Kenya Public Expenditure Analysis 2019 Creating fiscal Space to deliver the Big 4 while undertaking a needed fiscal consolidation June 2019 Macroeconomics, Trade & Investment (MTI) Kenya Contents Foreword....................................................................................................................................................... 4 Acknowledgements....................................................................................................................................... 5 Executive Summary....................................................................................................................................... 6 Chapter 1 Introduction ................................................................................................................................. 9 Chapter 2 Macro-Fiscal Developments ....................................................................................................... 12 A. Macroeconomic context ................................................................................................................. 12 B. Overall fiscal balance, financing and stock of public debt .............................................................. 15 C. Fiscal and Debt Sustainability risks ................................................................................................. 20 D. Conclusion and Recommendations................................................................................................. 25 Chapter 3 Trends and Performance of Government Revenues.................................................................. 27 A. Is the Government mobilizing enough Tax Revenue? .................................................................... 27 B. How close is Kenya’s revenue mobilization comparable to its peer countries?............................. 28 C. What explains the decoupling of tax revenue from their tax bases? ............................................. 31 D. Conclusion and Recommendations................................................................................................. 36 Chapter 4 Trends and Profile of Central Government Expenditures .......................................................... 38 A. Size and trends in Central government expenditures for the last five years ................................. 38 B. What are public resources being spent on? ................................................................................... 40 C. How easy is it, to re-allocate public resources towards national priorities? .................................. 47 D. How effective is spending of public resources? .............................................................................. 50 E. Is Kenya getting value for its money ............................................................................................... 53 F. Conclusion and Recommendations................................................................................................. 55 References .................................................................................................................................................. 58 Annex 1: National Fiscal Operations (FY2013/14-2017/18) ....................................................................... 60 Annex 2: Fiscal and debt sustainability risk scenarios ................................................................................ 61 1 List of Figures Figure 2-1: The Kenyan Economy has rebounded ...................................................................................... 12 Figure 2-2: Contribution to Real GDP growth ............................................................................................. 12 Figure 2-3 : poverty head count against GDP per capita ............................................................................ 13 Figure 2-4 : International poverty comparison ........................................................................................... 13 Figure 2-5: Contribution to real GDP growth (in percentage points) ......................................................... 13 Figure 2-6 : Contribution of private investment remains weak (in percentage terms) .............................. 13 Figure 2-7 : Inflation remains within the target range ............................................................................... 14 Figure 2-8 : Private sector credit growth remains subdued ....................................................................... 14 Figure 2-9 : The current account deficit has narrowed .............................................................................. 15 Figure 2-10 : Government and corporate capital flows financed the current account deficit ................... 15 Figure 2-11 : Remittance inflows ................................................................................................................ 15 Figure 2-12 : Real and nominal exchange rate ........................................................................................... 15 Figure 2-13 : Overall fiscal balance was expansionary over the review period (% of GDP) ....................... 16 Figure 2-14 : Fiscal balance is wider relative to EAC peers (% of GDP)....................................................... 16 Figure 2-15 : Fiscal Impulse and the output gap for Kenya (% of GDP) ...................................................... 17 Figure 2-16 : Public debt has stabilized after a rapid rise in previous years (% of GDP) ............................ 18 Figure 2-17 : Pubic debt moderation is driven by a decrease in the primary balance ............................... 18 Figure 2-18 : External Public Debt by instrument at end-December 2018 (% of total) .............................. 19 Figure 2-19 : Domestic Public Debt by instrument at end-December 2018 (% of total) ............................ 19 Figure 2-20 : External Public Debt, by currency at end-December 2018 (% of total) ................................. 19 Figure 2-21 : Projected Debt Redemption over the medium to long term (in Ksh.million) ....................... 20 Figure 2-22: Drivers of Debt to GDP (% of GDP and percentage points) over the medium term .............. 22 Figure 2-22: Business as usual scenario-Fiscal deficit and public debt ....................................................... 23 Figure 2-22: Lower growth scenario-Fiscal deficit and public debt ............................................................ 24 Figure 2-22: Higher domestic borrowing costs scenario: Fiscal deficit and public debt ............................ 25 Figure 3-1: Tax revenue collection across main tax items is falling ............................................................ 28 Figure 3-2: Underperformance driven by challenges in both Income taxes and VAT ................................ 28 Figure 3-3: Tax revenue vs. GDP per Capita, selected countries ................................................................ 29 Figure 3-4: Tax revenue relative to selected countries (2013-2017) .......................................................... 29 Figure 3-5: Tax revenue (in % of GDP) Kenya and peer countries .............................................................. 30 Figure 3-6: Tax revenue (in % of GDP) tracks closely what happens to taxes on incomes. ........................ 30 Figure 3-7: Taxes on incomes, Kenya and selected countries (2013-17).................................................... 31 Figure 3-8: Taxes on goods and services, Kenya and selected countries (2013-17 .................................... 31 Figure 3-9: Growth in real tax revenue vs. growth in the real tax base ..................................................... 32 Figure 3-10: Tax revenue over-time, customs revenue vs. domestic revenue ........................................... 33 Figure 3-11: Real provisions for refunds (Ksh. million) ............................................................................... 34 Figure 3-12: Access to duty remissions and VAT refunds over time .......................................................... 34 Figure 4-1: Fiscal policy has been expansionary over the review period ................................................... 38 Figure 4-2: Government spending was elevated over the review period .................................................. 38 Figure 4-3: Central government expenditure over time (% of GDP) .......................................................... 39 Figure 4-4: Average Central government expenditure, over 2013-17(%of GDP) ....................................... 39 Figure 4-5: The share of development expenditure relative to total government spending increased over time ............................................................................................................................................................. 40 2 Figure 4-6: The quality of infrastructure has improved .............................................................................. 40 Figure 4-7: Capital stock per capita relative to peers (2015, in constant 2011 international dollars) ....... 41 Figure 4-8: The share of capital stock, Labor and TFP in GDP growth (1990-2017) ................................... 41 Figure 4-9: General government expenditures by economic classification ............................................... 42 Figure 4-10: General government Wage bill in percent of GDP ................................................................. 43 [Figure 4-11: Wage bill, 2014-16 (Percent of GDP)] ................................................................................... 43 Figure 4-12 : Composition of wage bill by economic category, in Ksh. billion (2013-18) ........................... 43 Figure 4-13:Composition of wage bill expenditure 2013-18(%) ................................................................. 43 Figure 4-14: Share of general government employees (in % of head count) 2014 compared to 2018 ..... 44 Figure 4-15: Public service head count of employees (in ‘000) .................................................................. 44 Figure 4-16: Average annual salary (in Ksh. ‘000) ....................................................................................... 44 Figure 4-17: Kenya’s spending priorities in the last five years were tilted towards infrastructural and human capital development ....................................................................................................................... 46 Figure 4-18: Composition of rigidity by Economic Classification ................................................................ 50 Figure 4-19: Execution of development budget remains a challenge mainly at the county level ............. 51 Figure 4-20: Budget execution was especially lower in Energy, Infrastructure and ICT............................. 52 Figure 4-21: Allocated budget to Transport infrastructure v. actual expenditure (2014/15-18) ............... 53 Figure 4-22: Drivers of Transport Expenditure Growth 2016/17 ............................................................... 53 Figure 4-23: Secondary school completion against expenditure per capita, secondary (% of GDP per capita) (2017) .............................................................................................................................................. 54 Figure 4-24 : Infant survival per 1000 live births against per capita public health expenditure per capita (constant 2011 int’l $) 2017 ........................................................................................................................ 54 Figure 4-25: Kenya could achieve faster growth in its GDP per capita, given its GFCF level ..................... 55 Figure 4-26: Kenya’s public investment and access to electricity (2014-2016) .......................................... 55 List of Tables Table 2-1: Domestic Public Debt, by maturity at end-December 2018 (% of total/billions of US$)........... 19 Table 2-2: Medium-term baseline scenario 2019-2024.............................................................................. 21 Table 3-1: Priorities for enhancing revenue collection and improving tax system efficiency .................... 37 Table 4-1: Classification of rigidity by object of expenditure ..................................................................... 47 Table 4-2: Rigidity by Economic Classification (Share of total expenditure) .............................................. 49 Table 4-3: Priorities for Improved Public Expenditure Management ......................................................... 57 List of Boxes Box 3-1: Steps for estimating elasticities and results from existing literature ........................................... 35 Box 4-1: Rigidity Analysis ............................................................................................................................ 47 3 Foreword The Kenyan government remains committed to a planned fiscal consolidation pathway, which should help contain public debt stock at a sustainable level. Nonetheless, there are significant challenges stemming from a slowdown in revenue collection, a growing demand for transfers to county governments, and the need to fund the Big 4 agenda. These issues raise the probability for fiscal slippages, requiring adequate mitigation to safeguard macroeconomic stability. This report provides an overview of the challenges in revenue mobilization, the size and composition of the national government expenditures, and the efficiency of this spending over the last five years. The analysis identifies options for supporting ongoing fiscal consolidation and creating fiscal space for the Big 4 and broad public services delivery. There are three key messages. First, Kenya’s tax revenue as a share of GDP has decreased and decoupled from the growth in the economy, suggesting some important constraints to enhancing revenue collection. There is need for a step change in Kenya’s tax revenue mobilization effort. In the short-run, the government could review and update elasticities as well as tax bases used to project tax revenues to reflect the changed economic structure, ensure promptness in payments of value added tax (VAT) refunds which could boost tax payment moral and compliance, and undertake a review for a revenue neutral rate under corporate income tax (CIT) to allow elimination of multiple preferential rates. Over the medium reforms could focus on reducing the revenue forgone through multiple exemptions in VAT and CIT. Second, government spending is allocated well (to infrastructure and human capital) but there is scope to improve outcomes from the use of these resources. Public expenditure has had a good impact—even better than peers (as evidenced by the improved quality of infrastructure, and improved outcomes on education and health). Nonetheless, weaknesses in the public investment management system reduce the technical efficiency of capital spending and could lead to delays in project implementation. Further, holding up payments to vendors could be delaying business to business transaction and making the cost of delivering projects very high, as contractors inbuilt risks of delayed payments into their bids. Delays in public payments can affect private sector liquidity, profits and ultimately economic growth. There is need, therefore, to clear any pending bills/arrears, and reduce costs and time over-runs for development projects. Third and Finally, ensuring efficiency and effectiveness of public spending is critical given tight fiscal space and the expenditure needs under the Big 4. Public sector resources devoted to the Big 4 would need to be contained within a fiscally sustainable resource envelope and should seek to reduce inefficiencies in spending in order to maximize impact. In the short-run the government could prioritize completion of ongoing investment/development projects, clearance of any pending bills and arrears owed to suppliers. Over the medium term, there is need to safeguard the medium-term expenditure framework (MTEF) to restore credibility of the multi-year budgeting process and improve budget execution. The world Bank remains committed to working with key stakeholders in Kenya to identify policy and implementation issues that could enhance revenue mobilization, efficiency and effectiveness of spending as well as equity concerns in fiscal management goals. This public expenditure analysis offers a platform for such a discussion and contributes to this policy debate. 4 Acknowledgements The Kenya Public Expenditure Analysis (PEA) was made possible through invaluable contributions from a talented team of staff, whom we gratefully acknowledge. The overall task of drafting and synthesizing the report was led by Peter W Chacha (Senior Economist, GMTA3) and Allen Dennis (Program Leader, AFCE2). Chapter 2 and 3 on macro-fiscal developments and the performance of government revenues, was led by Peter W Chacha. The team included Angelique Umutesi, Sarah Sanya, Patrick Nderitu, Ariel Melamud, and Juan Pradelli. Chapter 4 on trends and profile of government expenditures was led by Peter W Chacha and Angelique Umutesi. The team included Patrick Nderitu, Aaron Thegeya, and Alex Giron Gordillo. Peer reviewers for the PEA were Friederike Norma Koehler and Arvind Nair. Review comments were also gratefully received from Utz Pape, and James Origa Otieno. The team received overall guidance from Abebe Dadi (Practice Manager, Macroeconomic Trade and Investment), Philip Schuler (Lead Economist for Kenya, Rwanda, Uganda, and Eritrea), Allen Dennis (Program Leader for Kenya, Rwanda, Uganda, and Eritrea), and Felipe Jaramillo (Country Director for Kenya, Rwanda, Uganda, and Eritrea). Their guidance is greatly appreciated. They are not, however, responsible for any of the conclusions in this report. Any errors are the responsibility of the authors. The team would like to thank Anne Khatimba, Evelyn Kagwanjah and Carolyne Odicko for providing logistical and administrative support and Robert Waiharo for design and layout of the report. We are also grateful to Paul Clark for excellent editorial support. Partnership with key Kenyan policy makers was instrumental in the production of this report. We are grateful to Kenya National Bureau of Statistics (KNBS), Central Bank of Kenya (CBK), and the National Treasury for most of the data used in this report. The preliminary findings from this report was shared by the National Treasury, Kenya Revenue Authority, and the Central Bank of Kenya. 5 Executive Summary The government of Kenya remains committed to fiscal consolidation, which is expected to contain public debt stock at a sustainable level but also create space to support spending for service delivery and the Big 4 agenda. Nonetheless, the main challenge is how the government’s fiscal management goals can be achieved in the context of slowing revenue collection, a growing demand for transfers to county governments, and the need to fund the Big 4 agenda and broad public services delivery. This report provides an overview of the macroeconomic context, the sufficiency of revenue mobilization, the size and composition of the national government expenditures, and the efficiency of spending over the last five years. The analysis identifies options for improving revenue mobilization and for promoting allocative and technical efficiency in government spending to create fiscal space. The main policy messages are organized per chapter. Chapter 2: Macro-Fiscal Developments This chapter finds that the performance of the economy has been robust with real GDP growth of about 5.7 percent for the last five years (2013-2018). The macroeconomic environment remains stable (inflation is within target) and the current account deficit has narrowed (driven by lower imports) and is adequately funded by the surplus in capital and financial inflows (corporate and government debt flows). Nonetheless, growth has been state-led rather than private sector driven (as reflected in ramp-up in public sector investment and relatedly slowdown in private sector investment and limited private sector credit growth). This has pushed Kenya close to a point where fiscal sustainability would be in question, resulting in fiscal consolidation to rein in rising public debt stock and create fiscal space for the Big 4. Over the short-run to medium term horizon, fiscal policy remains adequately anchored towards achieving sustainable public to debt ratio but slowdown in revenues and emerging spending needs, presents a significant risk of fiscal slippages. Policy to crowd in private sector participation, including in the implementation of some of the Big 4 projects remains critical. Specifically, policy should strengthen the regulatory environment for the mainstreaming public private partnership (PPPs) initiatives including a feasible risk sharing mechanism, ensuring promptness in the payment of vendors to support liquidity among businesses trading with the government, and clearance of any pending bills perhaps in a phased- out approach given a tight fiscal space-which could stimulate private sector activity and create jobs. Improving debt management by putting in place a transparent and regular platform for primary issuance of debt instruments could also help. Adopting an electronic platform could improve the primary auction of government securities. This could promote transparency and enhance efficiency in the management of government debt. Adoption of this technology could, for instance, hasten the settlement period after every auction and reduce liquidity management challenges. With a growing inclination towards foreign debt, a clear communication strategy on the government’s preparedness to tackle upcoming debt repayments (interest and principal), including refinancing strategies, remains critical to sustaining market confidence. Debt management strategy could also focus on rebalancing the mix of expensive and shorter maturity commercial loans by taking advantage of available concessional debt, which tends to be more affordable. 