Report No. ACS4446 Poland: Reforming Government Pay Setting Practices June 11, 2013 Poverty Reduction and Economic Management Unit Central/South Europe and Baltics Country Unit Europe and Central Asia Region Document of the World Bank CURRENCY EQUIVALENTS Currency Unit 1 US$= PLN 3.18 WEIGHTS AND MEASURES The Metric System is used throughout the report. FISCAL YEAR January 1 to December 31 ABBREVIATIONS AND ACRONYMS BESTA Public Sector Pay Statistical System (Sweden) DG Director General (Poland) ESA European System of National and Regional Accounts MOF Ministry of Finance NACE European Classification of Economic Activities SACO Confederation of Professional Associations (Sweden) SAGE Agency for Government Employers (Sweden) SAMV Agency for the Public Sector as Employer (Finland) TCO Federation of Professional and Salaried Employees (Sweden) Vice President: Philippe H. Le Houerou Country Director: Mamta Murthi Sector Director: Yvonne Tsikata Sector Manager: Satu Kahkonen Task Team Leader: William Dillinger and Ewa Korczyc . Standard Disclaimer: . This volume is a product of the staff of the International Bank for Reconstruction and Development/ The World Bank. The findings, interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. . Copyright Statement: . The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/ The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. 2 This report is based on a series of consultations with the Government of Poland in 2012. The report was prepared by William Dillinger, Ewa Korczyc and Knut Rexed (consultant). The peer reviewers were Nick Manning (PRMPS), Robert Sierhej (Senior Economist, IMF Regional Office for Central and Eastern Europe and Baltics) and Zsuzsanna Lonti (Head of the Statistics and Indicators Unit in the Budgeting and Public Expenditures Division in the OECD Governance Directorate). The report benefitted from the assistance of a range of Polish counterparts, including staff of the Ministry of Finance and the Civil Service Department at the Chancellery of the Prime Minister. Table of Contents Executive Summary ................................................................................................4 Introduction .............................................................................................................6 The Current Wage Determination Process .............................................................7 Does the System Succeed at Controlling the Aggregate Wage Bill? ...................13 Are Salaries High Enough to Attract and Retain Qualified Staff? .......................15 Is Wage Setting Fair? Does it Motivate Staff? .....................................................17 Directions for Reform ...........................................................................................17 Implementation .....................................................................................................24 Annex 1. Glossary .................................................................................................28 Annex 2: Distribution of Wage Coefficients by Title, Polish Civil Service ........29 Annex 3: Examples of Position Classification Systems .......................................30 Cover photo: the building of the Ministry of Finance in Warsaw. The photo was obtained from the Ministry of Finance. 3 Executive Summary 1. This report is a response to a request from the Polish Ministry of Finance (MOF) for technical advice on Poland’s system of public sector pay. The Ministry’s primary aim is to contain the public sector wage bill as part of an overall effort to reduce the fiscal deficit. But the Ministry is also concerned with more fundamental structural issues in the system of pay setting: that individual pay levels for some positions may be higher than is necessary to attract and retain qualified staff and, in other cases, may be too low. And that the absence of clear rules for job grading, pay setting, and promotion may undermine staff morale. 2. This initial report looks only at a subset of public employees—the so-called budget sphere. This includes all staff paid from the central government budget—including central government civil servants and military personnel—but excludes staff paid by local governments. These include the majority of teachers and medical personnel. A more detailed analysis of the wage setting process looks at a subset of the budget sphere: only civil servants. Note that the decision to focus on the civil service was dictated by the resource constraints of this study and the scarcity of data on public employment outside the civil service. It does not imply that problems in the civil service are more severe than elsewhere in the public sector. 