loans from bilateral and multilateral for Guatemala to reach investment grade is estimated sources. Figure 2.7: Projected public debt service* 14,000 12,000 10,000 8,000 Q. million 6,000 4,000 2,000 0 Public Expenditure Review 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 Loans (other public sector) Eurobonds Local bonds (Central Administration) Loans (Central Administration) *Based on existing stock of debt as of end-2011. Source: Public Credit Of�ce, MOF. 64. Debt intolerance is defined as the phenomenon of certain countries falling into repeated debt defaults at relatively low levels of debt. Fiscal rigidities, economic inefficiencies and exposure to external shocks make these economies vulnerable to fiscal and debt crises at rela- tively low levels of debt to GDP (Bannister & Barrot 2011). 42 65. Bannister & Barrot (2011). E. CONCLUSIONS AND RECOMMENDATIONS 2.42 The approval of the recent tax reforms is an important milestone, but the fiscal impact will de- pend on the effectiveness of Guatemala’s tax ad- ministration. Given the low estimated rates of tax ef- fort, significant tax expenditures and rampant evasion, significant gains in revenue collection can only be real- ized through a strengthening of the tax and customs administration. 2.43 The transparency of public expenditure needs to be strengthened further. Guatemala has good information systems (such as the SIAF, SICOIN and Guatecompras) that are not fully utilized yet. It would be important to ensure their compulsory use through- out the entire public sector and strengthen controls to prevent misuse. The new government has shown commitment towards strengthening transparency through the creation of the Secretariat of Transpar- ency and the submission of a package of reforms to Congress. This package includes important measures, such as the strengthening in the oversight of public trust funds (fideicomisos), and changes to the Budget Law, which would replace the important but tempo- rary provisions on budget implementation included in the annual regulations of the budget. 2.44 Better information is needed to improve ex- penditure quality and adequately assess the region- al distribution of public spending. Given the existing deficiencies in the SIAF and SICOIN systems, it is difficult to obtain detailed and comprehensive infor- mation on resource allocations by municipality or de- partment. Availability of sectoral outcome indicators at the sub-national level is even more limited. In order to allocate scarce resources in the most effective way, better information is required and needs to be linked to budget preparation. 2.45 Finally, investment in infrastructure should be Public Expenditure Review boosted. The new government is committed to link- ing its medium-term debt management strategy more closely to its investment plans, which would help uti- lize financing in a more efficient way. However, given the limited public resources available, Guatemala’s investment needs cannot be met by the government alone; private sector participation will be important (see Chapter 3). 43 Chapter 3 The Quality of Public Expenditure in Selected Sectors 3 » A. Introduction 3.1. As mentioned in Chapter 1, concerns over the quality (as well as the transparency) of public expenditure have been at the forefront of the public debate on fiscal reforms in Guatemala. Given the scarce resources available, as well as the perceived lack of progress with regards to a number of the Peace Accords targets (especially in the social sectors), im- proving the quality of public spending in key sectors has emerged as a priority. But assess- ing expenditure quality requires a significant amount of information, both on inputs (budget allocations, personnel numbers etc.) and on outputs and outcomes, which is currently not available for all sectors. 3.2. The purpose of this chapter is to analyze the quality of public spending in education, health and the transport sector. The three sectors were chosen because they represent a large share of total public expenditure: in 2010, they represented 45 percent of central government and 25 percent of municipal government spending. Using the methodology de- veloped under the World Bank BOOST program,66 a data platform was developed, consoli- dating all spending and outputs at the central and sub-national level,67 to explore linkages between government spending and development outcomes, as well as tracking spending of individual sectoral programs. The objective is to gain a better understanding of how and how well public funds are spent, in order to help start a conversation between budget pro- viders and budget implementers on why similar combinations of spending and inputs result in different outcomes, and how better or similar outcomes can occur for budget centers that receive lower spending and non-financial inputs. Public Expenditure Review 66. The BOOST initiative is a new program launched by the bank to deepen the knowledge base of public spending by supporting cre- ation, use and dissemination of data platforms that consolidate disaggregated spending and output data for exploration of new frontier areas in public expenditure analysis. Currently the program has been launched in more than 20 countries worldwide. 67. The geographical breakdown of the budget in 2010 allows identifying to which municipality goods and services are oriented, as well as the unit responsible for the execution of programs, sub-programs, projects, activities and works that integrates its budget. 45 B. QUALITY OF PUBLIC EXPENDITURE IN has raised its education spending from 2.6 percent THE EDUCATION SECTOR of GDP to 3 percent69, reflecting a strong focus on strengthening social expenditure by the previous ad- Allocative efficiency ministration (Figure 3.1). The increase can primarily be attributed to an expansion in primary education 3.3. Despite the progress in increasing primary spending, from 1.2 to 1.5 percent of GDP. Primary school enrollment, Guatemala continues to lag be- education now accounts for 44 percent of total edu- hind in the overall performance of its education cation spending, with pre-primary, secondary and sector. For instance, the Guatemalan population has tertiary receiving around 10 percent each. While the an average of 5 years of schooling, compared to a increase in primary education expenditure is a posi- Central American average of 7.1 years, 8.3 years in tive trend, total public spending (as a share of GDP) Costa Rica and 9.4 years in Panama. The situation in the sector continues to be significantly less than is even worse for the indigenous population: Gua- neighboring countries, as well as the Latin American temalans whose native language is not Spanish only (4.4 percent) and lower middle income (4 percent) have 4 years of schooling, compared to 7 years for country average. native Spanish speakers. Pre-primary and secondary enrollment rates are particularly low, reaching only 3.5. Education remains a highly centralized func- 30 and 40 percent of the school-aged population, tion. The bulk of public spending in the sector takes respectively.68 Outcomes also vary significantly from place at the central government level (96 percent), region to region: the department of Guatemala has reflecting the centralized nature of the sector (Figure the highest literacy and total gross enrollment rates 3.2). Any spending taking place at the municipal level in the country, while the departments in the north- is to complement the deconcentrated services pro- west of the country (Quiché, Alta Verapaz and Hue- vided by the central government, through the hiring huetenango) rank at the bottom. of contract teachers, improvements of school build- ings and similar activities. While there is some varia- 3.4. Education spending, especially at the pri- tion in education spending among municipalities, mary level, has increased in recent years, in line municipal shares in total spending do not exceed 11 with government priorities. Since 2004, Guatemala percent in any department. Excluding Guatemala, Figure 3.1: Central government spending on education, by function (% of GDP)* 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% Public Expenditure Review 0.0% 2004 2005 2006 2007 2008 2009 2010 Pre-primary Primary Secondary Tertiary Administration Other *Excludes transfers to the Program Mi Familia Progresa. Source: BOOST database using SICOIN data. 68. World Bank (2011c). 46 69. Excludes transfers to the Program Mi Familia Progresa. the department with the highest total combined third of total spending. In the case of Huehuetenan- public expenditure on education spends 1.6 times as go (one of the poorest departments in the country), much as the department with the lowest per capita the school meal program represented 74 percent of spending. the total spending on the five selected programs; a similar situation can be observed in Quiche, another 3.6. Incidence analysis suggests that education poor department. expenditure is overall poverty neutral. While prima- ry education expenditure is found to be progressive Technical efficiency (i.e. disproportionately benefiting the poor), spend- ing on secondary and tertiary education is regressive. 3.7. Comparisons between financial inputs and The high share of education expenditure allocated to key outputs provide a first cut of the efficiency the primary level is therefore beneficial for poverty analysis. Levels of municipal spending based on reduction.70 In the case of individual programs, the budget data from the Ministry of Finance71 are com- picture also varies considerably, especially when tak- pared to qualitative outputs and outcomes at dif- ing account the geographic distribution of the ben- ferent levels measured by key output indicators in efits. The school supplies program is a good example education such as student’s assessments results and of a transfer that reaches an important segment of completion rates, while controlling for the influence the population (12.9 percent) and is pro-poor (64.5 of socio-economic indicators such as adult literacy percent of its benefits are received by the poor); and chronic malnutrition rates. By controlling for another is the school meal program, which reached proxies capturing differences in initial endowments some of the poorest regions in the country includ- among municipalities, this analysis provides a first ex- ing the departments of Huehuetenango and Quiche ploration of linkages between spending, outputs and in 2010. However, spending on textbook programs outcomes in basic education to identify areas where was concentrated in the capital city and surround- efficiency of spending can be improved. ing areas, and is thus more likely to benefit non-poor populations. In the case of teacher training, spending 3.8. Preliminary analysis suggests that there is a is limited across the entire country, although in some weak relationship between the level of spending departments (such as Jalapa) it represents up to a and student achievement at the municipal level. Figure 3.2: Distribution of education expenditure by department (average 2008-09) 800 700 600 Q. per capita 500 400 300 200 100 0 Guatemala El Progreso Zacapa Baja Verapaz Chiquimula Santa Rosa Solola Retalhuelen Suchitepequez Jutiapa Jalapa San Marcos Quetzaltenango Izabal Huehuetenango Toto nicapan Quiche Peten Sacatepequez Chimaltenango Escuintla Alta Verapaz Public Expenditure Review Central Municipal Source: BOOST database using SICOIN data. 70. World Bank (2009). 71. The 2010 national budget allows for geographical classification by municipalities, disaggregating all economic and financial transactions carried out by public sector institutions by different regions of the country. However, spending through the MiFaPro program (representing 11 percent of total education spending in 2010) could not be assigned to particular municipalities and as such is not included in this analysis. 47 For instance, higher public spending on education (Alta Verapaz) that have very low internal efficiency is not associated with higher school completion in terms of number of graduates and completion rates. The two municipalities of San Sebastián and rates, but similar performance of those that actually Tectitán in the Department of Huehuetenango (Fig- graduate compared with municipalities with more ure 3.3) show very different 9th grade completion number of graduates and higher completion rates, rates (12.4 percent and 35.5 percent, respectively) such as Pasaco (Jutiapa) (Map 3.1). Overall, the corre- even though both record very low adult literacy lation between completion rates and graduates test rates and high malnutrition rates, and even though scores for 12th grade at the municipal level is only the municipality of San Sebastián spends more per 0.15 and for 9th grade only 0.05. In 30 percent of the pupil than Tectitán. Likewise, the municipalities overall sample, municipalities ranking in the top quar- of San Martin Zapotitlán and Tectitán show similar tile in completion rates are in the bottom quartile of completion rates despite the fact that the former test scores. These patterns emphasize how high per- spends almost twice as much as the latter and has formance in completion rates is not necessarily cor- considerably higher adult literacy rates and lower related with high quality, as measured by higher test malnutrition rates. It is important to highlight that scores. these variations do not represent isolated cases of severe outliers but rather the norm within the entire 3.11. The analysis suggests that demographics and territory, affecting the large majority of Guatemalan other socio-economic factors – rather than the lev- municipalities. el of public education spending – may be respon- sible for driving the variations in outcomes across 3.9. Similarly, the relationship between educa- municipalities. Here we propose two approaches tion spending and test scores appears to be weak. to try to capture the combined influence of these The Municipality of Guatemala, for instance – with factors in determining variations in outputs. First, higher per pupil spending, higher adult literacy and we implement the data envelopment analysis (DEA) lower malnutrition rates than any other municipality methodology73 to derive efficiency scores using mul- – performs relatively worse than the municipalities tiple outputs and multiple inputs (including “control� of San Sebastián Huehuetenango, Santa Lucia Utat- variables). Second, we run basic OLS regressions to lán and Villa Canales, Guatemala (Figure 3.4). Similar estimate determinants of student achievement at the analysis can be done by comparing, for instance, the municipal level. municipality of Pasaco (Jutiapa), which has similar per pupil spending but higher adult literacy and low- 3.12. Technical efficiency is not necessarily higher er malnutrition than Guatemala, with Senahú (Alta among richer municipalities. Technical efficiency is Verapaz), which has much lower adult literacy and defined as the degree to which the government ob- higher levels of malnutrition, but higher test scores. tains the greatest possible output with a given set Once again these patterns can be found across the of inputs.74 To illustrate, Map 3.2 shows technical ef- entire landscape of Guatemalan municipalities indi- ficiency in education across Guatemalan municipali- cating the need to further investigate the underlying ties. The majority of municipalities within the Depart- determinants of these variations.72 ments of Baja Verapaz and Chiquimula in particular exhibit the lowest technical efficiency scores of the 3.10. In addition, higher graduation rates and high- country while surprisingly several poor municipali- er test scores are not necessarily correlated. There ties within the Huehuetenango Department and the Public Expenditure Review are a substantial number of cases, such as Senahú Franja Transversal del Norte75 territory reported high 72. One potential explanation for these findings may, for instance, be migration to the capital city of adults educated in other parts of the country. 73. For more information on DEA, please see Annex 2. 74. It should be noted that the concept of technical efficiency has some limitations in cases like Guatemala’s. In a context where overall public spending is very low, municipalities may be found to be very efficient at using the limited inputs available to achieve overall poor outcomes. 48 75. The area along the East-West transport corridor running from the Gulf of Honduras to the Mexican border in the West. Figure 3.3: Relationship between completion rates (grade 9), adult literacy, spending on education per pupil (bubble size) and malnutrition rates (color) in selected municipalities (2010) Completion Rates 1 Guatemala 0.8 San Juan 0.6 Tecuaco Mazatenango Tectitan 0.4 San Martín Cubulco Zapotitlán 0.2 San Sebastián Purulha Huehuetenango San Rafael las Flores 0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Adult literacy (above primary) Average High Malnutrition Low Malnutrition Source: BOOST database, Ministries of Education and Health. Figure 3.4: Relationship between test scores of graduates, adult literacy, spending on education per pupil (bubble size) and malnutrition rates (color) in selected municipalities (2010) Test Scores 2 Fraiianes 1.8 Santa Lucas 1.6 Sacatepequez 1.4 Santa Lucía Villa Canales San Sebastián Utatlán Huehuetenango Guatemala 1.2 Santa Cruz el Chol Santa Cruz Pasaco 1.0 Balanya Senahu 0.8 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 Adult literacy (above primary) Average High Malnutrition Low Malnutrition Source: BOOST database, Ministries of Education and Health. Map 3.1: 9th grade completion rates vs. graduate test scores at municipal level (2010) Public Expenditure Review Completion rates Test scores < 0.196900 < 2.00847458 0.196901 - 0.320500 2.00847459 - 2.25000000 0.320501 - 0.447100 2.25000001 - 2.45967742 0.447101 - 0.676800 2.45967743 - 2.67721519 > 0.676800 > 2.67721520 Missing data Missing data Source: BOOST database, Ministries of Education and Health. 49 Map 3.2: Technical efficiency in education linkages to spending. It would be relevant to carry across Guatemalan municipalities (2009) out further research to explain what other factors are driving these results. 3.14. Regression analysis suggests that public edu- cation spending does not explain student achieve- ment at the municipal level. Multiple regressions were run to further investigate determinants of varia- tions in completion rates and test scores. One of the main results – summarized in Table A1.1 in Annex 1 – is Ef�ciency that spending is never statistically significant with re- <0.000001 0.000001 - 0.697000 spect to explaining student achievement, as opposed 0.697001 - 0.800000 0.800001 - 0.914000 to the majority of demographic and socio-economic > 0.914000 Missing data indicators. While extreme poverty rates help explain Source: BOOST database, Ministries of Education and Health. student achievement both in terms of completion rates and test scores, adult literacy rates, malnutri- tion rates and presence of large indigenous popu- technical efficiency scores once initial endowment lation have a significant negative relationships to were taken into effect. This finding is consistent with educational outcomes (test scores) but do not have the conclusions of the 2005 Public Expenditure Re- much influence over completion rates. Conversely, view, which found that there was room for improving the number of teachers and population densities efficiency, i.e. improving outcomes with the current are good predictors of student achievement in terms expenditure envelope. of completion rates but have no effect over student performance in national tests at any level. 3.13. Municipalities’ efficiency scores appear to be unrelated to whether they spend a lot or whether C. QUALITY OF PUBLIC EXPENDITURE IN their needs are high. The analysis of technical effi- THE HEALTH SECTOR76 ciency can be extended by mapping efficiency scores against indices of need (extreme poverty rates) and Allocative efficiency against overall spending at the municipal level. Two interesting lessons emerge: first, several munici- 3.15. This section focuses on describing the ef- palities appear to be inefficient whether measured ficiency of health expenditure at the department against need or spending, or a combination of both. level from the allocation point of view. Contrary Second, in all the three variations, there appears to to the education sector, most health indicators are be clustering among efficient and inefficient munici- not available at the municipal level, and the use of palities. For example, most municipalities near the technical efficiency frontier calculation techniques is capital city of Guatemala and the Department of therefore limited. Sacatepéquez present high technical efficiency but low per-pupil spending. On the other hand, those in 3.16. Despite a noticeable increase in public ex- the Departments of Baja Verapaz and Chiquimula penditure on health in Guatemala from 1.2 to 1.6 present low technical efficiency and high low per- percent of GDP between 2004 and 2010, public Public Expenditure Review pupil spending (Map 3.3). It is important to note that spending levels remain low by international and these findings and methodologies should only be in- regional standards. Despite the increase in recent terpreted as an entry point for a deeper analysis into years, Guatemala continues to spend less (as a share the determinants of student achievement and their of GDP) on health than other countries in the re- 76. This section focuses on describing the efficiency of health expenditure at the department level from the allocation point of view. Con- trary to the education sector, most health indicators are not available at the municipal level, and the use of technical efficiency frontier 50 calculation techniques is limited. We also look at the coverage and accessibility of public health services across socioeconomic groups. Map 3.3: Technical efficiency in Guatemalan municipalities (a) Technical ef�ciency vs. spending in education (b) Technical ef�ciency vs. need in education Ef�ciency vs. spending Ef�ciency vs. need Low / high Low / low High / low High / high All others All others Missing data Missing data (c) Technical ef�ciency vs. spending and need in education raise preventive health spending, which would have a positive impact on reducing maternal and child mor- tality rates; however, its share of total public health spending (39.2 percent in 2010) is still below the 50 percent target set by the government, which may ex- plain why progress towards meeting health outcome targets has been slower than expected. 3.18. Health expenditures are heavily concen- Ef�ciency vs. spending and need Low / high / low trated at the central level. For 2008 and 2009, data High / low / high All others from the BOOST database captures both expendi- Missing data ture at the central and the municipal level.78 On aver- age, the central government accounts for 96 percent Source: BOOST database, Ministry of Education. of total public health spending. However, there is a significant amount of geographic variation: municipal gion, and less than the average for middle income shares in total public health expenditure range from countries worldwide. As a result, the share of public 1.1 percent in the department of Retalhuleu to 18.8 spending in total health expenditure is the lowest: 35 percent in Suchitepéquez. In fact, there appears to percent in Guatemala compared to a LAC average be a mild positive correlation between expenditure of 50 percent and between 60 and 75 percent in El in health at the municipal and the central level: more Salvador, Honduras, Costa Rica and Panama.77 central spending in the sector is associated with more municipal spending. 3.17. Despite the scarce resources, public spend- ing remains skewed towards curative care. Table 3.19. Health insurance coverage among the poor is 3.1 presents central government spending on health, extremely low. According to the ENCOVI 2011, 89 Public Expenditure Review including those executed by the Ministry of Health percent of the population does not have health in- and Social Assistance, as well as other ministries and surance, with the share reaching 97 percent in the state secretaries (but excluding the Guatemalan So- lowest consumption quintile. However, outside of the cial Security Institute). There has been an effort to formal insurance mechanism, the Ministry of Public 77. Data is for 2010. Source: World Development Indicators. 78. Municipal data includes spending on both health and social assistance, as further disaggregation is not available. However, the bulk of this spending is likely to be on health. 51 Table 3.1: Central government spending on health* 2005 2006 2007 2008 2009 2010 (share of GDP) Preventive care 0.3% 0.3% 0.3% 0.3% 0.4% 0.4% Curative care 0.4% 0.4% 0.4% 0.4% 0.5% 0.5% Public health 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Health administration 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% Others* 0.1% 0.1% 0.1% 0.1% 0.1% 0.1% Total 0.9% 1.0% 1.0% 0.9% 1.1% 1.1% (share of total) Preventive care 31.8 33.9 35.4 33.2 36.1 39.2 Curative care 48.3 44.8 42.1 44.0 44.4 41.9 Public health 0.6 0.7 0.6 0.8 0.7 0.6 Health administration 11.5 13.3 13.4 11.8 12.0 10.6 Others* 7.9 7.3 8.4 10.2 6.9 7.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 *Excluye el Instituto Guatemalteco de Seguridad Social. Fuente: SICOIN. Health and Social Assistance (MPHSA) is entitled to efficiency frontier analysis. Considering these cave- provide free health services to the entire popula- ats, the analysis in this section remains at a stylized tion, so the poor can receive care at public health facts level, mainly describing observed trends in the centers even without insurance coverage. Neverthe- allocation of health spending and outcomes, and the less, monetary issues remain the main reason why potential correlation between them. the poor do not use public health services.79 Other reasons are on the supply side (transportation, dis- 3.21. There is a significant concentration of health tance, waiting times or insufficient health workers), spending in the department of Guatemala. At Q.287 but there appear to have been some improvements per capita, Guatemala received in 2008 and 2009 al- in this respect between 2006 and 2011, according to most double the amount of the average department household survey results. (Q.151), accounting for 36.6 percent of total central government health spending. Sacatepéquez (Q.212) Technical efficiency and Zacapa (Q.200) follow Guatemala in terms of ex- penditure per capita. Departments in the southwest 3.20. This section looks at some of the traditional- of the country tend to have lower health spending ly considered outcome indicators in the health sec- per capita (Chimaltenango, Q.87; San Marcos, Q.98; tor, such as infant mortality rates and immunization Suchitepéquez, Q.99; Sololá, Q.107) (Map 3.4). The Public Expenditure Review rates (see Herrera and Pang, 2005). However, unlike distribution of health workers is also heavily con- in the case of education, health outcome informa- centrated in Guatemala (with 30 doctors per 10,000 tion is not available at the municipal level. Working at people), Quetzaltenango (15) and Sacatepéquez (10), the department level (22 observations) coupled with compared to the national average of 4.7 (Figure 3.5). data limitations prevent the development of a robust There are also more nurses in Guatemala (20 per 79. This may be due to the fact that many health care facilities lack the required personnel, medicines and other inputs to provide free 52 services, thus requiring users to seek care outside the public health system where payment is required. 10,000 people) and Sacatepéquez (13), compared to 3.23. Health outcomes also vary greatly from de- the national average of 7.7. On the other hand, there partment to department. For instance, vaccination is no clear relationship between moderate poverty rates are highest in the departments in the center, levels and per capita expenditure on health services west and southeast of the country, even though some across departments. of the departments have low per capita health spend- ing. Baja Verapaz (87.5 percent) and Chiquimula (85 3.22. This unequal distribution of resources is due percent) are the departments with the highest vac- to the criteria used in the allocation of health ex- cination rates, compared with Izabal (56.9), Sacate- penditures, which reinforce historical trends. Until péquez (57.3) and Sololá (59.8). With respect to infant recently, public resources dedicated to health were mortality rates, Chiquimula (55 per thousand), Santa distributed mainly based on historical spending Rosa (51), Totonicapán (51), Sololá (49) and San Mar- trends, with a slight increase for inflation. Regional cos (48) present the highest rates, whereas mortality allocations were based on human resources, num- is lower in Guatemala (16) and Quetzaltenango (19) ber of beds and number of health establishments. As a result, poorer regions with inadequate health service coverage were assigned the lowest amount Map 3.4: Per capita expenditure on health of funds (regardless of population increase or epide- services by department (average for 2008-09) miological profile). Few incentives existed to improve service delivery and performance.80 It is hoped that the introduction of a pilot of results-based financ- ing in health and nutrition in selected municipalities in 2010 will improve the resource allocation among sub-national governments. In addition, the expan- sion of MiFaPro may also have created pressure on local governments to improve the availability of Yearly health expenditure per capita health services so that families can comply with their 87.27 - 107.81 112.53 - 135.79 co-responsibilities. However, deficiencies in budget 143.07 - 172.03 186.61 - 212.43 management continue to present problems for the 287.65 result-orientation of public expenditure (see Chap- Source: BOOST database using SICOIN data. ter 5 for more information). Figure 3.5: Distribution of central government health expenditure and doctors by department (average 2008-09) 35 350 30 300 25 250 20 200 15 150 10 100 Public Expenditure Review 5 50 0 0 GUA SAC ZAC ROS QUE BVE PET PRO IZA CQU QUI ESC RET JUT JAL AVE HUE TOT SOL SUC MAR CMA PC central exp. (right axis) Number of doctors per 10,000 people (left axis) Source: BOOST database using SICOIN. 80. World Bank (2010c). The 2005 PER also found deficiencies in the way resources are allocated and targeted in the sector, as well as in the strong centralization of planning,budgeting and management functions. 53 Map 3.5: Vaccination and infant mortality rates by department Vaccination rate (12-23 months) Infant mortality rate (per 1,000) 56.9 - 59.8 16 - 19 61.8 - 65.9 24 - 29 67.7 - 71.5 30 - 37 77.0 - 81.6 40 - 46 85.0 - 87.5 48 - 55 Source: ENSMI 2008-09. Figure 3.6: Relationship between vaccination coverage, poverty, per capita spending on health (bubble size) and malnutrition rates (color) by department Vaccination rate (12-23 months) 90 Chiquimula C B j Verapaz Baja V 85 Retalhuleu Huehuetenango 80 Santa Rosa San 75 Chimaltenango Escuintla Zacapa l Verapaz Quiche Alta Q i h Marcos M 70 Jutiapa Guatemala Suchitepequez h Peten Jalapa 65 J p El Progreso Totonicapan 60 Sacatepequez Izabal Quetzaltenango Solola 55 0 10 20 30 40 50 60 70 80 90 Moderate poverty rate Average High Malnutrition Low Malnutrition Sources: BOOST database 2008-09, ENSMI 2008-09, INE, 2009. Figure 3.7: Relationship between infant mortality, poverty, per capita spending on health (bubble size) and malnutrition rates (color) by department Infant mortality rate (per 1,000) 60 S Santa Rosa Chiquimula Solola Totonicapan 50 Jalapa J l Suchitepequez 40 Peten San Marcos El Progreso Quiche Q ih Jutiapa Public Expenditure Review Chimaltenango 30 E scu ntla sc Escuintla Retalhuleu Alta Huehuetenan Sacatepequez Baja Verapaz 20 Izabal Verapaz V Guatemala Zacapa 10 Q Quetzaltenango 0 0 10 20 30 40 50 60 70 80 90 Moderate poverty rate Average High Malnutrition Low Malnutrition Sources: BOOST database 2008-09, ENSMI 2008-09, INE, 2009. 