6 Chapter 3: Trends and Performance of Government Tax Revenues Analysis in this chapter shows that Kenya’s tax revenue as a share of GDP has decreased and decoupled from growth in the economy, suggesting some important constraints to enhancing revenue collection. Tax revenue as a share of GDP has decreased from 18.1 percent of GDP in FY2013/14 to 15.7 percent in FY2017/18. The lower buoyancy of tax revenue relative to expansion in nominal GDP over the past five years could be attributed to changes in structure of the Kenyan economy, with Agriculture gaining a sharp increase relative to other sectors (industry and services). This suggests that despite recent tax policy and administrative measures to support domestic revenue mobilization, including integration of iTax with other databases (e.g. IFMIS), roll out of integrated customs management, and expansion of tax bases - all of which are positive developments - additional time may be required (through structural transformation) for the envisioned revenue increases to materialize. There is need for a step change in Kenya’s tax revenue mobilization effort to support fiscal adjustment and fund critical service delivery. In the short-run, the government could review and update elasticities as well as tax bases used to project tax revenues to reflect the changed economic structure, ensure promptness in payments of VAT refunds which could boost tax payment moral and compliance, and undertake simulations to settle for a revenue neutral rate under corporate income tax and eliminate multiple preferential rates once a suitable rate is established. Over the medium reforms could focus on reducing the revenue forgone through multiple exemptions in VAT and CIT. Revenue forgone due to VAT exemptions and zero rating is estimated at about 3.5 percent of GDP (World Bank, 2017). Similarly, revenue forgone from CIT exemptions, accelerated depreciation allowances and preferential rates are estimated at about 1.9 percent of GDP (World Bank, 2017). There is need, therefore to review VAT and zero-rating incentive schemes as well as generous deductibles and investment allowances under CIT to reduce revenue forgone. Further, for improved transparency on the costs of such exemption, prepare a repository of all tax expenditures and present an annex to the budget policy statement and/or the annual budget-showing the costs of such expenditures to the tax payers and rationale for their introduction. Chapter 4: Trends and Profile of Central Government Expenditures Government spending is allocated well (to infrastructure and human capital) but there is scope to improve outcomes from public spending. Public expenditure has had a good impact—even better than peers (as evidenced by the improved quality of infrastructure, and improved outcomes on education and health). Nonetheless, weaknesses in the public investment management system reduce the technical efficiency of capital spending and could lead to delays in project implementation. Further, holding up payments to vendors could be delaying business to business transaction and making the cost of delivering projects very high, as contractors inbuilt risks of delayed payments into their bids. Delays in public payments can affect private sector liquidity, profits and ultimately economic growth1. There is need, therefore, to clear any pending bills and reduce costs and time over-runs for ongoing development projects and enhance efficiency of public spending. 1 Checherita et al.2015 7 Further, a growing share of recurrent spending is non-discretionary, which could undermine government’s ability to re-allocate resources to priority sectors. Approximately 68 percent of the central government budget is on items of high to medium rigidity (items that cannot easily be adjusted due to high judicial, political or social costs). Spending on recurrent items such as transfers (to counties, SoEs, Parliament, and the Judiciary), interest payments, and compensation of employees presents a large portion of expenditures that is implicitly non-discretionary and undermines ability of the government to re-allocate resources to say the Big 4 priority sectors. To support the ongoing fiscal consolidation agenda, while creating space to implement the Big 4, policy could focus on the following: • Prioritize completion of ongoing investment/development projects, clearance of any pending bills and arrears owed to suppliers; • Ensure promptness in payments for government supplies to de-risk contracts and general supplies to government; • Improve budget execution and absorption of development budget by building staff capacity for project appraisal, selection, budgeting and implementation across all levels of government; • Improve the regulatory environment and put in place appropriate incentive structure (or a risk sharing mechanism) to achieve some of the flagship projects under the Big 4 through public private partnership initiatives; • Contain growth in wage bill by cleaning the payroll register (to ensure it reflects current staff establishment), ensuring that personal allowances are in the base wage, and restricting new hires to critical and technical services (i.e. teaching, security and health care); • Contain growth in transfers to counties by supporting counties with capacity to step up own source revenue mobilization; and • Improve outcomes realized relative to inputs in terms of public spending by reducing wastages, ensuring value for money in procurement, and tightening public financial management systems (to reduce opportunity for resource leakages). 8 Chapter 1 Introduction 1.1. The broad fiscal policy agenda in Kenya is to support devolution, implementation of the Big 4 agenda and rebuild resilience to shocks. The main challenge is how the government’s fiscal management goals can be achieved in the context of slowing down revenue collection, a growing demand for transfers to county governments, and the need to support implementation of the Big 4 agenda. Devolution has changed how Kenya spends its public resources, with at least 15 percent of ordinary revenue earmarked for transfers to support service delivery at the counties. At the same time the government of Kenya is implementing the Big 4 agenda2, which coincides with the Medium-Term Plan (MTP III) over the period 2018-2022. This will require more resources to support attainment of set targets over the medium term. 1.2. The Big 4 agenda set an ambitious development goals whose achievement calls for support from both the public and the private sector. The Big 4 includes food and nutritional security for all, affordable housing (at least 500 000 affordable housing units), increased share of manufacturing (from 9 percent of GDP to about 15 percent in 2022), and universal health coverage by 2022. The attainment of these goals should help advance the Vision 2030 agenda – helping Kenya to move towards a middle-income economy with a high standard of living. For the Big 4 to be realized support from both the public and the private sector will be required. Given narrowing of fiscal space and the extent of resources needed to achieve the Big 4, the public sector could play the important role of creating a conducive environment to catalyze private sector resources to achieve the Big 4. Public sector resources devoted to the Big 4 would need to be contained within a fiscally sustainable resource envelope and should seek to reduce inefficiencies in spending to maximize impact. 1.3. A lack of fiscal space could hamper Kenya’s ability to absorb economic shocks while delivering on the Big 4 agenda. Shocks could include a recurrence of adverse drought conditions like the one of 2016/17 that could negatively impact agricultural output and growth prospects. Estimates show that delivering the Big 4 agenda will require a significant re-allocation of resources to the four priorities areas, while also ensuring that enabling sectors are adequately funded. Continued wage agitations from among other nurses, teachers, doctors, members of parliament, and award of police housing allowances could add fiscal pressures and raise the wage bill. These shocks raise probability for fiscal slippages from planned fiscal consolidation, potentially adding risks to macroeconomic stability. 1.4. Kenya could consider expanding its fiscal space through enhanced revenue mobilization, reallocation from low to high priority spending across functions, and improving efficiency in spending. Tax revenue mobilization remains below Kenya’s potential (in terms of its income level) and has been declining as a share of GDP. Reversal of this trend is essential to rebuild fiscal space. In addition, the country has ramped up spending on social sectors(education) and infrastructure, while transfers to counties, salary payments to employees, and rising interest payments have made recurrent budget adjustment quite rigid. A combined policy action to enhance revenue mobilization and improve efficiency of spending could help rebuild fiscal space, enabling the public sector to contribute more effectively to the advancement of the Big 4. 2Government of Kenya has embarked on implementation of the Big 4 agenda focusing on delivering food and nutrition security, affordable housing, increased manufacturing, and universal health coverage by 2022. 9 1.5. In line with this, the public expenditure analysis (PEA) looks at options to create fiscal space and build resilience, which are essential for macroeconomic stability and catalytic to achievement of inclusive growth agenda of the government. The analysis answers three key questions, namely: (i) Is revenue mobilization sufficient? (ii) How well is spending aligned with national priorities (allocative efficiency)? and (iii) Is Kenya getting value for its money (technical efficiency)? 1.6. On the sufficiency of revenue mobilization, Kenya’s tax revenue as a share of GDP has decreased and decoupled from the growth in the economy, suggesting some important constraints to enhancing revenue collection. There is need for a step change in Kenya’s tax revenue mobilization effort. In the short-run, the government could review and update elasticities as well as tax bases used to project tax revenues to reflect the changed economic structure, ensure promptness in payments of VAT refunds and undertake simulations to settle for a revenue neutral rate under corporate income tax. Over the medium, reforms could focus on reducing the revenue forgone through multiple exemptions in VAT and CIT. 1.7. Government spending is allocated well (to infrastructure and human capital) but there is scope to improve outcomes from public spending. Public expenditure has had a good impact—even better than peers (as evidenced by the improved quality of infrastructure, and improved outcomes on education and health). Nonetheless, weaknesses in the public investment management system reduce the technical efficiency of capital spending and could lead to delays in project implementation. Moreover, holding up payments to vendors could be delaying business to business transaction and making the cost of delivering projects very high, as contractors inbuilt risks of delayed payments into their bids. Delays in public payments can affect private sector liquidity and profits and ultimately economic growth. There is need, therefore, to clear any pending bills and reduce costs and time over-runs for ongoing development projects and enhance efficiency of public spending. 1.8. Further, a growing share of recurrent spending is non-discretionary, which could undermine government’s ability to re-allocate resources to priority sectors. Approximately 68 percent of the central government budget is on items of high to medium rigidity (items that cannot easily be adjusted due to high judicial, political or social costs). Spending on recurrent items such as transfers (to counties, SoEs, Parliament, and the Judiciary), interest payments, and compensation of employees represents a large portion of expenditures that is implicitly non-discretionary and undermines ability of the government to re-allocate resources to say the Big 4 priority sectors. 1.9. Finally, ensuring efficiency and effectiveness of public spending is critical given tight fiscal space and the expenditure needs under the Big 4. Public sector resources devoted to the Big 4 would need to be contained within a fiscally sustainable resource envelope and should seek to reduce inefficiencies in spending to maximize impact. In the short-run the government could prioritize completion of ongoing investment/development projects, clearance of any pending bills and arrears owed to suppliers. Over the medium term, there is need to safeguard the medium-term expenditure framework (MTEF) to restore credibility of the multi-year budgeting process and improve budget execution. In addition, there is need for continued improvement of the regulatory environment and appropriate incentive structure (or a risk sharing mechanism) for crowding in private sector contribution into achieving some of the flagship projects under the Big 4-through PPP initiatives. 1.10. This public expenditure analysis (PEA) complements several other Bank led reports that contributes to the discussion on policy options for improving public expenditure in Kenya. In FY18, upon request by the GoK, the Bank supported the government’s Comprehensive Public Expenditure Review 10 (CPER)3 2017 by providing technical assistance to the process. That report highlighted the challenges, institutional weaknesses, as well as efficiency, equity and effectiveness of public spending over the period 2013/14-2016/17. Further, to understand equity considerations in fiscal measures (taxation and spending), the MTI and Poverty teams carried out a fiscal incidence analysis in 2018 (KEU18)4 that examined the income distributional consequences of spending and taxation using the commitment to equity methodology. Beyond these, the Agriculture Sector Expenditure Review and the Social Protection and Jobs programs PER, both deliverable in FY19 analyses the efficiency and effectiveness of spending in agriculture and the social protection sectors, respectively. 1.11. To understand the size, trends, and composition of spending at the counties, a study on “Making Devolution Work for Service Delivery� is planned for delivery in FY2020. This is a deep dive into the contribution of devolution to changes in service delivery and identifying policy options to address challenges and opportunities at the county level of government. The analysis is expected to complement the government CPER 2017, in which the “Evolution of Devolved Fiscal Governance� was covered as a full chapter. As such, this PEA excludes the analysis on county spending and efficiency issues at that level of government, only making occasional mention where county activity has implications to the central government’s expenditure. Finally, given that some of the equity concerns of fiscal measures have been recently addressed through the fiscal incidence analysis, it is equally excluded from this PEA. 1.12. A key value added of this report, therefore, is to provide an overall storyline of the challenges faced in public expenditure management in the context of a slowdown in revenue collection, tight fiscal space, and the government’s commitment to implement the Big 4 agenda. The report presents potential policy options to expand the fiscal space from both the revenue and expenditure fronts. The rest of this report is organized as follows: Chapter two discusses the recent macro-fiscal developments and implications for fiscal policy. Chapter three examine the sufficiency of government revenue mobilization, while chapter four discusses recent trends and profiles in national government expenditure. Each of the chapter contains conclusion and policy recommendations. 3 Comprehensive Public Expenditure Review (2017)-National Treasury and Planning. 4 Kenya Economic Update (KEU18). 11 Chapter 2 Macro-Fiscal Developments A. Macroeconomic context 2.1. Kenya has achieved strong economic growth contributing to expansion in GDP per capita over the last five years. The economy expanded at an average of 5.7 percent over the period 2013-2018 (Figure 2-1). Growth in real GDP was broad based, coming from agricultural output which accounted for about 1 percentage points of the average growth, industry with an average contribution to growth of about 1.2 percentage points and services sector which accounted for about 3.0 percentage points(or 52.6% of the average GDP growth) over the review period (Figure 2-2). More importantly, growth is translating into improved livelihoods as reflected by the increase in per capita GDP from US$ 1,130 in 2013 to US$ 1,620 in 2018. This places Kenya within the low middle-income group of countries. Further, the latest Kenya Household Integrated Budget Survey (KHIBS) 2015/16 shows a 10-percentage point decrease in poverty head counts from (46.1) percent in 2005/06 to (35.6) percent in 2015/16 (Figure 2-3 ). While this represents significant progress, inequality still remains high in Kenya and poverty head counts are also high compared to comparator countries such as Uganda and Ghana. These challenges call for fiscal policy measures (taxes and spending) that are pro-poor and promotes inclusive growth. Figure 2-1: The Kenyan Economy has rebounded Figure 2-2: Contribution to Real GDP growth 7 6.3 5.9 5.7 5.9 6 5.4 4.9 5 Percentage points 4 2.8 3.3 3.2 2.5 2.8 3 3.1 2 1.4 1.0 1.0 1.1 1.1 1 1.2 1.2 0.7 1.4 1.0 1.0 0 0.4 2013 2014 2015 2016 2017 2018 Agriculture Industry Services Net taxes GDP Source: KNBS and World Bank; *Notes: excludes statistical discrepancy and changes in inventory. 2.2. Private consumption accounts for most of the growth on the demand side of the economy, while the contribution of private investment remains subdued. Private consumption remains a key driver of growth, accounting for at least 78 percent of growth in the last five years (Figure 2-5). This reflects a growing middle class, favorable harvests and rising remittances (both diaspora and internal). This is contrasted by subdued private sector investment. Weak private investment is reflected in the lack of adequate credit to the private sector (since mid-2016), slow demand for industrial imports, and weak profitability by corporates. Consequently, the contribution to growth from investments by Kenya’s private sector has been falling and contracted by 2.0 percent in 2016 (Figure 2-6 ). 12 Figure 2-3 : Poverty head count against GDP per Figure 2-4 : International poverty comparison capita Source: The World Bank, KEU 2018 Figure 2-5: Contribution to real GDP growth (in Figure 2-6 : Contribution of private investment percentage points) remains weak (in percentage terms) 4 2 Percentage points 0 -2 -4 -6 2012 2013 2014 2015 2016 2017 2018e Government Investment Private Gross Fixed Investment Source: KNBS and World Bank; *Notes: excludes statistical discrepancy. 