3. The report focuses on three questions. The first is: Does the System Succeed in Controlling the Aggregate Wage Bill? 4. To answer this question, the report examines trends in the wage bill over time (as a percent of GDP) and recent Government policies that have affected the wage bill. It concludes that the current system does succeed in controlling the aggregate wage bill, largely as a result of the tight control that the MOF exercises over the wage bill envelope of each ministry. Recently this has been reinforced by an EU-imposed ‘excessive deficit procedure’ (EDP). To comply with the EDP, the Government has imposed a comprehensive wage freeze. Once Poland meets the EDP deficit targets, however, the EDP will be withdrawn and this source of external control will no longer exist. To replace it, the Government proposes to impose its own fiscal responsibility rule, which will limit the growth of central government expenditures to the growth rate of GDP. Since the ceiling is self-imposed, however, it can be readily overturned. The report recommends instead that the Government index wages to inflation or a combination of wages and GDP. Are Salaries High Enough to Attract and Retain Qualified Staff? 5. To answer this question, the report compares wages in the public sector with those in the private sector (for equivalent positions). It also examines key indicators: the number of applications per advertised vacancy (which would indicate the attractiveness of public employment to potential recruits) and the turnover rate among existing staff (which would indicate their dissatisfaction with current employment conditions). The report finds that, for most positions, public sector compensation tends to match private sector levels. High-level technical and managerial positions, however, tend to be underpaid relative to their private sector equivalents. Nevertheless, the public sector has no difficulty attracting and retaining such staff. 4 This is in part due to the current economic slowdown, which limits employment opportunities in the private sector. As the economy recovers, however, the government may have difficulty attracting and retaining such staff, unless it increases salaries for such positions. Is Wage Setting Fair? Does it Motivate Staff? 6. Although the current wage setting system does succeed in controlling the aggregate wage bill and attracting and retaining qualified staff (at least for the time being) more fundamental problems remain. These derive from the wage setting process itself. At present, the salary of a given staff in the civil service is derived from: (1) a base reference wage, which applies to all civil servants and is adjusted annually, and (2) a so-called multiplier which applies to the individual staff and is applied to the base reference wage to yield a base monthly salary. Within a given ministry, the manager of each unit has the authority to propose a change in the wage coefficients (the multipliers) of individual employees, subject only to the approval of the ministerial personnel office (the director general) and very broad floors and ceilings on the coefficients. 7. There are, in theory, certain advantages to this highly flexible system of wage setting. Each unit manager can adapt individual salary levels to local labor market conditions and can reward high-performing staff. But this arrangement places a high degree of reliance on the ability of the unit managers to make these judgments—to decide who receives an increase in multiplier and who does not. The report analyzes the dispersion of coefficients for specific positions and concludes that this discretion may be being abused. If so, this is likely to harm staff morale and the ability of government to attract people who hope to pursue a career in the public sector. 8. To identify solutions, the report examines the wage setting practices of selected countries in the region, focusing in particular on the degree of autonomy enjoyed by managers at the lowest level of administration. Practices in these countries run the gamut from near total autonomy (e.g. Sweden) to no autonomy at all (Germany). Based on the specific characteristics of Poland’s situation, the report recommends a middle ground, similar to that of Finland and Lithuania. 5 Introduction 1. The Ministry of Finance of the Government of Poland has requested technical assistance from the World Bank to assist it in reforming the system of public sector pay. The Ministry’s primary aim is to contain the public sector wage bill as part of an overall effort to reduce the fiscal deficit. But it also sees the current fiscal situation as an opportunity to address more fundamental structural problems in the wage setting process. 