54 (Map 3.5). High per capita health expenditure and a 3.26. At the same time, most of the departments strong concentration of medical facilities and staff in presenting high infant mortality rates are also char- the department of Guatemala may be an important acterized by a low percentage of births attended determinant for the lower mortality rates observed by doctors. This is the case of Sololá, Totonicapán or in that department. Chiquimula; Santa Rosa is an exception that combines high infant mortality rates with large share of births 3.24. Vaccination coverage does not seem to be attended by a doctor and above average health per correlated with health care expenditure or needs capita expenditure. The poorest departments, most indicators.81 There are no substantial differences in of them in the northwest of the country, are clearly vaccination rates across departments with dissimilar lagging behind in terms of percentage of births at- poverty levels, except for departments with poverty tended by a doctor. Since these departments do not rates under 40 percent which, against what would be present an average ratio of doctors per capita ratio expected, exhibit low vaccination coverage (Guate- that is significantly lower than other regions, a pos- mala, Escuintla, Sacatepéquez). On the other hand, sible explanation for this puzzle is the existence of a group of departments in the northwest, character- accessibility obstacles for the poor (distance to hos- ized by high malnutrition rates, significant poverty in- pitals and care centers, lack of information, etc.). cidence and average or low spending per capita, has high vaccination coverage (Quiché, Huehuetenengo, 3.27. As for the coverage of services at the first San Marcos, Alta Verapaz) (Figure 3.6). level of primary health, there does not seem to be a clear relationship between average health ex- 3.25. Mortality under five and infant mortality penditure per capita and coverage. The exception seem to be inversely correlated to public expen- is the department of Guatemala, where both levels diture on health care services, but only to a lim- are extremely high compared to other departments. ited extent. Chiquimula and Santa Rosa present In the range of 85-95 percent coverage we find very relatively high spending per capita levels and high dissimilar departments in terms of poverty levels, mortality rates. Some western departments of the malnutrition, adult literacy and public expenditure country (Sololá, Totonicapan, San Marcos) show high per capita. For instance, San Marcos has low spend- mortality rates, clearly correlated to high malnutri- ing, low literacy and high malnutrition rates; on the tion and an extremely high poverty incidence. On other hand Quetzaltenango displays relatively high the other extreme, Guatemala, Sacatepéquez, and public spending, above average adult literacy rates Zacapa spend more on health, and have lower pov- and average malnutrition, with lower primary health erty rates, better malnutrition figures and low infant care coverage than San Marcos. mortality rates. Similar to vaccination rates, infant mortality rates do not seem to be correlated to the D. QUALITY OF PUBLIC EXPENDITURE IN number of doctors per capita, except for the outliers THE ROAD TRANSPORT SECTOR82 Guatemala, Sacatepéquez and Quezaltenango, were the number of doctors is more than twice that of the 3.28. Guatemala’s road network has deteriorated in average department, and infant mortality rates are part due to consecutive natural disasters. The reg- thus much lower. This may suggest a threshold effect, istered road network has a total extension of 15,780 whereby departments with 10 or more doctors per kilometers, with about 11,600 kilometers of main, sec- 10,000 people have substantially improved health ondary, and tertiary network (60 percent paved), and Public Expenditure Review outcomes. 4,181 kilometers of rural roads.83 According to an as- 81. It would be necessary to determine the exact amounts spent on vaccination campaigns in order to more consistently assess the relation- ship; however, budget lines for vaccination programs are not available in the SICOIN data used to construct the BOOST database. 82. Due to the lack of information about relevant outputs for the sector (e.g. number kilometers built, rehabilitated or worked on, and the condition of the road network) at the sub-national level, an efficiency analysis similar to that conducted for the social sectors could not be carried out. This section therefore presents an overview of the transport outcomes at the national level, and a detailed analysis of public expenditures in the sector. 83. However, DGC estimates that the total road network including municipal roads is about the double of the classified network. In 2000 the total road network length was 26,000km, thus DGC currently estimates the total road network to be between between 28,000 and 30,000km. 55 sessment carried out by the Road Conservation Ex- hit Mexico and the Central American region, leav- ecuting Agency (Unidad Ejecutora de Conservacion ing significant damages in these countries, includ- Vial – COVIAL)84 before Hurricane Agatha in 2010 of ing Guatemala. The disasters of 2010 (Agatha and 77 percent of the paved network (5,362 kilometers), Pacaya) caused an estimated US$305 million (0.74 the condition of the road network is as follows: 62 per- percent of GDP) in damages in the transport sector cent in good condition, 14 percent in average condi- alone, while 12-E added another US$35 million (0.09 tion, and 24 percent in poor condition. It is important percent of GDP).86 Reconstruction investment needs to highlight that the assessment was carried out for were estimated at US$372 million (close to 1 per- the paved network in best condition and before two cent of GDP) over a 5-year period (2010-2014), put- events that significantly damaged the road network, ting additional strains on the already limited budget Agatha (2010) and the storm 12-E (2011). Therefore, for the sector (see Annex 4 for more information). these estimates do not present a complete picture of the network. According to estimates by the General 3.31. Guatemala budgets about 1.4 percent of GDP Directorate of Roads (Direccion General de Caminos – per year for the road sector. As a share of GDP, DGC) after the storm 12-E in October 2011, 65-70 per- Guatemala’s investment in road transport is lower cent of the paved network was in bad condition.85 than that of Panama (1.9 percent) and Nicaragua (1.8 percent), but above Honduras (0.9 percent) and El 3.29. In international comparisons, the condition Salvador (0.3 percent) (Table 3.3). of the Guatemalan road network is about average. However, in the last years the score of the Guate- 3.32. Compared with other sectors, municipalities malan road network has deteriorated based on in- carry a larger share of public expenditures in the ternational standards and in comparison with Latin transport sector. In 2008-09 (the years for which American and Caribbean (LAC) countries (Table 3.2). municipal expenditure data on transport is available), In 2008, only three LAC countries (Chile, El Salvador, 79 percent of total public expenditures for transport and Uruguay) were better ranked than Guatemala. was channeled through the central administration, The following years Panama and Mexico improved and 21 percent by the municipalities. This ratio of and in 2011 Ecuador and Dominican Republic were public spending through the municipalities is higher also better placed than Guatemala. for the transport sector than for total public spend- ing (14 percent).87, 88 The municipal share also varies 3.30. Given Guatemala’s vulnerability to natural from department to department – from 8 percent in disasters, investment needs in the transport sec- Zacapa to 54 percent in Totonicapan. On a per capita tor are also driven by damages caused by hurri- basis, public expenditures on transport range from canes, earthquakes and volcanic eruptions. Every Q.845 in the department of Retalhuleu to Q.203 in year Guatemala is hit by the negative impacts of the Chimaltenango (Figure 3.8). In general, there is rela- rainy and stormy seasons, and during the last ten tively little correlation between central and munici- years has suffered the negative impact of four strong pal government spending in the sector: those depart- natural disasters: Hurricanes Stan (2005) and Ag- ments that receive most resources from the central atha (2010), and the eruption of the Volcano Pacaya government do not necessarily spend more from the (2010). In October 2011, the Tropical Depression 12-E municipal resources in the sector. Public Expenditure Review 84. The annual assessment of the road network by COVIAL is through a simulator that includes subjective factors. The last study carried out to assess the structural condition of the road network was in 2004. 85. Currently, there is an initiative in place to update the road inventory and to assess the condition of the road surface with financing from the Japanese Agency for International Cooperation (JICA). The evaluation will include around one third of the network (5,600 kilometers) in two phases.The first phase that includes 4,100 kilometers will be completed before the end of December 2011 and the second phase of 1,500 kilometers will be completed by the end of the first quarter of 2012. 86. Government of Guatemala (2011) and CEPAL (2011). 87. Municipal data was available for 2008-2009. Even though municipal data does not subdivide transport spending by transport type (e.g. road transport), given the nature of municipal spending and transport services in Guatemala, most transport spending goes to road trans- port, instead of air, water, railway, and other types of transport. 88. One reason for this higher share of spending in the transport sector is that municipalities are required to spend transfers originating 56 from the fuel tax on roads and highways. Table 3.2: Condition of the road network in selected Latin American countries 2008-09 2009-10 2010-11 2011-12 Chile 5.5 5.8 5.9 5.7 El Salvador 5.0 5.3 5.4 4.8 Mexico 3.5 4.0 4.1 4.3 Panama 3.8 3.8 4.1 4.2 Ecuador 2.6 2.8 3.5 4.2 Uruguay 4.1 4.4 4.4 4.0 Guatemala 4.0 4.0 4.1 3.9 Dominican Republic 3.6 3.6 3.6 3.9 Honduras 3.7 3.5 3.5 3.4 Nicaragua 2.5 2.9 3.2 3.3 Argentina 2.9 3.2 3.3 3.2 Peru 2.6 2.9 3.3 3.2 Colombia 2.9 2.8 2.9 2.9 Brazil 2.5 2.8 2.9 2.8 Costa Rica 2.2 2.7 2.8 2.5 Guatemala's position among all countries 52 55 61 72 Number of countries 134 133 139 142 Source : The Global Competitiveness Report – World Economic Forum Note: Scores ranges from a scale from 1= underdeveloped, 7=extensive and efficient Table 3.3: Public expenditure on road transport in Central American countries* 2006 2007 2008 2009 2010 % of GDP Panama 1.2% 1.3% 1.6% 2.3% 1.9% Nicaragua 1.9% 1.7% 1.5% 1.7% 1.8% Guatemala 1.5% 1.7% 1.4% 1.3% 1.4% Honduras 0.9% 1.1% 1.5% 2.0% 0.9% Public Expenditure Review El Salvador 0.4% 0.3% 0.3% 0.3% 0.3% *Data based on approved budgets. Source: National Ministries of Finance. 57 3.33. Central government expenditures on trans- executed by DGC (7 percent through FSS in 2010) port are mostly channeled through the Ministry and the second one Maintenance by Contract of the of Communications, Infrastructure and Housing Road Network92 executed by COVIAL. Expenditures (Ministerio de Comunicaciones, Infraestructura under the Maintenance Program by COVIAL during y Vivienda, MICIVI),89 and the vast bulk of public the last three years have surpassed the amount of expenditures on transport by the central admin- the Road Infrastructure Program by DCG. The road istration are allocated to the road network (96 investments of the Reconstruction Program are chan- percent). MICIVI’s expenditures during the last nel through DGC and COVIAL, where DGC has ex- five years on average amounted to Q.4.5 billion, of ecuted the larger amount (79 percent as an average, which Q.3.5 billion were channeled to the road sec- reaching 93 percent in 2009). tor. Most road transport expenditures of MICIVI are channeled through DGC (56 percent), although CO- 3.35. Both DGC and COVIAL have programs in VIAL has also played a significant role in investments road maintenance and construction. Table 3.5 pres- in the road sector in the past 5 years (43 percent). ents the investment subprograms under the main COVIAL’s participation in investments reached its MICIVI program 11 - Development of the Road Net- highest level in 2008 and 2009, even surpassing work. Three sub-programs under the DGC together DGC’s investments (58 and 52 percent, respectively). accounted for 81 percent of the total: Pavement of Even though COVIAL’s expenditures were still high the Road Network (31 percent), Rehabilitation of the in 2010, it is expected that with the new decree ap- Road Network (27 percent), and Construction and proved in January 2010 that reinforced COVIAL’s re- Expansion of the Road Network (23 percent). The sponsibilities for maintenance only, the percentage expenditures under maintenance activities by DGC of expenditures channeled through COVIAL should average 11 percent of DGC subprograms.93 COVIAL, decrease to the levels observed in 2006-2007 (32 whose exclusive responsibility should be the main- and 28 percent, respectively). The Social Solidarity tenance of the road network, has been increasingly Fund (Fondo Social de Solidaridad, FSS) has thus far channeling public funds for new investments under executed only a small percentage of expenditures the budget project line Improvement of the Road Net- for road transport in 2010 (3 percent). work.94 This project line channels the highest amount of funds executed by COVIAL. It accounted for only 3.34. MICIVI’s capital investments for the past 34 percent of COVIAL’s total spending in 2006, but 5 years have represented 90 percent of the min- increased up to 60 percent in 2010. Thus, the divi- istry’s total expenditures. The main programs ad- sion between the two main subprograms under DGC ministered by MICIVI in the road sector are: De- and COVIAL is not synonymous with the distinction velopment of the Road Network (which receives 82 between construction and maintenance. Whereas percent of the funds) and Reconstruction Programs the average of the last three years of MICIVI funds (18 percent) due to natural disasters.90 The Devel- channeled through COVIAL for the road network opment of the Road Network Program has two sub- amount to 60 percent, the percentage channeled for programs, the first one Road Infrastructure91 mostly maintenance is only 29 percent. 89. There are other institutions that execute road works in Guatemala, such as the National Fund for Peace, the National Fund for Develop- ment, the Secretary of Executive Coordination and the Rural and Urban Development Councils. These institutions contribute between 10 and 16 percent of annual central government spending in the road sector. 90. Given the geographical location of Guatemala, there have been reconstruction programs due to three specific natural disasters for the Public Expenditure Review analyzed period – since 2005. 91. This subprogram includes pre-investment studies and designs. 92. This subprogram includes the supervision. 93. DGC maintenance program is carried out by force account through DGC’s regional offices; whereas COVIAL maintenance programs are all under contracts with firms. COVIAL maintenance program covers the paved network and some unpaved secondary roads (Transversal del Norte, 7E, 7W). DGC maintenance program covers the rest of the classified network that is not covered by COVIAL (in the pace that the current personnel and equipment allow them). 94. The programmatic budgetary allocation divides the budget first by programs, second by subprograms and third by projects. Improve- ment of the road network means increasing the road design level to a higher category (usually increase the width) so the road can serve a greater level of transit; Rehabilitation means rebuilding road sections or structures to reestablish the quality of the road surface when this has been deteriorated; Reconstruction means rebuilding not only the road surface, but also the road base and sub-base when they have 58 been already destroyed. Figure 3.8: Central and municipal expenditure in road sector by department (in quetzales per capita, 2008-09) 900 800 700 600 Q. per cápita 500 400 300 200 100 0 Retalhuleu Zacapa Jalapa Solola Peten Jutiapa Escuintla Guatemala Santa Rosa Suchitepequez San Marcos Izabal Chiquimula Quetzalengo Quiche Baja Verapaz Huehuetenango El Progreso Sacatapequez Totonicapan Alta Verapaz Chimaltenango Central Municipal Source: BOOST database using SICOIN data. Table 3.4: MICIVI expenditure in the road transport sector (Q. millions) 2005 2006 2007 2008 2009 2010 By institution DGC 1,000 1,852 2,849 1,540 1,606 1,937 COVIAL 697 882 1,081 2,084 1,707 1,741 FSS - - - - - 102 Total 1,697 2,734 3,929 3,623 3,313 3,781 By program Development of the Road Network (Progr.11) 1,685 2,135 3,369 2,985 2,692 3,090 Road reconstruction (Prog.97+98) 12 599 561 638 621 691 Total 1,697 2,734 3,929 3,623 3,313 3,781 Source: BOOST database using SICOIN data. 3.36. Financial needs in the sector are generally during the budget year from inter-ministry transfers, much higher than budgetary allocations. For exam- primarily from ministries that are unable to execute ple, DGC’s estimation of required financial resources all their allocated resources. In recent years, the ad- for 2012 is around four times the executed budget ditional resources allocated during the budget year of 2010 (Q.8.2 billion), including signed contracts for through inter-ministry or intra-ministry transfers reconstruction works after Agatha. In general, finan- were allocated to the Reconstruction Program of cial requirements include commitments made under DGC and its three main subprograms (constructions Public Expenditure Review contracts that have been signed but for which work and expansion, rehabilitation and pavement) and the has not been executed, but also for work that has improvement program of COVIAL. been executed under signed contracts but has not been paid.95 Generally the former is significantly 3.37. Budget overruns are rampant in road sector larger than the latter. As a result of the chronic defi- projects. Within DGC and COVIAL, it is general cit in budgetary allocations, MICIVI tends to benefit practice that contracts are extended in terms of time 95. Such payment arrears currently reach around Q.1.6 billion, roughly equivalent to the total budget executed by DGC annually. 59 Table 3.5: MICIVI’s investment and maintenance expenditures under Prog. 11 (Q. millions) 2006 2007 2008 2009 2010 DGC – Subprogram Road Infrastructure 1,468 2,478 1,047 1,031 1,320 Maintenance of the Road Network by force account * 122 120 109 134 132 Construction and Expansion of the Road Network 534 915 250 244 286 Rehabilitation of the Road Network 416 790 278 301 345 Pavement of the Road Network 347 613 340 247 455 Other 49 40 70 105 102 COVIAL – Subprogram Maintenance of the Road Network by Contract 668 891 1,938 1,661 1,668 Maintenance of the Paved Road Network 172 201 242 160 127 Maintenance of the Un-Paved Road Network 143 155 200 164 119 Improvement of the Road Network ** 230 407 1,072 1,086 980 Other 123 127 424 250 442 TOTAL 2,135 3,369 2,985 2,692 2,988 Source: BOOST database using SICOIN data. and cost increases. Overruns should be more the in part due to the reduction in allocated resources exception than the rule. The review undertaken by to MICIVI and its units for 2007 and 2008, but are CoST in the baseline study found that among the 16 mostly due to inefficiencies in the planning and man- projects reviewed in 8 agencies, the most significant agement of public resources and the contracts for variations of works contracts were within DGC and works and services within the sector. COVIAL, with average overruns of more than double of original terms, 123 percent in time and 103 per- E. CONCLUSIONS AND RECOMMENDATIONS cent in cost (see Chapter 4 for more information). Overruns of time and cost produce an additional 3.39. Better targeting of social spending is need- burden to administrative procedures and cause de- ed. The analysis suggests that public education lays or even stoppages in project execution. When and health spending does not benefit regions with roads projects are not terminated according to the the greatest need. While some education programs contracts, there is a risk that the works undertaken actually reach poor areas such as the school meal could suffer damage, causing transit delays, and road program, some others such as text books are con- accidents because usually signaling and protection centrated in urban and semi-urban areas. Spending works are among the last things to be executed. A allocation is also carried out without considering stu- considerable amount of work may also have to be re- dent performance and educational outcomes, such done, which implies a waste of public resources. as completions rates in 6th and 9th grades. Similarly, Public Expenditure Review in the health sector, there is a high concentration 3.38. As a result, budget overruns have been com- of public expenditure in the departments of Gua- mon in MICIVI in recent years. In the past two years, temala, Sacatepéquez and Santa Rosa. In the case budget execution rates for MICIVI, DGC and COVI- of the department of Guatemala, this has to do to AL were close to 100 percent, while there was a sig- some extent with the fact that the country’s two larg- nificant over-execution during the years 2006-2008 est hospitals are located there. It may also reflect (rates between 120 percent and 200 percent of bud- the fact that health remains a centralized sector and 60 geted resources). These higher execution rates are some of its expenditures continue to be registered in the department of Guatemala (due to the loca- lic expenditure across departments; this could be tion of the Ministry of Health and Public Assistance), pointing to potential deficiencies in preventive care even though the recipient or beneficiary department services. Results from the household survey confirm may be a different one. On the other hand, except that access to health care services remain a problem, for Quetzaltenango, less money per capita is allocat- especially affecting pregnant women in the poorest ed to departments in the southwest of the country. quintile living in rural areas. Thus, efforts could be The lack of clear prioritization in the allocation of aimed at increasing access to health facilities and resources in the social sectors does not allow for a doctors in the less-favored areas. In both education maximization of the benefits for the most vulnerable and health, outcomes appear to be primarily driven parts of the population. by demographics and other socio-economic indica- tors. This suggests that in order to increase the ef- 3.40. Monitoring and evaluation systems should be fectiveness of expenditures in these sectors, closer strengthened to facilitate deeper analysis of the coordination with other social programs is needed. determinants of sectoral outcomes. A key finding from the analysis is that there exist large disparities 3.42. Greater results orientation of the budget in spending performance in education across the ter- could help ensure that resource allocations are ritory. Disparities in the technical efficiency of public aligned with sectoral priorities. It appears as though education spending exist even after controlling for inequities in resource allocations currently reinforce a variety of socio-economic factors, suggesting that historical trends and existing endowments. Regional better data and more analysis would be needed to allocations of the health budget, for instance, have delve into the determinants of student achievement been based on the number of beds and health ser- and their linkages to education spending at the mu- vice entities in each location. As a result, the poor- nicipal level. In the transport sector, the lack of out- est regions – which tend to have fewer health facili- come and output information at the municipal level ties – receive less funds, regardless of demographic represents a challenge for decision makers to ade- or epidemiologic needs.96 The recent progress to pilot quately measure the impact of public investment in results-based budgeting in the Ministry of Health is the sector. Systematic information is needed to allow promising and represents a step in the right direction. making informed decisions among programs and sub- programs and among departments and municipalities 3.43. In the transport sector, the need to improve in the country. road infrastructure and reduce the net require- ment of fiscal resources for the road sub-sector 3.41. The analysis revealed a weak relationship continue to be a priority. Investments in the road between education and health spending and out- sector are important to keep up the provision of comes, suggesting that greater coordination among road infrastructure according to the demands of social programs could be important. In the health transit of goods and people for the competitive- sector, immunization rates do not show a clear-cut ness of the economy and trade. The constraints of relationship either with public spending on health or availability of resources calls for the need to use with poverty levels across departments. Only infant the scarce resources more effectively, improving mortality and mortality under five indicators show a road assets management, preparing good designs negative (but not significant) relationship with public and build infrastructure with sound technical stan- spending on health. The percentage of children that, dards that allow the infrastructure to resist the de- Public Expenditure Review according to the 2008/09 National Maternal Infant mands of transit and weather conditions of natural Health Survey (Encuesta Nacional de Salud Materno events, but also guarantee adequate maintenance Infantil, ENSMI), has respiratory diseases, diarrhea of the road infrastructure to extend its life cycle. and other illnesses appears to be much more related to living conditions and mortality rates in the differ- 3.44. There are a number of options that could be ent geographic areas, rather than variations in pub- analyzed in more detail to improve the condition 96. World Bank (2012). 61 Table 3.6: Expected impact of alternative financing options Term on fiscal Impact on fiscal Impact on impact resources users Concession Escuintla-Puerto Quetzal +++ + + Bypasses ++ ++ ++ Urban road networks + +++ +++ CREMAs +++ + ++ Source: Barbero (2009). of the road network that could also reduce the fis- right direction. However, some aspects of the law cal pressure for the sector. These options include still need to be further developed in order to pre- (i) including private participation for the financing pare and carry out PPP projects, such as the insti- of new infrastructure (or significant work improve- tutional arrangements. In addition, the sector would ments) and charging user tolls; and (ii) combining benefit from a capacity building program given the rehabilitation and maintenance activities under per- limited experience of PPPs in Guatemala. In this re- formance-based multi-year contracts. A preliminary gard, there are important experiences and lessons analysis in 200997 identified the concession of the learned of PPPs in the transport sector in the Latin route Escuintla-Puerto Quetzal and the Multi An- American Region that could be used to strengthen nual Contracts for Rehabilitation and Maintenance the sector if Guatemala decides to move forward (CREMAs) as the options that could be implemented with the agenda. in the short term (see Table 3.6). The CREMAs are likely to have a higher impact on lowering the public 3.46. Likewise, better classification of transport fiscal pressure and improving quality for the users. A spending into maintenance and investment would third option is the construction of bypasses in spe- help in monitoring the use of resources. Cur- cific points of the road network (along the CA-2 west rently, COVIAL’s investments are highly focused and CA-2 east, and eventually on the CA-9 and CA-1) on road improvements rather than maintenance. that could be built through concessions and would Even though the decree 5-2010 restated COVIAL’s reduce the congestion in these routes. The fourth responsibilities for maintenance in January 2010, option that would contribute even more to reducing investment for road improvement in 2010 under the fiscal pressure on public resources for the sector COVIAL was still high. Maintenance is important to is the construction of urban road networks. A com- maximize the life cycle of road investments and also prehensive project like this would have the highest extremely important to minimize disaster risks of impact in public financing and for users given that natural events. A revision of MICIVI’s subprograms would include the participation of the private sec- to better classify the investment and maintenance tor or public private partnerships for the financing projects under the correct classifications may help and would provide higher standard roads for transit. clarify and monitor the resources that are meant to The benefits for the road users as a result of the im- be channeled for new construction (including reha- Public Expenditure Review proved road infrastructure through the mentioned bilitation and improvements) and for maintenance. options include the reduction on travel time, travel Also, channeling resources for improvements of the cost, and traffic accidents. road network under DGC will allow using the pro- curement processes through the Law on Contract- 3.45. The approval of the Law on Public-Private ing by the State which increases the transparency of Partnerships in 2010 is an important step in the the procurement process. 62 97. José Barbero (2009). 3.47. A third element that currently limits the quality of spending in the sector is related to bud- get and contract management. DGC and COVIAL need to change their institutional culture and proce- dures to improve contract management and reduce to the minimum contract over-runs in terms of time and costs. Maintaining original contract designs and terms is important for improving budget manage- ment for projects in execution. It would also elimi- nate the waste of public financial resources related to projects that are not fully executed, which require additional public resources to restart works and complete the projects. 