2.3. Inflation remains within the government’s target range but temporary acceleration on account of a spike in food prices was experienced over the review period. With the exception of 2017 (first three quarters), overall inflation rate has remained within the government’s target band of 5±2.5 percent (Figure 2-7 ). Overall inflation has dropped from the peak of 11.0 percent in May 2017 to about 4.1 percent in February 2019. The spike in 2017 is associated with higher food prices in that year occasioned by drought. Nonetheless core inflation, which excludes food and energy prices, has remained below mid target of 5 percent reflecting an economy where underlying demand pressures are still benign. The low inflationary pressure has also been supported by a stable local currency against the US dollar thereby serving as a nominal anchor to inflationary expectations. 13 Figure 2-7 : Inflation remains within the target Figure 2-8 : Private sector credit growth remains range subdued Source: KNBS and CBK 2.4. Private sector credit growth remains subdued after decreasing from a peak of 25.8 percent in June 2014. As of December 2018, private sector credit growth stood at 2.4 percent, well below its peak of 25.8 percent in June 2014 (Figure 2-8 ). In real terms, credit growth in Kenya is actually negative. Some of the factors behind the slowdown in credit growth in Kenya include a sharp depreciation of the Kenyan shilling in 2015, earlier bank liquidations that created uncertainty in the banking sector and imposition of interest rate caps in the last quarter of 2016. 2.5. The current account deficit has steadily narrowed over the past 5 years and remains adequately financed. In 2018, the current account deficit narrowed to 5.0 percent of GDP (from 10.4 percent of GDP in 2014) due to stronger diaspora remittance inflows, improved exports of tea and horticulture, strong receipts from tourism, and a slowdown in imports (Figure 2-9 ). The current account deficit continues to be adequately financed by resilient capital flows (government and corporate loans) resulting in a 9.3 percent increase in official foreign reserves to US$8,131 million (or 5.3 months of import cover) in 2018 relative to 2017(Figure 2-10 ). These capital flows continue to provide a comfortable buffer against external short-term shocks. 2.6. The shilling has remained generally stable over the review period. A relatively lower import bill, strong remittance inflows (Figure 2-11 ), a rebound in tourism, and government borrowing in foreign currency have continued to support a stable exchange rate for the Kenyan shilling against the US dollar (Figure 2-12 ). 14 Figure 2-9 : The current account deficit has Figure 2-10 : Government and corporate capital narrowed flows financed the current account deficit Source: The World Bank, WDI Figure 2-11 : Remittance inflows Figure 2-12 : Real and nominal exchange rate Source: The Central Bank of Kenya B. Overall fiscal balance, financing and stock of public debt 2.7. The overall fiscal deficit widened in the last five years due to expenditure growth and revenue shortfalls. The overall fiscal deficit was an average of 8.1 percent of GDP during the review period. In particular the deficit widened from 6.1 percent of GDP in FY2013/14 to 8.8 percent in FY2016/17 before easing to 6.3 percent of GDP in FY2017/18 (Figure 2-13 ). At least 45.2 percent of the deficit was funded through net domestic borrowing, while 46.4 percent was financed through the net external borrowing, with the balance being funded through other financing. The implication is that as Kenya continues to rely more on net external funding sources, this has also raised vulnerability to developments in the international markets and exchange rate risks. 15 Figure 2-13 : Overall fiscal balance was Figure 2-14 : Fiscal balance is wider relative to expansionary over the review period (% of GDP) EAC peers (% of GDP) Source: World Bank based on National Treasury 2.8. Reflecting government’s recent commitment to fiscal consolidation, the overall fiscal deficit decreased for a second year. The overall fiscal deficit (including grants) was reduced to 6.8 percent in FY2017/18 from 8.8 percent of GDP in FY2016/17, surpassing the targeted budget deficit of (7.2) percent of GDP and is projected to be reduced further to 6.3 percent in FY2018/19. Notwithstanding progress in consolidation, Kenya’s fiscal deficit is elevated relative to EAC peers (Figure 2-14 ). 2.9. Looking simply at the year on year changes in the overall fiscal balance, without accounting for the business cycles, may lead to a misleading conclusion on the orientation of fiscal policy 5. A solution in the literature is to obtain a cyclically adjusted fiscal balance and calculate the structural fiscal balance by taking the difference between the unadjusted fiscal balance and the cyclically adjusted fiscal balance6. The year on year changes in the structural balance yields the fiscal impulse. A positive impulse suggests a loose fiscal stance, while a negative one implies a tighter stance. Figure 2-15 presents Kenya’s fiscal impulse together with the output gap. Since the trends in the fiscal impulse and the output gap time moves in the same direction, Kenya’s fiscal stance has been procyclical over the review period. 5 See Van der Nord (2000) and Girouard & Andre (2005). 6 We are aware of the challenge of estimating the potential output in African countries and this result should be interpreted as only indicative. 16 Figure 2-15 : Fiscal Impulse and the output gap for Kenya (% of GDP) Source: World Bank based on National Treasury data 2.10. Over the medium term, there is need to ensure fiscal policy remain countercyclical rather than procyclical. This will make it possible for the government to spend proactively during a recession to help in the recovery of the economy and create jobs, while cautious to unwind this stimulus once the output gap closes. Thus, efforts to monitor the performance of the economy should be enhanced and to ensure that once it is operating at full potential, the windfall in revenue collection is saved for a rainy day and to create space for future countercyclical policy. 2.11. With expansionary fiscal policy, Kenya’s public and publicly guaranteed debt grew rapidly over the review period, but ongoing fiscal consolidation has halted this trend. Overall public debt rose from about (47.8) percent of GDP in FY2013/14 to (57.6) percent of GDP in FY2016/17 before stabilizing at (57.2) percent of GDP in FY2017/18 (Figure 2-16 ). This is partly attributed to a narrowing of the fiscal deficit in FY2017/18, but also due to growth in GDP and a relatively stable exchange rate. The drop in primary deficit from an average of 5.0 percent of GDP in FY2015/16 to an average of about 3.0 percent in FY2017/18 (Figure 2-17 ) slowed the pace of debt accumulation. However, interest payments’ contribution to debt stock increased from an average of 2.9 percent of GDP in FY2015/16 to an average of 3.4 percentage points of GDP over the FY2017/18 period. 17 Figure 2-16 : Public debt has stabilized after Figure 2-17 : Pubic debt moderation is driven a rapid rise in previous years (% of GDP) by a decrease in the primary balance 80 12% 60 57.6 57.2 56.5 55.5 8% Percent of GDP 48.8 47.8 42.1 4.2% 40 4% 3.1% 2.2% 2.6% 1.5% 1.4% 0% -0.6% 20 -4% 0 2012/13 2014/15 2016/17 2018/19e -8% 2012 2013 2014 2015 2016 2017 2018 Domestic External Total public debt (Gross) Primary bal. Seignorage Growth effect Interest Revaluation Residual Change in debt_t Source: World Bank based on National Treasury data 2.1 The accumulation of total public debt included both external and domestic components, as government borrowed widely to finance large infrastructure projects . The split between external and domestic debt in the total debt stock is about 50:50. However, reflecting higher domestic interest rates, debt servicing charges on the domestic debt stock are about three times higher than from the external debt stock. Kenya continues to access international markets to refinance its external debt. For example, in February 2018 it successfully issued a US$2 billion Eurobond (US$1 billion for 10 years and US$1 billion for 30 years at 7.25 and 8.25 percent, respectively) and in May 2019 it issued US$2.1 billion (US$900 million for 7 years and US$1.2 billion for 12 years at 7.0 and 8.0 percent respectively). The proceeds from these recent issuances are expected to help refinance upcoming bullet payments on external debt obligations. However, vulnerability to developments in international markets and exchange rate volatility have also increased while the cost of interest rate payments is likely to be large. 2.12. As at the end of 2018, external public debt stock comprised primarily loans from multilateral lenders (34 percent), bilateral partners (31 percent) and commercial creditors (34 percent). In terms of the creditors, of the multilateral debt, IDA remains the largest creditor accounting for 66 percent, followed by the ADB/ADF, whose debt accounts for 23 percent. Of bilateral debt, China accounts for 68 percent and with regard to commercial creditors, fixed coupon debt reached 55 percent, while the commercial banks were at 45 percent (Figure 2-18 ). Domestic debt comprises mainly of Treasury bills and bonds (36 percent). The short-term debt with maturities of 91, 182, and 364 days is approximately US$ 8.8 billion which account for 36.2 percent of domestic debt while the medium- and longer-term domestic debt, primarily Treasury bonds, is equivalent to 62.5 percent of domestic debt (Figure 2-19 ). 18 Figure 2-18 : External Public Debt by Figure 2-19 : Domestic Public Debt by instrument instrument at end-December 2018 (% of total) at end-December 2018 (% of total) Source: World Bank based on National Treasury 2.13. The total government debt portfolio is diversified across various currencies. The main exposure is to the U.S. dollar at 71.3 percent, followed by the Euro at 15.4 percent, with the Chinese Yuan and Japanese Yen (JPY) at 6.1 percent and 4.3 percent, respectively (Figure 2-20 ). While the government is working towards reducing its exposure to the US dollar through contracting new debt in other currencies, and by matching external liabilities with the currency composition of Kenya’s forex inflows and international reserves, there remain significant exposure to the Ksh/U$S exchange rate fluctuation. A significant depreciation of the local currency to the US dollar could exert immense fiscal pressures from drastic shift in debt service profile. Figure 2-20 : External Public Debt, by currency at Table 2-1: Domestic Public Debt, by maturity at end-December 2018 (% of total) end-December 2018 (% of total/billions of US$) Remaining Maturity In Billion in Years In Percent of Total US$ <1Y 43% 10,376 between 2-3Y 10% 2,419 4-5Y 13% 3,225 6 - 10 Y 21% 5,118 > 15 Y 13% 3,168 Total 100% 24,306 Source: World Bank based on National Treasury 2.14. The bunching of upcoming repayments together with shortening maturities of debt could impose fiscal pressures on account of debt service obligations. While 2019 will see significant debt redemptions due to upcoming principal repayments on short-term domestic debt, it is projected that over the medium term the level of redemptions will decline (Figure 2-21 ). Domestic debt maturing in less than a year is 43 percent of total domestic debt (Table 2-1). Higher external debt payments for 2024 and 2028 correspond to principal payments of the Eurobond debt. Thus, the bunching of upcoming repayments together with shortening maturities of debt could impose fiscal pressures on account of debt service 19 obligations and has raised Kenya’s risk of external debt distress from low to moderate7. Figure 1: Projected Debt Redemption Profile over the medium to long term Figure (In 2-21 : Projected millions Debt Redemption of Kenyan over Shillings) the medium to long term (in Ksh.million) Source: Source:World Bank, based National on National and Treasury Treasury and CBK Central Bank of Kenya 27. The total portfolio has 91.9 per cent of debt at fixed interest rate debt while Fiscal C.the and total Debt Sustainability domestic debt is 100 risks percent fixed interest rate. Out of the total debt, 32.4 per cent will re-fix in FY2019/20. The average time to re-fixing for prudent 2.15. For (ATR) external fiscal management and domestic debtthe of risks, PFM Act 2012 portfolio is 9.7 requires years theand preparation of a 4.7 years respectively. statement of specific fiscal risks. This is simply an assessment of fiscal risks that the Kenyan economy is exposed to that may affect the achievement of the macroeconomic targets and objectives detailed in its medium-term budget policy statement. The fiscal risks arise from assumptions that underlie fiscal projections, the dynamics of public debt, state corporation operations, contingent liabilities, financial sector vulnerabilities and natural disasters. This section analyses fiscal and debt sustainability risks over the medium term under a baseline scenario, business as usual scenario, lower growth scenario, and a high domestic borrowing costs scenario. 8 2.16. The baseline scenario replicates macro, fiscal, and debt forecasts reported in the February 2019 Budget Policy Statement until FY2022/23. This assumes current fiscal policies by the government would continue as planned over the medium term. The forecast is extended until 2024 using constant ratios to nominal GDP in 2023-2024. Information on public and publicly guaranteed (PPG) debt is taken as at end- 2018 and associated amortization and interest payments is consistent with the MTDS March 2019. 2.17. The Gross Borrowing Requirements (GBR) result from four elements: budget deficits, amortizations of old and new debts, other funding needs (e.g., contingent liabilities), and accumulation of financial assets (e.g., Government deposits). It is assumed that GBR are met with: • short-term domestic debt (1-year securities carrying a 12% interest rate and denominated in KSH) covering 50% of GBR (NB: This assumption reflects a permanent rollover of the nominal amount of short-term liabilities identified in the MTDS March 2019); • medium-term domestic debt (5-year securities carrying a 12% interest rate and denominated in KSH) covering 30% of GBR (NB: This assumption basically reflects using this debt instrument 7 According to the latest Debt Sustainability Analysis (DSA, October 2018), the debt sustainability indicators show that Kenya faces a moderate risk of debt distress. 20 to fund the budget deficit—excluding 40% of the CAPEX that is assumed to be funded with external loans—and the amortizations of old and new medium-term domestic debts); and • long-term external loans (40-year loans with 5-year grace period, carrying a 2% interest rate and denominated in USD) covering 20% of GBR (NB: This assumption implies a positive net flow of external financing—i.e., new disbursements associated to CAPEX project execution exceed the amortizations of old and new foreign loans). Table 2-2: Medium-term baseline scenario 2019-2024(in % of GDP unless stated otherwise) Historical Projections 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 GDP at current prices (KSH billion) 7,194 7,749 8,252 9,411 10,688 12,071 13,636 15,434 17,407 19,632 GDP at constant prices, annual growth rate (%) 5.7 5.9 4.9 6.3 6.2 6.3 6.6 6.9 6.5 6.5 GDP deflator, annual growth rate (%) 10.0 8.1 8.4 8.0 7.0 6.2 6.0 5.9 5.9 5.9 Exchange Rate KSH/USD 102.3 102.5 103.2 101.3 -- -- -- -- -- -- Real Exchange Rate, Index 2015=100 (1) 100.0 93.8 88.5 82.3 -- -- -- -- -- -- Revenue 16.7 17.5 18.3 18.2 18.8 18.9 19.0 19.1 19.1 19.1 Direct Taxes (Tax on Income & Profits) 7.4 7.7 7.7 7.5 7.8 7.9 8.1 8.3 8.3 8.3 Indirect Taxes (Taxes on G&S and Int'l Trade) 6.7 7.1 7.3 7.4 7.7 7.8 7.9 7.9 7.9 7.9 Grants (2) 0.4 0.4 0.3 0.4 0.5 0.4 0.4 0.4 0.4 0.4 Other Revenue (3) 2.2 2.4 2.9 2.9 2.9 2.8 2.7 2.6 2.6 2.6 Expenditure 23.4 25.1 25.8 24.8 24.4 23.5 22.7 22.5 22.3 22.2 Wages and Salaries 4.2 4.2 4.4 4.3 4.2 4.0 3.7 3.4 3.4 3.4 Goods and Services 6.1 6.9 7.2 7.1 6.9 6.8 6.9 7.3 7.3 7.3 Transfers to Counties 3.7 4.1 3.8 3.7 3.4 3.1 2.8 2.5 2.5 2.5 Other Expenses 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Interest Payments 2.7 3.2 3.6 3.8 3.8 3.8 3.6 3.5 3.4 3.3 Development Expenditure (4) 6.7 6.9 6.7 5.8 6.1 5.8 5.7 5.7 5.7 5.7 Primary Balance -4.0 -4.5 -3.9 -2.8 -1.9 -0.8 0.0 0.2 0.2 0.2 Overall Balance -6.7 -7.6 -7.5 -6.5 -5.6 -4.6 -3.7 -3.3 -3.2 -3.1 Gross Borrowing Requirements 17.9 19.3 20.4 19.3 17.7 13.7 11.7 10.7 9.5 12.9 Overall Balance (+ indicates deficit) 6.7 7.6 7.5 6.5 5.6 4.6 3.7 3.3 3.2 3.1 Amortizations Payments 9.8 10.4 12.7 13.0 13.1 9.8 7.8 7.2 6.1 9.6 Other Funding Needs 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Net Acquisition of Fin.Assets (e.g., Deposits) (5) 1.4 1.2 0.2 -0.3 -1.1 -0.7 0.2 0.2 0.2 0.2 Borrowing Sources 17.9 19.3 20.4 19.3 17.7 13.7 11.7 10.7 9.5 12.9 Issuance of Domestic Debt (% share) (6) 80.0 80.0 80.0 80.0 80.0 80.0 80.0 Issuance of External Debt (% share) 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Public Debt 44.9 49.4 55.5 56.1 55.6 54.0 52.2 50.1 48.3 46.5 Domestic Debt 28.7 29.5 28.5 27.7 26.8 26.0 25.7 External Debt 27.3 26.1 25.5 24.4 23.3 22.3 20.8 Notes: (1) Real exchange rate defined as the KSH/USD bilateral exchange rate times the ratio between int'l prices and domestic GDP deflator. (2) Includes program grants, project grants, and other grants. (3) Includes … (4) Includes net lending. (5) Excludes net lending. (6) Includes domestic debt issued to rollover maturing short-term liabilities. Source: World Bank projections. 2.18. Under the baseline scenario, the macroeconomic outlook reflects robust economic expansion and stable macro environment. Real GDP is expected to ease slightly from 6.3 percent in 2018 to 6.2 percent in 2019, before picking up to 6.9 percent in 2022. Prudent fiscal and monetary policies are expected to contain inflation within the government’s target of 5+2.5 percent and the GDP deflator is projected to decrease from 8.0 in 2018 to about 6.0 percent over the medium term. Revenue are expected to increase from 18.2 percent in 2018 to stabilize at about 19.1 percent in 2021, while total expenditures are projected to decrease from 24.8 percent of GDP in 2018 to about 22.2 percent of GDP in 2024. Consequently, the overall fiscal deficit is projected to decrease from 6.5 percent of GDP in 2018 to about 3.1 percent of GDP in 2024. 2.19. With the ongoing fiscal consolidation, the public debt to GDP ratio is expected to decline over the medium term. The baseline reflects the government’s commitment to fiscal adjustment, which if 21 implemented successfully could halt the rapid acceleration of debt to GDP level and maintain its sustainability. The level of public debt is expected to decrease from 56.1 percent in 2018 to about 46.5 percent of GDP in 2024(Table 2-2). This decrease is primarily driven by fiscal consolidation (as reflected by improved primary deficit) and strong real GDP growth (Figure 2-22). Figure 2-22: Drivers of Debt to GDP (% of GDP and percentage points) over the medium term 4 60 Primary Deficit (p.p.) 2 50 Other Funding Needs & Net Acquisition of Fin.Assets (p.p.) Valuation Effect Due To 40 Depreciation Against USD (p.p.) 0 Real GDP Growth (p.p.) 30 -2 Real Interest Rate (p.p.) 20 Other Non-Identified Factors (p.p.) -4 10 Change in Public Debt-to-GDP Ratio (p.p.) -6 0 Public Debt (% of GDP, right) 2018 2019 2020 2021 2022 2023 2024 Source: World Bank projections Shocks Scenarios 2.20. The decline in debt-to GDP under the baseline scenario is predicated upon robust growth in the economy and lack of fiscal slippages-that could potentially raise the costs of borrowing in the domestic market. An obvious risk scenario is to impose shocks on these key determinants, as well as a business as usual scenario. Risk Scenario 1: Business as usual Scenario 2.21. Business-as-usual (BAU) Scenario assumes no major policy changes relative to current situation, and thus keeps revenue and primary expenditures stable relative to nominal GDP at the ratios observed in 2018—with the exception of interest payments. Allowing for possibility of increased expenditure on interest payments is relevant for Kenya given the likely high refinancing costs in the event of a more than anticipated tightening of global financial markets (the need to refinance upcoming bullet repayments on its Eurobonds). In addition, any further fiscal slippages would necessitate a recourse to increased domestic borrowing, whose interest obligation has been rising more rapidly. 