2. Overall, the size of employment in public administration and wage spending in Poland is not particularly high by European standards. As shown in Figure 1, the number of staff employed in government functions (excluding front-line service providers such as teachers and health workers) is about 28 per 1000 inhabitants in Poland, at the low end of the European range and somewhat below the average for the EU27. As shown in Figure 2, the wage bill in Poland, at about ten percent of GDP, is again toward the lower end of the European distribution. Over time, spending on the wage bill has remained fairly stable as a percent of GDP, slightly exceeding it in the recession years early in the decade and then steadily falling during most of the subsequent boom years. Wage increases and slowing economic growth in 2008-2009 prompted an increase in the wage bill (relative to GDP) but this has been reversed through a combination of a wage freeze imposed in 2010 and the return of moderate economic growth. Figure 1. Employment in Public Administration Figure 2. Public wage bill as percent of GDP, per 1,000 inhabitants, 2011 2011 60 20 18 50 16 14 40 12 30 10 8 20 6 4 10 2 0 0 NL EE ES EL EE ES EL NL LU PL LT HU IT PT IE IE IT PT HU SI FI RO LV PL AT LU EU27 SK AT LV MT LT DK SK CZ RO DE BG UK BE FR SE DK FI SI SE CY DE BE CY EU27 UK BG CZ MT FR Source: Eurostat, 2012. Source: Eurostat 2012 Notes: Based on labor force surveys. The term ‘public Notes: Based on government expenditure data. The figure administration’ includes staff employed in public includes wages and salaries paid by both central and administration, defense and compulsory social security. It subnational governments, as well as their respective social excludes teachers, health care workers, and members of security contributions. the judiciary. The figure includes staff employed by both the central government and subnational governments. For further details, see glossary in Annex. 3. The Ministry is nevertheless concerned that with the easing of the country’s fiscal situation, pressures will arise to increase wages again. It is also concerned with more fundamental structural issues in the system of pay setting: that individual pay levels for some positions may be higher than is necessary to attract and retain qualified staff and, in other cases, 6 may be too low. And that the absence of clear rules for job grading, pay setting, and promotion may undermine staff morale. 4. This initial phase of the work covers only a subset of public employment: those employees working in what is termed the budget sphere. As such, it excludes two large blocks of public employees: those in decentralized units of administration (including the health care system) and those working for local governments (including the majority of teachers). It also excludes employees of social security funds and units of higher education. What it does cover is the staff of central government ministries and agencies, including managerial, administrative, technical, and support staff, as well as staff in the uniformed services (professional soldiers, police, firefighters and other emergency personnel) and the judiciary (judges and prosecutors). As shown in Table 1, the staff covered by this report constitute about 25 percent of total public employment. The Current Wage Determination Process 5. The process of determining wages for these groups is complicated. Poland’s legal framework makes the conventional distinction Table 1. Public Sector Employment, between civil servants and non-civil servants (termed (thousands, 2010) ‘non-multiplier’ staff). But the legislation also designates other groups of staff for separate systems State budgetary sphere 602 of wage determination: so-called high ranking army, police 250 government officials, the uniformed services, and the civil service 122 judiciary (judges and prosecutors). chancellery 30 judges, prosecutors 70 6. Civil Service The base salary of staff in the customs 10 civil service is derived from: (1) a base reference non-multiplier staff 120 wage, which applies to all civil servants and is adjusted annually, and (2) a so-called multiplier Subnational governments 760 of which teachers 510 which applies to individual staff and is applied to the Social security funds 60 base reference wage to yield a base monthly salary. State agencies 20 In addition to the base wage, staff also receive Higher education 80 compensation in the form of: (1) a length-of-service Other agencies, inc. health system 1040 bonus (a percentage increase in monthly salary for each year of service up to 20 years), (2) an TOTAL 2562 anniversary bonus (a one-time bonus granted to staff Source: Central Statistical Office, KPRM, after 20 and 30 years of employment), and (3) a World Bank staff estimates thirteenth salary. Taken together, the base salary and these three bonuses constitute about 85 percent of total compensation. Staff may also receive: (1) an award for particular achievements in professional work, (2) a bonus for belonging to the civil service cadre (see Box 1), (3) a task bonus (for temporarily taking on additional work, typically on behalf of staff on leave), and (4) hazard pay (which, in the case of civil servants, applies to tax collectors in charge of delinquent accounts and staff working in adverse environments, such as mines). These latter bonuses are not a large part of compensation. The award for professional achievements constitutes about eight percent of compensation and the remaining bonuses, about seven percent. 