3.48. In short, these findings indicate that Guate- mala can improve both the efficiency and equity of spending. One way of doing so is to look more in depth at the determinants of performance variations across the Guatemalan territory to promote a more strategic approach towards improving the allocation of resources and strengthening comprehensive geo- graphic targeting and prioritization mechanisms. Public Expenditure Review 63 Chapter 4 Citizen Security and Public Expenditures 4 » A. Introduction 4.1. Crime and violence have become, once again, the most important destabilizing fac- tor for political, social, and economic life in Guatemala. The murder rate in 2010 reached 37 per 100,000 inhabitants, a level seen only during the worst years of Guatemala’s brutal civil war, and the seventh highest in the world.98 4.2. Today, however, the manifestations and drivers of the violence are significantly dif- ferent. The civil war that killed 200,000 people between 1960 and 1996 took place mostly during a military regime, in rural areas, and was driven primarily by internal stresses related to deep inequality and social exclusion. Today’s violence happens under a democratic re- gime and engulfs both rural and urban areas. Chronic, internal stresses such as persistent inequality and lack of opportunities continue as important drivers of the violence, but newer, external sources of stress – namely the rise of Guatemala as a major trafficking route for drugs – now compound the problem. The Guatemalan state now finds itself overwhelmed in the face of growing urban violence, and increasing activity by organized crime and drug trafficking organizations (DTO), with more and better trained members, highly sophisticated equipment, and exponentially more resources at their disposal. 4.3. Guatemala’s rising crime and violence exact a high cost on development. Estimates suggest that violence cost the country an estimated US$2.4 billion, or 7.3 percent of GDP in 2005.99 This figure represents more than double the damage caused by Hurricane Stan in the same year, and more than double the budget for the Ministries of Agriculture, Health and Education for 2006. A World Bank study estimated that bringing Guatemala’s homicide rate down by just 10 percent could boost per capita income by 0.7 percent.100 Public Expenditure Review 98. UNODC (2011). 99. Acevedo (2008). 100. World Bank (2010a). 65 4.4. The costs of crime and violence to the business sector represent a serious drag on growth and development. The 2011 Global Competitiveness Report ranked Guatemala as the worst performer, out of 142 countries, in terms of the costs that crime imposes on businesses.101 Earlier data from the Enterprise Survey suggested that the direct costs to the business sector averaged about 3.9 percent of sales in 2006, compared to 3.7 percent for the region.102 4.5. These economic costs pale in comparison to the immeasurable human costs of Gua- temala’s violence. In a 2010 report on Central America, the World Bank estimates that the health costs, measured in Disability-Adjusted Life Years (DALYs), are around 1.43 percent of GDP, the second highest in the region after El Salvador.103 The bulk of the costs of violence are incurred by the health sector, including services to victims, lowered productivity, and trau- ma.104 In comparison, costs to the legal and security sector were estimated at 1 percent of GDP, while the costs of private security to households and businesses were 1.5 percent of GDP. Material costs and private damages were estimated at 0.8 percent of output. 4.6. This chapter provides a preliminary assessment of the quality of public expenditures on Guatemala’s security sector. The analysis is necessarily limited by the availability of data. The chapter begins with a brief discussion of trends in crime and violence, followed by key drivers of violence. An analysis of the institutional response to crime and violence is given next, followed by an analysis of expenditures on security and justice, and their relationship with the high impunity rate. The final section offers recommendations for improving the state response to crime and violence. B. TRENDS IN CRIME AND VIOLENCE 2009 to 38.6 per 100,000 inhabitants in 2011. Al- though it is too early to call this a trend, it is impor- 4.7. Violence in Guatemala has been rising dra- tant to note that Guatemala is the only country in the matically in the post-conflict period. The homicide northern triangle that reported a persistent decline in rate – the most commonly used indicator of levels of the homicide rate during this period (Figure 4.1). violence – nearly doubled during the 2000s, from 24.2 per 100,000 inhabitants in 1999 to 46.3 in 2009. With an 4.9. Homicide rates only tell part of the story, how- annual average of 6,000 homicides per year (equal to ever. While homicide rates appear to have stabilized the number of homicides in the entire European Union, or possibly decreased in recent years, overall victim- where the population is 36 times larger) Guatemala ization levels are high and on an upward trend in Gua- now sees more murders annually than during the most temala. A survey conducted by the Latin American brutal years of its armed conflict. In 2010 Guatemala Public Opinion Project (LAPOP) reported that 23.3 had the seventh highest homicide rate in the world, percent of the population had been a victim of violent fifth in Latin America, and third in Central America.105 crime in Guatemala in 2010; the second highest rate Public Expenditure Review in Central America, following El Salvador. Previous 4.8. However, over the past two years, the homi- LAPOP surveys document a rising trend in violence cide rate saw an important decrease, from 46.3 in victimization in Guatemala, from 12.8 percent in 2004, 101. World Economic Forum (2011). 102. World Bank (2010a). 103. World Bank (2010a). 104. Acevedo (2008). 66 105. UNODC (2011). Figure 4.1: Central American homicide rates (2000-2011) 90 Honduras 80 Homicide rate per 100,000 70 El Salvador 60 50 40 Guatemala 30 Belize 20 Panama 10 Nicaragua Costa Rica 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011* Belize Honduras Costa Rica Nicaragua El Salvador Panama Guatemala Source: UNODC (2011) for 2000-2010; PNC for 2011. to 18.9 percent in 2006, 17 percent in 2008 and 23.3 trafficking of drugs), and white collar crime such as percent in 2010.106 money laundering. The boundaries between many of these types of crimes are fluid. Forms of Violence Coexistence-related crimes 4.10. Violence in Guatemala today is not the prod- uct of any easily identifiable group such as gangs 4.12. The most common coexistence-related crime or DTOs. Violence occurs in a range of arenas, from in Guatemala is intra-family violence. There is a long homes to schools to the streets, and is perpetrated history of violence against women and children in Gua- by a range of actors: from family members, to small temala, which intensified during the civil conflict as gangs, to private security and vigilante groups and both guerrilla and state armed actors targeted women organized crime. and children for physical and sexual abuse. Available data suggest that this violence continues to be high in 4.11. Crime and violence can be broadly catego- the post-conflict period, with some observers identi- rized in two types, based on the main motivations fying it as the most prevalent form of violence.107 Be- and factors driving them. Coexistence-related tween 2008-2011, an average of 60,000 cases per year crimes are motivated by social conflicts, and include of domestic violence against women were reported, in a range of crimes from intra-family violence, violent a population of 13 million.108 conflicts between neighbors and in schools, to femi- cide, lynchings, and social cleansing. Economically- 4.13. The most extreme form of violence against motivated crimes are motivated by economic gains. women (femicide)109 has reached alarming rates in They vary from petty crimes involving individuals (for Guatemala, prompting new legislation to protect example assault, cell phone theft and bus robbery) women. According to National Civil Police data, from Public Expenditure Review to those involving small groups (such as kidnappings 2000 to 2010, over 4,000 Guatemalan women and or extortion by gangs) to organized crime (such as girls were murdered. The country registered 708 femi- 106. Since 2004, LAPOP includes a question that asks the respondents if they have been victims of a crime in the past 12 months. The question was slightly modified in 2010 by adding some examples of crimes such as robbery, assault, extortion or threats. See: http://www. vanderbilt.edu/lapop/ 107. Adams (2011). 108. Guatemala Human Rights Ombudsman (2011). 109. The term ‘femicide’ is contested. Some definitions consider all murders of women to be femicide; others narrow the use of the term to the misogynistic murder of women, as a hate crime. The term here refers to the murder of women generally. 67 cides in 2009 and over 630 in 2010.110 In 2008 a law in the hundreds in recent years.118 Many of the victims to protect women against violence was enacted111, yet of such crimes are youth, which some observers at- femicide cases were on the rise, with over 720 femi- tribute to a frustration with the limited capacity of the cides reported that year.112 Because of the low rates juvenile justice system to deal effectively with criminal of prosecution and investigation of femicide in Guate- activity by youth. mala, it is difficult to know what is driving these crimes. Hypotheses include: male partners feeling threatened Economically motivated crimes by women taking on new roles both during and after the conflict; turf wars between rival male gang mem- 4.16. Much of Guatemala’s violence is driven by bers;113 a strategy by DTOs to distract attention from economic incentives, and is perpetrated by a variety other illicit activities, or to destabilize the government; of different groups. That is to say, the problem of vio- and attempts to warn women who transgress tradi- lence cannot be attributed to one group in particular, tional social norms.114 and likewise, all sectors of society seem to be affected by violence. 4.14. Lynching appears to be rising in Guatemala. According to a 2011 report from the Ombudsman’s 4.17. Petty theft in Guatemala is common, and of- Office,115 these crimes have been on the rise since ten involves violence. Of the individuals surveyed by 2004, going from 25 attempted lynchings in that LAPOP in 2010, 12.5 percent reported being a victim year (4 deaths) to 147 attempted lynchings in 2011 (47 of armed robbery in the previous year, the highest rate deaths). There appears to have been a particularly in Central America. This represents a slight increase large jump between 2008 (18 deaths) to 2009 (43 from 2008, when it was 11.5 percent.119 deaths). According to police records, in 2011 there were 49 deaths by lynching in Guatemala. These oc- 4.18. Kidnapping appears to be a significant prob- curred in 10 departments out of a total of 22, and in lem, and is likely under-reported. According to Na- 32 municipalities out of a total of 333.116 The causes of tional Civil Police data, there were 163 reported kid- the lynchings are complex, but seem to be motivated nappings in 2009, about half of which occurred in by frustration at the perceived inability of the state urban areas. Figures are only available for 2007 (96) to control crime and violence since they tend to con- and 2008 (213) making it difficult to identify trends. centrate around the border with Mexico where the Victims range from business owners, politicians, pub- state is mostly absent. lic servants and their families, to common citizens of all socioeconomic classes.120 “Express� kidnappings 4.15. Social cleansing (limpieza social) also appears are the most common form, in which the victim is ob- to be an important problem, and allegedly linked to ligated to visit ATMs and extract cash before being the rise – or in some cases, the resurgence – of ex- released. These crimes are likely under-reported be- tra-legal security groups. 117 Some of these groups are cause of difficulties in defining kidnapping, such that directed by ex-military officers and politicians – a re- these crimes are often classified as robbery or assault. sponse to ineffective security forces and a sense that In addition, reporting of kidnapping tends to decrease the security forces cannot protect people. One analy- as incidence of kidnapping increases, as people be- sis estimates that extrajudicial executions increased come more fearful or cease to rely on police to find by 60-70 percent from 2001-2005, with annual deaths victims, negotiate ransom, etc.121 Public Expenditure Review 110. Guatemala Human Rights Commission (2011). 111. Law 49-82 and 69-94. 112. Guatemala Human Rights Ombudsman (2011). 113. Musalo (2010). 114. Sanford (2008). 115. Human Rights Ombudsman (2011). 116. Mendoza (2012). 117. While both social cleansing and lynching are illegal extrajudicial executions, the former is a private, clandestine act while the latter is a public act that involves the community to produce some type of exemplary punishment. 118. Agner (2008) and Replogle (2005). 119. LAPOP (2008). 68 120. Ministerio de Gobernación (2009). 121. UNODC (2007). 4.19. Extortion is a frequent problem, especially for ported; thus their involvement in violence should not transport sector workers. These workers are vulner- be disregarded. able to regional organized crime groups, who extort periodic payments from them according to how many 4.22. The extent to which youth gangs are linked to passengers they transport, the number of kilometers larger DTOs is unclear. Some analysts suggest youth they travel and other variables. Resistance by trans- gangs are involved in extortion and kidnapping, and port workers, or their failure to comply with some of that some individual members may have begun hir- the demands, often results in homicide. In addition to ing themselves to Mexican DTOs as hitmen.127 Others these dynamics, there are also violent disputes be- have argued that the latter activities are mostly car- tween rival groups about who has access to certain ried out by organized crime, involving older and more buses or routes.122 experienced members than the youth gangs.128 4.20. Rates of car theft are high in Guatemala, and C. DRIVERS most car theft involves violence. According to the Prosecutor’s Office, the most common forms of car 4.23. Today’s violence in Guatemala is a result of theft are dismantling (50 percent), resale of the ve- the confluence of several factors, primarily weak hicle after changing the documentation or identifica- state institutions and the rise of Guatemala as a tion of the vehicle (30 percent) or theft of items left main trafficking route for drugs. With the signing of inside the car (20 percent). Violence is used in an es- the Peace Accords, just as Guatemala’s authoritar- timated 90 percent of these crimes. An estimated 25 ian institutions were moving toward democratization, groups specializing in car theft operate in Guatemala, control of the drug trade was shifting from South with between 4-6 members each. They operate with a America to Mexico, with Guatemala as a key transit certain degree of impunity given the difficulties in ap- country. Guatemala’s nascent democratic institutions prehending them and the fact that thieves can arrive have been vulnerable to the rising drug trade, which at and cross the national borders quickly, putting them makes profits dwarfing the GDPs of the region. Wide- beyond the jurisdiction of the police.123 spread availability of firearms left over from the war has contributed to this, and persistent inequalities 4.21. Violence and crime are also connected to and social exclusion have resulted in a large pool of youth gang activity, though the evidence on the ex- potential recruits for gangs and DTOs, creating a per- tent of this activity is thin and contradictory. Inde- fect storm for an escalation of violence in the post- pendent sources estimate there are 14,000 gang mem- conflict period. bers in Guatemala, and 434 distinct youth gangs.124 A 2006 report by USAID estimated that the majority 4.24. In Guatemala, as throughout Central Ameri- of these small gangs, called clikas, are loose affiliates ca, the increase in drug trafficking is probably the of the larger, transnational gangs Mara Salvatrucha single most important driver of violence. The per- (MS-13) and Calle 18. The extent to which these clikas centage of cocaine destined for the United States collaborate or are under the authority of the larger crossing through Guatemala has doubled between gangs is unclear.125 While gangs are often blamed for 2006 and 2010, from 9 to 17 percent.129 With this rise rising violence, there has been little empirical analysis in drug traffic has come a dramatic increase in homi- of gang activity in Guatemala, and the available data cides along trafficking routes. Just seven percent of casts doubt on the perception of gangs as the main the municipalities (25 out 333), mostly located along Public Expenditure Review instigators of violence.126 At the same time, it is likely drug routes, concentrated 60 percent of all homi- that many violent activities by youth gangs go unre- cides in 2009. 122. International Crisis Group (2010). 123. Central American and Caribbean Commission of Police Chiefs. Cited in World Bank (2010a). 124. Congressional Research Service (CRS) (2009). 125. Ranum (2006). 126. World Bank (2010b). 127. Guatemalan Human Rights Commission Gang Fact Sheet http://www.ghrcusa.org/Publications/GangFactSheet.pdf 128. Dudley (2010). 69 129. International Crisis Group (2011). 4.25. Weak state institutions have been over- tors to the problem of insecurity, through increasing whelmed by crime and violence, and when cap- employment of private security and vigilante justice. tured by criminal organizations have become part Depending on the estimate, the size of private secu- of the problem. With the lowest tax base in the Latin rity personnel is four to six times larger than that of American region, the Guatemalan state is under-re- the police.133 These security firms are often better- sourced to confront the problem of rising violence equipped than the police, and subject to limited, if and crime. The police remain under-funded and any, formal regulation.134 poorly equipped. The legal system – with a dearth of adequately trained investigators, prosecutors and 4.27. The widespread availability of firearms com- forensic examiners, and lacking modern equipment pounds the problem. Guatemala has the highest es- – is similarly overwhelmed. There is evidence of infil- timated rate of gun ownership in Central America. tration of some criminal elements into state institu- There are approximately 16 guns per 100 people in tions charged with maintaining law and order, further Guatemala, compared to 2.8 in Costa Rica and more undermining institutional capacity to deal with crime than double the rates of Panama (5.4), Honduras (6.2), and violence. Former Guatemalan Police Chief Er- El Salvador (7) and Nicaragua (7).135 A 2008 study of win Perisen estimated that, in 2007, a full 40 percent firearms in Guatemala found that firearms were over- of the PNC had been involved in corruption.130 Or- whelmingly present in reported incidents of violent ganized crime groups routinely receive early warn- crime. In the same survey, 31 percent of participants ings of planned raids by police forces, and recently, said they already owned some kind of firearm, while a security officials were alleged to have helped drug third of them said they had handguns specifically. The traffickers steal weapons from military armories. The proliferation of firearms in Guatemala is both a legacy extent of the infiltration of these elements into state of its armed conflict and a reflection of the growing institutions contributes to a culture of corruption presence of the drug trade in the region. The lucra- throughout the Guatemalan state system. Public of- tive drug trade helps finance the purchase of firearms, ficials are increasingly reluctant to report abuses in which then contribute to more violent crime. This their agencies for fear of retribution. problem is aggravated by the flow of arms from the Southern United States through Central America, es- 4.26. The weak institutional capacity of the Guate- pecially Honduras and El Salvador. malan state to stem the growing violence contrib- utes to a crisis of confidence in the state, and in de- 4.28. An inequitable pattern of socio-economic de- mocracy generally. Chronic violence and a growing velopment has created an increasingly disaffected culture of corruption fuel fear, distrust and cynicism segment of the civilian population, vulnerable to in Guatemalan society. In a 2006 survey, 86 percent criminalization. The lack of access to economic and of Guatemalans reported feeling that the level of in- educational opportunities – especially for poor fami- security in the country poses a serious threat to the lies – is a key factor underlying Guatemala’s crime and future well-being of the country.131 This sense of in- violence problem. In 2006, 44 percent of youth aged security contributes to a tendency toward more au- 14-17 in the bottom quintile (the poorest 20 percent of thoritarian governance systems. A Latinobarometro households) are both out of school and economically survey reported that only 8 percent of Guatemalans inactive. Even in urban areas, where education access felt democracy works better in their country than is better, 15 percent of the population is both out of in the rest of the region, the lowest figure among school and economically inactive. With few available Public Expenditure Review all Latin American countries. Equally troubling is a alternatives, youth are more easily lured into activities tendency for Guatemalans to look to non-state ac- with organized crime or urban gangs. 130. Bureau of International Narcotics and Law Enforcement Affairs (2009). 131. USAID (2006). 132. Latinobarómetro (2008). 133. A 2010 UN report documented at least 107,000 private security agents while another report from ICG estimated around 150,000 in 2007. The PNC has 25,000 policemen. 134. Ortiz (2009), Porras Castejón (2008). 70 135. Karp (2008), Arias Foundation (2005), cited in World Bank (2010a) D. THE INSTITUTIONAL RESPONSE violence control and prevention. The security sector is led by the Ministry of Interior (Ministerio de Gober- 4.29. While Guatemala has made progress in its in- nacion) which has to coordinate with a wide range of stitutional environment, creating institutions and ministries and agencies that perform critical functions passing legislation, the implementation track record for the provision of security. This includes several min- has been very poor. Progress has been hindered by istries involved in crime prevention, the police and a number of factors, especially the lack of a reliable penitentiary system (which are part of the Ministry of system to collect and manage information about vio- Interior), and the Guatemalan justice system which in- lence, and the lack of coordination across the differ- cludes the Judicial Branch (Organismo Judicial), the ent institutions and sectors that share responsibility Office of the Attorney General (Ministerio Público), for violence control and prevention. These are also and the Public Defender’s Office. Personnel policies, complicated by high turnover in key institutions, es- strategic planning, and data collection and analysis are pecially within the Minister of the Interior and the Na- conducted by the different institutions in isolation, re- tional Civil Police (PNC). sulting in much inefficiency. 4.30. The disconnect between the ambitious aspira- 4.33. Several ambitious initiatives have fallen short tions of national policy and the reality of conditions of their stated goals. The 1997 Accord to Strengthen throughout Guatemala is striking, and reflects the Civilian Power and the Role of the Military in a Dem- scale of the task the government faces in providing ocratic Society was signed as part of the Peace Ac- meaningful security to its citizens. For example, of cords. In its wake, the government established a Com- the ten national policies relating to violence preven- mission to Strengthen Justice in order to support the tion identified by the Secretariat for Planning and reforms envisioned in the Peace Accords. By 2007, Programming (SEGEPLAN), only three have defined important inroads had been made in addressing the budgets.136 There is little buy-in from the private sec- needs of a multicultural, in many cases illiterate popu- tor at the national level, and perhaps even less from lation (through the introduction of oral arguments and local-level stakeholders. translators); transitioning from an inquisitorial to an adversarial system; the creation of an Attorney Gen- 4.31. The most significant challenge to implement- eral; a school for judges; and the establishment of a ing policy on violence control and prevention is argu- democratic procedure to name high-ranking judges ably the lack of a coordinated system of information and magistrates, among others. However, in spite of collection and analysis. In Guatemala, information on this progress, many positive reforms were found to be crime and violence is highly unreliable even for most short-term and dependent on international support. A basic indicators, such as homicide. Data on crimes is system that privileged the powerful and repressed the not disaggregated by sex or age, or by other key in- weak was still in place, and judicial officials struggled dicators such as ethnicity, type of crime, or location with threats to their lives, as well as co-option by pow- of the crime. Different institutions employ different in- erful forces.137 dicators, formats and recording systems. Civil society organizations have led important efforts to construct 4.34. Similarly, the 2009 National Agreement for indicators and monitor trends, but these have been the Advancement of Security and Justice promised isolated and uncoordinated. Under these conditions to increase accountability of institutions. This law it is impossible to construct a solid evidence base on holds the State and its institutions responsible for Public Expenditure Review which to build effective policy. “guaranteeing security and the administration of jus- tice for the well-being of the population, obligating 4.32. Overall, the institutional response has been that they protect the individual and family…�.138 How- complicated by the lack of coordination mechanisms ever, a 2010 appraisal showed that its implementation across institutions which share responsibilities in had produced scarce results. According to the Grupo 136. Partners for Democratic Change (2012). 137. Calvaruso et al (2008) 138. Op cit, 25. 71 Garante (G-4), a multi-sectoral group in charge of mon- 4.38. Investment in rehabilitation and reintegration itoring the 2009 agreement, of the 101 commitments of criminal offenders has also been lacking, especial- contained in the Agreement only 12 were fulfilled after ly for youth. The Guatemalan government has experi- nine months, 45 showed some progress and 44 no ac- mented with a variety of innovative prevention strate- tivity. The causes of this failure are rooted in the con- gies, especially for youth, but these have been mostly text of a weak state and the absence of political will or financed by donors and have not been scaled up.139 political leadership to ensure its full implementation. The DIGEX program, financed by the Education Min- istry, which offers vocational training to at-risk youth, 4.35. Guatemala approved a law on gun control in is often cited as one successful example. Similarly, in 2009, but has faced challenges in enforcing it due to terms of juvenile rehabilitation services, Guatemala limited resources. Only an estimated one quarter of has only four juvenile detention centers, all of which the estimated one million guns in circulation are for- are located in the department of Guatemala. An in- mally registered. People cite various reasons for not dependent assessment found all the centers failed to registering their guns, from concerns about confisca- meet the mandated standards related to reintegration tion, to the fear that, if they do use the gun (even in and support services and to assure full realization of self-defense) they could face prosecution. the rights of youth, especially minors, in detention.140 4.36. The disconnect between policy ambitions and 4.39. On the bright side, one of the most promising their implementation is also apparent at the local initiatives, the International Commission against Im- level. To begin with, the role of local governments in punity in Guatemala (Comisión Internacional Con- citizen security is not well established in the legislative tra la Impunidad en Guatemala, CICIG) has made framework. While security is clearly a national level important gains, but formidable challenges remain. mandate, the Constitution (article 259) authorizes The CICIG was created in December 2006 at the municipalities to create municipal police and justice request of the Guatemalan Government and began systems (juzgados municipales) under the authority of functioning in January 2008. The establishment of the the mayor. There is space in the Municipal Code for CICIG was the result of the strong political will of the taking on additional responsibilities at the municipal Government, private sector and civil society. Funding level, which could potentially include an expanded is provided by various donors, including private funds role in citizen security. for the Commission’s two-year mandate, an extension of which will be endorsed by incoming President Per- 4.37. At the municipal level, the Municipal Boards ez Molina. CICIG works with the Public Ministry and for the Protection of Children and Adolescents other justice-sector institutions to “investigate and (MBP) are charged with some violence prevention dismantle violent criminal organizations believed re- responsibilities, but these are poorly funded and are sponsible for the widespread crime and the paralysis not formally linked to the justice system. Established in the country’s justice system.� With donor support, in 1997, the MBPs were designed to protect children the CICIG has been trying to develop and fortify cells and adolescents from violence, and as such consti- of highly trained, vetted prosecutors and police that tute an important mechanism for community outreach can eventually lead reform from within their respec- and programming. Currently 160 MBPs are organized tive institutions. throughout the country, covering approximately half of all municipalities. These bodies are coordinated by 4.40. CICIG made important progress in pursuing Public Expenditure Review departmental or municipal assistant of the Office of some high profile cases. In less than a year and a half, Human Rights and thus lack formal jurisdiction within partly as a result of CICIG’s work with state agen- the justice system. Though their role as a liaison with cies, 1,700 people have been expelled from the po- communities is recognized, they are severely under- lice force, including 50 senior officials and the deputy resourced. director of national policy; 10 prosecutors had been 139. See Cooper and Ward (2008) for a review, and IADB (2008). 