2.22. Under BAU scenario, the public debt to GDP ratio could remain high into the medium term, with fiscal policies failing to dent the ratio to ensure sustainability. Interest payment is increasing as a ratio of GDP and this could accelerate unless debt management tools are deployed to reverse this. Revenue are expected to stabilize at 18.2 percent (i.e. the 2018 level), while total expenditures are projected to increase from 24.8 percent of GDP in 2018 to about 25.1 percent of GDP in 2024 (Table A2-1 in the annex). The main driver of increased spending is the interest payments. As a result, the overall fiscal deficit is projected to widen from 6.5 percent of GDP in 2018 to about 6.9 percent of GDP in 2024. Similarly, public debt could increase from 56.1 percent in 2018 to about 60.3 percent of GDP in 2024 (Figure 2-23). 22 Improved debt management could help reduce the interest rate obligations through a plan to retire short- term and expensive commercial debt. One way to achieve this, is by raising the share of concessional borrowing relative to non-concessional debt in future financing of the deficit, even as the government stays the course in its announced fiscal consolidation pathway. Figure 2-23: Business as usual scenario-Fiscal deficit and public debt (% of GDP) Fiscal Deficit 8 Fiscal Deficit - Baseline 60 Public Debt (right) WBG Projections Public Debt (right) - Baseline 6 55 4 50 2 45 0 40 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 Source: World Bank projections Risk Scenario 2: Lower growth scenario 2.23. Lower growth scenario assumes real GDP growth will mimic historical average of 5.5 percent (2009-2018). This scenario reflects frequent shocks (especially from drought), which thwart efforts to accelerate economic growth in Kenya but also over-estimation of real GDP growth and tax revenues. Lower GDP growth and revenue would result into a relatively slower fiscal adjustment, particularly in the context of tight fiscal space and spending needs under the Big 4 (Table A2-2 in the Annex). As a result, the overall fiscal deficit is projected to decrease gradually from 6.5 percent of GDP in 2018 to about 4.3 percent of GDP in 2024. The level of public debt would decrease from 56.1 percent in 2018 to about 51.6 percent of GDP in 2024 (Figure 2-24). 23 Figure 2-24: Lower growth scenario-Fiscal deficit and public debt (% of GDP) Fiscal Deficit 8 60 Fiscal Deficit - Baseline WBG Projections Public Debt (right) Public Debt (right) - Baseline 6 55 4 50 2 45 0 40 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 Source: World Bank projections Risk Scenario 3: High domestic borrowing costs scenario 2.24. Higher domestic borrowing costs scenario introduces higher than anticipated interest rates on new government securities. We assume an increase in interest rate on new domestic debt (short-term and medium-term government securities) to 15% (compared against 12% in the Baseline Scenario). Given the sizable gross borrowing requirements covered by domestic sources of funds, an increase in interest rates can increase the interest bill and absorb a large share of the budget (Table A2-3 in the Annex). The overall fiscal deficit is projected to decrease from 6.5 percent of GDP in 2018 to about 3.8 percent of GDP in 2024, while the level of public debt would decrease from 56.1 percent in 2018 to about 48.8 percent of GDP in 2024 (Figure 2-25). 24 Figure 2-25: Higher domestic borrowing costs scenario (% of GDP) Fiscal Deficit 8 Fiscal Deficit - Baseline 60 WBG Projections Public Debt (right) Public Debt (right) - Baseline 6 55 4 50 2 45 0 40 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 Source: World Bank projections D. Conclusion and Recommendations 2.25. To conclude, this chapter provides an update on the recent economic developments and the macroeconomic context for which the general government’s spending and revenues have evolved over the last five years. It also examines the government’s fiscal deficit and the financing of the same and analyzes dynamics in public debt stock. Finally, using government’s own macro and fiscal projections as the baseline, the chapter examines fiscal and debt sustainability risks over the medium term underlining the importance of maintaining fiscal discipline and policy options to support this. The main messages are as follows: 2.26. First, the performance of the economy has been robust despite a slight deceleration in 2017 with real GDP growth of about 5.7 percent for the last five years (2013-2018). The macroeconomic environment remains stable (inflation is within target) and the current account deficit has narrowed (driven by lower imports) and is adequately funded by the surplus in capital and financial inflows. Nonetheless, growth has been state-led rather than private sector driven (as reflected in ramp-up in public 25 sector investment and relatedly slow down private sector investment and limited private sector credit growth). This has pushed Kenya close to a point where fiscal sustainability would be in question, resulting in fiscal consolidation to rein in rising public debt stock and create fiscal space for the Big 4. 2.27. Second, the authorities should stay the course in implementation of the announced fiscal adjustment plan and reduce chances for fiscal slippages. The baseline reflects the government’s commitment to fiscal adjustment, which if implemented successfully could halt the rapid acceleration of public debt to GDP and maintain its sustainability. This is expected to be achieved primarily through fiscal consolidation (as reflected by improved primary deficit) and strong real GDP growth. Enhanced revenue mobilization and ensuring efficiency and effectiveness in public spending will continue to be key fronts to sustainable fiscal operations. Alternative scenarios assuming lower real GDP growth and higher domestic borrowing costs represents a significant risk to achieving fiscal targets of reducing overall deficit to 3.3 percent of GDP by 2022 and keeping public debt at a sustainable pathway. 2.28. Third, crowding in private sector participation, including in the implementation of some of the Big 4 projects remains critical. The ongoing fiscal adjustment plan is counteracted by the need to fund the Big 4 agenda. There is a significant risk of fiscal slippage if the funding of these projects is to be undertaken using mainly public finances. Not only won’t there be enough public money in light of other core spending needs (education, infrastructure, debt service, and county transfers) but also some of the investments could be misdirected and end up being wasteful. Commercial activities should be left to the private sector, with government providing supportive environment. Specifically, policy could focus in strengthening the regulatory environment for the mainstreaming of PPPs including a feasible risk sharing mechanism, ensuring promptness in the payment of vendors to support liquidity among businesses that trade with the government, and clearance of any pending bills to restore liquidity, stimulate private sector activity and create jobs. 2.29. Fourth and finally, improve debt management by putting in place a transparent and regular platform for primary issuance of debt instruments and effective communication strategy. Adopting an electronic platform could improve the primary auction of government securities. This could promote transparency and enhance efficiency in the management of government debt. Adoption of this technology could, for instance, hasten the settlement period after every auction and reduce liquidity management challenges. With a growing inclination towards foreign debt, a clear communication strategy on the government’s preparedness to tackle upcoming debt repayments (interest and principal), including refinancing strategies, remains critical to sustaining market confidence. Debt management strategy could also focus on rebalancing the mix of expensive and shorter maturity commercial loans by taking advantage of available concessional debt, which tends to be more affordable. 26 Chapter 3 Trends and Performance of Government Revenues 3.1. This section reviews the trends and performance of general government revenues over the last five years. Improvement in Kenya’s domestic revenue mobilization (DRM) is key to supporting ongoing fiscal consolidation and funding public service delivery. Kenya’s DRM strategy aims at widening the tax base, strengthening tax administration and redesigning tax instruments. Preliminary results from the reform process remain positive with revenue targets under recent revenue enhancement initiative (REI) broadly achieved, although the overall revenue collection target still falls behind. This suggests that improvement and reversal of the downward trends in revenue buoyancy with respect to nominal GDP could take some time. The section analyses the revenue profiles in light of recent tax policy and administrative reforms and highlights potential short-term and medium-term priorities for enhancing revenue collection and improving tax system efficiency. A. Is the Government mobilizing enough Tax Revenue? 3.2. Tax revenues account for most of the government’s total revenue and their stable contribution is key to anchoring spending decisions. Having a stable contribution from tax revenues is essential for anchoring spending decisions against expected tax receipts. Tax revenues accounts for approximately (85) percent of the government’s total revenue over the last five years, while non-tax revenue (investment income, traffic and land revenues, fines and forfeitures, road maintenance levy, railway and petroleum development levies) account for the balance (about 15 percent). This then suggests that the performance in tax revenue (relative to GDP) is expected to have a strong correlation with performance in economic activity, which has been robust as pointed out in the previous section. 3.3. Nonetheless, the decoupling of tax revenue performance from GDP growth suggests some important constraints to enhancing revenue collection. Tax revenue as a share of GDP has decreased from (18.1) percent of GDP in FY2013/14 to (15.7) percent in FY2017/18 (Figure 3-1). More concerning, however, is the apparent low buoyancy of tax revenue relative to expansion in nominal GDP over the past five years. This has been attributed to changes in structure of the Kenyan economy, with Agriculture gaining a sharp increase relative to industry and services. The contribution of agriculture 8to tax revenue is about 2 percent, while over 98 percent is attributed to industry and services sectors. This may suggest that despite recent tax policy and administrative measures to support domestic revenue mobilization, including integration of iTax with other databases (IFMIS, and NHIF), roll out of integrated customs management, and expansion of tax bases - all of which are positive developments - additional time may be required (through structural transformation) for the envisioned revenue increases to materialize. 3.4. The decrease in tax revenue as a share of GDP reflects mainly challenges in the collection of income taxes and value added taxes (VAT). The slowdown in revenue is broad based, coming from both direct taxes (income tax) and indirect taxes (import duty, excise taxes, and value added taxes (VAT)). However, income tax and VAT accounted for 84 percent of the decline in tax revenue. Income tax decreased by 1.6 percentage points of GDP from 8.9 percent in FY2013/14 to 7.3 percent in FY2017/18, while VAT shaved off 0.5 percentage points to 4.1 percent of GDP in FY2017/18 (Figure 3-2). In addition, import duty contributed to a 0.2 percentage points decline in tax-revenue to GDP over the two periods. 8 Tax effort is low in jurisdictions with relatively high share of agriculture in the total economy-as is characterized by very high level of informality. 27 Excise taxes and other revenue9 detracted jointly at least 0.2 percentage points of GDP from tax revenue between FY2013/14 to FY2017/18. Figure 3-1: Tax revenue collection across main tax Figure 3-2: Underperformance driven by challenges items is falling in both Income taxes and VAT 19 18 0.1 17 1.6 Percent of GDP 16 0.5 0.2 0.1 15 14 18.1 15.6 13 12 11 10 2013/14 Excise Income tax VAT Import Other 2017/18 Duty Duty Revenue Negative Postive Source: World Bank based on National Treasury 3.5. Within income tax-there are challenges in personal income taxes, corporate income taxes, and capital gains taxes. The government introduced significant changes in the PIT rate structure and other personal tax provisions in FY2016/17 and FY2017/18 (of which KRA estimated a revenue loss of about 8.5 billion or 0.08% of GDP). The fall in the banking sector profits in FY2017/18 (due to among others introduction of interest caps), increased claims for investment deduction and accelerated depreciation on account of wear and tear, are some of the factors explaining slowdown in CIT. Estimates suggest that tax exemptions and differentiated tax rates led to forgone CIT revenue of 1.6-1.7 percent of GDP in 2015 (World Bank, 2017). 3.6. Tax revenue from income taxes could be increased significantly by limiting discretionary and non-discretionary tax exemptions, and by aligning growth in the tax base to the underlying development in economic growth. While the mechanics of Kenya’s Corporate Income Tax (CIT) is reflective of its economic structure, the pervasiveness of tax incentives could be affecting its buoyancy. Authorities might consider various tax policy and tax administrative measures, including rationalization and better management of the awarding and monitoring of CIT exemptions, to ensure they are meeting their objectives and safeguarding the CIT tax base. Thus, the most challenging revenue policy issue going forward may entail raising PIT collection and progressivity, simplifying the corporate tax code (with the new Income Tax Act), and closing remaining loopholes related to investment allowances and capital gains taxes. B. How close is Kenya’s revenue mobilization comparable to its peer countries ? 9Non-tax revenues comprise of investment income, traffic and land revenue, fines and forfeitures, other miscellaneous receipts, road maintenance levy, railway and petroleum development levies, among others. 28 3.7. There is potential for improvement in Kenya’s tax revenue mobilization as it is currently below revenue levels attained by countries at similar incomes (in terms of GDP per capita) . Kenya’s revenue mobilization is lower compared to a ratio of tax revenue to GDP of between 20-23 percent for countries at similar levels of income (Figure 3-3 and Figure 3-4). Vietnam and South Africa have a higher revenue mobilization of about 19.1 and 26.7 percent of GDP respectively. A revenue mobilization effort to the level reported for middle-income countries remains essential to create the needed fiscal space for adjustment to shocks while maintaining a growth friendly expenditure profile to achieve Kenya’s development objectives. 3.8. Tax revenue to GDP for Kenya is higher relative to regional peers (Rwanda, Uganda, Tanzania, Ethiopia, and Ghana) but is lower than of aspirational peers (Vietnam, Thailand, India, and South Africa). Over the period 2000-2017, Kenya’s tax revenue to GDP has remained above that of regional peers but below that of middle-income aspirational peers (Figure 3-5). It is also clear that performance in total tax revenue tracks closely with developments in taxes on incomes, profits, and capital gains. In particular, the slowdown after 2014 is clearly reflected in the slow-down in income taxes and in collection of taxes on goods and services. On the other hand, taxes on international trade and transactions has been relatively stable as a share of GDP (Figure 3-6). Figure 3-3: Tax revenue vs. GDP per Capita, selected Figure 3-4: Tax revenue relative to selected countries countries (2013-2017) Tax/GDP vs. GDP per capita(2013-2017) Tax/GDP, selected countries(2013-2017) 40 30 26.8 Tax revenue in percent of GDP 30 ZAF 20 19.1 % of GDP 20 16.3 15.8 VNM KEN 13.9 THA 13.2 RWA 11.9 UGA 11.6 TZA LKA 10 10 ETH BGD 8.8 8.7 0 0 2000 4000 6000 8000 10000 GDP per capita (constant 2010 US$) 0 Data Source: World Bank WDI ZAF VNM KEN THA RWA UGA TZA LKA ETH BGD Source: The World Bank, WDI 29 Figure 3-5: Tax revenue (in % of GDP) Kenya Figure 3-6: Tax revenue (in % of GDP) tracks closely and peer countries what happens to taxes on incomes. 20 20 15 10 15 5 10 0 2000 2003 2005 2007 2009 2011 2013 2015 2017 2000 2003 2005 2007 2009 2011 2013 2015 2017 year Kenya Regional Peers** tax revenue taxes on goods and services Aspirational Peers* taxes on incomes taxes on international trade Source: WEO, IMF, * aspirational peers: Vietnam, Thailand, India and South Africa. ** regional peers: Tanzania, Uganda, Rwanda, Ethiopia and Ghana. 3.9. Income tax accounts for about 40 percent of government revenue, underscoring its importance in revenue efforts to fund government expenditures. The performance in income tax depends on income tax from individuals (PAYE) and CIT. Revenue collection from the latter decreased from 3.9 percent in FY2014/15 to 3.2 percent of GDP in FY2017/18, largely on account of generous tax exemptions and differentiated CIT rates (World Bank, 2017), slowed profitability in the banking sector and investment deductibles. Nonetheless, Kenya collects about 8.2 percent of GDP (over 2013-17) in income taxes, which places it second to South Africa amongst its regional and income level comparators (Figure 3-7). 3.10. Taxes on goods and services comprising of excise taxes and value added taxes account for one half of tax revenues in Kenya. The ratio of VAT revenue to GDP for Kenya decreased from 4.5 percent of GDP in FY2014/15 to 4.1 percent in FY2017/18. The slowdown is partly attributed to proliferation of exemptions and compliance challenges resulting in foregone VAT tax revenues of about 3.6-3.7 percent of GDP (World Bank, 2017). Excise taxes have been broadly stable due to some of excises being specific in nature and therefore inelastic but also leakages in lieu of fake excise stamps. Comparing Kenya’s revenue collection through VAT and excise taxes, the country has on average (over the period 2013-2017) lagged the performance levels in Rwanda and Uganda, as well as levels attained in aspirational economies such as South Africa, Vietnam, and Thailand (Figure 3-8). 30 Figure 3-7: Taxes on incomes, Kenya and Figure 3-8: Taxes on goods and services, Kenya selected countries (2013-17) and selected countries (2013-17 14.8 10.3 15 10 8.8 8.5 8 7.3 7.2 10 6.5 6.4 % of GDP % of GDP 8.2 6 7.7 6.5 5.6 4 3.3 5 4.5 2.2 2.7 2 2.2 2.1 0 0 ZAF KEN VNM THA RWA UGA BGD ETH LKA ZAF VNM THA RWA UGA KEN LKA BGD ETH Source: The World Bank, WDI C. What explains the decoupling of tax revenue from their tax bases? 3.11. Comparing real tax revenue growth for Kenya’s main tax heads (income, VAT, excise, and import duty) relative to their respective tax bases indicates a weak correlation. For example, the real growth rate of income tax is lower than the growth in real GDP and historically the difference has been large (Figure 3-9a). This could suggest that growth in Kenya’s economy could be emanating from sectors that are hard to net in in terms of income tax (agriculture and informal sectors), which is in line with the hypothesis that structural change in Kenya’s economy in favor of agriculture has led to erosion of the effective tax base. The growth in real VAT is quite volatile relative to its tax base (Final consumption) reflecting the effect of numerous discretionary changes (in exemptions and zero-rating) (Figure 3-9b). The growth in excise and import duties appears to closely trace the growth in their tax bases. The weak relationship between tax revenues and their tax bases (especially incomes and VAT) reflects the challenges of attaining realistic revenue projections and a regular update of elasticities (at least after every 5 years) is needed to capture changes in the structure of the economy10. 