7 Box 1. The Civil Service Cadre Within the civil service, Polish legislation provides for a ‘civil service cadre’. Status as a member of the cadre is assigned to individuals, not to positions. Thus one civil service position may be filled by a member of the cadre while another, identical, one is filled by an ordinary civil servant. When established in the late 1990s, the “civil service cadre� was supposed to become a elite core, constituting around 20 percent of the total civil service. Currently, the cadre has only 8,000 members, constituting about six percent of all civil servants. Entrance to the cadre is by examination and nomination. Nomination to the civil service cadre is made upon successful completion of a qualification exam or graduation from the National School of Public Administration (KSAP: Krajowa Szkola Administracji Publicznej). Inflow into the civil service cadre is, in addition, constrained by quantitative limits set out in the annual budget law. The limits are determined by the Cabinet of Ministers based on the recommendation of the Civil Service Head and the Ministry of Finance. In the mid-2000s, the limits were usually set at around 3,000 but have since dropped to 2001. While graduates of the KSAP receive priority, many people who have merely passed the test are still not Table 2. Civil service cadre bonus in 2012 admitted to the cadre2. (in PLN) civil service cadre civil service bonus multiplier Cadre members enjoy certain benefits which are not rank (PLN) 3 available to other civil servants . One of these is a I 0,47 881 salary bonus. The size of the bonus depends on the II 0,65 1 218 professional rank of the employee. There are 9 III 0,85 1 593 professional ranks in the civil service cadre, which IV 1,05 1 968 correspond to multipliers 0.47-2.05 (Table 2). Upon V 1,25 2 342 nomination to the civil service cadre, the first rank is VI 1,45 2 717 assigned automatically. Progression through the ranks VII 1,65 3 092 requires members of the cadre to receive a VIII 1,85 3 467 performance rating of at least 3.5 on a scale of 1-5 IX 2,05 3 841 during biennial performance evaluations. The Source: Resolution of the Prime Minister from majority of cadre members easily meet this December 9, 2009 requirement. In addition to the bonuses, cadre members enjoy longer annual leave and greater protection against dismissal than ordinary civil servants. Because membership in the cadre is assigned to individuals, rather than positions, it has resulted in considerable pay inequities. Staff performing identical jobs are paid different amounts, depending on whether or not they are cadre-members. To address this problem, the Government may wish to switch to a position-based system, in which special benefits and pay are reserved for specific high level positions. 7. In a more conventional system, each position in the civil service would be assigned to a grade and each grade would have a single multiplier. The Polish system is considerably more flexible. Until 1999, the Polish civil service had no grade structure per se. Positions had titles, 1 For 2013 Budget Draft the Ministry of Finance suggested that the limit for the entry to the civil service cadre be set at 0 vs. the proposal of 600 from the Civil Service Head. 2 The graduates of the National School of Public Administration have priority as regards nomination. This means that they decrease the limit, which is available for the officials who took and passed the exam. 3 The civil service cadre members also have certain obligations, e.g. they cannot be members of any political parties and cannot perform any paid activities without consent of the director general. 8 but there were many different titles, set out in many different laws, each pertaining to some subset of the civil service. In 1999, the Government attempted to group position titles within the civil service into grades and continued those efforts in 2009 following the approval of the new Civil Service Law from 2008. (See Government Ordinance 211 of December, 2009.) As shown in Table 3, there were five such grades, ranging from ‘civil service support position’ to ‘high level civil service position’. 