72 140. Partners for Democratic Change (2012). asked to leave and the Attorney General has been re- the lowest in terms of GDP. As Figure 4.3 shows, El placed. However, the fight against impunity in Guate- Salvador leads both categories. It spent more than 20 mala remains a daunting task, exemplified by the low percent of its budget on security compared to Guate- homicide conviction rate. mala, Nicaragua and Honduras that spent between 10 and 15 percent. El Salvador spent about 3.5 percent E. FINANCING OF SECURITY AND JUSTICE of its GDP compared to 2.3 percent in Guatemala and 2.8 in Honduras.141 4.41. This section analyzes the main trends and sources of funding for security and justice, the bal- 4.44. Public expenditure on security and justice ance and composition of these expenditures across is concentrated in three main institutions: the the core pillars of the sector, and the degree of pre- Ministry of Interior in the case of security, and the dictability and flexibility of expenditures. Judicial Agency and entities not affiliated with spe- cific ministries (such as the Constitutional Court, Trends and sources of funding Human Rights Office and others)142 in the case of justice. Between 2006 and 2010, there were no 4.42. Between 2006 and 2010, the budget for se- significant changes in the distribution of resources curity increased by fifty percent and the budget among executing branches; during these years the for justice by thirty percent in nominal terms. As of Ministry of Interior and the Judicial Agency and 2010, the Guatemalan government allocated Q.7.5 bil- other entities controlled more than 80 percent of lion, or 15 percent of the total budget. As a share of security and justice expenditures. Allocations to GDP and of the overall budget, however, the alloca- key security institutions such as the National Civil tion for security and justice has remained relatively Police and the Judicial Agency remained around flat: in 2010, it represented 2.3 percent in terms of 4.4 percent and 4.2 percent of the total budget, re- GDP, the same share as in 2006. spectively, between 2008 and 2010 (see Annex 4). 4.43. Compared to its Central American neighbors, 4.45. At the municipal level, security and justice Guatemala’s security and justice expenditure is the expenditure account for 8-9 percent of municipal second highest in terms of overall budget share, but budgets.143 The bulk of this spending (7.8 percent) Figure 4.2: National Government expenditure on security and justice (2006-2010) 4,500 18% 4,000 16% 3,500 14% 3,000 12% 2,500 10% Q. millon 2,000 8% 1,500 6% 1,000 4% 500 2% 0 0% Public Expenditure Review 2006 2007 2008 2009 2010 Security % of national budget (left axis) Justice % of GDP (right axis) Source: BOOST database using SICOIN data; UNDP (2011). 141. There is a slight discrepancy between UNDP’s and our own calculations of the share of national budget that corresponds to security. This is probably due to the inclusion of some different programs in either process. As a share of GDP, however, the numbers coincide. 142. Allocated through the budget line “obligations of the state in charge of the treasury�. 143. Budget data for the security function at the municipal level is only available for 2008-2009, while budget data for the justice function at the municipal level is only available for 2010-2011. 73 Figure 4.3: Security budget in Central American countries(2006-2010) (a) as a share of national budget (b) as a share of GDP 25 4 El Salvador 3,5 El Salvador 20 Nicaragua 3 Panama 15 Guatemala Belize Honduras Nicaragua 2,5 % % Honduras Costa Rica 10 Belize Guatemala Costa Rica 2 Panama 5 1,5 0 1 2006 2007 2008 2009 2010 2006 2007 2008 2009 2010 Source: UNDP (2011). Figure 4.4: National Government expenditure Figure 4.5: National Government expenditure on security, by institution (2006-2010)* on justice, by institution (2006-2010)* 100% 91.2% 56% 89.0% 89.0% 87.6% 88.0% 54.1% 90% 54% 80% 52.3% 51.7% 51.5% 51.6% 52% 70% 60% 50% 48.5% 48.8% 48.3% 47.7% 50% 48% 40% 45.9% 46% 30% 44% 20% 11.0% 11.0% 12.4% 12.0% 8.8% 42% 10% 0% 40% 2006 2007 2008 2009 2010 2006 2007 2008 2009 2010 MoI Other Judicial Agency Other entities *Allocated budget. Source: BOOST database, UNDP (2011), authors’ calculations. *Allocated budget. Source: BOOST database, UNDP (2011), authors’ calculations. Figure 4.6: External funding for citizen security (2005- 2011) Canada 2,6% EU 12,2% Germany Denmark 17,7% 0,4% Public Expenditure Review Loans (IDB) Grants 10,3% 89,7% USA Sweden 21,0% 10,2% UN 6,6% Spain Holland 9,9% 8,5% Other Foundations (Kellog, Ford, Soros, OSF) 0,8% 74 Source: Authors’ calculations based on IDB and WOLA (2011). is allocated to the justice sector, with most of it fi- cient, for example if jails are unable to accommodate nanced through central government transfers (70 new inmates, or worse, if jails are not able perform percent). Security spending, on the other hand, is their function of rehabilitating inmates. If the jails are primarily financed with municipal own source rev- only ‘universities for crime’ and recruiting centers for enue. Municipal security spending is concentrated in organized crime, the recidivism rates will be high and the department of Guatemala, the only subnational systems for prevention and control of crime will be entity with its own Municipal Police force, which ac- overwhelmed. For this reason, each pillar is connect- counts for over 70 percent of the total; expenditures ed with the pillars that precede and follow it. on justice, on the other hand, are more evenly dis- tributed. 4.48. A broad definition of the type of programs that contribute to crime and violence prevention 4.46. International development cooperation fi- suggests that at least twenty percent of the security nances an average of US$43.5 million per year, budget is dedicated to this purpose in Guatemala.145 which represents about one-third (28.7 percent) of However, a closer look at the content of some of these the average annual national budget for security, ex- programs raises questions about their real impact in cluding wages and salaries, or 6.9 percent including terms of preventing violence. For instance, more than salaries and wages. Guatemala is the country in Cen- three quarters of that 20 percent comes from three tral America with the highest share of funding from programs which do not have violence prevention as international development cooperation for citizen se- an explicit objective and may affect it only indirectly: curity.144 The country received US$304.2 million from (i) FONAPAZ, which invests in poverty reduction proj- 2005 to 2011, which is about one third of the total for ects in areas that were affected by conflict but where the region, mainly oriented to institutional strengthen- currently there are very low rates of crime and vio- ing and prevention. Ninety percent were grants and lence; (ii) the Human Rights Prosecutor, which is re- the rest loans. Funding comes mainly from the United sponsible for investigating crimes committed by gov- States government (21 percent), German government ernment officials which abuse their authority; and (iii) (17.7 percent), European Union (12.2 percent) and IDB follow up to the Peace Agreements, which include a (10.4 percent). In total there were 90 projects identi- large number of institutional commitments in a num- fied, with an average size of US$3.4 million. Ninety-six ber of different areas (not only prevention). This sug- percent of these projects are in implementation and gests the need for a follow up study on the nature of the rest are being designed. each program and ways in which they may contribute to the prevention of violence. Composition and balance of expenditures 4.49. The budget for is fragmented across multiple 4.47. For the security system to function properly, ministries with a weak coordinating role of the Min- there needs to be a balance across the different pil- istry of Interior (MOI). Most of the budget for pre- lars of the system (prevention, control, justice, and vention is executed by more than five other ministries rehabilitation). A solid preventative system will not including the Ministries of the Presidency, Education, be effective if it is not accompanied by an efficient Health, and FONAPAZ. While the presence of pre- system for control and repression of violence. At the vention programs across sectors is positive since it same time, a developed public security sector will fall speaks to the mainstreaming of the issue, the Ministry short of its objectives if the Prosecutor’s Office is not of the Interior only allocates 1.37 percent of its secu- Public Expenditure Review able to prosecute cases effectively. A strong Prosecu- rity budget for prevention activities. Budget fragmen- tor’s Office cannot be a prudent steward of resources tation in absence of effective coordination mechanism if the courts are not able to take the cases forward and lack of policy directives results in the impossibil- effectively. In turn, the sentences given by the courts ity of integral policy on security. The incapacity to be will be of little use if the penitentiary system is defi- able to coordinate the programs for primary, second- 144. This data is based on own analysis of data generated through a mapping exercise done by IDB and WOLA (2011). 145. UNDP (2011). 75 ary and tertiary prevention that fall within its purview ministries. Even though the MOI has the mandate is impeding its broader role of defining public security for citizen security, and is responsible for the Nation- policy in Guatemala. The MOI does not have a pres- al Civil Police, other ministries such as the Ministry of ence, participation, or in some cases even knowledge Defense, the Presidency of the Republic, and others of the prevention-oriented activities of other govern- control about half of the budget. In addition, the lack ment agencies. The Ministry, however, as the main of budgetary instruments for coordination makes it policymaking body on prevention, should oversee the very difficult to plan, implement and evaluate secu- coordination of these programs and ensure a degree rity policies. Good practices in inter-agency coordi- of coherence and consistency between them. nation suggest that the resources available for the execution of certain public policies should be man- 4.50. Similarly, the MOI manages only half of the aged by the agency responsible for the design and budget for crime control, and lacks budgetary in- coordination of those policies.146 In addition, the im- struments to coordinate with the other spending pacts of the activities carried out by other agencies Figure 4.7: Spending on prevention and crime control (2006-2010) (a) prevention expenditure (b) prevention and crime control expenditure 20% 19.1% 18.5% 7,000 18% 17.8% 17.3% 16.9% 6,000 16% 14% 5,000 12% 4,000 Q. millon 10% 7.2% 3,000 8% 6% 2,000 4% 2.8% 2.7% 1,000 1.3% 1.5% 1.1% 1.4% 2% 0.5% 0.5% 0.6% 0% 0 2006 2007 2008 2009 2010 2006 2007 2008 2009 2010 % of total security budget allocated to prevention Crime control outside MoI Crime control controlled by MoI % of total prevention budget controlled by the MoI Prevention controlled by MoI Prevention outside MoI % of total budget of the MoI allocated to prevention Source: Authors’ calculations based on data from UNDP (2011). Table 4.1: Budget allocated to security and justice functions (Q. million, 2011) Fixed Costs Operational Costs Total % Fixed Costs % Operational Costs Ministry of Interior 300 96 396 75.7% 24.3% National Police 1,994 458 2453 81.3% 18.7% General Direction 1,909 312 2221 85.9% 14.1% General Sub Direction of Studies 63 84 147 43.0% 57.0% Public Expenditure Review General Sub Direction of Police Health 13 6 18 69.0% 31.0% Transit Department 9 56 66 14.4% 85.6% Other institutions 237 379 616 38.5% 61.5% Source: BOOST database. 146. The Ministry does not oversee budget reporting and evaluation of the use of resources assigned to these activities managed by other 76 Ministries. Since the coordinator agency should in the chain of management and budgeting. When this is not the case, the result is that implementing agencies maintain their own priorities and follow their own bureaucratic culture, with the consequence that their activities are not necessarily in line with those of the Ministry. that are not coordinated with the Ministry are not 4.54. Transfers from the MOI to other ministries measured and cannot be attributed to any program happen across all budget lines and may contribute in particular. to perverse behavior by public servants. As Table 4.2 shows, all budget items are affected. Adding both 4.51. Resources allocated by the international modifications and under-execution, the Ministry loses development cooperation show also some imbal- 17.3 percent of its budget. It is hard to imagine how the ances. They are primarily destined to strengthening Ministry can function when it loses a share which is the managerial capacities within the security and jus- comparable to its budget for operational costs, which tice sectors (57.4 percent), as well as crime control may be a structural incentive that leads regular street (24 percent) and prevention programs (18.6 percent). workers (like policemen) to petty corruption. Again the glaring absence is rehabilitation and rein- sertion. F. ASSESSING THE OUTCOMES Flexibility and predictability of the budget 4.55. The lack of resources to establish a com- prehensive and coordinated policy for primary, 4.52. There is an inadequate balance between fixed secondary and tertiary preventive programs is af- and operating costs. Because many recurrent costs fecting the government’s capability to deter crime are inelastic, the distribution of resources for pro- and reduce violence. Elsewhere, effective primary grams and operations is crucial. In Guatemala, the preventive programs have demonstrated important process of resource allocation does not allow for the reductions of motivation for crime and violence; ef- effective functioning of the security policy. As shown fective secondary prevention and policing programs in the graph, the MOI spends less than 1 of every 4 have been successfully reducing opportunities for dollars of its budget on operations (non-wage current crime and violence; and well established tertiary expenditure). This ratio of 1 (operation) to 3 (fixed prevention programs have been effective in reduc- costs) differs substantially from best practice guide- ing recidivism. Such programs are neither properly lines, which recommend a ratio of 2 to 1 for an efficient funded nor articulated under Guatemala’s security public security system. The situation for government policy, and, as a consequence, underperform. The agencies that implement security programs (including lack and poor quality of data on crime and violence the Ministry of Justice and the Prosecutors Office) makes it almost impossible to establish proper pre- and those agencies that do not fall under the MOI ventive policies, strategies, programs and actions to have a somewhat better ratio (but still far from the address and evaluate them. High impunity rates com- ideal) of 3 to 2. More worrisome is the situation in the plete the combination of factors that prevents the National Civil Police, where, on average, the ratio is 1 Guatemalan state from sustainably lowering crime to 4. A police force cannot operate with only 19 per- and violence rates. cent allocated for this purpose. Analyzing the impunity rate 4.53. Budget under-runs, transfers, and modifica- tions frequently affect the resource availability for 4.56. The “impunity rate� measures the percentage public security in Guatemala. These variations affect of crimes that do not have an adequate institutional significant and relevant parts of the budget, as well as response (whether through the police and justice the normal and logical performance of project plan- system or alternative conflict resolution mecha- Public Expenditure Review ning and implementation. The national budget in secu- nisms). As such it provides a summative assessment of rity decreases on average 14 percent between alloca- the performance of a security and justice system. It has tion and execution. This reduction is mainly explained implications both in terms of preventing future crime by a decrease in the budget of the MOI (Table 4.2) through deterrence, as well as bringing justice to the that has to transfer resources to other ministries and victims of crimes, and maintaining an enabling environ- governmental programs. Between 2004 and 2010, the ment for the rule of law. In terms of deterrence, one MOI lost between 13 and 21 percent of the budget that of the findings of the American empirical criminology was initially allocated. literature is that the most important factor to deter 77 Table 4.2: Modifications and execution of security budget (Q. million, 2010) Descripción Approved Actual Modifications Execution Budget Budget Modified % Executed % Ministry of Interior 3,275 2,708 -567 -17.3% 2,629 97.1% National Police 2,170 1,902 -268 -12.4% 1,873 98.5% Planning 139 130 -8 -5.9% 125 96.1% Civil Intelligence Service 85 72 -13 -14.8% 58 80.7% Anti-Narcotics Research 10 8 -2 -16.0% 8 95.2% National Police Academy 200 73 -127 -63.4% 67 91.7% National Police Hospital 23 17 -6 -26.7% 16 95.8% Criminal Investigation 6 0 -6 -100.0% 0 Crime Prevention 6 0 -6 -100.0% 0 Migratory Control 193 111 -82 -42.5% 103 92.9% Community-based Prevention of 16 24 8 52.9% 20 84.0% Violence Prison System 249.2 245.7 -4 -1.4% 244.8 99.6% Transit Department 54.9 11.7 -43 -78.7% 9.1 77.8% Departmental Governments 61.6 64.6 3 4.9% 60.4 93.5% Register of Persons 9.5 0 -10 -100.0% 0 Official Gazette 53.5 49.4 -4 -7.7% 44.7 90.5% Source: SICOIN. crimes is not the severity of the punishment but the fectiveness of the state which manifests itself either in certainty of it, more precisely the perception that of- the low reporting of crimes or the incapacity to inves- fenders have about the likelihood of being punished if tigate and process criminal cases. they break the law.147 One way to assess the deterrent capacity of a system is to assess the level of impunity 4.58. The rate is only slightly better for homicides. that exists. The higher the level of impunity, the lower In 2009, only 3.5 percent of homicides ended up in a the capacity to deter crimes. High impunity rates do trial. There were 6,498 homicides, of which 4,626 en- not allow victims to have closure, incites vigilantism, tered the judicial system, of which only 230 ended up and lowers trust in the state. in a verdict.148 4.57. For every one hundred crimes that are com- 4.59. About eighty percent of crimes are not re- Public Expenditure Review mitted in Guatemala, only 1.2 end up in a trial and ported to the police. Victimization surveys are one only 1 in a conviction. Some cases do not go to trial way to find out the percentage of crimes that are not for good reasons - i.e., the offense was minor, or was reported. According to a POLSEC study, underre- settled through alternative conflict resolution mecha- porting rates per type of crime for the whole coun- nisms. However, the large majority is due to the inef- try hover around 80 percent for all crimes expect car 147. Kennedy (2008). 78 148. Organismo Judicial (2009). Figure 4.8: Impunity rate for all crimes Prevention/ Of�ce of Trial Sentences Prison Police the Attorney System General • 100 crimes • 9% resolved • 33% �led • 17% absolutions • Recividism • 80 not reported • 91% �led • 1.2 sentences • 1 conviction • 20 police • 1.8 cases reports Source: POLSEC; UNDP; Instituto de Estudios Comparados en Ciencias Penales de Guatemala (2007). theft (since a police report is required by insurance deprives government from gaining crucial empirical companies). Another study analyzes a set of six vic- information about the circumstances that lead peo- timization surveys conducted in the municipality of ple to commit crimes, which results from the opera- Guatemala, and shows that rates consistently average tion of the criminal justice system and which is vital around 75 percent. for effective policy making. 4.60. Under-reporting is symptomatic of the high Institutional factors that contribute to the levels of distrust of the police and the criminal jus- impunity rate tice system. When asked about the reasons for not reporting, 59 percent said that they did not think it 4.63. The disfunctionality reflected in the high lev- would have helped for anything, 10 percent said they els of impunity is a result of multiple factors. In this were scared of retaliations, 6 percent argued compli- section we focus on how the performance of the police cated bureaucratic procedures, and 13 percent said and justice system contribute to this overall outcome. that the offense was minor. Police 4.61. Of the crimes that are filed with the Pros- ecutor’s Office for investigation, only nine per- 4.64. Guatemalans’ trust in their police is the lowest cent are sent to be tried in court.149 The other 91 in Latin America and the Caribbean, after Argentina percent is archived or dismissed due to multiple rea- (Figure 4.10). Allegations of police involvement in cor- sons that include lack of evidence, offender has not ruption, human rights violations, participation in ‘social been identified, and so on. Of the cases that enter cleansing’, drug trafficking, robbery, kidnapping for the courts for trial, one-third is archived. If only 1.8 extortions and assassinations have brought frequent percent of crimes were sent for trial, and 30 percent changes in staffing, even at the highest levels, and in are archived, this would mean that only 1.2 percent of the organization of the PNC. Within this unstable con- crimes end up in a trial. text, it is hard to build discipline, and as a result the im- age of the police as an institution has suffered. In turn, 4.62. High impunity rates send the message to so- this situation makes police officers and government Public Expenditure Review ciety that crime has no cost or consequence to the officials vulnerable to corrupt influences. The loss of offender. It eliminates the deterrent effect of crimi- staff from firing of police personnel, which fluctuates nal justice for future offenders and the healing or between 300 to 2,500 a year, many times cannot be compensatory effect for victims of crime. Finally, it offset by recruiting of new personnel. 149. Of the crimes that are reported to the police, not all make it to the Prosecutor’s Office for investigation. Most likely causes for this include lack of suspects (no suspects, the product of the crime is not identifiable, small claims, etc.). Quantitative evidence is not available for this, but the suspicion is that it is relatively high. 79 Figure 4.9: Unreported crimes in Guatemala 100% 13% 14% 90% 19% 9% 20% 0% 27% 80% 43% 70% 60% 83% 50% 88% 86% 82% 8 80% 40% 73% % 30% 57% 20% 10% 17% 1 0% Car- property Fraud Assault Injuries House robbery Threats Car robberies robbery Not reported Reported Source: POLSEC (2005). 4.65. Guatemala has the lowest police presence of ing in the national police academy. The academy was all countries in Latin America (Figure 4.11). There are almost closed in 2000 due to the lack of resources. about 25,000 police officers, for a ratio of about 180 The government is considering increasing the basic per 100,000 inhabitants. When the number of officers training period to 9 months and adding a special train- employed in activities such as bodyguards for govern- ing course for officers. Some officers are able to take mental officials, judges, prosecutors, and politicians, courses in other countries. However, these cannot border patrol, or with duties in penitentiaries or jails is compensate for the real lack of specialized training taken into account, only about 35 percent of this num- in line with Guatemala’s legislation and culture that is ber is actually left to patrol the streets. The Central needed for intermediate and high ranking officers. American average is 312 (almost double) and the South American average is 547 (triple). The peace accords 4.68. The Ministry of Interior recognized in 2010 mandated a police force of 20,000 officers, but in that 70 percent of police officers live in conditions 2005 the United Nations Verification Mission in Gua- of poverty. They are located far from their nuclear temala (MINUGUA) recommended raising this num- families as part of a policy to avoid corruption and fa- ber to 44,000. Levels of crime and violence are higher voritism. Various South American countries adopted today not only compared to 1996, but also to 2004. the opposite approach, based on the premise that po- The government of Perez Molina has announced its lice officers working within their own social networks intentions to hire an additional 10,000 police officers. would face more social pressure against corrupt be- havior, and that officers would function more effec- 4.66. The resources available to the PNC (which tively in neighborhoods they know best. In contrast, represent 70 percent of the budget of the Ministry the opinion of many Central American police is that of Interior) are dwindling. More than 82 percent of it is better for officers to be based apart from their these are fixed costs to the institution, leaving only families because of the threats by organized crime and a small margin for operations. Best practices, estab- drug-related crime to their families. As a result, Guate- lished by the experiences of other countries that have malan police officers live in the same areas where they Public Expenditure Review advanced in this area, suggest that fixed costs should work (some even sleep in the cells or hallways of the not account for more than 35 percent of the total. This police stations) and spend a lot of their free time trav- suggests that the PNC would need an additional 150 eling to visit their families. With low-quality weapons, percent of its current budget, assuming that this entire and without effective modes of communication, such increase would go toward operations. as cell phones, police officers often feel they have no authority. The creation in the last decade of special 4.67. The qualifications of existing police officers units (Special Police Forces and Rapid-Action Groups) 80 are very low. Police receive three months of train- to address highly complex crimes and corruption es- Figure 4.10: Trust in the National Police (2010) Argentina 6 30.6 Guatemala 31.0 Trinidad and Tobago 6 31.6 Jamaica 6 32.6 Venezuela 35.5 Mexico 6 36.4 Paraguay 6 36.7 Belize 37.9 Dominican Republic 8 39.8 Peru 41.3 Bolivia 8 42.8 Guyana 42.3 Costa Rica 49.1 El Salvador 49.3 Brazil 53.1 Haiti 53.7 Honduras 54.1 Nicaragua 54.7 Uruguay 6 55.6 Colombia 6 56.0 Panama 57.9 USA 8 58.1 Suriname 59.5 Canada 6 6 62.6 Chile 6 70.6 0 10 20 30 40 50 60 70 80 Source: LAPOP, Latin American Opinion Project (2010). Figure 4.11: Police force per 100,000 habitants Uruguay 756 Colombia 635 Argentina 547 Venezuela 495 Average S. America 469 Panama 455 Ecuador 449 Belize 417 Peru 417 Mexico 406 Chile 403 Paraguay 382 Bolivia 344 Average rest of C. America 312 El Salvador 299 Costa Rica 291 Brazil 261 Nicaragua 207 Public Expenditure Review Honduras 204 Guatemala 180 0 100 200 300 400 500 600 700 800 Source: LAPOP, Latin American Opinion Project (2010). 81 tablished an elite caste of officers, who are granted the state’s capacity to conduct forensic research. This better equipment and a sense of superiority toward would be consistent with the creation of INACIF (Insti- their less fortunate colleagues. tuto Nacional de Ciencias Forenses) in 2007 and the allocation of a substantial budget for its functioning. 4.69. PNC has one-third the number of criminal investigators needed for the volume and type of 4.71. There is an insufficient number of public de- crimes that it investigates. This has been a problem fenders with an ineffective geographical distribu- for some time: on 2001, MINUGUA reported only tion. In 2005 Guatemala had only 255 public defen- 742 investigators in the PNC headquarters. In 2003, dants for more than six million people living in poverty. this number had decreased to 637, in contrast to the That is about 4 public defendants for every 100,000 1,500 recommended by MINUGUA. According to the poor people. The geographical distribution shows the Minister of Interior of the Colom administration, the urban bias: 49.5 public defendants in the metropolitan government needs to triple the number of criminal in- region versus 10.5 public defendants on average for vestigators that it currently has from 2,000 to 6,000 the other departments. The allocation of public de- to reach minimum international standards. fendants does not correspond to the number of cases entered in the system. In 2005 the metropolitan area Prosecuting crimes had a surplus of 13 percent while the other regions had a deficit of 7.8 percent.150 4.70. The turnover rate is low, but it is improving. De- spite the sustained increase in the volume of cases 4.72. The excessive duration of the judicial process that were filed in the judicial system between 2007 is another significant shortcoming in the perfor- and 2010, the percentage of cases that were tried has mance of the criminal justice system. Since the ju- slightly increased during the same period. A possible dicial system does not measure the time that accused explanation for this improvement is the increase in citizens are under investigation or trial, the ICCPG (In- Table 4.3: Effectiveness in the resolution of cases filed in first instance, according to judicial categories, (2005 & 2009) Costa Rica El Salvador Guatemala Honduras Nicaragua b/ Panama 2005 2009 2005 2009a/ 2005 2009 2005 2009 2005 2009 2005 2009 Effectiveness c/ Penal 112 99 62 65 65 72 56 32 69 77 72 104 Civil 63 37 148 199 21 21 95 29 36 20 105 104 Labor 108 113 188 142 69 48 88 52 31 31 111 97 Family 98 97 52 79 16 16 40 36 46 50 87 103 Work Load d/ Penal 65 100 21 20 25 34 2 2 9 12 16 22 Public Expenditure Review Civil 52 100 19 21 12 14 6 8 27 38 19 18 Labor 74 100 26 22 28 40 7 11 10 8 10 11 Family 91 100 83 94 215 296 26 33 39 85 91 97 Notes: a/ Data for El Salvador is for 2008; b/ Data for Nicaragua only includes cases for civil, labor and family filed in the first instance, and penal filed through the Code of Criminal Procedure. The methodology used only accounts for the cases closed the same year they were filed; c/ Proportion of cases closed by year to the cases filed the same year; d/ Proportion of cases filed by category to the total of cases filed that year. 82 150. Instituto de Estudios Comparados en Ciencias Penales de Guatemala (2007). stituto de Estudios Comparados en Ciencias Penales the approach the government has been following de Guatemala) conducted in 2004 a research project so far. Despite progress in some aspects or areas, across the country, providing empirical evidence of overall, the government’s response has been frag- the abuse that exists in this area. The law stipulates mented, discontinuous, and unbalanced. Confront- that the maximum amount of time a person could ing the challenge of criminal violence in Guatemala be detained for investigation and trial is 164 days in requires a comprehensive approach that addresses case of preventive prison and 254 days if there is no the multiple dimensions of the problem through pre- preventive prison. The study showed that the actual vention and law enforcement initiatives, coordinated time is 73 percent more of the maximum time in case by a clear authority within government, and with the of preventive prison (284 days instead of 164) and 47 capacity to oversee implementation of policies in a percent more in case of non preventive prison (374 in- consistent manner. stead of 254).151 4.76. Expenditures seem to be low for the size of G. CONCLUSIONS AND RECOMMENDATIONS the problem. Major rethinking is required to deal with the problem given the Guatemalan scarcity of 4.73. Crime and violence in Guatemala is a multi- resources. There is a need to boost resources along faceted problem that calls for a comprehensive the security and justice value chain, especially for: policy response given its multiple forms and pres- preventive programs and community policing; law ence across territories, demographic groups, and enforcement’s manpower and equipment; policing income levels. In addition to being the country with operations; improved access to justice; improved the 7th highest homicide rate in the world, Guate- prosecutors office capabilities; and penitentiary and mala is afflicted by many other forms of crime and rehabilitation systems. violence, from domestic violence and femicides to extortions, kidnappings, armed robbery, and all sorts 4.77. Better coordination of prevention programs of illegal trafficking. While each type of event tends to is also needed. The MOI should develop a system to be clustered in specific geographic areas of the coun- monitor the execution of all programs related to crime try, when taken as a whole it is difficult to find places and violence prevention, including those implemented that lack some form of violence. Similarly, while male by other sectoral ministries. The government should youth tend to be particularly affected by homicides, create an inter-ministerial coordination mechanism to women and other age groups are also affected by oth- review this information and identify areas for perfor- er forms of violence. Finally, while victimization tends mance improvement. A useful exercise for this group to increase with income level, poor people who can- would be to assess whether the main risks for crime not afford private security often have their neighbor- and violence have a matching program or set of in- hoods controlled by criminal gangs, and are subject to terventions to address them and the extent to which everyday forms of violence. these programs are deployed in an integrated fashion in the hot spot communities that concentrate these 4.74. Governance institutions are being over- risks. The government should also consider whether whelmed by the ability of criminal groups to infil- (i) it is desirable to give formal authority for coordinat- trate and capture them. Weak institutions have not ing interventions to one of the ministries, for instance, been able to cope with external pressures brought authorizing it to be present in the decision making by increased drug trafficking and this, combined with process and even resource allocation of other minis- Public Expenditure Review deep inequality and lack of opportunities for youth at tries’ programs when they have an important citizen risk and other vulnerable groups, has produced the security function; and (ii) decide whether this role explosive situation that the country finds itself in. should be given to the Ministry of Interior. 