3.12. Improved tax revenue mobilization is dependent on strong economic growth, stable tax policy, restricted tax expenditures and other sector specific issues. For instance, tax collection from customs depends on the demand for imported oil (wet volume), and merchandise imports (dry volume) by the Kenyan economy. A decline in import volumes is associated with a decrease in customs revenue. In real terms, both VAT on imports and duty on imports have decreased after a peak in 2014 (Figure 3-9a). This trend is also reflected as a share of GDP, where both import duty and VAT on imports are falling (Figure 3-9b). This period coincides with relatively weak private sector investment, as growth was driven largely by public investment in physical infrastructure (some of which could have benefited from tax exemptions). A re-ignition of private sector investment (including tackling issues of access to credit) remains key in restoring aggregate demand and customs revenue. Recent administrative measures for enhanced 10 The World Bank is providing a revenue modelling and forecasting Technical Assistance (TA) to the authorities to improve realism in revenue projection. 31 scanning to detect concealment, pre-export verification of conformity utilization, as well as improved cargo evacuation are likely to contribute to improved customs collection. Figure 3-9: Growth in real tax revenue vs. growth in the real tax base a) Growth in real income tax revenue b) Growth in real VAT c) Growth in real excise taxes d) Growth in real import duty Source: World Bank based on National Treasury and KRA 3.13. Tax revenue from income taxes (PAYE and other incomes) are falling in real terms and as a share of GDP. Tax collection from these tax heads depends on wage employment (PAYE), profitability of firms (CIT), and final consumption (VAT and excise domestic). Starting in 2013/14 both PAYE and other income taxes have decreased both in real terms and as a share of GDP ( Figure 3-10 c,d). This reflects the fact that wage employment may not be increasing at a faster rate or that increases in PAYE (and wages themselves) are much weaker mainly due to significant changes in the PIT rate structure and other personal tax provisions in FY2016/17 and FY2017/18. The decline in other income taxes reflects lower firm profitability, especially for the banking sector following the introduction of interest rate caps in FY2017/18. Nonetheless, there is a general lack of access to firm and sectoral data to examine in detail all the sectors affected by lower profitability and hence lower disbursements in corporate taxes11. Domestic VAT appears to increase steadily in real terms and as a share of GDP. Further administrative reforms to enhance the VAT auto reconciliation facility could help detect non-declaration of output VAT, which is claimed as input VAT by other taxpayers, and thus lift performance of Domestic VAT even higher. 11 This provides possible area of extending the current diagnostics on revenue underperformance in Kenya. 32 Figure 3-10: Tax revenue over-time, customs revenue vs. domestic revenue a) Real tax revenue from Customs (Ksh. million) b) Customs revenue (in % of GDP) c) Real tax revenue from Domestic (Ksh. million) d) Domestic revenue (in% of GDP) Source: World Bank based on KRA data 3.14. VAT exemptions account for a substantial amount of forgone revenue. Previous research has shown that tax expenditures on VAT (exemptions and zero rating) accounted for at least 3.5 percent of GDP in 2015 (World Bank, 2017). Furthermore, while VAT tends to be regressive, removal of VAT exemptions and providing affected poor households with targeted social transfers is a potentially welfare raising alternative as it helps household overcome poverty (World Bank, 2018). In real terms, VAT refunds have been decreasing, falling from Ksh.7,758 million in 2007 to Ksh. 4,617 million in 2017 (Figure 3-11). This in turn reflects unadjusted provisions for VAT refunds, which has resulted in large backlogs owed to firms and entrepreneurs. Furthermore, the nominal trend in expenditure on duty remissions and VAT refunds shows a large swing over-time (Chacha, 2018) and the number of beneficiary firms is very small (on average about 302 firms (Figure 3-12). The ratio of refund paid relative to refund claimed remains low and requires significant improvement to boost tax moral of business and discourage tax avoidance. 3.15. Authorities could review the relevance and efficiency of tax expenditures to reduce forgone VAT revenue, as well as broaden the tax base, and improve tax administration. Through the Finance Act of 2018, the Government of Kenya (GoK) removed the exemption status on petroleum products by introducing 8 percent VAT on petroleum products. Prior to the Finance Act of 2018, petroleum products had been exempted from the VAT rate. In addition, the government has introduced a governance framework for tax exemptions requiring more detailed information before exemptions are granted. Further, under the new framework, more stringent reporting requirements for the use of exemptions and regular internal audits have been introduced. While these are quite positive reforms, it could also help 33 over the medium term to conduct a review of exemption and zero-rating regimes to assess whether they satisfy the reasons they were introduced in the first place. It is also best practice to reflect the costs of such exemptions annually as an annex in the budget policy statement / and /or the annual budget. Figure 3-11: Real provisions for refunds (Ksh. million) Figure 3-12: Access to duty remissions and VAT refunds over time 500 330 450 319 434 320 400 310 350 TREO in US$(mn) 317 300 # of firms 300 290 290 250 287 280 200 278 177 270 150 100 260 99 50 250 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 TREO in US$ mn(LHS) #Firms(RHS) Source: World Bank based on KRA Source: Chacha, 2018 3.16. More realistic tax revenue forecasting is key to the preparation of a credible budget and its subsequent implementation. Optimistic revenue forecasts can lead to unjustifiably large expenditure allocations that will eventually require either an in-year and potentially disruptive reduction in spending or an unplanned increase in borrowing to sustain spending. According to the Kenya PEFA 2018 review, actual revenue fell short of budgeted revenue by at least 5.3 percent in the last five years. Although revenue performance in the last decade has been on an upward trend, overall revenue mobilization relative to the target has continued to underperform. For instance, in FY2017/18, total revenue performance fell 12.6 percent short of the target (i.e. Ksh.1,375.3 billion against a target of Ksh.1,555.3 billion) despite a 5.3% increase in revenue collection (Figure 3-13). Enhancing the level of revenue mobilization but also ensuring that realistic revenue projection informs expenditure allocations is critical in building credibility in the budget process and in supporting the government’s inclusive growth agenda. Figure 3-13: National government Revenue Figure 3-14: County government Revenue Performance Performance Source: World Bank based on National Treasury and KRA 34 Box 3-1: Steps for estimating elasticities and results from existing literature Elasticity measures growth in tax revenues after netting out the revenue impact of any discretionary changes in the tax structure (adjusted tax revenue) relative to actual growth in GDP from year to year, similar to the measurement of tax buoyancy. However, there is crucial difference in that revenue is calculated as it would have been if there had not been any change in the tax laws (no discretionary changes). Revenue is calculated as if the same tax structure would have been in place over the years. Two ways for adjusting tax revenue series: 1. Constant Rate Structure Method: Apply current year’s rates to previous years’ tax bases and construct an adjusted tax revenue series 2. Proportional Adjustment Method: Construct revenue series by adjusting for the effects of discretionary changes. (The proportional adjustment method requires information on the revenue impact of discretionary changes, which are not always available, thus a dummy variable can be used in the regression analysis to control the effect of a major tax reform. Using a dummy variable will not give a good result when the government makes frequent tax policy changes. There should be at least 4-5 years’ data both before and after a reform for a reliable dummy coefficient). Steps and data requirements • Compile actual revenue collections and data series for discretionary changes (for the data series to be statistically representative of a reasonable number of years, e.g. 10 years) • Convert all categories of taxes from nominal to real terms • Estimate Adjusted Tax Revenues (e.g. Proportional Adjustment Method) • Calculate the adjusted coefficient and cumulative coefficient to reflect discretionary changes • Multiply real tax revenues with the cumulative coefficient of discretionary changes to get an adjusted tax series • Estimate tax elasticity by regressing the respective adjusted tax revenues with the corresponding tax base. • Interpret the results for elasticity using T-stat and R square Source: World Bank, 2017 3.17. Authorities are improving their analytical tools as a first step towards realism in revenue projections and improvement in revenue mobilization. Previously, the World Bank carried out tax gap studies (World Bank, 2017) on VAT and CIT aimed at pointing out areas of revenue improvements. The Bank also undertook an own-source revenue (OSR) potential and tax gap study of Kenya’s County governments on behalf of the Kenya National Treasury12. Building on this work and recent request to improve revenue modelling and forecasting tools, a Technical Assistance (TA) is underway to support the government in enhancing the realism of its revenue projections through an update to the tax models and forecasting tools used by the authorities. The update to the KRA/NT model includes re-estimation of parameters such as elasticities (in line with Box 3-1) and agreeing on further improvements to monitor performance of revenue and to carry out policy simulations. 12 These analytical pieces have been useful in informing tax policy at the National level, including the modernization of the income tax act that is undergoing legal drafting and review of legal/policy framework for own source revenue mobilization by counties. 35 D. Conclusion and Recommendations 3.18. To conclude, Kenya’s tax revenue as a share of GDP has decreased and decoupled from GDP growth, suggesting some important constraints to enhancing revenue collection. Tax revenue as a share of GDP has decreased from 18.1 percent of GDP in FY2013/14 to 15.7 percent in FY2017/18. More concerning, however, is the apparent lower buoyancy of tax revenue relative to expansion in nominal GDP over the past five years. This could be attributed to changes in structure of the Kenyan economy, with Agriculture gaining a sharp increase relative to other sectors (industry and services). The contribution of agriculture13 to tax revenue is about two percent, while over 98 percent is attributed to industry and services. This suggests that despite recent tax policy and administrative measures to support domestic revenue mobilization, including integration of iTax with other databases (e.g. IFMIS), roll out of integrated customs management, and expansion of tax bases - all of which are positive developments - additional time may be required (through structural transformation) for the envisioned revenue increases to materialize. 3.19. Policy could undertake several reforms, both in the short-run and medium-term horizon to improve tax revenue mobilization. In the short-run, review and update elasticities as well as tax bases used to project tax revenues to reflect the changed economic structure. This is important to bring in realism in revenue projection14 and anchor spending decisions over the medium term. Specific to the value added tax (VAT): ensure promptness in payments of refunds whenever tax payers have a net VAT credit as this is essential for boosting tax payment moral of businesses and promoting private sector activity. On personal income tax (PIT) undertake a review of the impact on revenue from recent rate adjustment and special reliefs under PIT to stem further loss of revenue, and on corporate income tax (CIT) undertake simulations to settle for a revenue neutral rate and eliminate multiple preferential rates once a suitable rate is established. 3.20. Over the medium-term focus on reducing the revenue forgone through a review of the relevance and costs of multiple exemptions on VAT and CIT. Revenue forgone due to VAT exemptions and zero rating is estimated at about 3.5 percent of GDP (World Bank, 2017). Similarly, revenue forgone from CIT exemptions, accelerated depreciation allowances and preferential rates are estimated at about 1.9 percent of GDP (World Bank, 2017). There is need, therefore to review VAT and zero-rating incentive schemes as well as generous deductibles and investment allowances under CIT to reduce revenue forgone. Further, for improved transparency on the costs of such exemption, prepare a repository of all tax expenditures and present an annex to the budget policy statement and/or the annual budget-showing the costs of such expenditures to the tax payers and reasons for their introduction. 13 Tax effort is low in jurisdictions with relatively high share of agriculture in the total economy-as is characterized by very high level of informality. 14 The Kenya Tax modelling and forecasting TA re-estimated elasticities for all of the exchequer revenues to account for the recent changes in the structure of the economy. 36 3.21. The authorities could consider the following recommendations on improving the administration and performance of tax revenue (Table 3-1). Table 3-1: Priorities for enhancing revenue collection and improving tax system efficiency Reform Priority Short-run (6 months-1 year) Medium-term (2-3years) VAT • Fast-track payments of refunds • Reduce the revenue forgone whenever the taxpayer has a due to exemption and zero net VAT credit. rating through a review of • Clean the VAT taxpayer register these schemes to assess to ensure it includes an whether they have achieved accurate number of taxpayers the goals behind their (on those with turnover of over introduction and associated Ksh.5 million). costs. • Reflect the costs of such exemptions annually as an annex in the budget policy statement and /or the annual budget. Personal Income Tax (PIT) • Review the revenue impact of • Review the rate of capital gains the changes in PIT rate taxes (currently at 5 percent) to structure and from special disincentivize reclassification of reliefs under PIT. personal income as capital gains tax. Corporation Income Tax (CIT) • Finalize the review of the • Review the CIT rate structure, Income Tax bill 2019, which including modifying reduced streamline tax exemptions and preferential CIT rates for new deductibles under CIT. companies, and uniformly • Undertake simulations to apply a consolidated CIT rate. examine the optimal tax rate • For established companies for a neutral revenue way and benefiting from reduced rates, eliminate preferential rates consider maintaining current once a suitable provisions or negotiating a fade-out scheme. All tax heads • review and update • Prepare a repository of all tax elasticities plus tax bases to expenditures for all tax reflect changed economic sources. • Prepare a medium-term tax structure. revenue strategy in line with the new PFMR strategy. 37 Chapter 4 Trends and Profile of Central Government Expenditures 4.1. This chapter discusses the size and composition of expenditures for the national government15 in Kenya over the last five years (2013/14 -2017/18). It shows that that government spending is allocated well (to infrastructure and human capital) but there is scope to improve outcomes from the use of these resources. Public expenditure has had a good impact—even better than peers (as evidenced by the improved quality of infrastructure, and improved outcomes on education and health). Nonetheless, weaknesses in the public investment management system reduce the technical efficiency of capital spending and could lead to delays in project implementation. Further, holding up payments to vendors could be delaying business to business transaction and making the cost of delivering projects very high, as contractors inbuilt risks of delayed payments into their bids. Delays in public payments can affect private sector liquidity, profits and ultimately economic growth. There is need, therefore, to clear any pending bills/arrears, and reduce costs and time over-runs for development projects, and enhance efficiency of public spending. A. Size and trends in Central government expenditures for the last five years 4.2. Kenya’s fiscal policy has been mostly expansionary during the past five years. Overall, total expenditure by the national government increased from an average of (23.4) percent of GDP during 2007/8 – 2011/12 to an average of (26.6) percent during 2013/14 – 2017/18 (Figure 4-1). The expansion was driven by several factors: First, the implementation of the new 2010 constitution required funding of a devolved system of government and new constitutional commissions since FY2012/13. Second, the government committed to significantly improve and bridge Kenya’s economic and social infrastructure deficit. Third, in FY2016/17, the economy faced multiple shocks (drought and prolonged electioneering) necessitating large importation of food and the need to make provision for a second presidential election re-run. Figure 4-1: Fiscal policy has been expansionary over Figure 4-2: Government spending was the review period elevated over the review period Source: World Bank based on National Treasury 15In this section expenditure by the central government includes all ministries and departments, and independent constitutional offices. Local government is used interchangeably with county governments especially if a comparison with peers is needed. 38 4.3. Recognizing the narrowing of fiscal space, since FY2017/18 a process of fiscal consolidation has been underway. After rising from 24.9 percent of GDP in FY2013/14 to 27.5 percent in FY2016/17, government spending decreased by (3.2) percentage points of GDP to 24.3 percent in FY2017/18 (Figure 4-2). This was supported by a moderation in development expenditure following completion of some of the major projects (i.e. the phase one of the Standard Gauge Railway (SGR) and link roads) that started in FY2013/14 as well as adjustments for once-off expenses (i.e. food importation and election-related expenses). Nonetheless, the scope for achieving fiscal adjustment through expenditure reduction without hurting priority spending and growth is narrowing. This calls for enhanced revenue mobilization (expanding the tax base and rationalizing exemptions), but also ensuring solid governance systems to prevent wastages and improve efficiency and effectiveness of spending. 4.4. Central government expenditure in Kenya is on average larger than that of regional peers and has surpassed the country’s aspirational peers over the last five years (Figure 4-3). The expansionary fiscal policies over the last five years has catapulted the size of government as a share of GDP in Kenya to be higher than that of its regional peers. While government spending can contribute to economic growth, there is a growing literature on the deleterious impacts of excessive government spending can have on growth. Simulation results for Kenya suggests that, over the longer term, government consumption spending has had a negative impact on growth, whereas government infrastructure spending has been positive for Kenya’s long-term growth. Compared against aspirational peers16, Kenya is on average spending more than Thailand (21.6% of GDP) but below South Africa, Vietnam and India. Scaling up expenditure to levels attained by aspirational peers is possible but needs to be accompanied by enhanced revenue mobilization, as well as efficiency and effectiveness of government spending17. Figure 4-3: Central government expenditure Figure 4-4: Average Central government over time (% of GDP) expenditure, over 2013-17(%of GDP) 32.3 28 30 29.0 26.8 26 26.7 26.5 21.6 24 20.1 20 18.7 % of GDP 17.8 17.4 22 20 10 18 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Kenya Regional Peers 0 Aspirational Peers ZAF VNM IND RWA KEN THA GHA UGA ETH TZA Source: World Bank Calculations based on WEO-IMF 16 Regional peers include Uganda, Tanzania, Rwanda, Ethiopia, and Ghana. Aspirational peers comprise of South Africa, Thailand, Vietnam, and India 17 In the literature analysis of efficiency, effectiveness and equity dimensions are covered in most PERs. A fiscal incident analysis has covered the equity aspect of government spending. 