8. In the following year, the grading system was refined. Each ministry was required to evaluate its civil service positions according to a common methodology, assigning each position a number of points, and grouping its positions accordingly. In principle, the classification should have been based on the job requirements of each position: the technical background required, the scope of management responsibilities, etc. Some ministries, however, tended to classify positions according to the characteristics of the current occupant, rather than according to the requirements of job itself. A position, for example, could be said to require ten years of relevant experience (reflecting the characteristics of its current occupant) whereas in fact it requires only two. As a result, current position classifications do not necessarily reflect the actual requirements of the jobs to which they apply. 9. At the same time, no attempt was made to link grades to specific salaries. Instead, the resolution authorizing the new system merely incorporated the existing range of multipliers for all the positions in a given grade. Thus if the current multipliers for staff assigned to a particular grade ranged from 2 to 8, then range for that grade was fixed at 2 to 8. As a result, the range of multipliers in each grade is extremely wide. As shown in Table 3, multipliers for support staff (in the original five-grade classification system) ranged from 0.8 to 2.7. For high level civil servants, they ranged from 2.2 to 8.0. Under this system, a support staff could have a higher multiplier than a high level civil servant. In Table 3. Civil Service Categories and Multipliers the second round of grading, some Category Range of Multipliers ministries expanded the number of grades High level civil servant 2.2-8.0 and narrowed the range of multipliers in Mid level management 1.5-7.0 each grade. The Ministry of Education, for Coordinating and ‘independent’ example, now has eight grades. The range civil servant 1.0-6.0 within each grade nevertheless remains Specialist 0.9-3.5 extremely wide. Support level civil servant 0.8-2.7 Source: Government Ordinance 211 of December, 2009 10. In practice, the range of coefficients Notes: Table applies to staff of ministries and voivodships. assigned to a given position tends to be There are separate, but similar, tables for tax and customs fairly narrow. As shown in Table 4 (and officers. elaborated in a series of graphs in Annex 2) the vast majority of staff with a given position title receive approximately the same coefficient. This is particularly true for lower level positions. For example, 82 percent of secretaries and 93 percent of clerks are assigned a coefficient between 1.7 and 2.3. Eighty-eight percent of specialists and 86 percent of senior specialists are assigned a coefficient between 2.0 and 2.6. The range is wider in more senior positions, as might be expected. Only 46 percent of department directors receive multipliers within a similarly narrow range (5.3-5.9). 9 Table 4. Concentration of Multipliers within Civil Service Positions Common Range % of Staff within Position Title of Multipliers Common Range Department director 5.3-5.9 46% Deputy director 4.1-4.7 69% Division head 3.2-3.8 84% Chief specialist 2.3-2.9 85% Senior specialist 2.0-2.6 86% Specialist 2.0-2.6 88% Clerk 1.7-2.3 93% Secretary 1.7-2.3 82% Inspector 1.4-2.0 87% Source: Chancellery of the Prime Minister, World Bank staff calculations. Notes: Complied from individual data from five largest ministries. 11. In principle, annual changes in the base reference wage are derived through the deliberations of a tri-partite commission. (The three parties in the commission are the Government, the labor unions, and private sector employers. Discussions of public sector wages are only one item in the tripartite commission’s agenda. The commission also discusses other major social issues such as increases in the retirement age. (Hence the presence of private sector employers.) In practice, the Government’s proposals have prevailed in recent years. From 2003- 2009, the base wage increased each year. In 2010, the base wage was frozen at its 2009 level and remained frozen through 2013. 12. High ranking government officials and staff in uniformed services: There is a similar multiplier system for so-called ‘high ranking government positions’ and civil servants in the uniformed services. Each group has its own base wage. (In the 2011 budget, the base salary for civil servants was PLN 1,873. For uniformed services, it was PLN 1,523. For key positions, it was PLN 1,766.) In normal years (i.e., when a comprehensive wage freeze is not in effect) the base salary of each group is adjusted independently of the others. During the 2011 election campaign, the army and police were promised an increase in base salary, despite the freeze. This was granted in the 2012 budget, to the chagrin of staff in other uniformed services (e.g., firemen) who are now pressing the Government for wage increases as well. 13. Judges and prosecutors: For judges and prosecutors, there is a variant of the multiplier system. The base reference wage for the judiciary is the average wage in the economy as a whole (as measured in the second quarter of each year). Salaries for individual judges and prosecutors are multiples of this base. Thus wages for judges and prosecutors rise automatically with wages in the economy as a whole. 