4.75. A review of government policies, budgets, 4.78. A short-term priority for government should outcomes and outputs suggests the need to rethink be to improve its information systems on (a) the 151. Ibid. 83 incidence of crime and violence and (b) the per- ress and ensure greater dialogue and coordination formance of the security and justice institutions. across the multiple actors involved in the citizen se- The decision making process cannot perform prop- curity sector. Indicators could include: share of public erly without quality and appropriate data and data expenditures per pillar; coverage in terms of number analysis. This review found many gaps in this area. of personnel per capita assigned to each of the pil- Information systems are also critical for improved co- lars, whether police, public prosecutors, public defen- ordination and accountability. Below are some recom- dants, etc; managerial capabilities; recidivism rate; and mendations on how to approach this task. citizen confidence in public institutions. 4.79. Guatemala should invest in a second genera- tion crime and violence information system. This system should (a) capture the range of crime that goes unreported, (b) validate the data that is generated, (c) produce key descriptive information about the char- acteristics of the crime, and (d) feed all this data into a geo-referenced, user-friendly system that is useful for pro-active, results-oriented decision making on citizen security. Violence observatories are an institutional “technology� that has been proven useful to address this challenge, usually combining a national level unit that works to coordinate with, and improve national government information systems with, a network of municipal observatories in hot spot municipalities that validate and use the data for local citizen security strategies. Key descriptive information needed for de- cision making include: type of incident (homicide, rob- bery, assault, car theft, rape, etc); date, time, day of the week, address of where incident takes place, neigh- borhood; victim and perpetrator information (name, age, sex, occupation, literacy level, home address); type of weapon and context of event (interpersonal violence, domestic violence, organized crime, gang related, confrontations with law enforcement); and relationship of perpetrator to victim. The objective in- formation captured by government systems should be complemented with regular standardized surveys that produce subjective and situational information on vic- timization, perception of insecurity, violent attitudes, coping behaviors to reduce victimization, and urban and social characteristics of hot spot communities. Public Expenditure Review 4.80. Guatemala should also strengthen its informa- tion system on the performance of security and jus- tice institutions. To assess the performance of citizen security institutions, government should build a core set of indicators for each of the pillars of the value chain (prevention, law enforcement, criminal investi- gation, criminal justice, penitentiary, rehabilitation). 84 These indicators should serve to benchmark prog- Chapter 5 Budget Process, Public Financial Management and Procurement 5 » A. Introduction 5.1. Sound Public Finance Management (PFM) systems help countries manage their pub- lic resources to achieve key strategic goals. The political process determines the allocation of resources within the overall budget envelope. Then, Ministries and Agencies decide how to spend them, subject to control and accountability mechanisms. The processes through which those decisions are made and carried out are known as PFM systems. Thus, PFM seeks the fulfillment of strategic objectives through interactions of systems, institutions and practices, with three main objectives: (i) to promote strategic priorities; (ii) to ensure fiscal discipline; and (iii) to deliver value for money.152 5.2. The annual budget process, which is the central component of PFM, tends to be strongly influenced by political interests and institutional arrangements. It is therefore important to anchor that process to clear and transparent rules. The Public Expenditure and Financial Accountability (PEFA) assessment measures the core budget processes based on best practices within six critical dimensions: (i) budget credibility; (ii) comprehensiveness and transparency of the budget; (iii) policy-based budgeting; (iv) predictability and control in budget execution; (v) accounting, recording and reporting; and (vi) external scrutiny and audit. The PEFA assessment measures PFM system performance by scoring 28 key indica- tors and uses them to compare systems over time and across countries. 5.3. Guatemala has a strong Financial Management Information System (FMIS) in place. FMIS support the government in carrying out financial processes and transactions in an automated manner. In addition, FMIS help governments produce reports on the financial situation of the overall public sector to facilitate and make accountable decisions over the use of public resources.153 In the case of Guatemala, the FMIS is known as the Sistema de Administración Financiera (SIAF) within a core accounting functionality called Sistema de Contabilidad Integrada (SICOIN). The first phase of the SIAF was launched in 1997; since then, it has been expanded and improved with World Bank assistance.154 The SIAF meets im- portant requirements: (i) it covers the core PFM systems (i.e. budget, treasury, and account- Public Expenditure Review ing); (ii) it covers most of the public sector entities including the municipal governments; (iii) it operates under a Web platform which enable users to interact in real time with the system; (iv) it produces on-line reports, many of which are publicly accessible; (v) it enables sectors 152 See DfID (2001) and PEFA (2005). 153. See World Bank (2011b). 86 154. For a more detailed reference of the Guatemala experience on FMIS reforms see World Bank (2011b). to track and assess the situation of some financial as well as administrative processes through the Sistema de Informacion y Gestion (SIGES) system; and (vi) it is interlinked (albeit not yet completely) with the e-procurement system (Guatecompras). 5.4. Part A of this chapter is drawn primarily from the PEFA report. The report was jointly prepared by the World Bank, Inter-American Development Bank and European Commission in 2009. Box 5.1 lists the main strengths and weaknesses of the Guatemalan PFM systems found by the PEFA report. It also includes some sensitive topics identified by the PEFA analysis that are affecting the system’s overall credibility and represent broader governance challenges. Based on these findings, the section is organized as follows: On budgeting, Section 1 describes the budgeting process; Section 2 looks at the predictability and control in budget execution; Section 3 examines accounting, recording and reporting; and Section 4 evaluates external scrutiny and audit. On PFM practices, Section 5 presents the alternative vehicles to execute the budget outside the SICOIN system; Section 6 analyzes the quality of expenditures of the alternative vehicles; Section 7 looks into the accountability mechanisms of the alternative vehicles; Section 8 examines floating debt issues; and Section 9 summarizes the main conclu- sions and recommendations. 5.5. Part B draws on a new set of detailed procurement data, as well as a number of analyti- cal studies conducted in recent years. It includes examples of the specific challenges faced in the procurement of pharmaceuticals, as well as in the procurement of public works for infra- structure projects. Box 5.1: PFM strengths and weaknesses identified by the Guatemala PEFA (2009) Strengths • general public has access to fiscal information in a user-friendly way and without restrictions; • a debt sustainability analysis (DSA) is undertaken annually for both domestic and external debt; and • the tax administration plans and carries out annual audits based on integral risk management system as well as is able to maintain the taxpayers registry updated. Weaknesses • considerable variation of the expenditures composition from the original budget; • high level of arrears that cannot be confirmed accurately and opportunely; Public Expenditure Review • budget classification not fully aligned to international standards especially on functional classification critical to track spending by purpose; • incomplete reporting on budget execution; • lack of consolidated fiscal information of the overall public sector; • omission of some account balances; • lack of a single human resources database, incomplete descriptions of personnel profiles, and lack of audits of the human resources systems; • widespread use of non-competitive processes to procure goods and services; and • weak internal and external accountability mechanisms, including the role of Congress in overseeing the decisions of the Executive. 87 B. BUDGET PROCESS AND PUBLIC which affects the credibility of the fiscal targets; FINANCIAL MANAGEMENT and e) Weak controls and sanctions. The sanctions re- Key challenges in the budget cycle gime is weak to effectively deter practices that are not consistent with PFM systems performance and 5.6. Guatemala’s budget process is characterized credibility. by the co-existence of good practice and vulner- abilities. It is supported by detailed norms and regu- Budget preparation lations, an Access to Public Information Law and a good practice e-government system, the Integrated Process Financial Administration and Governmental Audit System (SIAF-SAG). Recent improvements to the 5.7. As with the other phases of the budget pro- SIAF, which supports the budget cycle, mean that cess, budget preparation is principally governed the main PFM systems are now internally coherent, by the 1997 Organic Budget Law and its regulation. sophisticated and powerful in recording and inte- Responsibility for budget formulation is assigned in grating financial management information for the these norms to the Technical Commission for Public entire public sector. The SIAF provides timely, accu- Finance (CTPF), a body presided by a Vice-Minister rate, and comprehensive financial and, increasingly, of Finance and composed of high-ranking Ministry of performance data to officials and public alike, and is Finance (MOF), National Planning Secretary (SEGE- a fundamental building block for the extension and PLAN), SAT and, upon invitation, Central Bank (BAN- consolidation of modern results informed budgeting GUAT) officials. Between January and May each year, within the context of multi-year framework. These MOF and SEGEPLAN elaborate the (i) strategic ori- formal advances are reflected in the country’s per- entations of public policy, (ii) budget norms, (iii) norms formance in the 2010 Open Budget Index, ranking for the elaboration of annual operating plans and, sub- highest in Central America (IBP 2010). At the same sequently, (iv) an exploratory budget with ceilings for time, the budget process in practice is rife with un- each public institution. Between April and May, tax sanctioned violations of legal norms, limited usage of revenue projections are calculated by SAT and MOF formal systems, de-linkages with sector needs, and on the basis of macroeconomic projections by BAN- limited expenditure control, to name a few. GUAT, and MOF undertakes public debt projections. Between May and June, a technical proposal with (i) The main remaining vulnerabilities of the budget pro- the total budget, (ii) budget ceilings for each public cess are the following: institution, (iii) estimated total revenues, (iv) revenue sources and (v) budget norms is presented to the a) Weak linkages between planning and budgeting. President and his cabinet. Budget ceilings frequently Also, there is no systematic evaluation of results lie below those demanded by the Ministries and Sec- attained; retaries. The Presidential approval of this proposal is b) Large budget reallocations. During budget execu- thus accompanied by intense lobbying by Ministers tion, sectors tend to modify the original budget allo- and other political operators with cabinet access, cation, which reduces the overall budget credibility; pressing for larger budget allocations or modifications c) Alternate vehicles to execute the budget. Fidei- to revenue sources and budget norms. While these comisos and NGOs manage public resources out- norms are generally regulated in the Organic Budget Public Expenditure Review side the budget, making difficult to assess the qual- Law, they are complemented by specific dispositions ity of expenditures and perform the corresponding in the annual budget. These have been important ve- external audits properly; hicles for enhanced budget transparency in recent d) Lack of control over the Floating Debt. Debt is cre- years, for instance by mandating the registry of public ated without the corresponding budget credits— expenditure operations in the SIAF system.155 88 155. Barrientos (2011). 5.8. Based on the approved budget ceilings, the in- terms and estimations of associated costs, including in stitutional programming phase of public institutions some cases estimations of unitary costs. These cost commences. Public investment projects are coordi- estimations form the basis for the preparation of the nated between central-level institutions, in particu- budget by linking the program structure of the annual lar SEGEPLAN, and the Departmental Development operating programs (Programas Annuales Operativos, Councils (CODEDEs), which are to aggregate projects POAs) with the program structure of the budget. Un- prioritized by planning processes in the community der the annual operational plans specific and measur- and municipal development councils. In July, draft pro- able objectives, outcomes, and indicators are added to gramming proposals are submitted by each institution the individual programs, in order to facilitate monitor- to MOF for its revision and integration. In August, a ing of results and allow for evaluation of performance. budget project is presented by MOF to the economic cabinet, and from there it advances to the general cab- 5.10. However, in practice, this approach is not fully inet for discussion and Presidential approval. Once the implemented because the FMIS cannot support dis- draft budget is sent to Congress, Guatemala’s Constitu- aggregated spending categories, limiting sectors tion stipulates that Congress is empowered to approve, to include in the budget only aggregate activities, reject or modify the budget proposal’s size, composi- which - once the budget is approved - facilitate real- tion and revenue sources at will. This amendment pow- locations among categories. Thus, the more detailed er places Guatemala in cross-country terms on the far activities and tasks of the operational plans are not end of legislative leverage over the budget process.156 fully incorporated in the budget. Also, sectors have no And, indeed, legislative amendments to the executive’s incentives to review the unitary costs of activities be- budget proposal – with inter-institutional modifica- cause if they found that fewer resources were needed tions of budget allocations amounting to an average to deliver services or procure goods, this could result of 4.9 percent157 of the total budget in 2008-12 – seem in lower budget allocations the following year. On the to be more pronounced in Guatemala than in other other hand, these revisions are critical to ensure the Latin American countries for which data is available.158 quality of program budgeting in order to keep it con- sistent with the public policy goals. In addition, and Policy orientation of the budget depending on the accounting capabilities of the FMIS, sectors in the POAs might only estimate the direct 5.9. The Government adopted a policy-based costs of their activities because the indirect costs budgeting approach, which is designed to ensure such as salaries, utilities, and others cannot be disag- due regard to government policies in the budgeting gregated. Also, if the unitary cost of health services process.159 In operational terms, policy-based budget- does not include the depreciation of a health facility, ing requires that budget preparation be carried out funding for maintenance may not be budgeted and through a programmatic structure where the stated services would deteriorate over time. programs reflect the policy intentions. A more de- tailed description of those policy intentions are stated 5.11. To overcome these problems, budget prepa- in the annual operational plans which describe the ration has been evolving from program budgeting programs, subprograms, activities and tasks to be car- (with no effective links to planning) towards a more ried out during the annual fiscal year mainly in physical performance-oriented budgeting.160 These efforts Public Expenditure Review 156. See, for instance, Wehner (2006). 157. As in subsequent calculations, the percentage refers to the sum of the modifications of budget ceilings for public entities. Following the country’s legal dispositions, only inter-institutional modifications – i.e. those which would require in the execution stage a presidential rather than only ministerial decree to be undertaken – are considered (see annex A for the resulting level of disaggregation). 158. The 4.9% average exceeds the 1997-2005 averages for Colombia, Peru, Venezuela, Brazil, Ecuador and Bolivia presented in Hallerberg et al. (2009). 159. PEFA (2005). 160. The evolution from program to performance budgeting comprises at least four stages. The first stage consists of setting up the program structure within a wider context of budget planning and policy, often as part of a Medium Term Expenditure Framework (MTEF), which is considered a strategic element to connect policy objectives with detailed annual budget plans. In a second stage it is critical to provide budget managers with more flexibility by removing the institutional and information constraints in order to make activities, programs, and strategic goals more consistent among each others. The third stage foresees the improvement and expansion of budget-costing systems. Finally, the fourth and most complex stage foresees the introduction of new accountability mechanisms and budget incentives to complete 89 a policy-based budgeting that ensures an effective and efficient use of public resources management. began in 1997 when the Ministry of Finance (MOF), resistance to the change, ownership of those indica- in conjunction with sectoral ministries, started to re- tors remains weak, and there is no formal evaluation view the programmatic structure of the budget and of the resulting indicators. It is expected that once the to make the budget processes more comprehensive. pilot initiative ends and the methodology has been Although these revisions have enhanced the program- adjusted, more sectors will go through the process based budgeting focus, by enabling sectors to allocate of reviewing their program structure, defining results resources to programs, subprograms, projects, activi- indicators, improving the budget preparation, and get- ties and works, the association between activities and ting better results from the use of public resources. programs still needs to be improved. In some cases, activities are defined very broadly (i.e. administrative 5.14. In particular, the MOF is interested in carrying cost or vaccination) which makes it difficult to assess out impact evaluation studies of public expenditures their relevance for the programs they are part of. in cooperation with sector ministries. In this context, the MOF is planning to implement on a pilot basis an 5.12. Since 2007, the MOF has implemented – on a expenditure review of some critical programs, begin- pilot basis – a budget based on results, with support ning with the child nutrition program executed by the from the European Commission (EC). The objec- Ministry of Heath. Although Guatemala has allocated tive is to improve the links between the planning with significant resources to combat child malnutrition, the budgeting processes by defining results indicators and country has one of the worst malnutrition indicators in the program structure both in the operational plans the world. The objective of the revision is to establish as well as in the budget. The critical element of the if the service has been delivered efficiently and in a new methodology is the definition of the results to be timely fashion. achieved with the available resources. The Ministries of Health, Education, and Interior were selected as Predictability and control in budget execution pilots, and since 2008, their budget proposals have been incorporating the new methodology. The Minis- 5.15. PFM systems are crucial in ensuring compli- try of Health has made significant progress: its 2012 ance during budget execution. Principally regulated budget defines two strategic objectives based on the by the 1997 Organic Budget Law, budget execution MDGs – (i) percentage of children under five whose falls under the sway of the CTPF. Control of compli- height are under the average of the reference popu- ance has been facilitated since 1998 by the SIAF-SAG lation, and (ii) the proportion of births attended by system for which more recently specialized subsys- medical staff or nurses (institutional delivery), and it tems for accounting (SICOIN), purchasing and con- allocates resources to programs aimed at achieving tracts (GUATECOMPRAS), and personnel adminis- those indicators. Although the pilot initiatives seem tration (GUATENOMINAS), among others, have been promising, the risk is that they may not be sustainable created. In parallel, a National Public Investment Sys- because the legal framework remains unchanged, in tem has been developed by SEGEPLAN. SIAF and its particular in relation to the Medium-Term Expenditure subsystems are operating in the central government Framework (MTEF). The General Budget Law does as well as all of the country’s municipalities. Thanks to not provide a reference framework for the medium these systems and the 2008 Access to Public Informa- term, and neither do its guidelines. Its Article 8 makes tion Law, access to budget execution details is largely a broad reference to the production and presentation available to the public.161 of a multiannual budget, but it’s not clear whether it Public Expenditure Review refers to two, three or more years. 5.16. Reallocations within the budget during execu- tion are highly discretionary and significant in scope. 5.13. To expand the use of the new methodology, While constrained by budget rigidities such as legisla- the MOF has encouraged sectors to use a de facto tive earmarks and demands by public sector unions, MTEF by proposing sectoral indicators associated the executive may regain control over budget execu- to budget preparation. However, there is internal tion through intra-institutional (via ministerial decree) 90 161. Barrientos (2011). and inter-institutional (via presidential decree) budget predictability in this situation also presents an obsta- transfers. In view of the country’s exposure to natural cle to effective results-oriented management. One of disasters, imperfections in budget planning and po- the reasons why the quota mechanism does not work tential fluctuations in its (limited) revenues requiring as expected is because of the lack of ex-ante controls, spending adjustments, this executive power is well resulting from the simultaneous (rather than chrono- warranted.162 However, budget transfers lend them- logical) entry of commitments, accruals and payments selves to political bargaining, and make it difficult to in the information system, which does not allow the follow through on the policy objectives supposedly treasury to predict cash needs in advance. represented by the budget. In fact, the average vari- ance in excess to the total deviation, which measures Accounting, recording and reporting the overall budget reallocation, was 11 percent in 2005- 2010. This means that the composition of actual expen- 5.18. The third element that affects budgeting and ditures compared with the original budged exceeded PFM systems performance is the accounting, re- at least twice the international standard.163 Thus, most cording and reporting of financial transactions. In sectors thus have little certainty over whether the particular, the Government faces challenges in pro- full budgeted resources will be available to fund their ducing its financial statements and reports according programs and projects. Reallocations among programs to international standards. Those challenges are main- within the same ministry do not require Cabinet ap- ly related to the fact that it is possible to execute part proval, but reallocations from one ministry to a depen- of the budget outside the SICOIN system. Neverthe- dent Secretary or to another ministry required Cabi- less, in 2010, the MOF presented for the first time the net approval. Once the approval has been obtained, General Government financial statements including the beneficiary ministry can start to use the additional the Balance Sheet Statement.164 This is an important resources while the MOF updates the information in step towards making the public accounts more trans- the SICOIN accordingly. Table 5.1 presents the actual parent and accountable. expenditures by ministry, compared to the original ap- proved budget for the period 2008-11. 5.19. On the accounting front, the government faces two important challenges. First, accounting 5.17. Although the MOF has implemented a quar- for land and buildings, where the book value has re- terly spending quota mechanism to support budget mained unchanged for more than 40 years and has execution, the system does not yet worked as ex- not been adjusted by either market prices or depre- pected. As the 2009 PEFA showed, the quarterly al- ciation.165 However, in many cases the MOF does not location system works well during normal times. How- have the information required and consequently can- ever, in times of a tightening fiscal situation, such as not properly register (in accounting terms) some fixed during the global economic crises of 2008-2009, the assets. quota system breaks down. Spending quotas were re- stricted to monthly limits, and any balances at the end 5.20. The second challenge is related to the ac- of the month had to be reimbursed to the treasury. In counting and registry of works and public infra- this situation, budget execution was affected because structure built through trust funds (fideicomisos). In sectors could not predict if enough liquid resources general, these works and infrastructure are property would be available to pay commitments and accruals. of the fideicomisos and are kept in their accounting The inability to pay accruals in some entities contrib- books. But before transferring those accounts to the Public Expenditure Review uted to the generation of arrears. The lack of resource General Government financial statements, it is neces- 162. As a caveat to this power, the executive is prohibited from undertaking transfers which shift investment to current expenditures. As a consequence, budget bills tend to overestimate current expenditures – for instance through a deliberate overestimation of debt service obligations – with a view to providing the executive with some leeway to augment current expenditures if need be (Barrientos 2011). 163. In that regard the PEFA methodology for PI-2 outlines that the country might get a D when the variance in expenditure composition exceeded overall deviation in primary expenditure by 10 percentage points in at least two out of the last three years. 164. Although there still were some accounts below the line equivalent to 10 percent out of the total balance in 2010, the financial statements were formally signed by MOF. 165. Since the accounting systems are not based on accrual basis the depreciation of fixed assets is not incorporated in the balance sheet. 91 Table 5.1: Deviations of actual from originally approved budget (2008-2010) 2008 2009 2010 2011 Public Institution Initially End of % Initially End of % Initially End of % Initially End of % approved year variation approved year variation approved year variation approved year variation budget budget budget budget budget budget budget budget Presidency of the 164 175 7% 191 213 11% 191 208 9% 185 231 25% Republic Ministry of Foreign 250 283 13% 220 304 38% 220 285 30% 278 302 8% Relations Ministry of 2,610 2,369 -9% 3,275 2,700 -18% 3,275 2,708 -17% 3,275 3,160 -4% Governance Ministry of National 1,265 1,261 0% 1,301 1,309 1% 1,301 1,278 -2% 1,555 1,543 -1% Defense Ministry of Public 227 235 3% 213 228 7% 213 246 15% 309 271 -12% Finance Ministry of Education 6,500 5,899 -9% 7,588 8,344 10% 7,588 9,350 23% 9,323 10,251 10% Ministry of Public 3,000 2,764 -8% 3,738 3,342 -11% 3,738 3,840 3% 3,930 4,454 13% Health Ministry of Labor 352 266 -24% 358 505 41% 358 632 77% 594 591 -1% Ministry of 230 352 53% 237 205 -13% 237 276 17% 275 257 -7% Economics Ministry of 1,231 1,185 -4% 828 664 -20% 828 978 18% 631 820 30% Agriculture Ministry of 3,000 4,447 48% 4,594 5,293 15% 4,903 6,733 37% 4,359 8,185 88% Communications, Infrastructure and Housing Ministry of Energy 46 48 4% 47 50 7% 47 119 154% 100 87 -13% and Mining Ministry of Culture 332 285 -14% 444 362 -18% 444 407 -8% 356 355 0% and Sports Secretaries and 2,020 2,553 26% 2,742 3,119 14% 2,742 3,127 14% 2,190 2,437 11% Other Executive Dependencies Ministry of the 41 52 25% 98 78 -21% 98 89 -9% 139 153 10% Environment and Natural Resources State Obligations 14,594 15,295 5% 16,986 16,689 -2% 16,986 16,744 -1% 18,083 18,195 1% in Charge of the Treasury Public Debt Service 6,626 6,425 -3% 6,812 6,577 -3% 6,812 7,212 6% 8,754 7,825 -11% General Attorney’s 47 43 -8% 53 53 0% 53 52 -1% 55 58 6% Public Expenditure Review Office Total 42,535 43,936 3% 49,723 50,032 1% 50,032 54,283 8% 54,391 59,175 9% Source: Barrientos (2011) and Ministry of Finance. 