39 B. What are public resources being spent on? Spending on physical infrastructure was a key priority over the review period 4.5. Reflecting the commitment to close the infrastructural gap, the ratio of development spending to central government budget is within the PFM Act of 2012 requirements of at least 30 percent of the budget. As a share of GDP, development and net lending rose from (6.3) per cent of GDP in FY2013/14 to (8.4) per cent in FY2016/17, before easing to 5.3 percent in FY2017/18(Figure 4-2). In terms of total budget, the share of development expenditure increased from (24.5) percent of total expenditure in FY2013/14 to (30.3) percent in FY2016/17 (Figure 4-5). The government prioritized infrastructure development with the aim of creating a conducive environment for trade while raising the ability of firms to respond to global growth opportunities. Figure 4-5: The share of development expenditure Figure 4-6: The quality of infrastructure has relative to total government spending increased over improved time Source: The National Treasury. Note: recurrent expenditure includes transfers to counties, parliamentary and judicial services 4.6. As a result, the ranking in quality of Kenya’s infrastructure has improved relative to its peers. Overall, Kenya’s quality of infrastructure ranks second to Rwanda amongst regional peers and ranks higher compared other peer countries such as Bangladesh and Vietnam (Figure 4-6). Additionally, Kenya ranks highest amongst both its regional and structural peers in terms of quality of railroad infrastructure and quality of port infrastructure. 4.7. Increased development spending has been associated with the completion of flagship infrastructure projects in the transport and energy sub-sectors. These include completion of the construction of 472 kilometers of the Standard Gauge Railway (SGR), increases in the paved road network by 3,300 kilometers between 2013 and 2016, expansion of the JKIA arrival and departure terminals to accommodate 7.5 Million passenger capacity per annum, and the rehabilitation of the runway to enhance serviceability and the number of aircraft parking bays from 21 to 37. Within the port of Mombasa, berth number 19 was completed under Phase I of the Mombasa Port Development Program (MPDP) and the acquisition of the Rail Mounted Gantry cranes (RMGs) to operationalize cargo evacuation using the SGR. On energy, between 2008 and 2017, the capacity of electricity generation has increased from 1,268MW to 2,336MW. Additionally, the number of households connected to the national electricity grid has increased to almost 6.5 million in 2017(or about 70 percent of total population). 40 4.8. Spending on infrastructure is also reflected in the contribution to growth, with the share of capital stock to GDP growth rising over the last five years. The share of capital stock in yearly real GDP growth remains high, followed by labor and total factor productivity (“TFP�). Nonetheless, capital stock per capita for Kenya remains low relative to comparator countries (Vietnam, South Africa, Thailand, Malaysia) but also below some regional comparators such as Ghana and Tanzania (Figure 4-7). The contribution of Total factor productivity to real GDP growth was negative before turning positive in 2010 onwards (Figure 4-8). The modest gains in total factor productivity suggest higher quality investment could help realize additional growth of the economy. Figure 4-7: Capital stock per capita relative to Figure 4-8: The share of capital stock, Labor and peers (2015, in constant 2011 international TFP in GDP growth (1990-2017) dollars) 60,000 50,000 40,000 30,000 20,000 10,000 0 LKA KEN TZA ZAF MYS BGD GHA VNM ETH IND THA RWA UGA Private Public Source: IMF Capital and Investment Database, 2018 and WDI database. 4.9. Further improvements to TFP and labor productivity call for enhancing the quality of human and physical capital. Kenya stands to enjoy gains in total factor productivity from investments in human capital, including implemented government policies for free primary and secondary school education in 2003 and 2008 respectively, as well as policies currently under implementation to provide free universal health care coverage. Kenya’s human capital index18 ranks higher than the average for lower-middle income economies and suggests that Kenya is positioned to enjoy relatively higher gains in total factor productivity growth. Nonetheless, additional analysis of this topic could shed light on gains that Kenya could enjoy in productivity and potential growth. 4.10. Turning to classifying general government expenditures by economic sectors, expenditure on grants (or transfers) has accounted for a significant share of the budget. Grants expenditures include transfers to county governments, Parliament, Judiciary, semi-autonomous agencies, and other SOEs. As a share of total central government expenditure, grants averaged about 39.1 percent over the period (2013/14-2017/18) (Figure 4-9). Reflecting a growing role of counties in public services delivery (i.e. health, agriculture, markets, and other devolved social infrastructure functions), county transfers accounted for more than a half of the expenditure on grants (averaging 3.8 percent of GDP). 18The human capital index is a measure of how much capital a child born today will expect to attain by the age of 18, given prevailing risks to education and health 41 4.11. The share of compensation of employees in total central government expenditure has almost doubled over the period under review - thanks to wage agitations and implementation of the resulting collective bargaining agreements (CBAs) with labor unions. Expenditure on compensation of employees increased from (14.3) percent of total expenditure in FY2013/14 to (25.0) percent in FY2017/18. The growth in wage bill was driven by increases in the average public sector salary per worker (i.e. salary adjustments for doctors, nurses, university lectures, and teachers) and new recruitments. In addition, salaries have been high in constitutional commissions and independent offices as well as in the county- level governments due to the need to attract critical skills with differentiated wage grids. Figure 4-9: General government expenditures by economic classification Central government expenditure by economic classiification (% of total central government expenditure) 100 2.6 2.6 2.5 1.7 3.5 5.0 4.9 6.1 4.7 5.0 10.5 10.5 10.2 10.4 10.9 80 13.2 14.0 12.0 16.3 26.6 15.8 12.5 60 17.7 15.3 14.6 14.3 19.8 25.0 40 21.3 11.3 20 36.1 33.8 29.4 28.8 31.2 0 2013/14 2014/15 2015/16 2016/17 2017/18 Grants Compensation Of Employees Interest Payments Acquisition Of Non- Financial Assets Use Of Goods And Services Others Subsidies Source: World Bank Calculations based on National Treasury 4.12. Kenya’s wage bill is high relative to its EAC peers but slightly lower relative to emerging economies average. For the period 2014-2018, Kenya’s wage bill as a share of GDP was higher relative to the EAC peer countries (Uganda, Rwanda, and Tanzania) (Figure 4-10). However, compared against emerging economies average, Kenya’s expenditure on wages and salaries was lower by at least 2.7 percentage points of GDP ([Figure 4-11). 42 Figure 4-10: General government Wage bill in [Figure 4-11: Wage bill, 2014-16 (Percent of percent of GDP GDP)] 9.0 7.0 Percent of GDP 5.0 3.0 1.0 20152014 2016 2017 2018 Kenya Tanzania Uganda Rwanda Source: The WDI and IMF 4.13. In nominal terms, all categories of the wage bill have increased between FY2013/14 and FY2017/18. Basic salary, on average, constituted the largest share of total payroll costs, followed by employment allowances, and pension benefits (Figure 4-12). Basic salaries grew on average by about 8.1 percent per year over the last five years, while employment allowances have increased by at least 11.0 percent per year. Looking at the shares, employment allowances has increased from 28 percent of wage bill expenses in FY2013/14 to 31 percent in FY2017/18 (Figure 4-13) -indicating growing significance of allowances in the remuneration structure and to some extent difficulties in ascertaining true staffing costs of public administration. Figure 4-12 : Composition of wage bill by economic Figure 4-13:Composition of wage bill category, in Ksh. billion (2013-18) expenditure 2013-18(%) Source: The National Treasury 3.22. The share of employees at the county level of government has increased faster relative to others. The head count of general government employees increased by 20.1 percent from 702,000 in 2014 to 843,000 in 2018, out of which the share of employees at the county level increased from 14 percent in 2014 to 21 percent in 2018 (Figure 4-14). This is expected as devolution of functions from the national to the county level of governments takes full effect (for example the health sector employees have all been fully deployed at the counties). Nonetheless, the drop in the share of employees at MDAs has not been as fast, even as functions (and jobs) moved to the counties. The share of employees at the MDAs decreased to 24 percent in 2018 (from 26 percent in 2014). 43 Figure 4-14: Share of general government employees (in % of head count) 2014 compared to 2018 Source: The National Treasury and KNBS 4.14. The key drivers of growth in wage bill is the change in the growing size of employees, especially at counties and increase in average wage per worker. The level and growth in the size of the civil service is greater than what is typically needed to address natural attrition and for attracting critical skills needed for effective public service delivery. The number of government employees rose by 9.6 percent from 682,200 in 2013 to 747,900 in 2018 (Figure 4-15) while the average annual salary for counties has been generally higher than rest of government (Figure 4-16). The average growth of the public service head count was about 2.4 percent per year, peaking at 8.8 percent in 2015, mainly due to accelerated recruitments. Figure 4-15: Public service head count of Figure 4-16: Average annual salary (in Ksh. ‘000) employees (in ‘000) Source: KNBS (Economic Survey 2018) and the National Treasury 4.15. With numerous wage agitations being awarded (from teachers, doctors, nurses, civil servants, among others), average wages and salaries per employee have increased, contributing to an expanded wage bill. The unit cost of labor grew by 61 percent in the county governments, 30 percent in Ministries, Departments and Agencies (MDAs) and 17 percent in the Teachers Service Commission between 2013 and 2018 (Figure 4-16). This resulted in an overall 37 percent upward adjustment to salaries and allowances. 44 4.16. Improving public wage bill management is imperative to create needed fiscal space to fund public programs, although most of this spending is somewhat rigid. Some of the drivers of persistent growth in the wage bill include faster increases in the head count of employees at the counties, rising personnel allowances (which is less rigid relative to basic salary), and wage adjustments through CBAs. Since personnel allowances make-up a significant share of remuneration structure, there is need consolidated them to the base wage to prevent distorting the remuneration structure. 4.17. Interest payment expenditure as a share of the budget has increased, reflecting a rapid accumulation of the stock of public debt, while expenditures on use of goods and services, subsidies and other expenditures remained broadly stable over the review period. Interest payments as a share of total expenditure went up by 2.0 percentage points between FY2013/14 and FY2015/16 but declined in FY2017/18 in line with decreasing public debt levels. Expenditures on use of goods and services, subsidies, and other expenditures were broadly stable over the last five years. Nonetheless, access to more disaggregated data such as the amount spent on professional services-e.g. consultancy and cleaning services (generally lamped into goods and services) could help us better quantify the cost of staffing in public administration. These issues are deferred for a fully-fledged public expenditure review. Infrastructure and human capital accounted for the largest share of central government spending 4.18. Similarly, central government expenditures by functional classification shows increased spending on infrastructure and human capital (education and health). The energy, infrastructure, and ICT sectors have accounted for about 21.1 percent of the central government’s expenditure (one fifth over the last five years). Equally, spending on education and health have accounted on average for about 16.2 percent and 3.3 percent, respectively of total central government expenditure. Increased expenditure in the social sector (education and health) is associated with improved rankings on the human capital index (HCI). Kenya’s HCI of 0.52 implies that expected work productivity of a child born today is about 52 percent of what it could be with complete education and full health. Health is a fully devolved function, with the National government retaining only the sector’s policy formulation functions (Figure 4-17). 45 Figure 4-17: Kenya’s spending priorities in the last five years were tilted towards infrastructural and human capital development Central government expenditure by sectors (% of total central government expenditure) 90 80 1.0 0.7 2.0 1.9 1.7 1.1 2.0 3.5 3.0 70 5.1 1.1 2.0 3.1 3.3 3.5 3.0 3.4 3.1 3.5 4.6 2.1 3.1 60 4.6 4.0 3.8 10.5 11.1 14.4 13.8 50 13.4 10.8 11.8 40 10.7 10.1 12.8 10.1 18.2 30 23.8 18.4 10.4 20 26.2 22.9 10 18.8 17.5 16.5 0 2013/14 2014/15 2015/16 2016/17 2017/18 Energy, Infrastructure And ICT Education Governance, Justice, Law and Order Public Administration And International Relations Agriculture, Rural & Urban Development Environment Protection, Water And Natural Resources Health Social Protection, Culture and Recreation General Economic and Commercial Affairs Source: World Bank Calculations based on National Treasury 4.19. Expenditure on agriculture, rural and urban development as well as on social protection, culture and recreation has been moderate and stable over the review period. Expenditure on social protection as well as on agriculture tend to be well targeted and are in general pro-poor. Spending on agriculture was on average about 3.8 percent by the central government budget19. This excludes the amount spent at the county level of government. Spending on social protection has been broadly stable at about 2.0 percent of the central government budget. 4.20. Public administration, governance, justice, and law and order still accounts for a significant but stable share of public resources. Strengthening governance systems and institutions, including security, remains a key focus of the national government. Resources to these two sectors have stabilized at about 20.6 percent of central government expenditure over the last five years. Finally, resources for environment protection, water, and natural resources have contracted from 4.6 percent of expenditure in 2013/14 to about 3.3 percent in FY2017/18. Proper management of water resources use remains key in expanding irrigable acreage in Kenya and in achieving food and nutrition security agenda. 4.21. Nonetheless, the biggest driver of fiscal consolidation over the last two years has been the slowdown in the pace of development spending. At the heart of most fiscal consolidation packages is the need to increase tax revenues and to rein in expenditures, both development and recurrent expenditures20. However, from a growth perspective, the latter should take a greater weight. Further, to 19 The FY19 Ag PER and the SPJ PER covers in detail the size and composition of spending on agriculture and social protection, respectively. 20 Nauschnigg (2006), Nauschnigg (2010) 46 counteract the drag from fiscal consolidation, policy measures could be put in place to stimulate the private sector’s contribution. For instance, while the state slows down on development spending a regulatory environment and incentive structure could be enhanced (within the PPP unit at the NT) to achieve flagship infrastructure projects through public private partnership 21initiatives. However, this needs to be undertaken within a fiscal sustainable way since PPPs are primarily a risk-sharing mechanism and not a mechanism to reduce government fiscal burden. If the risks are not well shared and the government takes on excessive contingent liabilities, then the impact on overall fiscal space could be negative rather than positive C. How easy is it, to re-allocate public resources towards national priorities? Rigidity of spending hinders reallocation towards national priorities 4.22. The classification of public expenditures by their rigidity is a useful way to quantify the growth and acceleration of expenditures. Knowing which budget items are driving expenditure dynamics and how discretionary these key components are, can be very useful for policymaking. Specifically, caution is needed before considering positive expenditure force22 in spending that will be relatively more difficult to adjust in the future. Moreover, the ease in which certain expenditures can be adjusted is important for considering policy options for fiscal tightening. 4.23. Different components of public budgets vary within a spectrum of flexibility degrees since not all of them can be easily modified in the short term by budgetary authorities. Those items that cannot be reassigned without high costs (either judicial, economic or political) are generally termed “rigid expenditures� (Box 4-1). To estimate the degree of rigidity of expenditures for Kenya, the following categories were used to classify expenditures (Table 4-1) Table 4-1: Classification of rigidity by object of expenditure High rigidity Personnel Expenditures (Permanent Staff), Social Security Benefits, Interest payments, Tax Co-participation with other levels of government, Capital Transfers to Municipalities High Medium rigidity Personnel Expenses (Temporary Staff), Earmarked Transfers to Institutions Medium rigidity Non-Personnel Services, Other Current Transfers Low rigidity Materials and Supplies, Other Current Expenses, Direct Real Investment, Financial Investment, Other Capital Transfers Source: Cetrángolo et al, 2010 Box 4-1: Rigidity Analysis 21 This remains an area for further consultations with authorities on reforms needed to strengthen the PPP unit at the NT and fast-track the pipeline of PPP projects towards implementation phase. 22 Merotto, et al. in 2015. The paper draws from the concepts of velocity, acceleration, momentum and force from physical science, to devise a measure of “force� for public finance economics. 