14. Wages for teachers paid from the government budget (a relatively small number of staff, because most teachers are employed by local governments) are determined according to the methodology specified in the teachers’ charter. 10 15. Non-multiplier staff: Until 2010, the wages of non-multiplier staff were determined entirely at the discretion of managers, subject only to minimum wage legislation and the aggregate budget constraints on each ministry. Government Order 27/134 of 2010 established a grade system for non multiplier staff. The Order sets out 21 base salary grades for non-multiplier staff. Each grade is assigned a range of salaries expressed in PLN. At the lower end, the range is fairly narrow. In the lowest salary grade (Grade 1) the top of the range is only 17 percent higher than the lowest. At the upper grades, however, the range is considerably broader. In grade 21, the top of the range (for certain categories of staff) is seven times that of the bottom4. Thus the range of salaries within a given grade is, in principle, as broad as it is with civil servants. Individual positions, moreover, may be assigned of range of grades. Thus the grade of a powiat inspector of construction supervision may range from 18 to 20. On top of this, there is a wide range of so- called functional bonuses. The functional bonus for a powiat inspector of construction supervision, for example, can range up to 150 percent of the minimum base salary5. Thus in principle, the wage for a that particular position may range by a factor of 2.3: from PLN 2,560 to PLN 5,880. 16. Calculating the Salary Envelope Aggregate control over the wage bill is exercised through the budget. At the start of the budget cycle, the budget circular sets out the aggregate budget ceiling for each first tier budget user (e.g., each ministry). While it does not specify wage bill ceilings for each ministry, it does specify the methodology that is to be used to calculate the wage bill, so that in practice the wage bill for each ministry is fixed at the start of the budget process. Prior to the 2008 economic downturn, there appears to have been considerable game playing between the sectoral ministries and the ministry of finance in the estimation of each ministry’s wage bill requirements. For each ministry, wage bill estimates were based on the number of authorized positions (head counts) in each ministry, multiplied by the average salary of staff in that ministry in the previous year. As not all authorized positions were filled (some were temporarily vacated due to maternity leave, for example) wage bill estimates exceeded actual requirements. The surplus was used to increase salaries of existing employees. Since the wage freeze went into effect, the use of head counts has been abandoned. Instead the Ministry of Finance merely allocates each ministry the same amount it was allocated in 2009. 17. The budget, once reviewed by the Ministry of Finance (MOF), adopted by Government and enacted by Parliament sets out an overall budget ceiling for each ministry in the form of ‘parts’ (‘czesci’ in Polish). One or more czesci comprise the wage bill ceiling of a given ministry.6 Within each czesc, each ministry is free to allocate its wage bill ceiling among budget management units as it sees fit. 4 The regulation sets out four separate schedules of grades, each with a different salary range for each of its 21 grades. There is, for example, one schedule of grades for internal auditors, employees of the office of forestry seeds, the chancellery of the Polish Academy of Science, the treasury of the state solicitor’s office, and the farmers’ social insurance institution. There is another schedule for managers and physical laborers employed in regional water authorities; another for employees in the institute of national remembrance and a fourth schedule for all others. 5 As specified in Government Order 27/134 of 2010, the minimum base salary is the lowest salary for the grade. For non-managerial positions, the size of the functional bonus is further restricted. 6 See Attachment 10 of the annual budget law. The accounting system further breaks down expenditures by ‘dzial’ (a term that, in this case, refers to sector of the economy) and then by economic category (such as personnel expenditure) and by economic subcategory (such as type of personnel expenditure). Only the breakdown by dzial is specified in the published budget. The other breakdowns are used in monitoring execution. 