92 sary to carry out a reconciliation of accounts. In the go largely unpunished.167 Instead, the CGC is primar- past the MOF transferred resources from the budget ily engaged in process audits, in particular ex post to those fideicomisos. In accounting terms the trans- audits. While also having a corresponding mandate, action was registered as a reduction of cash accom- the CGC’s involvement in public spending quality and panied with an equal increase of accounts receivable. impact evaluations has been limited to-date. Similarly, The fideicomisos have little incentives to regularize ex- SEGEPLAN, while also legally tasked with the evalu- penditures (e.g., through invoices and receipts registry ation of public investments, has in practice assumed into the SICOIN) in a timely manner. Therefore, as the only a limited role. fiscal space allows and the works and infrastructure are in progress or concluded, the MOF activates a 5.23. The Constitution establishes that the CGC regularization account named “Construction in Prog- has to audit annually the General Government finan- ress� as part of the fixed assets to reconcile accounts. cial statements to inform Congress about the col- As “Construction in Progress� is increased, the “ac- lection of revenues as well as expenditures.168 The counts receivable� is decreased accordingly. However, CGC evaluates the budget execution and the way “Construction in Progress� represents 41 per cent out it was carried out by the corresponding entities (i.e. of the total General Government assets and 60 per the associated procurement processes). These audits cent out of total Central Government Assets, which are performed through the revision of the financial represents a huge portion of the public assets subject reports generated by the SICOIN system and docu- to regularization. While this account appears in the ments presented by the Accountant General Office to public financial statements, a complete reconciliation the CGC. In general, the auditors’ opinion has been process is still pending. qualified, which means that the Central Government entities followed the international accounting rules 5.21. Regarding the adoption of international ac- appropriately with some minor exceptions. Thus, the counting standards, the current information tech- financial statements are deemed to represent fairly nology (IT) systems have serious limitations that their financial situation. make it difficult to incorporate the standards. There- fore, the MOF is interested in exploring options for a 5.24. On the other hand, decentralized and auton- new IT system more oriented to support administra- omous entities are audited based on the financial tive as well as financial processes. In this context the statements and reports generated by the respec- MOF is also interested in options of integral packages tive PFM system (SIAF-Muni in the case of municipal that might be found in the market, including the pos- governments). In general, the auditor’s opinion has sibility of a Government Resource Planning (GRP). been mixed with clean opinion in some cases and with caveats in others. A clean opinion is equivalent to a External scrutiny and audit qualified opinion used in the case of Central Govern- ment entities. Figure 5.1 presents the type of the au- 5.22. Budget control and evaluation are both nor- ditor’s opinion for the decentralized and autonomous matively and substantively the least developed entities in 2006-2010. phases of the budget process in Guatemala.166 Each public entity has a unit of internal auditing with ex- 5.25. Beyond the auditor’s opinion, the audit comes ternal oversight provided by the General Auditor’s along with a set of sanctions that the auditor rec- Office (Contraloria General de Cuentas, CGC). The ommends for public officials who are accountable Public Expenditure Review 2003 Law of the CGC provides it with modern audit for the misuse of resources. However, the amounts procedures, but – due to legal ambiguities and limited of those resources are difficult to estimate. Further, resources – constrained abilities for sanctions. Viola- sanctions can be appealed to the General Auditor’s tions of legal norms related to the budget process Office for reductions of up to 90 percent, and most of 166. Barrientos (2011). 167. Next to the CGC, the legislature can request inquiries and, when offenses are detected, submit cases to the Attorney General’s Office. In practice, however, this has not been an important pathway for budget control. 168. The corresponding audits outcomes for 2006-2010 are posted at the following link http://www.contraloria.gob.gt/i_inf.php. 93 the reductions are granted. Therefore, the sanctions is the primary legislation that regulates fideicomisos.169 structure is relatively weak, which affects the credibil- A fideicomiso is constituted through a contract that ity of PFM systems. involves three parties: (i) the Constituent (fideicomi- tente) who transfers goods and rights to constitute the Key challenges in PFM practices fideicomiso; (ii) the Trustee (fiduciario) who receives and manages goods and rights until the fideicomiso’s 5.26. The key challenges in PFM are (i) the fact that objectives are achieved; and (iii) the Beneficiary (fidei- it is possible to execute the budget outside the SI- comisario) who will benefit from the activities financed COIN system; and (ii) the practice of generating ar- by the fideicomiso. The beneficiary might or might not rears or “floating debt�. As indicated, the FMIS is well be specified in the fideicomiso’s charter. equipped to carry out the overall budget execution and ensure automated controls to make the public 5.29. The application of legislation originally de- sector more accountable and transparent. However, signed for private sector fideicomisos creates a alternate vehicles can be used to execute the budget, number of accountability issues. While the use of thereby reducing transparency and accountability. trust funds per se is not inadvisable, the circumvention This part describes and comments on two types of of the Organic Budget and Contracting laws, among alternative vehicles – the fideicomisos and the agree- others, raise transparency concerns in Guatemala.170 ments with NGOs– as well as the practices that create While public fideicomisos are similar to private ones the problem of floating debt. in that the Government (as fideicomitente) transfers funds and rights to the fiduciario, there are a number Alternative vehicles to execute the budget of important differences. First, public expenditure is outside the PFM systems generally required to follow public financial manage- ment and procurement procedures; however, these Trust Funds (fideicomisos) are not applied to public fideicomisos because they are regulated by private sector law. Similarly, monitoring 5.27. Trust funds as a parallel budget execution and accountability mechanisms (including audits) that mechanism first emerged in response to the 1976 would generally apply to public spending are not ap- earthquake. In light of the emergency, a fund was set plied equally to public fideicomisos. As a result, there is up following private sector statutes, rather than by limited information on the execution of public spend- a new regulation. The fideicomiso, acting as a single ing through fideicomisos, compared to regular public coordinating mechanism across different public en- spending carried out through the SICOIN system. tities, enabled the Government to quickly disburse funds for reconstruction. The lack of a specific legal 5.30. Public fideicomisos have been an increasingly disposition implied that trust funds represented an popular modality for executing spending, especially agile execution mechanism in the context of oth- by Central Government entities. Today, there are erwise onerous regulations, in particular regarding more than 60 fideicomisos operating in a wide range procurement. Allowing for multi-year expenditure of sectors. Figure 5.2 shows that the budget resources programming, trust funds also offered advantages for managed through fideicomisos more than doubled be- budget predictability. tween 2004 and 2010. 70 percent of this execution may be attributed to the country’s road maintenance fund 5.28. Although legislation regulating fideicomisos (COVIAL) and housing fund (FOGUAVI) – both under Public Expenditure Review refers only to private trust funds, public fideicomi- the Ministry of Public Works – as well as the national sos operate under the same legislation. The Com- peace fund (FONAPAZ) and national development mercial Code of January 1971 (Legislative Decree 2-70) fund (FONADES). Another part of the increase stems 169. In addition, there are new guidelines applicable to public fideicomisos, such as the Operational Manual (April 2010), which outlines the PFM procedures that the fideicomisos must follow to regularize expenditures and formulate the budget. This manual was approved trough a ministerial accord. In addition, the Budget Law of 2011sets out rules to regulate the fideicomisos’ functioning and to make them more account- able. However, this Law is only mandatory within the same budget year, which creates uncertainties over their sustainability in future years. 94 170. These were addressed in part in 2008, with the creation of detailed reporting requirements for executed public expenditures. Figure 5.1: CGC audit opinion regarding budgeted revenues and expenditures of decentralized and autonomous entities (2006-10) 50 45 40 35 30 15 24 18 19 22 26 20 15 10 20 1 18 18 15 13 5 0 2006 2007 2008 2009 2010 With limitations and caveats 1/ With clean opinion 3/ With caveats 2/ With adverse opinion 4/ No opinion 5/ Notes: 1/ an opinion with limitations and caveats establishes that, except of the pointed limitations and caveats, the �nancial statements presents fairly the �nancial situation and operations outcomes of the entity in accordance to the International Accounting Standards. 2/ an opinion with caveats establishes that, except of the caveats, the �nancial statements presents fairly the �nancial situation and operations outcomes of the entity in accordance to the International Accounting Standards 3/ a clean opinion establishes that the �nancial statements presents fairly the �nancial situation and operations outcomes of the entity in accordance to the International Accounting Standards. 4/ an adverse opinion, establishes that the �nancial statements do not presents fairly the �nancial situation and operations outcomes of the entity in accordance to the International Accounting Standards. 5/ no opinion establishes that the auditor cannot express an opinion because the limitations found cannot enable the auditor to audit the �nancial statements in accordance to the International Accounting Standards. Source: CGC Figure 5.2: Central government fideicomisos: Figure 5.3: Autonomous and decentralized resources allocated by primary purpose fideicomisos: resources allocated (2004-10) by primary purpose (2004-10) 3,500 350 3,000 300 2,500 250 2,000 200 Q. million Q. million 1,500 150 1,000 100 500 50 0 0 2004 2005 2006 2007 2008 2009 2010 2004 2005 2006 2007 2008 2009 2010 Total managing funds Total investment Total credit guarantees Total managing funds Total investment Source: Staff estimations based on MOF information Source: Staff estimations based on MOF information from the introduction and rapid expansion of the con- or (iii) a combination of the two. Fideicomisos can be ditional cash transfer program Mi Familia Progresa. executed either by the Central Government or by au- tonomous and decentralized entities. In 2004-2010, 69 5.31. Fideicomisos fulfill a number of different func- percent of Central Government fideicomisos were for Public Expenditure Review tions. In line with the different nature and objectives investment, 26 percent managed funds, and 5 percent of the fideicomisos, the MOF classifies them into three provided credit guarantees. Of the Autonomous and categories: (i) those that manage funds that are distrib- Decentralized fideicomisos, 61 percent managed funds uted in the form of targeted or general subsidies; (ii) and 39 percent were for investment. those that invest in infrastructure; and (iii) those that provide credits and guarantees. The financing provid- 5.32. Public fideicomisos have played a particular ed through fideicomisos can be (i) reimbursable (main- important role in the execution of capital expendi- ly credits and credit guarantees); (ii) non-reimbursable; ture. Since 2009, the MOF has registered and report- 95 ed the budget execution of all fideicomisos that have ber 1998 to finance the construction and maintenance received funds from the budget, in accordance with of the country’s highway network. The Peace Fund the transparency requirements laid out in the annual was established in 1991 to promote and consolidate budget laws. Based on this information, public fidei- the peace process after the end of the civil conflict. comisos have executed on average close to 18 percent Resources dedicated to this type of fideicomiso in- of the total capital budget and 7 percent of total cur- creased by 128 percent between 2004 and 2011, and rent expenditures in 2009-2010. represent 59 percent of total fideicomiso resources. 5.33. The fideicomisos classified as managing funds 5.35. The credit and credit guarantee fideicomisos mostly channel subsidies to beneficiaries. The four are aimed at improving access to credit and at di- fideicomisos in this category represented 94 percent recting credit to some sectors and small enterpris- of total non-reimbursable expenditures managed and es. Financing provided by these fideicomisos is largely executed by the fideicomisos in 2004-2011.171 Out of reimbursable. In addition to supporting the banking those four, the fideicomiso Mi Familia Progresa, which sector’s solvency, this kind of fideicomisos assists mi- finances the government’s conditional cash transfer cro and small enterprises as well as small agricultural program, represented more than 50 percent of these producers (including coffee). Resources dedicated to expenditures, followed by the fideicomisos manag- this type of fideicomiso have been decreasing, and ing subsidies for housing, agriculture and indigenous represent 1 percent of total fideicomiso resources groups. Mi Familia Progresa was formally created in (Table 5.2). April 2009 to assist extremely poor households with small children while requiring them to comply with 5.36. Arrangements with non-governmental orga- co-responsibilities with regards to health care, nutri- nizations are another commonly used vehicle to ex- tion and education. Since then, this fideicomiso has ecute public resources. Public contracting of NGOs received cumulative funds from the budget of around first originated in the health sector. With little public US$400 million. The fideicomiso for mortgage sub- capacity to provide health services in culturally and sides was created in 1996 to support a reduction in linguistically diverse rural areas, a large share of the the housing deficit, especially for the poorest people. interventions of the Integrated System for Health At- The fideicomiso for modernizing the agriculture sec- tention is carried out by NGOs. More recently, how- tor was created in 1994 to provide technical assis- ever, contracting of NGOs, often through political tance and funds to farmers in order to enhance the links, for public works has gained in prominence. In sector’s productivity. Finally, the fideicomiso for indig- part due to this practice, NGOs have come to claim enous people was created in July 1994 to support and an increasing share of the public budget, represent- strengthen the human development of the Mayan in- ing between 1 percent and 23 percent of sector min- digenous people. istries’ budget execution in 2009 and 2010. This is equivalent to 4 and 3 percent of the total budget, 5.34. The Public Investment fideicomisos are aimed respectively (Table 5.3). at complementing the public investment program by managing the project cycle of infrastructure proj- 5.37. Like fideicomisos, NGOs are not subject to ects. Six public investment fideicomisos represented regular budgetary oversight and controls. To ad- 99 percent of total non-reimbursable expenditures dress transparency concerns about the execution of of investment fideicomisos between 2004 and 2011. funds through NGOs, MOF introduced in 2008 norms Public Expenditure Review Within those six, the most relevant are the Roads which condition advances to NGOs on a prior detailed Fund (COVIAL) and the Peace Fund. The COVIAL accounting of executed expenditures. In 2010, slightly represented 60 percent of all expenditures managed less than 30 percent of the amount transferred to by non-reimbursable public investment fideicomisos. NGOs was accounted for through this practice; the It was created through a Government Accord in Octo- remainder was channeled to NGOs through congres- 171. The non reimbursable investments and expenditures, except the payments to the fiduciario, are expenditures to finance projects and programs such as roads, bridges, schools, and subsidies and others transfers to beneficiaries. The information is extracted from the fidei- 96 comiso’s financial statements by the MOF. Table 5.2: Expenditures of Central Government fideicomisos (2004-2011) (Q. millions) Tipo de fondo 2004 2005 2006 2007 2008 2009 2010 2011 Total % of total Administrative funds 119 271 420 487 137 978 1,550 1,144 5,105 19% Mi Familia Progresa 0 0 0 0 0 786 1,108 886 2,780 10% Fondo Guatemalteco de Vivienda 83 237 370 440 120 160 417 227 2,054 8% Investment funds 881 1,561 1,685 1,249 2,762 2,895 3,138 2,006 16,176 59% Fondo Vial (COVIAL) 640 689 873 879 1,924 1,646 1,678 1,416 9,744 36% FONAPAZ 240 323 569 350 723 847 1,229 409 4,691 17% Credit guarantees 61 23 6 6 108 0 0 0 204 1% Other trust funds 941 1,005 1,134 1,010 724 227 391 278 5,710 21% Total 2,002 2,860 3,243 2,753 3,731 4,100 5,079 3,428 27,194 100% Source: Staff calculations based on MOF data. Table 5.3: Expenditures executed by NGOs and other agreements as percentage of the sector’s total executed budget 2009 2010 Total NGOs/Total Ministry of Health 14% 12% Total NGOs/Total Ministry of Education 9% 4% Total NGOs/Total Infrastructure 2% 5% Total NGOs/Treasury Obligations 2% 1% Total NGOs/Total Ministry of Culture and Sports 23% 18% Total NGOs/Total Budget 4% 3% Source: Staff calculations based on MOF data. sionally-stipulated “direct contributions� to which Quality of expenditures carried out through the MOF norms may not be applied – a practice to alternate vehicles which legislators started resorting in response to the new norms.172 While the provision of health services 5.38. Although the use of alternate vehicles can through NGOs is not put into question, a call for the help deliver services, goods and works in a timely prohibition of public works execution through NGOs manner, there is an increased accountability risk. has been made by various initiatives.173 Incentive-align- This is because the processes related to the allocation Public Expenditure Review ment for such reforms, however, is challenging: the au- and execution of these resources are different from thority to prohibit congressionally-stipulated transfers those that are applied by public entities using the to NGOs lies with congressional members.174 SICOIN. Over time, the widespread use of these ve- 172. Barrientos (2011). 173. See CoST Guatemala (2011). 174. In this regard, the 2012 budget contains an article prohibiting the contracting of NGOs for public works. As it is unclear whether its for- mulation is sufficiently unambiguous to close legal loopholes, however, it remains to be seen whether it is effective in thwarting the continued contracting of NGOs (Plaza Pública 2011b). 97 hicles went beyond the specific objectives for which tensions of contracts (without the guaranteed avail- they were originally designed, such as multi-sectoral ability of resources). reconstruction programs following natural disasters, or the provision of public services in remote and post- 5.40. In the case of NGOs, the main issues affect- conflict areas. Once the budgetary transfers have ing the quality of expenditures are the following: (i) taken place, the fideicomisos and NGOs operate with NGO employees are not public employees and are significant autonomy but without adequate controls. therefore not covered by the sanctions regime under From the budget execution perspective, there is no the public procurement law, which is tougher than the registration of individual transactions to feed into the CGC sanctions; (ii) partial registry in the SICOIN sys- accounting system (SICOIN), the management system tem of the agreements with NGOs; (iii) lack of infor- (SIGES) or other information systems that would fa- mation of idle cash balances at bank accounts, which cilitate the CGC’s audit work. For instance, for the SI- are not transferred to the STA at the end of the year; COIN, the quality of accounting information is poorer (iv) partial registry of projects in the SNIP registry; (v) because expenditures have to be regularized manu- lack of monitoring by the SIGES management system; ally. Also, cash resources managed by these vehicles (vi) lack of transparency of fideicomisos to sub-con- are not within the Single Treasury Account (STA). Is tract NGOs to carry out projects and programs; and the use of these vehicles the best option to execute (vii) multi-year agreements inconsistent with annual the budget? Are the investment projects and services budget planning. provided fully consistent with the national priorities and public policies? These questions cannot be an- Accountability mechanisms of the alternative swered adequately because of the lack of transpar- vehicles ency and accountability. 5.41. According to public regulations, the CGC has 5.39. No specific legislation to regulate the opera- the responsibility to audit the above-mentioned al- tions of fideicomisos is in place; as a result, they have ternative vehicles. The CGC’s specialized units have managed significant amounts of resources without audited the fideicomisos and NGOs on an annual appropriate controls. This leads to lack of transpar- basis since 2008. In general, the audit objective is to ency in terms of procurement, quality of works, and establish whether they have used public funds in ac- impact assessment. In addition, activities financed by cordance with international accounting standards. fideicomisos may not be consistent with long-term development objectives because they take place out- 5.42. For most fideicomisos, the auditor’s opinion side the Public Investment Program (PIP). Since 2010, has been unqualified.176 However, it seems that there however, these activities have required the approval is little consistency in the auditor’s opinion because of the National Public Investment System (Sistema it differs substantially from one year to another. For Nacional de Inversion Pública, SNIP) and need to be instance, the COVIAL got an unqualified opinion in included in the budget. On the other hand, the Min- 2008 but in the next year it got a disclaimer opinion istry of Finance has recently been removed from the (i.e. no opinion at all). The following Table 5.4 presents COVIAL Technical Committee, which decides on proj- the CGC audits for a sample of fideicomisos for 2007- ects and the allocation of funds, undermining exter- 2010. nal control over its operations. Emergency situations are often used as justification for the extended use of 5.43. In particular, the CGC has audited COVIAL Public Expenditure Review emergency procedures, including the use of stream- and Mi Familia Progresa annually since 2008. In the lined procurement processes and extraordinary ex- case of the COVIAL audits of 2008-2010 resulted in a 175. The PIP is the list of public investments that were assessed by sectors following the SNIP and are ready to be incorporated to the budget. The SNIP comprises systems and regulation that enable sectors to complete the project cycle since the project profile up the completion assessment. The SNIP provide a unique number to identify projects that will be incorporated to the budget once the technical specifications have been fulfilled such as the estimation of yields and the present net value. The SNIP regulations should ensure that the proposed projects are aligned to national priorities and strategic goals. 98 176. This means that the financial statements of the entity adequately reflect the financial situation in accordance to international standards. Table 5.4: Fideicomiso audits performed by the CGC (2007-2010) Auditor's Opinion 2007 2008 2009 2010 No. % No. % No. % No. % Unqualified 1/ 16 73% 21 78% 22 71% 10 56% Qualified 2/ 5 23% 2 7% 4 13% 0 0% Disclaimer of opinion 3/ 1 5% 4 15% 5 16% 8 44% Total 22 100% 27 100% 31 100% 18 100% Total Fideicomisos 57 82 100 80 % Audited 39% 33% 31% 23% 1/ The entity followed all accounting rules appropriately and that the financial reports are an accurate representation of the company’s financial condition. 2/ The entity followed all accounting rules appropriately with some minor exceptions and that the financial reports are an accurate representation of the company’s financial condition. 3/ An auditor’s statement disclaiming any opinion regarding the company’s financial condition due to an inability to gather certain relevant facts. Source: Staff calculations based on MOF data. disclaimer opinion because the accrual accounts, pay- ments with NGOs is not clear: estimates suggest that ments and accounting did not match with the physical around 5,000 agreements are currently under opera- progress of works. The fideicomiso for the MiFaPro tion. The CGC has only focused on 300, of which 45 Program could not reconcile its accounts with infor- are regularly audited every year. Second, there is not mation on the transfers made to final beneficiaries, an official NGO registry and thus the number, fea- resulting in a disclaimer of opinion by the CGC. tures, location, and intrinsic risks are not evaluated by the CGC to plan audits. Third, complaints have been 5.44. The CGC carried out 36 special audits trig- the main source for choosing particular agreements to gered by particular outcomes found in the indicated audit rather than an evaluation based on risk criteria annual audits. For instance, in 2010, the CGC audited and cost analysis. Fourth, NGOs often argue that their the list of beneficiaries of MiFaPro. The audit verified employees are not civil servants and are therefore not not only the existence of the beneficiaries but also bound by the public sector legal framework, including whether they matched with the beneficiaries’ profile. the public procurement law. Because of that, the CGC The outcomes of this special audit were not conclusive cannot apply the sanctions regime established by that because the sample was small. In the case of COVIAL, law and instead applies the CGC law, which contains the special audits found that it has been creating float- less strict sanctions. ing debt (a commitment without guaranteed availabil- ity of resources) since 2002. Although the CGC estab- Payment Arrears (“Floating Debt�) lishes sanctions for persons who manage fideicomisos incorrectly, in general those sanctions are weak, since 5.46. The accumulation of payment arrears (“float- the CGC law caps fines at Q.80,000 (US$10,000 ap- ing debt�) is a long-standing problem.177 Floating proximately), and those fines are usually appealed and debt emerges when commitments or accruals are reduced by up to 90 percent. Therefore, in practice, generated without the corresponding budget credit.178 Public Expenditure Review no effective sanctions occur. However, this definition presents some limitations in the case of multi-annual projects, where the imple- 5.45. In the case of the NGOs, the CGC confronts mentation timeframe in general exceeds more than a number of limitations. First, the number of agree- one budget year. The IMF defines payment arrears as 177. The Government has contracted an international audit firm to conduct a comprehensive audit of floating debt; however, the audit could only be completed for a portion of Central Government ministries. 178. This definition is consistent with the General Budget Law (Article 26), which establishes that no commitment and accrual can be gener- ated without the corresponding budget credit. 99 the due payment obligations that the Government did ated without budget credit. In addition, for 2011, the not pay in a reasonable period of time (say 60 days) MOF has requested that SNIP certify projects before after the invoice was accepted and recorded.179 While incorporating them into de budget. The BAC and the a complete inventory of floating debt is not available, SNIP requirements are sound measures to discourage the problem is particularly acute in entities managing the creation of floating debt. However, it is too early multi-year infrastructure projects, including COVIAL. to assess whether contracts are still signed without a BAC and whether these get paid. In any case, it seems 5.47. The situation in Guatemala is complicated be- necessary to complement such administrative mea- cause parts of the budget can be executed outside sures with effective sanctions and penalties in order of the PFM systems, and because the MTEF and STA to significantly change behavior. are not fully consolidated. Thus, control over the use of resources is limited. During budget execution C. PUBLIC PROCUREMENT sectors register commitments and accruals when the payment order is processed. However, if there are not 5.50. Whilst in recent years there have been im- enough available funds, sectors may not register the provements in the Guatemalan public procurement invoice in the SICOIN system, resulting in an arrear system, there are nevertheless problems related to not registered in the SICOIN, which becomes floating efficiency, risks of corruption and low impact on na- debt. COVIAL generates floating debt in a different tional development. One of the major problems that way, by the repeated extension of contracts associ- has been identified in various analytical studies180 in is ated with infrastructure projects without the medium- the lack of coordination and fragmentation. This oc- term budget credits. The system does not alert public curs in spite of the fact that a single procurement law officials of the creation of an illegal accrual, and the exists; however, each entity applies the rules different- credibility of the PFM system is seriously undermined. ly, contracts are not standard, and a high level of sub- jectivity was applied. This fragmentation of regulation, 5.48. Despite the uncertain legal basis of the claims subjectivity and non-standard practices carried out by that are generated, they represent an important li- the different government entities results in a high level ability for the government and contribute to the low of inefficiency in budget execution and coordination, levels of trust in government. According to an ongo- limits participation from suppliers, makes oversight by ing CGC audit, the MOF has authorized payments to civil society and revelatory entities extremely difficult past invoices, even when they were not registered in and increases the risks of corruption.181 the SICOIN, in order to gradually eliminate these li- abilities. This has further contributed to the lack of 5.51. Perhaps one of the most telling results of this credibility of the government’s efforts to reduce float- fragmentation is that in 2009, 84 percent of the val- ing debt. ue of public procurement activities was contracted through exception mechanisms for which there is no 5.49. MOF has recently taken measures to deter the uniform approach to their application. It is interesting creation of floating debt, but with limited success. to note that the public procurement law in Guatemala The introduction of the Budget Availability Certificate does not have competition as an explicit objective, (BAC) in 2010 was designed to control the generation and exception mechanisms are by definition not com- of floating debt. The guidelines state that commit- petitive. This omission in the law is further exacerbat- ments (i.e. contracts) without BACs will not be com- ed by the lack of guidelines and control mechanisms Public Expenditure Review pulsory. Therefore, a contractor should request the to avoid discretion in procurement activities through corresponding BAC before signing a contract in order requirements for participation and evaluation criteria. to ensure that it will be paid. With the BAC, the MOF There are also several weaknesses in the capacity of intends not to recognize any arrears that were gener- the procurement directorate to ensure that there is ac- 179. IMF (2010). 