47 The different components of public budgets vary within a spectrum of flexibility degrees since not all of them can be easily modified in the short term by budgetary authorities . Those items that cannot be reassigned without high costs (either judicial, economic or political) are called “rigid expenditures�. Cetr ángolo et al. (2010) defines fiscal rigidity as “the institutional constraints that limit the ability to change the level or structure of public budgets in a specified period of time�. Thus, those components of the budget that are catalogued as rigid or in flexible are not subject to the immediate discretion of the authorities during the fiscal year. The distinction between discretionary and non-discretionary expenditures is not exhaustive, but rather an initial exercise aimed at identifying expenses gradually from the most rigid to the most flexible. The development of rigidities or inflexibilities can be seen as a reflection of the roles and functions that the public sector has assumed over time and can be analyzed from two perspectives: i) the negative, which emphasizes the limits to the maneuverability of fiscal policy, and ii) the positive ones, according to which rigidities arise from the need for institutional frameworks that contribute to achieving public policy objectives and that reflect the roles assumed by the State in different development models (Cetrángolo et al, 2010). These inflexibilities may have different regulatory roots and involve different legal instruments, contracts and institutional obligations of the public sector that have a budgetary impact. Without being exhaustive with respect to the typologies developed in this document, the following can be mentioned as origins of the inflexibilities23: • Rigidities that arise from policies or programs linked to the "benefit principle", that is, that are designed to benefit those who finance them (such as social security or health policies); • Rights and guarantees established in different various kinds of regulations, which give rise to specific allocations of resources in order to protect certain expenses; • Intergovernmental transfers, which are linked to the degree of decentralization of fiscal policy; • Rigidities linked to macroeconomic dynamics, such as interest payments on public debt, indexation of salaries and pensions, fiscal rules, stabilization funds; • Rigidities linked to tensions within the public sector, which encompasses both the conflicts between sectors due to the determination of public spending priorities, as well as the expenses inherent to the operation of the public sector and which turn out to be politically inflexible. 4.24. The analysis shows that, on average, about 68 percent of the national government expenditure is related to items of high to medium rigidity, and these components have been increasing. As shown in Table 4-2, spending on grants represents on average about one third (27.8 percent) of total expenditure. This is followed by compensation of employees (22.3 percent of total), acquisition of non-financial assets (18.8 percent of total) and interest payments (13.2 percent of total) in terms of degree of rigidity. Within compensation of employees, the single largest line of expenditure is compensation of permanent employees with a budget allocation of approximately 21 percent of total expenditure. 23 Cetrángolo et al. (2010). Tax expenditures and extraordinary income (extrabudgetary funds) are also origins of rigidities but are not included in the analysis for Kenya. 48 Table 4-2: Rigidity by Economic Classification (Share of total expenditure) 2013/14 2014/15 2015/16 2016/17 2017/18 Compensation of Employees 19% 16% 24% 23% 30% Permanent Employees (basic salaries and allowances) HR 17% 15% 22% 21% 27% Temporary Employees (basic salaries) M-HR 1% 1% 1% 1% 2% Other allowances M-HR 0% 0% 0% 0% 0% Social Contributions HR 1% 0% 0% 1% 1% Use of Goods and Services LR 13% 13% 13% 12% 13% Interest Payments HR 13% 12% 15% 14% 11% Subsidies 2% 2% 2% 2% 3% To Public Corporations M-HR 2% 2% 2% 1% 3% To Private Enterprises MR 0% 0% 0% 0% 0% Grants 31% 25% 25% 30% 28% To Foreign Govts./ Intl. Org LR 0% 0% 0% 0% 0% Current intergovernmental MR 12% 12% 10% 10% 13% Capital intergovernmental HR 15% 11% 12% 16% 12% Other and Emergencies MR 3% 3% 3% 3% 3% Social Benefits HR 3% 3% 4% 3% 0% Other Expenses 1% 0% 1% 0% 1% Contingencies MR 1% 0% 0% 0% 0% Other Expenses LR 0% 0% 0% 0% 1% Acquisition of Non-Financial Assets LR 18% 28% 17% 16% 14% Total Expenditure 100% 100% 100% 100% 100% High-Rigidity 49.4% 40.9% 53.5% 55.4% 51.0% Medium High-Rigidity 15.5% 14.5% 13.2% 12.4% 17.5% Medium Rigidity 3.8% 3.0% 3.2% 3.4% 3.4% Low Rigidity 31.3% 41.5% 30.1% 28.9% 28.1% Source: BOOST where: HR= High Rigidity; M-HR= Medium High Rigidity; MR= Medium Rigidity; and, LR= Low Rigidity. 4.25. By adding up the high rigidity to medium high rigidity items, the rigidity of budgetary items has increased by four percentage points over the review period. High rigidity budget represents about 68 percent of the budget. On the other hand, low rigidity items have decreased around three percentage points, representing around 30 percent of the budget. In sum, the increase in public spending between 2013/14 and 2017/18 has focused on high rigidity items such as permanent staff salaries, interest payments, and transfers to counties. Figure 4-18 presents a tight evolution of expenditure rigidity in Kenya over time. 49 Figure 4-18: Composition of rigidity by Economic Classification Source: BOOST database 4.26. A medium to long term strategy may be needed to shift the rigidity of spending. The target items that are tough to adjust without high costs (judicial, political, and social costs) include salaries for permanent employees, interest payments, and county transfers. For permanent employees, to contain growth in this expenditure, evidence from other jurisdictions shows that restricting new hires to critical and technical services such as teaching, security and health care may be necessary. Even in these critical areas, additional hires should be remitted to replacing retired officers and those transitioning out of the profession. Transfers to counties remains a constitutional obligation and given its dependency for services delivery at the counties, the only chance to stabilize transfers is to step up counties’ own source revenue mobilization. Previous estimates of OSR potential shows scope for additional revenue mobilization within counties that could reduce pressure on the exchequer, if realized. Finally, improved debt management to reduce the interest rate obligations through a plan to retire short-term and expensive commercial debt could help. A known alternative is to raise the share of concessional borrowing relative to non-concessional debt in future financing of the deficit, even as the government stays the course in its announced fiscal consolidation pathway. D. How effective is spending of public resources? Actual amounts spent are only a fraction of what is planned in the budget 4.27. Budget execution remains low, particularly for development spending. Over the period 2013-18 underspending on the development budget has been substantial (about 31 percent of the budget is not absorbed). Budget execution is also quite weak at the county level, making it difficult for the government to achieve its development objectives. While the national government has shown some improvement in execution of its development budget - with execution rates rising from 76 percent in 2014/15 to 87 percent in 2017/18 - county level government’s execution has actually declined, from 74 percent in FY2013/14 to 62 percent in FY2017/18 (Figure 4-19). Similarly, execution of recurrent budget varies, but is generally higher at both the national and county level of government. 50 Figure 4-19: Execution of development budget remains a challenge mainly at the county level Budget execution rate by central government (%) Budget execution rate by county governments (%) 100 100 90 90 80 80 70 60 70 50 60 40 50 30 2013/14 2014/15 2015/16 2016/17 2017/18 2013/14 2014/15 2015/16 2016/17 2017/18 Development Expenditure Recurrent Expenditure Development Expenditure Recurrent Expenditure Source: BOOST 4.28. Despite a larger share of the budget going to energy, infrastructure and ICT, its execution has lagged, suggesting absorption capacity constraints. Total infrastructure budget execution averaged 72 percent in the period 2013/14 to 2017/18. This execution deteriorated in 2015/16 with an execution rate for the sector reaching only 49 percent. This means that only Ksh.219 billion (3 percent of GDP) was utilized out of the Ksh.445 billion (6.2 percent of GDP) allocation. Budget execution in education, governance, justice, law and order, social protection, culture and recreation, general economic and commercial affairs, and public administration and international relations recorded higher execution rates, surpassing 80 percent (Figure 4-20). 4.29. Transport infrastructure is particularly experiencing underspending relative to allocated budget. As an example, the government allocated on average 5 percent of GDP to transport infrastructure during the period 2013-17 but budget execution was lower. On average, budget execution was approximately 4 percent of GDP (Figure 4-21). The difference between the final approved budget and actual expenditure (execution gap), averaged 1.3 percent of the GDP during the review period. The highest gap can be observed in 2015-16 when the budget for the sector was 4.9 percent of GDP and only 2.6 percent was executed, leaving a gap of 2.2 percent of GDP. 4.30. The slow execution of the development budget signals weaknesses in project appraisal, planning and sequencing of implementation. Projects implementation cycles and budgeting appear unsynchronized – in part due to uncertainty regarding multi-year budgets for projects. This practice tends to encourage over-estimation of budget needed for a given implementation of a project. This ultimately leads to stalled projects, delayed procurement, pending bills and delayed government payments. The associated delays increase the risk of supplying services and works to government, causing such costs and delays to be built into bids, resulting in expensive projects. 51 Figure 4-20: Budget execution was especially lower in Energy, Infrastructure and ICT Average approved versus executed sector spending by central goverment, 2013/14-2017/18 (% of GDP) 7 6.0 6 5.1 5 4 4.3 3.1 3.4 3 2.6 2.6 2 2.3 1.7 1.7 1.0 1.0 0.9 1 0.6 0.7 0.8 0.3 0.7 0.5 0.2 0 Energy, Education Public Governance, National Environment Agriculture, Health Social General Infrastructure Administration Justice, Law Security Protection, Rural & Urban Protection, Economic and and ICT And and Order Water And Development Culture and Commercial International Natural Recreation Affairs Relations Resources Average approved budget Average actual expenditure Source: World Bank, BOOST Portal 4.31. Poor planning and selection of project designs, lengthy procurement processes, contract management, cash and payment, financing structure (domestic, Loan, or grants) as well as land acquisition are key factors behind the execution gap. Acquisition of land and routine maintenance are among the fastest growing expenditures in transport infrastructure (Figure 4-22). Whereas an increase in expenditure for operation and maintenance is encouraged, meeting the growing demand for land required by infrastructure projects remains a great challenge. 4.32. The lack of alignment of procurement plans with disbursement of funds could be associated with a recent pick-up in pending bills. The 2018 enterprise survey for Kenya finds that approximately 12 percent of the 1,001 firms surveyed (or 120 firms) have had a contract with government that was in arrears (Kenya Enterprise Survey, 2018). The total value of pending bills is estimated to have increased from 0.9 percent of GDP in FY2015/16 to 1.6 percent in FY2017/18. This, if allowed to persist, could reduce firm liquidity and cause postponement of new investments. It could also increase firms’ default rate (in business to business transactions), which can be associated with a rise in non-performing loans for the banking sector. This trend underscores the importance of curbing pending bills and arrears as key measure of fiscal prudence, without which an economy could descend into weaker growth prospects as private sector activity and aggregate demand are curtailed. 52 Figure 4-21: Allocated budget to Transport Figure 4-22: Drivers of Transport Expenditure infrastructure v. actual expenditure (2014/15-18) Growth 2016/17 Source: World Bank, BOOST Portal E. Is Kenya getting value for its money 4.33. A discussion on public expenditure analysis is not complete without the analysis of both allocative and technical efficiency, as well as equity concerns of spending. Allocative efficiency could simply be understood to assess whether resources are allocated to priority sectors of the economy and execution rate of spending (actual versus planned budget). This aspect has to some extent been addressed in section B and D, where it was shown that over the review period resources were well targeted to infrastructure and human capital sectors, but there is room to improve budget execution rate. The analysis of equity of spending has also been previously examined in the context of the fiscal incidence analysis of spending on education, health, and social transfers-where they were found to be well targeted and pro-poor. Now, technical efficiency could be defined as getting more in terms of outputs for a given set of inputs (public spending). A decision-making unit is termed efficient if it is found to achieve the maximum output (at or near the frontier) for a given set of inputs. However, this type of analysis requires data at micro-level, which is beyond the scope of this PEA. Thus, the extent of the analysis of technical efficiency in this report is more illustrative from a cross-country benchmarking perspective24. There is scope for more outcomes relative to spending in education and health, which could be optimized or re-allocated within the sectors to achieve sector priorities 4.34. There is scope to improve efficiency of spending on education, as reflected by the amount spent per student relative to secondary school completion rate. Completion of secondary school is associated with high returns to education relative to primary education. It is also key for any country’s human capital development and improved labor productivity. A cross-country comparison between spending per secondary student (in per capita GDP percentage terms) and secondary completion rates shows that given what Kenya spends, there is room to obtain higher outcomes in terms of secondary school completion rates (in percentage terms of total people aged 25 years and above). A number of countries with similar levels of per student spending to Kenya, are able to achieve higher completion rates (For example: Hungary (HUN), Czech Republic (CZE), and Bulgaria (BGR)) (Figure 4-23). This suggests that education 24 There is need for a micro sector deep-dive to quantify fiscal savings and within sector optimal resource allocation. 53 outcomes could be improved at the current level of spending or alternatively, there are opportunities for fiscal savings by spending an amount that is proportional to attained secondary school completion rates (for example equal to what Ecuador (ECU) spends). Figure 4-23: Secondary school completion against Figure 4-24 : Infant survival per 1000 live births expenditure per capita, secondary (% of GDP per against per capita public health expenditure capita) (2017) per capita (constant 2011 int’l $) 2017 1000 100 SGP CYP EST HRV KOR BLR BIH SRB MNE LVA POL LTUHUN CEBCUB SVK MYS LBN ROU BGR ATG RUS MDV BHR CHL OMN CRI BHS URY Infant survival (survivors per 1000 live births) LKA UKR ALBTHA MKD SAU BRN GEO CHNMUS LTE KNA SYC BRB ARG ARM SYR TON TUN MDA GRD KAZ ECU UMCLBYCOL MEX ECATEC TUR BLZ JAM SLV LCA IRN PER BRA PLW Completed upper secondary, ppl 25yrs(% of total) WSM PAN CZEUSAEST GBR VNM NIC HND TEA VCT EAP CPV JOR VEN LCN EAS LAC TLA CSS MNG SLBFJI PRY SUR CHE LVA KGZ EGY PSS DZA DMA MEA 80 POL DEU VUTMAR MNA TMN TUV TTO PHL IDN UZB AZE IBD BMULTU ISR SWE GTM DOM DNK IRQ GUY CYM HUN NORAUT SVN MDA KHM TJK BTN FSM MHL NRU NPL STP BOL ARB LUX BGD MIC KOR 950 MNG FIN EAR BWA WLD BGR ZAF ERI IND IBT LMY NAM ROU SRB IRL CYP RWA KEN TKM SST SGP MDG SAS TSA YEMTLS 60 FRA COG SEN LMC GAB MMR PNG KIR OSS PHL GHA HKG CHL ZAF TZA UGA DJI SDN ZWE MYS LAO ETH GMB MWIZMB SWZ PER IDX COM BDI LDC QAT JOR ITA AFG IDA PAN ESP LIC TGO PAKFCS IRN HPC 40 OMN VEN LBR MRT MOZ HTI TSS SSA SSF COL BOL BLZ IDB PRE 900 CRI BRA ECU LSO GIN CMR CIV DOM MEX GNB BFA GNQ AGO IDN PRY URY KWT COD BEN NER THA SLV KEN MLT PRT SYR 20 GHA NGA CMR MLI LSO GTM TGO ETH SEN MLI RWA CAF TCD SLE BFA 850 0 0 20 40 60 80 0 200 400 600 800 1000 Expenditure per student, second (% of GDP per capita) Public health expenditure per capita Source: World Bank calculations using WDI data 4.35. Further improvement in the efficiency of spending on health exists, although the distance to the frontier is short in this sector. The amount of input is public health expenditure per capita relative to the infant survival per 1000 live births. Comparing Kenya’s spending relative to outcomes, Figure 4-24 shows that there is room for efficiency gains in terms of infant survival per 1000 live births, where Kenya records high infant mortality rates relative to Bangladesh (BLD), even though both countries spend about the same amount in public health expenditure per capita. 4.36. The above comparisons of spending (inputs) relative to outcomes suggests room for potential fiscal savings or re-allocation to obtain government objectives under the human capital development sector. Nonetheless, a detailed sectoral approach to examine efficiency and effectiveness of spending in both education and health could be warranted to quantify potential fiscal savings and re-allocation to improve outcomes. This could be undertaken in the context of a fully-fledged PER with access to decision making units data (or micro-level data). Weaknesses in public investment management system reduce the efficiency of capital spending 4.37. The challenges in project selection, procurement planning and implementation delays could be associated with relatively low public investment outcomes. Despite a ramp-up in development spending, public investment remains low relative to peer countries. Comparing Kenya’s Gross Fixed Capital Formation (GFCF) as a percentage of GDP relative to per capita GDP growth suggests room for improving the efficiency of investment spending. At about 20 percent of GDP, Kenya’s gross fixed capital formation is associated with per capita GDP growth of about 0.17 percent. This is lower compared to Israel (ISL) at the frontier that spends almost similar (20% of GDP) and gets a 12.2 percent growth in per capita GDP 54 (Figure 4-25). This suggests opportunities for efficiency gains in both public and private investment and ultimately in the total capital stock. Figure 4-25: Kenya could achieve faster growth Figure 4-26: Kenya’s public investment and access in its GDP per capita, given its GFCF level to electricity (2014-2016) 5 BEL CAN FIN 100 NZL SGP LCN TLA TEA LAC IDN JAM FJI LCA ESP IRL EAS EAP ATG SLV CSS DNK MLT SWE MYS SAU IBDPER LKA MDV NLD MHL BLZ GRD GTM MIC PHL GNQ HUN AUS KOR HND GDP per capita growth(in log) WLD NPL Access to electricity (% of population) LTU ZAF KIR LAO ISL THA IBT LMY JPN NIC AFG 80 MUS FRA MNG SLV LUX PSS NGA GBR URY BGR HRV COG LBY POL PER UKR CRI USA SYC CPV FSM ARG BRA MNG TUR BHR COL BWA MRT BGD CYP DEU BGD GNQ CHE BRN GRC BRBPRT ZAF NOR ECU CIV TLS BHS SEN 60 RUS IRQ KWT 0 GNBZWE GMB KHM ALB SEN CMR BWA ITA TUN MEX JOR AUT CZE LKA MDA NIC KGZ MMR BFABLZ BOL EST UGA TGO ISR DZA EGY NAM CHN PHL GHA STP IND OMN NAM KHM GIN LSO VUT SLB ERI COD PRY ARM LBN KAZ IDN ZMB TGO KEN MAR BLR GAB 40 MDG CIV CHL AGO KENJAM TJK LVAROU AGO COM FJI DOM HND MOZ GEO TZA MLI MLI IRN NER BTN SWZ PAK RWA TCD HTI ZMB CAF LSO MMR VNM DJIETH MWI TZA RWA PNG UGA 20 BEN BFA SLE SLE NPL MWI -5 BDI 0 10 20 30 40 50 60 GFCF(in% of GDP) 0 5 10 15 20 Net investment in nonfinancial assets(%GDP) Source: World Bank staff calculations based on WEO-IMF. 4.38. There is scope for improving the efficiency of development spending. Given limited fiscal space, improving the efficiency of spending provides the best chance to advance infrastructure investment needs in Kenya. For instance, relative to other lower-middle income economies with comparable levels of public investment as a share of GDP (5.8 percent of GDP), Kenya is achieving less in terms of access to electricity (about 44.5%) (Figure 4-26), although recent data shows that as of 2017, percent of electricity access was about 70 percent of the population. Comparators such as Senegal (SEN), Namibia (NAM) and South Africa (ZAF) spends way below what Kenya is spending on public investment but have achieved high access to electricity (%of population). This indicates presence of some scope to improve efficiency in development spending, but its actual quantification calls for detailed analysis, beyond this report. A sector deep-dive analysis of development spending at the project level could reveal ways to improve the efficiency and effectiveness of development spending. This could also provide an estimate on time and cost over-runs per project, project design, appraisal, selection and implementation issues. 