180. These include a country procurement assessment report carried out by the World Bank in 2005 and the OECD procurement indicators applied by the government in 2010 with the support of the World Bank and the Inter-American Development Bank. 100 181. The figures in this section and in the annex describe the procurement and contracting practices as of August 31, 2011. tual coordination between procurement and budget, the technical specifications, lack of available budget contract management, decision-making and associ- as well as appeals by the potential suppliers. ated accountability, further development of the e-pro- curement platform, professionalization of public offi- 5.55. One of the most significant efforts by the cials, as well as institutional and civil society oversight. government to address some of these weaknesses has been the development of the e-procurement Based on the OECD procurement indicators, the fol- platform Guatecompras. Guatemala was one of the lowing issues were identified: first countries in Central America to introduce an e- procurement platform, Guatecompras, and is one of Legal, regulatory and institutional framework the few countries that has linked the platform with its financial management system. The system requires 5.52. There are several weaknesses in the legal and that all steps in the procurement process be followed regulatory framework given that the regulations in- before any payments can be undertaken. Guatemala clude provisions that were not covered by the law. has also developed a contract management system At the same time, government entities established that captures the planned, awarded in final amounts their own processes and requirements in their internal of each individual contract.182 regulations, which in turn are contemplated neither by the law nor the general regulations. These practices Market extend to the selection process as well as the contract management phase. The lack of guidelines determin- 5.56. Like other countries in the region, Guatemala ing the scope and minimum content of bidding docu- lacks a professional stream of public procurement ments, as well as of standard contract clauses, results officials. The lack of standard profiles for procure- in many cases in sufficient information, which increas- ment officials results in a great disparity in the capac- es the the risks of ambiguity, error and corruption. ity from agency to agency; this compounded by the high level rotation of personnel significantly increases 5.53. There is a significant lack of coordination be- the risks associated with procurement. Public sector tween public procurement activities and budget regulations in Guatemala do not identify specific re- planning and execution. Whilst the government has sponsibilities associated with the procurement pro- undertaken commendable efforts to strengthen this cess; they also lack sanctions and any management link by requiring that entities obtain a certificate of standards. The delegation of responsibility is agency availability of funds prior to undertaking any procure- specific, but accountability remains centralized, thus ment activities, many entities fail to comply with this weakening the process of managing by results. requirement. By signing contracts without the prop- er prior budget authorization these entities assume 5.57. The OECD indicators also identified signifi- debts for which the state is liable. This practice is par- cant barriers for private sector participation in the ticularly prevalent in MICIVI where contract commit- public procurement system. These include, amongst ments far exceed the available budget. This not only others, the lack of trust in the transparency of the has a negative impact on budget management, but procurement processes, the impartiality of awarded also delays payments to suppliers. contracts, the lack of a government policy to increase competition, the fragmentation of the public procure- 5.54. One of the most challenging problems that af- ment system, the cost and restrictions of obtaining Public Expenditure Review fect procurement planning is the length of time of credit, the lack of certainty on timeframes for pay- procurement processes. The 2010 OECD indicators ments, the fear of changes in contract conditions and identified that on average procurement processes timeframes and the weakness of government contract ranged between 150 to 250 days due to problems of supervision. 182. These systems have generated a wealth of information on the behavior of the public procurement system and in particular the lack of compliance by government entities with government reporting requirements. However, the procurement directorate does not have a team dedicated to analyzing this information. 101 5.58. In 2009, there was a significant increase in the of the issues resulting from complaints and irregulari- procurement of goods and services. The central gov- ties in the procurement system. There is a very low ernment expenditure increased by around 6 percent level of convictions in relation to the number of pro- between 2004 and 2009. In 2009 the procurement cesses opened. of common use goods represented around 27.8 per- cent of the total procurement of goods and services 5.61. The role of civil society in overseeing public compared with 15.1 percent in 2004. In 2009, procure- procurement in Guatemala is limited, and there ment expenditure on goods, works and services was are few organizations dedicated to monitoring estimated in Q.17 billion, with the central government procurement and a national level. Nevertheless, procuring directly around 65 percent, and 35 percent there is a growing interest at the local level in social through transfers to decentralized institutions. In audits related to procurement activities. The vari- 2010, the figure was Q.12 billion, and it reached Q.11 ous professional associations do not have strate- billion by August 31, 2011. gies, nor do they regulate the ethical behavior of their members, as a rule they also don’t participate Integrity and transparency in anticorruption programs targeted at public pro- curement activities. The media is the only entity 5.59. As mentioned previously, the development of that plays a sustained and constant role in monitor- Guatecompras has significantly increased the trans- ing public procurement activities. This results in an parency of public procurement in Guatemala. Nev- overall weak procurement system as demonstrated ertheless, practices employed by agencies to either in Figure 5.5. bypass or avoid the existing system controls signifi- cantly weaken these efforts. Equally, the failure by the 5.62. These institutional weaknesses also apply to procurement directorate to utilize the existing data other elements of the public administration linked generated by the various governance systems further with the public procurement system (Figure 5.6). It undermines the transparency of the system. is recognized that in order for a country to have a well-functioning public procurement system, all the 5.60. The Guatemalan public procurement system interrelated components of the public administra- does not have an administrative body to address tion must also perform efficiently and effectively. It complaints and or investigate infringements in the would thus stand to reason that any reforms to the public procurement process. This function has to be public procurement system must also be accompa- carried out through the court system, and as such is nied by reforms and strengthening of other related subject the limitations of the system to address many government functions. Figure 5.4: Performance of the Guatemalan Figure 5.5: Performance of the Guatemalan public procurement system public procurement system 1. Legislative Framework 1. Legislative 12. Anticorruption and ethics Framework policies and implementation 2. Process 0.6 11. Access to 3. Linkage between information 0.9 procurement Public Expenditure Review 0.4 0.3 and budgeting 0.5 0 0.5 4. Integrity and 0.6 0.3 2. Institutional 10. Ef�ciency of 0.6 4. Management transparency and Management protest mechanism 0.3 0.8 0.3 0.5 Capacity 0.7 0.5 9. Ef�ciency of audit and control 5. Institutional mechanisms capacity 8. Contract Management 6. Procurement ef�ciency 7. Operation of the public procurement market 3. Procurement Operations and Markets Country Score Ideal Situation 102 Source: OECD (2010). Maximum Score Source: OECD (2010). Current Situation Table 5.5: Procurement by method (% of total) Year Direct Direct (resulting Quotation Tendering Open Exception Not subject to Unclassified from NV) contract Procurement Law 2009 1.89 0.27 11.86 8.74 6.67 25.31 8.57 36.69 2010 5.76 0.46 17.01 20.04 0.75 40.69 15.29 0.00 2011 4.80 0.35 11.98 13.33 0.04 48.96 20.54 0.00 Source: Guatecompras. Performance mission to review and explore what were the princi- pal opportunities and challenges of the current pro- 5.63. As noted previously, there is a high level of curement mechanism of open contracts utilized by fragmentation and use of non-competitive procure- the government to procure pharmaceuticals. ment practices. Both have increased over the last few years, and rather than moving towards international 5.66. In principle, the government’s use of open best practices of consolidation and increased compe- contracts is similar to the use of framework agree- tition, Guatemala has moved in the opposite direction. ments by other jurisdictions that use such mecha- nisms to procure goods, works, services and phar- 5.64. There has been a growing tendency to either maceuticals. Nevertheless, the use of open contracts procure by exception or limited competitive meth- in Guatemala is hampered by a regulatory framework od rather than through open tendering processes. that limits the potential of the tool by prohibiting the The use of exceptions is particularly prevalent in infra- use of many of its most beneficial characteristics. structure, whilst direct procurement tends to be as- The use of framework agreements to procure phar- sociated with goods and services. Likewise, there is a maceuticals generally is undertaken through the con- significant reduction in the use of open contracts; this solidation of demand, skill sets, and standards. That reduction is accompanied by an increase in requested is to say, a centralized unit highly specialized in the exceptions for the procurement of items outside the development and management of framework agree- open contract. In 2009, when the use of the open con- ments brings together all of the key stakeholders to tracts was at its peak and accounted for 6.6 percent of determine the specifications and protocols, before government procurement, there were 240 exceptions. going out to the market. In this context, framework In 2010, the value of the open contract was reduced agreements can be a very efficient and effective tool. to 0.75 percent of government procurement but there However, their use requires a very solid and techni- were 618 exceptions. As of August 31, 2011, the value of cally capable platform. Whilst framework agreements the open contract as a percentage of total spend had would not address all the problems and challenges been reduced to 0.04 percent and there were already encountered in the procurement of pharmaceuticals 631 exceptions. This trend indicates that not only is the by the public sector, they could nevertheless provide government missing out in terms of savings, efficiency many relevant opportunities to deliver more efficient and governance through the advantages of aggregat- and effective outcomes. Public Expenditure Review ing procurement, but also that the subsequent in- crease in direct procurement is reducing competition 5.67. The use of open contracts was first introduced and transparency. in Guatemala over 10 years ago. Prior to their intro- duction for the procurement of pharmaceuticals, the Case study: procurement of pharmaceuticals public sector used traditional approaches such as ten- dering, quotations, but also with increasing frequency 5.65. In 2010, at the request of the Ministry of Fi- direct procurement. As a result, there was a high level nance, a World Bank team carried out a technical of variation in prices of pharmaceuticals that led to a 103 necessity to identify mechanisms that would increase These and other deficiencies have resulted in a lack the efficiency of the procurement. Prior to 2009, there of clarity on the use of open contracts, their rules and were three open contracts with annual renewals of up exceptions, which, in turn, reduces competition trans- to 5 years. parency and value for money for the government. 5.68. In 2009, the Guatemalan government car- Case study: infrastructure ried out a new selection process for an open con- tract to procure approximately 861 pharmaceuti- 5.70. In general, the procurement of infrastruc- cals. Nevertheless, the government only awarded ture tends to be outside the scope of governmen- approximately 50 percent of the required pharma- tal procurement reviews. This often occurs due to ceuticals, due to a variety of challenges with the the high level of transactions associated with goods other 50 percent. The challenges ranged from pro- and services: procurement of these activities gen- tests and appeals from suppliers as well as disagree- erally accounts for 80 percent of all procurement ments between the evaluation committees and the transactions. Nevertheless, infrastructure (construc- Ministry of Finance. It is also important to note that tion, maintenance, and associated services) gener- the government had not defined the list of essen- ally accounts for 80 percent of value. Procurement tial medicines for each level of attention, neither for of infrastructure also tends to be the domain of a the Institute of Social Security nor for its own hospi- few specialized government entities, as opposed to tals, leading to a higher probability of duplication in goods and services which are common throughout all the requirements but nevertheless having different ministries. technical specifications. 5.71. There are 12 government agencies involved 5.69. The review identified several deficiencies and in the procurement of infrastructure in Guate- opportunities to improve the procurement of phar- mala, but the Ministry of Communications, Infra- maceuticals in Guatemala, such as: structure and Housing (MICIVI) accounts for 80 percent of total government expenditure in the • Limited coordination between institutions as well sector. As a result of this, this section provides only as lack of strategic information required to plan a short overview of general infrastructure procure- demand and supply chain logistics. ment, before focusing more specifically on procure- • Limited capacity at entity level and executing unity ment in MICIVI. level to understand the specific use of particular medicines. 5.72. Non-competitive procurement methods ac- • Lack of standard protocols for diagnosis and treat- count for a large share of procurement, especially ment, which in turn impacts on specifications and of infrastructure. Three main procurement methods inventory control. can be identified as the most commonly used across • Lack of inventory control, leading to both oversup- government agencies: quotations, public bidding and ply and undersupply. by exception. Figure 5.7 shows that there is a wide • Lack of clearer guidelines for the planning, prepa- variance in the use of the identified methods by the ration, awarded and management of open con- different ministries. This ranges from the exclusive tracts. procurement by exception by the Ministry of For- • Lack of operational guidelines to capture and sys- eign Affairs to the exclusive use of quotation by the Public Expenditure Review tematise lessons learned so that they can be fed Ministry of Economy. On the other hand, the Min- into subsequent procurement processes. istry of Education tends to divide its procurement • Lack of guidelines for the extension of existing between quotation and exception, with only a small open contracts, for example between 2005 and percentage being undertaken through public bidding 2010, the extensions were carried out in an ad hoc processes. MICIVI, on the other hand, contracts ap- manner without predefined mechanisms. Equally, proximately two-thirds of its infrastructure through there is no mechanism to update prices falling ex- public bidding and the remainder through exceptions 104 tensions of contracts. and quotations. This dataset indicates that, whilst some entities utilize public bidding to procure their 5.74. A key determinant for the level of interest required infrastructure a very large amount of the by potential bidders to participate in future pro- overall spending is nevertheless carried out through curement processes, as well as for their pricing non-competitive methods. structure, is the expectation to be paid on time. There is a wide range in the payment performance 5.73. An analysis of the methods used by MICIVI to of various ministries, with the majority having large procure works, maintenance and supervision, shows contract amounts outstanding. It would therefore the following: be unrealistic to expect that the delays in payment resulting from these outstanding commitments do • From a total of Q.34 billion in contracts awarded not have a significant impact on the number of par- by tender, ticipants and the prices offered to the government. o Works represent 57.3 percent; This situation may actually be contributing to the o Maintenance represents 3.2 percent; and high level of exceptions and contract extensions, o Supervision represents 39.5 percent. since firms who are currently engaged in govern- ment projects may be willing to undertake further • From a total of Q.599 million in contracts awarded work in the hope of recouping payments for previ- by quotation, ously performed work. o Works represent 93.7 percent; o Maintenance represents 0.2 percent; and 5.75. Not surprisingly, the largest outstanding o Supervision represents 6.1 percent. amounts to be paid are related to road construc- tion and maintenance. This applies both in terms • From a total of Q.8 billion in contracts awarded by of the number of contracts and the outstanding exception, amount. o Works represent 90.5 percent; o Maintenance represents 0.9 percent; and 5.76. The large amount of outstanding commit- o Supervision represents 8.6 percent. ments indicates an apparent disconnect between budget planning and budget execution. Figure 5.8 Tendering is thus used both for works and supervision, compares MICIVI’s budget from 2008 through 2011 but quotations and exceptions are almost exclusively in relation to the overall government budget and out- used to procure works. standing commitments. In 2008, MICIVI’s outstanding Figure 5.6: Institutional procurement practices – amounts contracted by method (2000-Aug 2011) Departments and other Agencies of the Executive (895 Contracts for Q. 2 bn) 12% 87% 1% Ministry of Labor and Social Assistance (12 Contracts for Q. 12 m) 20% 78% 2% Ministry of Public Health and Social Care (86 Contracts for Q. 265 m) 8% 90% 2% Ministry of Foreign Affairs (1 Contract for Q. 0.7 m) 0% 100% Ministry of the Interior (72 Contracts for Q. 129 m) 25% 73% 3% Ministry of Energy and Mines (3 Contracts for Q. 200 m) 0% 100% Public Expenditure Review Ministry of Education (16 Contracts for Q. 6 m) 33% 20% 47% Ministry of Economy (4 Contracts for Q. 1 m) 100% 0% Ministry of Culture and Sports (344 Contracts for Q. 349 m) 30% 64% 7% Ministry of Environment and Natural Resources (5 Contracts for Q. 57 m) 1% 99% 0% Ministry of Agriculture, Livestock and Food (13 Contracts for Q. 18 m) 30% 70% 0% Ministry of Communications, Infrastructure and Housing (7754 Contracts for Q. 43 bn) 3% 66% 31% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Source: Guatecompras. By quotation By public bidding By exception (Art. 44 and others) 105 Figure 5.7: Budgetary performance of MICIVI procurement (Q. million) 37,700 2011 58,374 7,475 34,548 2010 54,283 6,733 30,696 2009 50,032 5,293 13,202 2008 43,935 4,447 0 10,000 20,000 30,000 40,000 50,000 60,000 Current MICIVI Budget Current Government Budget To be paid for active contracts *2011 reflects data through August 31, 2011. Source: Guatecompras. Figure 5.8: MICIVI contracting practices - contracts amendments 125% 2011 1% 100% 2010 114% 63% 100% 2009 132% 95% 100% 2008 137% 93% Amount initially contracted 100% 2007 193% 103% Amount awarded 100% 195% 2006 103% Final/current amount of contract 100% 2005 299% 102% 100% 131% 2004 100% 100% 2003 258% 100% 100% 236% 2002 100% 100% 332% 2001 57% 100% 518% 2000 90% 100% 0 1 2 3 4 5 6 *Initial contract amount = 100%. 2011 reflects data through August 31, 2011. Source: Guatecompras. commitments were three times its annual operational pact that these shortfalls can have. Figure 5.9 identi- budget, and 30% percent of the government’s annual fies the amounts of contract variations by dividing the Public Expenditure Review operating budget. In 2011 that figure increased to 5 contracts into six thresholds. The top three thresholds times its annual operational budget and now repre- indicate the largest variations ranging from 199 to 358 sents 54 percent of the government’s annual budget. percent. Whilst it would be expected that there would be some variation between the amount contracted 5.77. Infrastructure projects are often associated initially and the final payment, the magnitude of the with cost overruns resulting from under-budgeting, observed variations indicate significant deficiencies, poor planning and inefficient contract management. ranging from planning and execution to contract man- 106 As a rule, the larger the project, the greater the im- agement. D. CONCLUSIONS AND RECOMMENDATIONS 5.80. Preliminary results of an audit by CGC show that the floating debt might be equivalent to 3 per- Budget Process and Public Financial cent of GDP (US$1.2 billion). In 2010, CGC identified Management floating debt created by the COVIAL fideicomiso, by contract extensions without budget credit authoriza- 5.78. Based on the 2009 PEFA, the Government tion by its Technical Committee. COVIAL argued prioritized several key issues to be addressed in or- that it was done to address emergency situations der to improve PFM, as follows: caused by natural disasters, and that they were regu- lated by the Commercial Code. In emergency situa- • Floating debt - because it was not registered and tions, two types of exceptions are usually granted: (i) it represents a potential burden for the public fi- the release of funds without following PFM regula- nances due its magnitude; tions and in particular the procurement law; and (ii) • Fideicomisos and other alternate vehicles - they the transfer of funds for execution by a fideicomiso. register only transfers, without detailed transac- Both of these exemptions, however, could under- tions in the SICOIN system. mine transparency and increase the risk of improper • Budget Formulation - there was no link between use of public resources. planning and budget, and there was no results-ori- ented budgeting; 5.81. In order to deter the generation of floating • Monitoring and fiscal risk – need to consolidate all debt, it is essential to enforce the law and sanction public sectors accounts to monitor fiscal risks; public officials who do not comply with it. The exist- • Accounting – to address substantial weaknesses in ing CGC Law must be amended to include sanctions this area; that are substantial and credible. That would also • Budget reporting - to improve the quality of infor- make public servants, NGOs’ officials, and managers of mation on budget execution; fideicomisos more accountable. Also, due to the fact • Municipal transfers - municipalities should be able that Guatemala has been confronting natural disasters to anticipate their transfers, to plan and prepare regularly, it is critical to better regulate the use of pub- their budgets appropriately; lic funds and their accountability in those situations. • Human recourses - to update the data base of the civil servants; Fideicomisos • Audits (internal and external) – should focus not only on auditing transactions, but rather on out- 5.82. Fideicomisos have enabled public entities to comes and results; and deliver services and infrastructure in a timely man- • Predictability of donor funds – enhance coordina- ner. However, because of political interference (e.g., tion between the MOF and donors to plan dis- COVIAL and Mi Familia Progresa.), their use should bursements of external funds in advance. be restricted only to special cases. As mentioned above, the CGC has been performing regular au- Floating Debt dits to the relevant fideicomisos, and most of them have been operating in accordance to international 5.79. Although the Government has implemented standards. Nevertheless, this is no guaranty of the a set of measures to deter the creation of floating correct use of resources. In particular, CGC audits debt, the practice continues to affect PFM perfor- outcomes have often been positive one year and Public Expenditure Review mance and credibility. In 2010, the IMF produced a negative the next. report recommending steps to deter the practice. Also, with the assistance of the European Commis- 5.83. The lack of a public fideicomisos law has se- sion, an audit on the floating debt is undergoing. How- verely limited CGC’s power to better oversee and ever, notwithstanding this analytical work and the ad- audit the fideicomisos’ performance. Currently, ministrative measures taken to date, floating debt is CGC depends on the willingness of the fiduciario to still being generated. hand over financial and operational information to ex- ecute the audits. Enforcement of credible sanctions 107 could help in that respect. In addition, a new law for 5.87. Based on the advantages of the pilot program, public fideicomisos would deprive those, such as the it is highly recommendable that the MOF drafts a COVIAL, which have argued that they are subject to new PFM legislation to improve the PFM systems the regulation of the Commercial Code, i.e., that they performance, drawing from the lessons learned can act as private entities even though they receive from the pilots’ experience. An improved MTEF will and spend public funds. support the revision of programs as well as all budget cycle processes. 5.84. The Government should continue its efforts to build consensus around a public fideicomisos Law. Public Procurement Most of the regulations that might be included in the draft law are stated in the 2011 and 2012 budget laws. 5.88. A key characteristic of a modern public pro- However, there is no guarantee that Congress will curement system is a strong accountability frame- include such regulations in future years. A new Law work. This means that managers are held account- would provide a permanent legal basis governing the able for their decisions and the resulting outcomes. management of fideicomisos. In order for Guatemala to have a procurement system that delivers the government’s investment program as Budget Formulation well as the goods and services demanded by society, substantial reform is required that shifts the objective 5.85. Currently, annual budgets are formulated of the procurement system from compliance with pro- based on past allocations. Traditional programs are cesses to compliance with outcomes. It is important to not reassessed and revised, and an inertial budget note that shifting the focus to outcomes does not im- execution has become common practice. Improved ply that compliance with procedures is not important; budget preparation would reduce the need to modify rather, it means that both outcomes and procedures allocations during the year and also enhance the cred- will become measurable results. ibility of the PFM systems. In addition, more realistic income projections would facilitate budget and cash 5.89. The analysis presented in this section clearly management. indicates the need to improve planning, consolida- tion, competition and contract management as key 5.86. The MOF, with support from the EC, has components in order to enhance the efficiency and launched an initiative to improve the budget for- transparency of Guatemala’s public procurement mulation by revising the program structure of three system. Whilst Guatemala was a pioneer in the intro- pilot ministries: Health, Education, and Interior. The duction of framework agreements in the Latin Ameri- progress achieved in the Ministry of Health shows can region, it nevertheless failed to capitalize on the that using a results oriented approach is possible potential benefits generated by the use of this tool. and useful.183 This approach links policy objectives Design flaws in the planning and operational model with allocation of available resources, and provides contributed to this. The decline of framework agree- a medium-term horizon for achieving strategic goals. ments is reflected in the decreasing use of this tool by Further, better estimations of unit costs would help public sector entities from a maximum of nearly 7 per- monitor and evaluate the quality of expenditures by cent of contracts in 2009 two less than 0.04 percent program and assess whether intended results have in 2011. been achieved. The program revision also includes Public Expenditure Review investment projects to improve infrastructure. Thus, 5.90. By redesigning the operational model for whether a ministry executes its capital budget by itself framework agreements, the government could le- or through a fideicomiso or NGO, this will make the verage its e-procurement platform, Guatecompras, public investment program more consistent with the to implement whole-of-government framework sector’s desired results and strategic goals. agreements for common use goods and services. 183. In particular, for the indicated pilots the MOF has produced a set of guidelines to support the implementation of a results-based bud- geting framework. This approach is based on the definition of results indicators for the short to medium term. Results indicators in the health 108 sector include, for instance, the reduction of child and/or maternal mortality. This would eliminate the use of direct contracting and exceptions for the procurement of common use goods and services, and thereby result in increased efficien- cy, reduced prices and increased transparency. This approach could also be applied to the procurement of pharmaceuticals. 5.91. In the area of infrastructure, one of the key deficiencies identified by this review is the appar- ent lack of accountability in the area of contract management. It appears that, whilst the government captures the entire contract management data relat- ed to infrastructure projects, it does not use this data as a management tool. The review identified a large number of contracts with significant overruns, some of which were many times the original contracted amount. In order to address these issues, a review should be undertaken to identify the causes of these overruns, and terminate any contracts which should have been re-bid rather than extended. Also, there is a need to implement stricter procedures in the contract management system to ensure that contracts cannot simply be extended or incur significant overruns with- out the proper public scrutiny. Another practice that should be avoided is the entering of contracts into the system after they have been complete (rather than when they are signed) only in order to make payments. 5.92. The large commitments that exist in signed contracts exceeding the institutional budgets many times over, point to a disconnect between budgeting and execution. These weaknesses point to a system where accountability is focused on compliance with process, rather than with a view to achieving results. 5.93. A key area for reform is the development of a cadre of procurement specialists whose qualifica- tions are founded on a competencies based system that is linked to career advancement as well as re- sponsibility. The adoption of such a certification sys- tem would ensure that public officials tasked with car- Public Expenditure Review rying out procurement activities are trained and thus knowledgeable on the application of a wide range of procurement methodologies. This initiative would also serve to create specialized units that would in turn have the skills and knowledge to address sector spe- cific issues such as the procurement of pharmaceuti- cals, setting up and administering framework agree- ments, as well as complex infrastructure projects. 109 Lista de referencias Acemoglu, Daron and James Robinson (2006). “Eco- www.monografias.com/trabajos-pdf3/guatemala- nomic Backwardness in Political Perspective.� monografia-partidos-politicos/guatemala-mono- American Political Science Review 100: 115–31, grafia-partidos-politicos.pdf. 2006. ASIES (2010). Monografía del Congreso de la Repúbli- Acemoglu, Daron and James Robinson (2008). “The ca de Guatemala. Asociación de Investigación y Role of Institutions and Development,� Working Estudios Sociales, Departamento de Investiga- Paper 10, Commission on Growth and Develop- ciones Sociopolíticas, Guatemala, viewed May 10 ment, 2008. 2011, http://www.asies.org.gt/sites/default/files/ar- ticulos/publicaciones/201011%20Monografia%20 Acevedo, Carlos (2008). Los costos económicos de la del%20Congreso.pdf violencia en Centroamérica. San Salvador: Con- sejo Nacional de Seguridad. ASIES (2011). Análisis del proyecto de presupuesto de ingresos y egresos del Estado para 2011. Ciudad de Achard, Diego and Luis E. Gonzalez (eds.) (2004). A Guatemala: ASIES Challenge for Democracy – Political Parties in Central America, Panama and the Dominican Re- Bannister, Geoffrey J. & Luis-Diego Barrot (2011). A public. Washington DC: Inter-American Develop- Debt Intolerance Framework Applied to Central ment Bank. America, Panama and the Dominican Republic. IMF Working Paper No. 11/220. Washington, DC: Adams, Tani (2011). Consumed by Violence. Cambridge, International Monetary Fund. MA: Collaborative for Development Action. Barbero, Jose (2009). Potencial para Reducir los Re- Afonso, Antonio, Ludger Schuknecht, & Vito Tanzi querimientos Financieros del Sector Vial. Mimeo. (2005). “Public sector efficiency: An international comparison�. Public Choice, Vol. 123(3), pages 321- Barreix, Alberto, Martin Bés and Jeronimo Roca 347, June 2005. (2009). Equidad Fiscal en Centroamérica, Panamá y República Dominicana. Washington, DC: IADB. Agner, Joy (2008). “The Silent Violence of Peace in Guatemala�. North American Congress on Latin Barrientos, Ricardo (2011). La Economía Política del America. Viewed March 2012. http://nacla.org/ Gasto Público en Guatemala. Background Paper node/4665 for the Guatemala Public Expenditure Review. Asociación de Investigación y Estudios Sociales Bowser, Diana M. & Ajay Mahal (2011). “Guatemala: (ASIES) (2005). Instituciones políticas, procesos de The economic burden of illness and health system formulación de políticas y resultados de las políti- implications�. Health Policy, vol. 100(2-3), pages Public Expenditure Review cas en Guatemala. Latin America Research Net- 159-166, May 2011. work, Inter-American Development Bank, viewed May 10 2011, http://www.iadb.org/res/publications/ Briscoe, Ivan and Martin Rodriguez Pellecer (2010). pubfiles/pubS-305.pdf A state under siege: elites, criminal networks and institutional reform in Guatemala. Netherlands ASIES (2009). Monografía de los Partidos Políticos, Institute of International Relations, Clingendael, 2004-08. Asociación de Investigación y Estudios viewed May 10, 2011, http://www.clingendael.nl/ Sociales, Departamento de Investigaciones Socio- publications/2010/20100913_cru_publication_ibris- 110 políticas, Guatemala, viewed May 10 2011, http:// coe.pdf Bureau of International Narcotics and Law Enforce- Dudley, Steven S. (2010). “Drug Trafficking Organiza- ment Affairs (2009). Counternarcotics and Law tions in Central America: Transportistas, Mexican Enforcement Country Program: Guatemala. Wash- Cartels and Maras�. In: Olsen, Eric L., David A. ington, DC: US Department of State. Shirk and Andrew Selee (eds.). Shared Responsi- bility: US-Mexico Policy Options for Confronting Cabrera, Maynor (2010). Guatemala: La tributación Organized Crime. Washington, DC: Woodrow Wil- directa, la evasión en el impuesto sobre la renta y son Center. sus desafíos. In Juan Pablo Jiménez, Juan Carlos Gómez Sabaini and Andrea Podesta (eds.), Evasión Estado de la Nación (2011). El Estado de la Región. San y Equidad en América Latina. Santiago de Chile: José, Costa Rica. CEPAL. Fuentes, Juan Alberto and Maynor Cabrera (2005). El Cabrera, Maynor (2011). El lento progreso de las refor- Pacto Fiscal de Guatemala: una oportunidad per- mas tributarias en Guatemala: elementos para la dida. Paper presented at the 17th Regional Semi- construcción de una reforma viable. Background nar on Fiscal Policy. Santiago de Chile: CEPAL. Paper for the Guatemala Public Expenditure Re- view. García Díez, Fatima (2001). Sistemas electorales de América Central y el Caribe: una introducción y Calvaruso, Andrea, Ricardo Stein and Fabrizio Feliciani algunas notas para el debate. In Martínez, R. (ed.). (2008). Transformaciones del Estado Guatemalte- Prácticas de Sistemas Políticos Comparados II: co a Diez Años de los Acuerdos de Paz: Un Ensayo. Centroamérica y Caribe. Barcelona: Universidad Guatemala: Vicepresidencia de la República. de Barcelona. P. 15-45. Comisión Económica para América Latina y el Caribe Gobierno de Guatemala (2010a). Evaluación de Da- (CEPAL) (2011). Resumen Regional del Impacto de ños y Perdidas Sectoriales y Estimación de Nece- la Depresión Tropical 12-E en Centroamérica. Cu- sidades Ocasionados por Desastres Naturales en antificación de Daños y Pérdidas sufridos por los Guatemala entre mayo septiembre de 2010. Gua- Países de la Región en el mes de octubre de 2011. temala City. Santiago de Chile. Gobierno de Guatemala (2010b). Evaluación de Da- Congressional Research Service (2009). Gangs in ños y Perdidas Sectoriales y Estimación de Nece- Central America. Washington, DC. sidades Ocasionados por el paso de la Tormenta Tropical Agatha y la Erupción del Volcán Pacaya, Cooper, Adam & Catherine Ward (2008). Prevention, Resumen Preliminar. Guatemala City. disengagement and suppression: a systematic re- view of the literature on strategies for addressing Gomez Sabaini, Juan Carlos and Juan Pablo Jimenez young people’s involvement in gangs. Cape Town: (2011). Tax structure and tax evasion in Latin Amer- Human Science Research Council. ica. Santiago de Chile: CEPAL. Cubero, Rodrigo and Ivanna Vladkova Hollar (2010). Grupo de Apoyo Mutuo (GAM) (2011). Informe de Equity and Fiscal Policy: The INcome Distribution Transparencia Presupuestaria del Ministerio de Effects of Tacation and Social Spending in Central Gobernación. Guatemala City. America. IMF Working Paper No. 10/112. Washing- ton, DC: International Monetary Fund. Guatemala Human Rights Ombudsman (2011). Informe Public Expenditure Review Estadístico Sobre Homicidios, Lesiones y Violencia Department for International Development (DfID) Intrafamiliar y Delitos Sexuales Contra la Mujer. (2001). Understanding and Reforming Public Ex- Guatemala. penditure – Guidelines for DfID. London. Herrera, Santiago & Pang, Gaobo (2005). Efficiency Dirección General de Caminos (DGC) (2007). Refor- of public spending in developing countries: an ef- mulación y Actualización del Plan de Desarrollo ficiency frontier approach. Policy Research Work- Vial 2008-2017. Guatemala City. ing Paper Series No. 3645. Weashington, DC: The World Bank. 111 ICEFI (2007b). Viabilidad de incrementar los im- Keefer, Phil (2011). “Collective Action, Political Parties, puestos al tabaco. Working Paper Instituto Cen- and Pro-Development Public Policy�. Asian Devel- troamericano de Estudios Fiscales (ICEFI), Guate- opment Review, vol. 28, no. 1, pp. 94−118. mala, Mimeo. Kennedy, David (2008). Deterrence and Crime Preven- ICEFI (2012). Análisis del Pacto Fiscal para el Cambio. tion: Reconsidering the Prospect of Sanction. Lon- http://www.icefi.org/categories/16?clas=5&deta don: Routledge. il=254. Viewed March 14, 2012. LAPOP (2010). Cultura política de la democracia: Iniciativa de Transparencia del Sector Construcción Informe de las Américas. Viewed December, (CoST) Guatemala (2010). Estudio de Línea Base. 2011. http://www.vanderbilt.edu/lapop/ab2010/ http://www.costguatemala.org/wp-content/up- Report_on_the_Americas_Spanish_Formatted_ loads/2010/12/CoST-Guatemala-Estudio-de-Trans- V3_4april2011.pdf parencia1.pdf. Accessed on March 6 2012. LAPOP (2011). Cultura política de la democra- Iniciativa de Transparencia del Sector Construcción cia en Guatemala, 2010. Viewed May 22, 2011, (CoST) Guatemala (2011). Informe de Aseguramien- http://www.kas.de/wf/doc/kas_21707-1522-4-30. to. http://www.costguatemala.org/wp-content/ pdf?110124190800 uploads/2011/07/INFORME-ASEGURAMIENTO- COST-JULIO-2011.pdf. Accessed on March 6, 2012. Lopez, Humberto (2001). Internal World Bank Memo. InSight - Organized Crime in the Americas (2011). Mendoza, Carlos (2010). Dinámicas políticas de las re- Cuerpos Ilegales y Aparatos Clandestinos de Se- formas fiscales en América Central. Working paper guridad - CIACS (insightcrime.org/criminal-groups/ for the 2nd Report of Fiscal Policy ICEFI, mimeo. guatemala/ciacs/item/1018-ciacs). Mendoza, Carlos (2012). http://linchamientos. Instituto Nacional de Estadisticas (2010). Informe Fi- blogspot.com/p/datos.html nal. V Encuesta Nacional de Salud Materno Infantil 2008-2009. Guatemala City. Menocal, Alina and Andrea Calvaruso (2008). Power and Change Analysis: Guatemala. Royal Nether- Instituto de Estudios Comparados en Ciencias Penal- lands Embassy, Guatemala City, Guatemala, Au- es de Guatemala (2007). Observatorio de Justicia gust 2008. Penal. Guatemala City. Moser, Caroline (2004). “Urban Violence and Insecu- Inter-American Development Bank (2008). Policymak- rity: An Introductory Roadmap.� Environment & ing in Latin America: How Politics Shapes Policies. Urbanization, Vol. 16 No. 2, October 2004. Washington DC. Musalo, Karen, Elisabeth Pellegrin and S. Shan Rob- Inter-American Development Bank and Washington erts (2010). “Crimes without punishment: Violence Office on Latin America (WOLA) (2011). Mapping against women in Guatemala�. Hastings Women’s Study of Internationally Funded Citizen Security Law Journal, July: 161-221. Projects in Central America. Washington, DC. Noriega, Arnoldo, Enrqieu �lvarez and Mario Chocoj International Crisis Group (2010). Guatemala: Asfixiada (2001). Cuando la sociedad guatemalteca se en- Public Expenditure Review entre el Crimen y la Impunidad. Informe sobre Améri- cuentra: la negociación política del pacto fiscal. ca Latina Nº33, 22 (crisisgroup.org/en/regions/ Guatemala: IPES. latin-america-caribbean/guatemala/033-guatema- la-squeezed-between-crime-and-impunity.aspx) OECD (2011). Revenue Statistics in Latin America (1990-2009). Paris: Organization for Economic Co- International Crisis Group (2011). Drug Trafficking and operation and Development. Violence, Latin American Report Nº33. Guate- mala City: Commission of Police Chiefs of Central Organismo Judicial (2009). Informe de Labores del Or- 112 America and Caribbean. ganismo Judicial 2009. Guatemala. Ortiz, Gerson (2009), “Empresas de seguridad privada la Gestión de Riesgo en la Inversión Pública. Gua- reflejan la debilidad del Estado� (“Private Security temala City. Firms Reflect the Weakness of the State�), La Hora, October 9, 2009. Schneider, A., Forthcoming, State-building and Tax Re- gimes in Central America. New York: Cambridge Partners for Democratic Change (2012). Adjustment University Press and Validation of the Juvenile Justice Assessment Tool. Guatemala Juvenile Justice System Assess- Schneider, Aaron (In Press). Socio-economic Change ment. Washington, DC. and Fiscal Challenges in Central America: Global- ization, Democratization and the Politics of Re- Peacock, Susan C. and Adriana Beltrán (2003). Hidden form. In Sánchez-Ancochea, D. and Morgan, I. (Ed.) Powers in Post Conflict Guatemala. Washington, Political Economy of the Budget Process in the DC: WOLA (cgrs.uchastings.edu/pdfs/Hidden- Americas. New York: Palgrave PowersFull.pdf). Schneider, Aaron and Annabella España-Najera (2011). Pessino, Carola and Ricardo Fenochietto (2010). Guatemala: Limited Advance within Advancing “Determining countries’ tax effort.� Revista de Limits. Unpublished Manuscript for the Interna- Economía Pública 195-(4/2010):65-87. Madrid: Ha- tional Budget Project cienda Pública Española. Schuster, Christian (2011). The Political Economy of Pico, Juan Hernández (2009). “The Rosenberg Case: Public Finance in Guatemala. Mimeo. A Guatemalan Labyrinth,� Revista Envío (English edition), June 2009, available from www.envio.org. Small Arms Survey (2008). Small Arms Survey 2008: ni/articulo/4015 Risks and Resilience. Geneva, Switzerland. Porras Castejón, Gustavo (2008). Guatemala: Ten Swiston, Andrew & Luis-Diego Barrot (2011). The Role Years After the Agreements on a Firm and Lasting of Structural Reforms in Raising Economic Growth Peace. Washington, DC: Inter-American Dialogue. in Central America. IMF Working Paper No. 11/248. Washington, DC: International Monetary Fund. Public Expenditure and Financial Accountability (PEFA) Secretariat (2005). Public Financial Man- Transparency International, Global Corruption Report agement Performance Measurement Framework. 2007: Corruption in Judicial Systems, New York: Washington, DC. Cambridge University Press, 2007, pp. 210-212 Ranum, Elin Cecilie (2006). Pandillas juveniles trans- UNDP (2006). El Costo Económico de la Violencia en nacionales en Centroamérica, México y Estados Guatemala. Guatemala City. Unidos. Diagnóstico Nacional. Guatemala: Instituto Universitario de Opinión Pública. UNDP (2011). Análisis de los presupuestos públicos de seguridad y justicia en Centroamérica. Proyecto Replogle, Jill (2005). “In Guatemala, a Rise in Vigilante AECID/SICA/PNUD. Panama. Justice,� Christian Science Monitor, October 6, 2005. UNODC (2011). Global Study on Homicide. Vienna: Sanchez, Omar (2008). Guatemala’s Party Universe: UNODC. A Case Study of Under-Institutionalization, Latin Public Expenditure Review American Politics and Society, Vol. 50, No. 1 (2008), USAID, Central America and Mexico Gang Assess- pp. 123–51. ment: Annex2: Guatemala Country Profile. April 2006 Sanford, Victoria (2008). “From genocide to femicide: Villalobos, Joaquin (2011). “Partidos Políticos y Feudal- Impunity and human rights in twenty-first century ismo Criminal,� El País, Guatemala, September 13, Guatemala�. Journal of Human Rights, No.7: 104-122. 2011. Secretaría de Planificación y Programación de la Pres- Villela, Luiz, Andrea Lemgruber and Michael Jorratt idencia (SEGEPLAN) (2010). Guía para incorporar (2010). Tax Expenditure Budgets – Concepts and 113 Challenges for Implementation. IADB Working Pa- World Bank (2011c). La política educativa como instru- per No.131. Washington, DC: Inter-American Devel- mento de desarrollo de largo plazo en Guatemala. opment Bank. Mimeo. Wehner, Joachim (2006). Legislative Institutions and World Bank (2012). Prioridades para Salud y Nutrición Fiscal Policy. PSPE working papers, 08-2006. De- en Guatemala: Un Camino Propuesto a Seguir. partment of Government, London School of Eco- Mimeo. nomics and Political Science, London, UK. World Economic Forum (WEF) (2008). Global Com- World Bank (2005a). Guatemala Public Expenditure petitiveness Report 2008-2009. Geneva, Switzer- Review. Washington, DC. land. World Bank (2005b). Guatemala Country Economic World Economic Forum (WEF) (2009). Global Com- Memorandum – Challenges to Higher Economic petitiveness Report 2009-2010. Geneva, Switzer- Growth. Washington, DC. land. World Bank (2007). “The Promise of Youth: Policy for World Economic Forum (WEF) (2010). Global Com- Youth at Risk in Latin America and the Caribbean.� petitiveness Report 2010-2011. Geneva, Switzer- land. World Bank (2008a). Guatemala – Investment Climate Assessment. Washington, DC. Report No.47193- World Economic Forum (WEF) (2011). Global Compet- GT. Washington, DC. itiveness Report 2011-2012. Geneva, Switzerland. World Bank (2008b). Local Gains from Global Oppor- tunities: Improving Central America’s Investment Climate. Washington, DC. World Bank (2009). Guatemala Poverty Assessment: Good performance at low levels. Report No.43920- GT. Washington, DC. World Bank (2010a). Crime and Violence in Central America. Washington, DC. World Bank (2010b). Guatemala Performance Report on Public Expenditure and Financial Accountabil- ity (PEFA). July 2010. Viewed October 12, 2011, http://www.minfin.gob.gt/archivos/prensa/pefa_ing. pdf World Bank (2010c). Prioridades para Salud y Nu- trición en Guatemala: Un Camino Propuesto a Se- guir. Mimeo. Public Expenditure Review World Bank (2011a). Doing Business 2011. Washington, DC. World Bank (2011b). Financial Management Informa- tion Systems: 25 Years of World Bank Experience on What Works and What Doesn’t. Washington, DC. 114 Annexes ANNEX 1: RESULTS OF OLS REGRESSION Table A1.1: Results of OLS Regression Test Scores Test Scores Test Scores Completion Completion level 3 level 6 level 12 Rates Level 6 Rates Level 9 Per pupil Spending in Education -0.11 0.03 -0.19 -0.03 0.05 0.69 0.23 1.56 0.31 0.46 Adult literacy Rates 0.45 0.66 -0.02 -0.09 0.10 2.34** 5.40*** 0.26 0.82 1.45 Teachers per 100 students 0.01 -0.03 0.02 0.05 0.08 0.39 1.19 0.98 2.41** 5.37*** Chronic Malnutrition rates 0.00 -0.01 0.00 0.00 0.00 2.43** 3.92*** 0.10 0.42 1.49 % population living in urban 0.08 0.08 0.04 -0.13 -0.04 0.75 0.91 0.73 1.90* 0.80 % population indigenous -0.15 -0.08 -0.01 -0.01 -0.03 2.80*** 1.62 0.39 0.33 0.82 Population density 0.05 0.07 0.03 0.07 0.05 1.13 1.97* 1.59 2.37** 2.10** Poverty rates -0.14 0.03 -0.21 -0.11 -0.01 1.00 0.21 2.28** 1.01 0.18 Extreme poverty rates -0.11 -0.08 0.10 -0.06 -0.05 3.19*** 3.05*** 7.81*** 1.51 1.91* Observations 231 233 185 233 233 Public Expenditure Review Adjusted R-squared 0.43 0.47 0.23 0.05 0.37 Robust t-statistics in italics. * significant at 10% level; ** significant at 5% level; *** significant at 1% level Source: Authors’ calculations. Notes: per pupil spending refers to spending in education function (excluding tertiary) in 2010 for central spending plus average spending at municipal level in 2008 and 2009. Test scores, completion rates, teacher pupil ration and adult literacy rates are from the Ministry of Education and Directorate of Planning (DIPLAN), Malnutrition rates were downloaded from SIGSA. Percentage population living in urban, indigenous, population density, poverty and extreme poverty rates taken from the Municipal Development Plans developed by the Secretariat of Planning (SEGEPLAN). 115 ANNEX 2: DATA ENVELOPMENT ANALYSIS (DEA) Data Envelopment Analysis (DEA) of the efficiency of public spending Data envelopment analysis is a methodology that, production frontier for Guatemala was modeled by a through linear programming, generates “efficiency combination of multiple output indicators – including frontiers� and estimates relative efficiency scores completion rates and test scores – and multiple as the difference from such frontier. Once derived, inputs which included per pupil spending in basic these efficiency scores can be mapped to detect education, teachers/pupil ratios, and class sizes, as possible geographical patterns or can be compared well as a host of socio economic indicators that are with approximations of need (i.e. extreme poverty known to play a large role on student achievement. rates, malnutrition, etc.) and per capita expenditure These control variables included adult literacy rates to reveal differences across the country. The and malnutrition rates 184. Gráfico A: Modeling the production frontier using data envelopment analysis Output: literacy, numeracy, completion, test results Input: wages, teachers, classrooms, capital Looking for relative ef�ciency as distance (ranking) from frontier of “good practice� Public Expenditure Review 184. Additional variables could have been added to this analysis including poverty rates, population density, etc., however were omitted due to the non availability of these indicators for all municipalities. These variables however were used in the multi-variate regressions discussed 116 in the next section. ANNEX 3: IMPACT OF RECENT NATURAL road sector include damages to the physical infra- DISASTERS ON ROAD SECTOR structure (slopes erosions, landslides, damages on the road surface, bridges, drainage, ancillary and hy- Guatemala, like most of the countries in Central draulic works). The damages in the road sector were America, is highly vulnerable to natural disasters due 51 percent of the total damages of US$595.8 and they to its geographical location. Every year Guatemala is also represent 67 percent of the executed budget hit by the negative impacts of the rainy and stormy of DGC and COVIAL in 2010. Table 10 presents the seasons, and during the last ten years has had the im- quantified damages by the type of network, showing pacts of four strong natural disasters that left signifi- that the rural road network was the most damaged. cant negative impacts in the country across sectors: The evaluation identified over 50 road points where Stan in 2005, Agatha in 2010, and the Eruption of traffic was suspended due to destroyed infrastruc- the Volcano Pacaya also in 2010. In October 2011, the ture (bridges and other structures). Tropical Depression 12-E hit Mexico and the Central American region leaving significant damages in these The Damages and Losses Assessment (DALA) carried countries including Guatemala. out after the storm 12-E assessed the damages in the road sector equivalent to US$ 35 million (Q.275.4 mil- A Post Disaster Needs Assessment (PDNA) carried lion). After this event the road sector again was the out after the eruption of the Pacaya Volcano (May most affected sector. The damages in the sector rep- 2010) and the Agatha Storm (June 2010) quantified resented 43percent of the total damages estimated damages of the road sector in the equivalent of in US$81.9 million (Q.643.7 million). As presented in US$304 million (Q.2,456 million). Damages in the Table 10 for Guatemala the storm 12-E left less dam- Table A3.1: Damages due to the natural disasters of Agatha and Pacaya Volcano (2010) and storm E-12 (2011) Sector Agatha and Pacaya Tropical depression 12-E Damages in the transport sector (US$ million) 304.8 35.0 % of total damages 51% 43% % of 2010 executed budget of DGC and COVIAL 67% 8% % of GDP 0.74% 0.09% Damages in all sectors (US$ million) 595.8 81.9 % of total damages 100% 100% % of 2010 executed budget of Central Administration 10% 1.3% % of GDP 1.45% 0.2% Source: PDNA 2010, DALA 2011, and own estimations. Public Expenditure Review 185. The PDNA also quantified the losses for the transport sector in US$20.9 (Q.168.4 million). Since the losses in the transport sector are related to the private sector or vehicle passengers - vehicle operating costs (gas, lubricants, wheels depreciation, and maintenance) and the time value of the passengers due to congestion, and transit detours - for the purposes of the PER analysis have been excluded. 186. The losses estimated due to the tropical depression 12-E for the road sector amount US$1.79 million (Q.14.08 million) and the losses including all sectors US$251.2 million (Q.1,974.6 million). 187. The PDNA also quantified losses in all sector for an amount of US$379 million (Q.3,055.3 million) for a total amount of damages and loss- es of US$974.9 million, 2.37% of GDP. Guatemala has had in the past natural disaster that left damages as a percentage of GDP higher than the events in 2010. The earthquake in 1979 left damages and losses as percentage of GDP of 17.9%, Mitch Storm 4.7% and Stan Storm 3.4%. 117 ages and losses than the natural disasters in 2010, for are built. For reducing the vulnerability of bridges both, the road sector and including all sectors. is important building higher and wider structures to avoid the increased water level during the rainy/ The PDNA also proposed an investment program storm season. For slopes, reshaping is important to in the sector to reconstruct damage infrastructure reduce the risk of landslides on the routes which as presented in Table 12 for the period 2010-2014. cause damages to the roads, cause removal costs, The objective of presenting separate investment and are a risk to vehicles and passengers. The esti- lines for bridges and slopes is to give importance mation for reconstruction investment needs is larger to these structures including to the preventive ac- for roads (74 percent), but the needs for bridges and tivities to reduce the vulnerability of bridges and slopes represent about 15 percent for each item line. slopes. Horizontal structures are highly vulnerable The reconstruction needs are estimated for annual to natural disasters. Recent designs are important amounts of around US$80 million for the first years, to consider changes in the terrain and changes in which is around 19 percent of the 2010 executed the flow of water sources where these structures budget for DGC and COVIAL. Table A3.2: Damages in the Road Network due to the natural disasters of 2010 - Agatha and Pacaya Volcano Road Network Type (within Guatemalan territory) Quantified Damages Quantified Damages % (in Q. million) (in US$ million) Centro-American 456.3 56.6 19% Principal National Network 511.0 63.4 21% Departamental Secondary Network 507.6 63.0 21% Rural Network 981.5 121.8 40% Total 2,456.4 304.8 100% Source: PDNA, 2010. Table A3.3: Reconstruction investments needs in the road sector (in US$ million) Total amount % 2010 2011 2012 2013 2014 Rehabilitation and reconstruction of Main Routes 186.2 50% 43.4 43.4 37.2 31.0 31.0 Rehabilitation of Secondary and Tertiary Routes 80.7 22% 24.8 18.6 18.6 12.4 6.2 Bridges designs – preventive 6.2 2% 2.5 3.7 0.0 0.0 0.0 Reconstruction of Bridges 43.4 12% 8.7 9.9 9.9 8.7 6.2 Public Expenditure Review Reshaping of slopes – preventive 55.8 15% 6.2 12.4 12.4 12.4 12.4 Total 372.3 100% 85.6 88.1 78.2 64.5 55.8 % of 2010 executed budget of DGC and COVIAL 81% 19% 19% 17% 14% 12% 81% Source: PDNA, 2010. 118 ANNEX 4: PUBLIC EXPENDITURES ON CITIZEN SECURITY Table A4.1: Allocated national budget by institution and type of expenditure (2004-2010) Q. million 2004 2005 2006 2007 2008 2009 2010 Security (Function 202) Ministry of the Interior 1,581 1,700 1,912 1,912 2,348 3,005 2,938 President's Office* 38 94 95 95 126 139 139 Ministry of Communications, Infrastructure and Housing 12 47 47 40 27 27 Secretaries and other dependence of the Executive 31 Power Obligations of State in charge of the Treasury 66 71 94 94 165 122 204 Total Security 1,685 1,877 2,148 2,148 2,679 3,293 3,338 Justice (Function 102) Ministry of the Interior 30 30 Obligations of State in charge of the Treasury 1,210 1,246 1,438 1,438 1,974 1,990 1,893 Public Defender’s Office** 37 34 37 37 47 53 53 Judicial Agency*** 1,616 1,825 1,974 1,990 1,980 Human Rights Office*** 37 74 41 48 224 Total Justice 1,247 1,280 3,127 3,373 4,065 4,110 4,150 Total Security and Justice Q. million 2,931 3,157 5,276 5,522 6,744 7,403 7,488 % current exp. 95.1 95.7 93.0 93.0 96.0 94.6 95.0 % capital exp. 4.9 4.3 7.0 7.0 4.0 5.4 5.0 US$ million 369 413 694 719 892 907 929 % of total exp. 9.5 9.7 14.0 14.6 15.9 14.9 15.0 % of GDP 1.5 1.5 2.3 2.1 2.3 2.4 2.3 *Includes: Secretaria de Asuntos Administrativos y de Seguridad, Consejo Asesor de Seguridad y Consejo Nacional de Seguridad; not officially classified as Security **Not officially classified as Justice ***Data from PNUD (2011), Organismo Judicial includes: Organismo Judicial, Ministerio Publico and Instituto de Defensa Publica Penal. Source: Official National Budget Database, and UNDP. Own calculations Public Expenditure Review 119 Table A.4.2: Allocated national budget on citizen security, by pillar and institution (2004-2010) Q. million 2004 2005 2006 2007 2008 2009 2010 CONTROL/ REPRESSION JUSTICE Ministry of Interior 30 30 Ministry of Communications, Infrastructure and Housing Secretaries and other dependence of the Executive Power Obligations of State in charge of the Treasury 1,210 1,246 1,438 1,438 1,974 1,990 1,893 Public Defender’s Office* 37 34 37 37 47 53 53 Judicial Agency*** 1,616 1,825 1,974 1,990 1,980 Human Rights Office*** 37 74 41 48 224 SECURITY Ministry of Interior 119 153 74 74 2,034 2,580 2,528 Ministry of Communications, Infrastructure and Housing 11 11 11 2 1 1 Secretaries and other dependence of the Executive Power 22 Obligations of State in charge of the Treasury 20 20 96 President's Office** 38 94 95 95 126 139 139 Total Control/ Repression 1,403 1,538 3,327 3,573 6,227 6,830 6,936 INSTITUTIONAL STRENGHTHENING SECURITY Ministry of Interior 1,308 1,398 1,670 1,670 122 100 109 Ministry of Communications, Infrastructure and Housing 36 36 35 26 26 Obligations of State in charge of the Treasury 66 71 74 74 165 122 6 Total Institutional Strengthening 1,374 1,468 1,780 1,780 322 248 141 PREVENTION SECURITY Ministry of Interior 8 2 16 16 21 76 52 Secretaries and other dependence of the Executive Power 9 Obligations of State in charge of the Treasury 101 Total Prevention 8 2 16 16 21 76 162 REHABILITATION SECURITY Public Expenditure Review Ministry of Interior 146 147 152 152 170 249 249 Ministry of Communications, Infrastructure and Housing 1 1 1 4 Total Rehabilitation 146 148 152 152 174 249 249 TOTAL 2,931 3,157 5,276 5,522 6,744 7,403 7,488 *Not officially classified as Justice. **Includes: Secretaria de Asuntos Administrativos y de Seguridad, Consejo Asesor de Seguridad y Consejo Nacional de Seguridad; not officially classified as Security. ***Data from PNUD (2011), Organismo Judicial includes: Organismo Judicial, Ministerio Publico and Instituto de Defensa Publica Penal. Source: Official National Budget Database, UNDP and own calculations 120 MAP Public Expenditure Review 121 The World Bank 1818 H Street, NW, Washington, DC 20433, USA