4.39. Further, investment in transport infrastructure faces several challenges. Land acquisition makes it expensive to develop transport infrastructure, especially due to the need to pay a market value for land easement. As a consequence, provisions for land acquisition and wayleave constitute a large share of the budget. Other challenges include encroachment on transport wayleave; inadequate financing; infrastructure vandalism; increased urbanization; lengthy procurement procedures; inadequate skilled manpower in transport management and planning and inadequate pathways for Non-Motorized Transport (NMT). Under the road sub-sector, the backlog on road maintenance, and delayed absorption of donor funded projects constitute critical spending impediments. F. Conclusion and Recommendations 4.40. What are the key messages from this chapter? The chapter shows that government expenditures are allocated well (to infrastructure and human capital) but there is scope to improve outcomes from the use of these resources. Public spending has had a good impact—even better than 55 peers (as evidenced by the improved quality of infrastructure, and improved outcomes on education and health). Nonetheless, weaknesses in the public investment management system reduce the technical efficiency of capital spending and could lead to delays in project implementation. Furthermore, holding up payments to vendors could be delaying business to business transaction and making the cost of delivering projects very high, as contractors inbuilt risks of delayed payments into their bids. Delays in public payments can affect private sector liquidity, profits, and ultimately economic growth. There is need, therefore, to clear any pending bills/arrears, and reduce costs and time over-runs for development projects, and enhance efficiency of public spending. 4.41. Improve the regulatory environment and put in place appropriate incentive structure to achieve some of the flagship projects under the Big 4 through PPPs. To counteract the drag from fiscal consolidation, policy measures could be put in place to stimulate the private sector’s contribution. For instance, while the state slows down on development spending a regulatory environment and incentive structure could be enhanced (within the PPP unit at the NT) to achieve flagship infrastructure projects through public private partnership initiatives. This could provide an avenue where the State does not have to act through public bodies but through private entities to advance spending on infrastructure such as roads, which are financed through tolls and vignettes. However, this needs to be undertaken within a fiscal sustainable way since PPPs are primarily a risk-sharing mechanism and not a mechanism to reduce government fiscal burden. If the risks are not well shared and the government takes on excessive contingent liabilities, then the impact on overall fiscal space could be negative rather than positive. 4.42. Further, with up to 68 percent of the central government budget being on items of high to medium rigidity, this could undermine government’s ability to re-allocate resources to priority sectors. spending on transfers (to counties, SoEs, Parliament, and the Judiciary), interest payments, and compensation of employees presents a large portion of expenditures that is implicitly non-discretionary and undermines ability of the government to re-allocate resources to say the Big 4 priority sectors. A medium to long term strategy may be needed to shift the rigidity of spending through either enhanced own source revenue of the counties, assessing the financial viability of some SOEs and improved debt management to reduce the interest rate obligations through a plan to retire short-term and expensive commercial debt. 4.43. Improve budget execution, absorption of development budget, and disbursement of funds from the exchequer. Actual amounts spent are only a fraction of what is planned in the budget. About 31 percent of the budget is not absorbed making it hard for the government to achieve its development objectives. Delays in budget execution suggests absorption capacity constraints which calls for capacity building in project appraisal, selection, planning, budgeting and execution across all levels of government. A synchronized project implementation cycles to budgeting and faster exchequer releases could contribute to enhanced credibility of the multi-year budget (and the medium-term expenditure framework). 4.44. Finally, there is scope for improvement on outcomes realized relative to inputs in terms of public spending in education, health and physical infrastructure. This result is derived from a cross- country comparison on the level of outcome achieved relative to inputs. Nonetheless, with resources availability, there is need for a sector- deep dive to quantify exact amount of fiscal savings and potential re-allocation both within and across- sectors to realize government objectives. 56 4.45. The following matrix provide a summary of some of the policy options that could be considered to improve public expenditure management and create needed fiscal space for the Big 4 (Table 4-3). Table 4-3: Priorities for Improved Public Expenditure Management Reform Priority Short-run (6 months-1 year) Medium-term (2-3years) Ensure that the ongoing projects • Project management • Curb rising pending bills and are completed before launching committee should ensure the arrears in payments at both new projects and clear any pending ongoing infrastructure project county and national bills and arrears owed to suppliers. are completed and suppliers governments25. This paid within the specified constitutes a prudent fiscal timelines for optimal returns surveillance program for any to investment and to spur country. private sector activity. • A phased-out approach could be adopted on clearing outstanding arrears, given tight fiscal space. Crowding in Private sector • Fast-track appraisal, selection, • Review and update the contribution to delivery of feasibility, and implementation regulatory environment and infrastructure projects. schedule for the list of pipeline incentive structure (risk projects under PPP. sharing mechanism) to achieve flagship infrastructure projects through public private partnership initiatives. Reduce Rigidity in recurrent • Improved debt management • Enhance own source revenue spending (especially interest could reduce interest mobilization at the county payments and transfers to payments through a plan to level to reduce dependency on counties, and SoEs; as well as retire short-maturity and transfers from the national compensation to employees). expensive commercial debt. government. • Clean and update the payroll register • Consolidate personnel allowances to the base wage to prevent distorting the remuneration structure, Improve budget execution and • Synchronized project • Build staff capacity in project absorption of development budget implementation cycles to appraisal, selection, planning, budgeting and fast-track budgeting and execution across exchequer releases. all levels of government. 25 The 2018 enterprise survey for Kenya finds that approximately 12 percent of the 1,001 firms surveyed (or 120 firms) have had a contract with government that was in arrears (Kenya Enterprise Survey, 2018). 57 References Cetrángolo, Oscar; Jiménez, Juan Pablo; Ruiz del Castillo, Ramiro. 2010. Rigidities and Fiscal Space in Latin America: A Comparative Case Study. Macroeconomía del Desarrollo Serie. Santiago, Chile April 2010 https://repositorio.cepal.org/bitstream/handle/11362/5470/1/S2009411_en.pdf Chacha,P. 2018. Firm level export dynamics and market access costs: evidence from Kenya. PhD thesis. University of Cape Town. South Africa. (https://open.uct.ac.za/browse?value=Chacha%2C+Peter+Wankuru&type=author) Checherita-WestPhal, C., Klemm, A, and Viefers, P. 2016. Government’s payment discipline: The Macroeconomic impact of public payment delays and arrears. Journal of Macroeconomics, 47, PP.147- 165. Girouard, N., & André, C. (2005). Measuring cyclically-adjusted budget balances for OECD countries, OECD Economics Department Working Papers, No 434. OECD Publishing. do i, 10, 787626008442. IMF. 2018. Kenya: Staff Report for the 2018 Article IV Consultation and Establishment of Performance Criteria for the Second Review Under the Stand-by Arrangement. Washington, DC. International Budget Partnership. 2018. Glass Half Full? Kenya’s County Budget Implementation in 2015/16. May 2017. Nairobi. Merotto, Dino Leonardo; Hayati, Fayavar; Stephan, David Andrew; Battaile, William G. 2015. Dismal science, accounting and Newton’s second law: identifying force and rigidity in public expenditure analysis. Policy Research Working Paper. The World Bank. (http://documents.worldbank.org/curated/en/738741467998792031/pdf/WPS7431.pdf) Nauschnigg, F. (2006). Macroeconomic policymaking–lessons from Austro-Keynesianism. Konferenzband, Delivering the Lisbon goals–The role of macroeconomic policy; Europäischer Gewerkschaftsbund und European Trade Union Institute. Nauschnigg, F. (2010). Better European financial architecture to prevent crises. After the crisis: towards a sustainable growth model, 87. Republic of Kenya. 2018. Third Medium Term Plan 2018-2022. Transforming Lives: Advancing socio- economic development through the “Big Four�. The National Treasury and Planning. Nairobi. Salary and Remuneration Commission Report 2016 Salary and Remuneration Commission; Public Sector Wage Bill Study-Lessons learnt from corporations and County governments 2018 The National Treasury. 2016. 2017 Medium Term Budget Policy Statement. November 2016. Nairobi The National Treasury. 2019. 2018 Medium Term Budget Policy Statement. January 2019. Nairobi Van der Waaij, D., & Nord, C. E. (2000). Development and persistence of multi-resistance to antibiotics in bacteria; an analysis and a new approach to this urgent problem. International journal of antimicrobial agents, 16(3), 191-197. 58 Vega, Susana. 2005. “Inflexibilidades: Rigideces y Afectaciones Presupuestarias y Otros Aspectos. Experiencia Argentina.� Estudio preparado para XXXII Seminario Internacional de Presupuesto Público, Lima, 25-28 de abril. http://www.cepal.org/ilpes/noticias/paginas/5/22145/inflexibilidades%20experiencia%20argentina%20 ultimo.pdf World Bank. 2017. Kenya Economic Update Edition 16: Poised to Bounce Back? Reviving Private Sector Credit Growth and Boosting Revenue Mobilization to Support Fiscal Consolidation. World Bank. 2017. Kenya Tax Policy Studies: Value Added Tax & Corporate Income Tax World Bank. 2018. Kenya Economic Update Edition 17: Policy Options to Advance the Big 4: Unleashing Kenya’s Private Sector to drive Inclusive Growth and Accelerate Poverty Reduction. World Bank. 2018. Kenya Economic Update Edition 18: In Search of Fiscal Space: Government Spending and Taxation: Who Benefits? World Bank. 2019. Kenya Economic Update Edition 19. Unbundling the Slack in Private Sector Investment: Transforming Agriculture Sector Productivity and Linkages to Poverty Reduction. 59 Annex 1: National Fiscal Operations (FY2013/14-2017/18) Table A1-1: National Fiscal Out-turn, 2013/14-2017/18 (in % of GDP, unless otherwise stated) 2013/14 2014/15 2015/16 2016/17 2017/18 Percent of GDP Revenue and Grants 19.7 19.5 19.2 18.9 17.6 Total Revenue 19.2 19.0 18.7 18.6 17.3 Tax revenue 18.1 17.7 17.7 17.1 15.5 Income tax 8.9 8.7 8.6 8.2 7.3 VAT 4.6 4.5 4.4 4.4 4.1 Import Duty 1.3 1.3 1.2 1.2 1.1 Excise Duty 2.0 2.0 2.1 2.2 1.9 Other Revenues 1.3 1.3 1.3 1.1 1.2 Railway Levy 0.0 0.0 0.0 0.2 0.2 Appropriation in Aid 1.1 1.3 1.0 1.3 1.6 Grants 0.5 0.5 0.5 0.4 0.3 Expenditure and Net Lending 25.6 28.1 27.2 27.6 24.4 Recurrent 14.8 14.8 15.6 15.2 15.4 Wages and salaries 5.5 5.1 4.7 4.4 4.4 Interest Payments 2.7 2.9 3.3 3.5 3.7 Other recurrent 6.6 6.7 7.5 7.3 7.2 Development and net lending 6.3 8.7 7.0 8.4 5.3 County allocation 3.8 3.9 4.1 3.7 3.5 Contingencies 0.0 0.1 0.1 0.1 0.1 Parliamentary Service 0.4 0.4 0.3 0.3 0.3 Judicial Service 0.3 0.2 0.2 0.2 0.1 Deficit (incl.grants cash basis) -6.1 -8.1 -7.3 -8.8 -6.8 Financing 6.1 8.1 7.3 9.1 7.2 Foreign Financing 4.0 4.4 3.1 4.1 3.1 Domestic Financing 2.1 3.7 4.1 5.0 4.0 Total Public Debt(gross) 47.8 48.8 55.5 57.5 57.2 GDP (Fiscal year, Ksh. billion) 5,074 5,828 6,566 7,658 8,793 Source: The National Treasury 60 Annex 2: Fiscal and debt sustainability risk scenarios Table A2-1: Business as usual Scenario (in % of GDP, unless otherwise stated) Historical Projections 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 GDP at current prices (KSH billion) 7,194 7,749 8,252 9,411 10,688 12,071 13,636 15,434 17,407 19,632 GDP at constant prices, annual growth rate (%) 5.7 5.9 4.9 6.3 6.2 6.3 6.6 6.9 6.5 6.5 GDP deflator, annual growth rate (%) 10.0 8.1 8.4 8.0 7.0 6.2 6.0 5.9 5.9 5.9 Exchange Rate KSH/USD 102.3 102.5 103.2 101.3 -- -- -- -- -- -- Real Exchange Rate, Index 2015=100 (1) 100.0 93.8 88.5 82.3 -- -- -- -- -- -- Revenue 16.7 17.5 18.3 18.2 18.2 18.2 18.2 18.2 18.2 18.2 Direct Taxes (Tax on Income & Profits) 7.4 7.7 7.7 7.5 7.5 7.5 7.5 7.5 7.5 7.5 Indirect Taxes (Taxes on G&S and Int'l Trade) 6.7 7.1 7.3 7.4 7.4 7.4 7.4 7.4 7.4 7.4 Grants (2) 0.4 0.4 0.3 0.4 0.4 0.4 0.4 0.4 0.4 0.4 Non-Tax Revenue (3) 2.2 2.4 2.9 2.9 2.9 2.9 2.9 2.9 2.9 2.9 Expenditure 23.4 25.1 25.8 24.8 24.7 24.9 24.9 25.0 25.1 25.1 Wages and Salaries 4.2 4.2 4.4 4.3 4.3 4.3 4.3 4.3 4.3 4.3 Goods and Services 6.1 6.9 7.2 7.1 7.1 7.1 7.1 7.1 7.1 7.1 Transfers to Counties 3.7 4.1 3.8 3.7 3.7 3.7 3.7 3.7 3.7 3.7 Other Expenses 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Interest Payments 2.7 3.2 3.6 3.8 3.8 3.9 3.9 4.0 4.1 4.1 Development Expenditure (4) 6.7 6.9 6.7 5.8 5.8 5.8 5.8 5.8 5.8 5.8 Primary Balance -4.0 -4.5 -3.9 -2.8 -2.8 -2.8 -2.8 -2.8 -2.8 -2.8 Overall Balance -6.7 -7.6 -7.5 -6.5 -6.5 -6.7 -6.7 -6.7 -6.8 -6.9 Gross Borrowing Requirements 17.9 19.3 20.4 19.3 18.6 16.2 15.7 15.9 15.4 19.5 Overall Balance (+ indicates deficit) 6.7 7.6 7.5 6.5 6.5 6.7 6.7 6.7 6.8 6.9 Amortizations Payments 9.8 10.4 12.7 13.0 13.1 10.2 8.9 9.0 8.4 12.4 Other Funding Needs 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Net Acquisition of Fin.Assets (e.g., Deposits) (5) 1.4 1.2 0.2 -0.3 -1.1 -0.7 0.2 0.2 0.2 0.2 Borrowing Sources 17.9 19.3 20.4 19.3 18.6 16.2 15.7 15.9 15.4 19.5 Issuance of Domestic Debt (% share) (6) 80.0 80.0 80.0 80.0 80.0 80.0 80.0 Issuance of External Debt (% share) 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Public Debt 44.9 49.4 55.5 56.1 56.5 56.8 57.7 58.4 59.4 60.3 Domestic Debt 28.7 30.3 30.7 31.9 32.8 33.8 35.1 External Debt 27.3 26.3 26.1 25.8 25.7 25.6 25.1 Notes: (1) Real exchange rate defined as the MKW/USD bilateral exchange rate times the ratio between int'l prices and domestic GDP deflator. (2) Includes program grants, project grants, and other grants. (3) Includes … (4) Includes net lending. (5) Excludes net lending. (6) Includes domestic debt issued to rollover maturing short-term liabilities. Source: World Bank projections. 61 Table A2-2: Lower Real GDP growth Scenario (in % of GDP, unless otherwise stated) Historical Projections 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 GDP at current prices (KSH billion) 7,194 7,749 8,252 9,411 10,695 11,988 13,406 14,978 16,734 18,696 GDP at constant prices, annual growth rate (%) 5.7 5.9 4.9 6.3 6.2 5.5 5.5 5.5 5.5 5.5 GDP deflator, annual growth rate (%) 10.0 8.1 8.4 8.0 7.0 6.3 6.0 5.9 5.9 5.9 Exchange Rate KSH/USD 102.3 102.5 103.2 101.3 -- -- -- -- -- -- Real Exchange Rate, Index 2015=100 (1) 100.0 93.8 88.5 82.3 -- -- -- -- -- -- Revenue 16.7 17.5 18.3 18.2 18.8 18.9 19.0 19.1 19.1 19.1 Direct Taxes (Tax on Income & Profits) 7.4 7.7 7.7 7.5 7.8 7.9 8.1 8.3 8.3 8.3 Indirect Taxes (Taxes on G&S and Int'l Trade) 6.7 7.1 7.3 7.4 7.7 7.8 7.9 7.9 7.9 7.9 Grants (2) 0.4 0.4 0.3 0.4 0.5 0.4 0.4 0.4 0.4 0.4 Non-Tax Revenue (3) 2.2 2.4 2.9 2.9 2.9 2.8 2.7 2.6 2.6 2.6 Expenditure 23.4 25.1 25.8 24.8 24.4 23.6 23.1 23.2 23.3 23.5 Wages and Salaries 4.2 4.2 4.4 4.3 4.2 4.0 3.8 3.5 3.6 3.6 Goods and Services 6.1 6.9 7.2 7.1 6.9 6.8 7.1 7.5 7.5 7.6 Transfers to Counties 3.7 4.1 3.8 3.7 3.4 3.1 2.8 2.6 2.6 2.6 Other Expenses 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Interest Payments 2.7 3.2 3.6 3.8 3.8 3.8 3.7 3.7 3.6 3.6 Development Expenditure (4) 6.7 6.9 6.7 5.8 6.1 5.8 5.8 5.9 6.0 6.0 Primary Balance -4.0 -4.5 -3.9 -2.8 -1.9 -0.9 -0.4 -0.4 -0.6 -0.8 Overall Balance -6.7 -7.6 -7.5 -6.5 -5.6 -4.8 -4.1 -4.1 -4.2 -4.3 Gross Borrowing Requirements 17.9 19.3 20.4 19.3 17.6 13.9 12.3 11.9 11.1 15.2 Overall Balance (+ indicates deficit) 6.7 7.6 7.5 6.5 5.6 4.8 4.1 4.1 4.2 4.3 Amortizations Payments 9.8 10.4 12.7 13.0 13.1 9.9 8.0 7.6 6.7 10.6 Other Funding Needs 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Net Acquisition of Fin.Assets (e.g., Deposits) (5) 1.4 1.2 0.2 -0.3 -1.1 -0.7 0.2 0.2 0.2 0.2 Borrowing Sources 17.9 19.3 20.4 19.3 17.6 13.9 12.3 11.9 11.1 15.2 Issuance of Domestic Debt (% share) (6) 80.0 80.0 80.0 80.0 80.0 80.0 80.0 Issuance of External Debt (% share) 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Public Debt 44.9 49.4 55.5 56.1 55.6 54.5 53.5 52.7 52.0 51.6 Domestic Debt 28.7 29.5 28.8 28.6 28.4 28.4 29.0 External Debt 27.3 26.1 25.7 24.9 24.3 23.7 22.6 Notes: (1) Real exchange rate defined as the MKW/USD bilateral exchange rate times the ratio between int'l prices and domestic GDP deflator. (2) Includes program grants, project grants, and other grants. (3) Includes … (4) Includes net lending. (5) Excludes net lending. (6) Includes domestic debt issued to rollover maturing short-term liabilities. Source: World Bank projections. 62 Table A2-3: High cost of domestic borrowing scenario (in % of GDP, unless otherwise stated) Historical Projections 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 GDP at current prices (KSH billion) 7,194 7,749 8,252 9,411 10,688 12,071 13,636 15,434 17,407 19,632 GDP at constant prices, annual growth rate (%) 5.7 5.9 4.9 6.3 6.2 6.3 6.6 6.9 6.5 6.5 GDP deflator, annual growth rate (%) 10.0 8.1 8.4 8.0 7.0 6.2 6.0 5.9 5.9 5.9 Exchange Rate KSH/USD 102.3 102.5 103.2 101.3 -- -- -- -- -- -- Real Exchange Rate, Index 2015=100 (1) 100.0 93.8 88.5 82.3 -- -- -- -- -- -- Revenue 16.7 17.5 18.3 18.2 18.8 18.9 19.0 19.1 19.1 19.1 Direct Taxes (Tax on Income & Profits) 7.4 7.7 7.7 7.5 7.8 7.9 8.1 8.3 8.3 8.3 Indirect Taxes (Taxes on G&S and Int'l Trade) 6.7 7.1 7.3 7.4 7.7 7.8 7.9 7.9 7.9 7.9 Grants (2) 0.4 0.4 0.3 0.4 0.5 0.4 0.4 0.4 0.4 0.4 Non-Tax Revenue (3) 2.2 2.4 2.9 2.9 2.9 2.8 2.7 2.6 2.6 2.6 Expenditure 23.4 25.1 25.8 24.8 24.4 23.8 23.2 23.0 23.0 22.9 Wages and Salaries 4.2 4.2 4.4 4.3 4.2 4.0 3.7 3.4 3.4 3.4 Goods and Services 6.1 6.9 7.2 7.1 6.9 6.8 6.9 7.3 7.3 7.3 Transfers to Counties 3.7 4.1 3.8 3.7 3.4 3.1 2.8 2.5 2.5 2.5 Other Expenses 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Interest Payments 2.7 3.2 3.6 3.8 3.8 4.2 4.1 4.0 4.0 3.9 Development Expenditure (4) 6.7 6.9 6.7 5.8 6.1 5.8 5.7 5.7 5.7 5.7 Primary Balance -4.0 -4.5 -3.9 -2.8 -1.9 -0.8 0.0 0.2 0.2 0.2 Overall Balance -6.7 -7.6 -7.5 -6.5 -5.6 -5.0 -4.1 -3.9 -3.8 -3.8 Gross Borrowing Requirements 17.9 19.3 20.4 19.3 17.7 14.1 12.3 11.5 10.5 14.0 Overall Balance (+ indicates deficit) 6.7 7.6 7.5 6.5 5.6 5.0 4.1 3.9 3.8 3.8 Amortizations Payments 9.8 10.4 12.7 13.0 13.1 9.8 8.0 7.5 6.5 10.1 Other Funding Needs 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Net Acquisition of Fin.Assets (e.g., Deposits) (5) 1.4 1.2 0.2 -0.3 -1.1 -0.7 0.2 0.2 0.2 0.2 Borrowing Sources 17.9 19.3 20.4 19.3 17.7 14.1 12.3 11.5 10.5 14.0 Issuance of Domestic Debt (% share) (6) 80.0 80.0 80.0 80.0 80.0 80.0 80.0 Issuance of External Debt (% share) 20.0 20.0 20.0 20.0 20.0 20.0 20.0 Public Debt 44.9 49.4 55.5 56.1 55.6 54.3 52.9 51.3 50.0 48.8 Domestic Debt 28.7 29.5 28.8 28.3 27.7 27.2 27.3 External Debt 27.3 26.1 25.6 24.6 23.7 22.8 21.5 Notes: (1) Real exchange rate defined as the MKW/USD bilateral exchange rate times the ratio between int'l prices and domestic GDP deflator. (2) Includes program grants, project grants, and other grants. (3) Includes … (4) Includes net lending. (5) Excludes net lending. (6) Includes domestic debt issued to rollover maturing short-term liabilities. Source: World Bank projections. 63