11 18. Calculating individual wages Within each ministry, the task of allocating the wage budget among individual management units involves a struggle between the director general (DG) of the ministry and the managers of each unit. Within the budgetary sphere as a whole, there are thousands of separate budget management units, each of which is, for legal purposes, a separate employer. Within a given ministry, the manager of each unit, in principle, has the authority to alter the wage coefficients (the multipliers) of individual employees. In principle, unit managers can use this discretion to reward high performing staff, or to offer an attractive wage to a potential new recruit. This discretion is, however, checked by the DG of each ministry.7 DGs are responsible for ensuring that the aggregate wage spending in the ministry remains within the ceiling fixed in the ministry budget. The interests of the two parties therefore differ. While a unit manager may wish to increase the salaries (i.e. the multipliers) of the staff in her domain, the DG must ensure that the ministry’s overall wage bill remains within the ceiling set in the annual budget. To this end, the DG must fend off demands from all the unit managers within the ministry. No clear rules govern this process. Individual unit managers must negotiate individual requests with their DGs. The likelihood of success is increased if a unit manager can show savings elsewhere in the wage bill. Thus the departure of a senior staff (with a high multiplier) may free up funds to increase the multipliers of the remaining staff in the unit. But there is no guarantee that the freed-up funds will remain in the unit. The DG can assign them elsewhere. 7 It is also checked by a provision in the labor law which requires the base reference wage for multiplier staff to be increased by the same proportion that the total wage bill for the budgetary sphere is increased. Thus managers have more flexibility over the wages of non-multiplier staff than multiplier staff. 12 Does the System Succeed at Controlling the Aggregate Wage Bill? 19. As a system for controlling the aggregate wage bill, the current system seems to work fairly well—so far. As noted earlier, aggregate wage bill spending has remained a fairly constant percentage of GDP over last ten years (Figure 3). This success reflects the ability of the MOF to enforce spending ceilings at the stage of budget preparation and wage ceilings (at the czesc level) on individual first-line budget users during budget execution. But it also reflects the Figure 3. Central Government Wage Bill, percent influence of the EU-imposed ‘excessive of GDP deficit procedure’ (EDP), which Poland must 5 observe as long as its deficit exceeds three 4 percent of GDP. Although the EDP does not set a ceiling on the government wage bill, it 3 does set out an aggregate general 2 government deficit target of three percent of GDP and requires the Government to present 1 an acceptable plan for reaching it. The plan presented by the Polish government includes 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 a wage freeze, which went into effect in 8 January of 2010. Once Poland drops below Source: Eurostat, World Bank staff estimates. the three percent deficit ceiling, however, the EDP will no longer be in effect. To replace it, the Government proposes to impose its own fiscal responsibility rule, which will limit the growth of central government expenditures to the trend growth rate of GDP, once a structural deficit of one percent of GDP is achieved. This is expected to occur in 2016. 20. As the fiscal responsibility rule is self-imposed, it will only be as strong as Parliament’s will to observe it. Moreover, in the period before it goes into effect, there would appear to be some risk that the MOF will be inundated with demands for wage increases by public employee unions. So far the government has relied on a freeze to fend off these demands. How should the government manage the thaw? 21. Different countries use different methods to address broad based wage demands. As discussed later in this report, the highly unionized Scandinavian countries tend to rely on collective bargaining agreements, in which the broad parameters of wage increases are agreed through negotiations with the major public sector employee unions. Within these parameters, individual administrative units may negotiate salary increases for their own staff. In Germany, government employees are divided into two distinct groups: (1) public service employees (Angestelle) who have no lifetime employment guarantee but are permitted to strike and (2) civil servants (Beamte) who are guaranteed lifetime employment and are not permitted to strike. In Germany, adjustments in wages for Angestelle are settled through collective bargaining agreements with public employees unions. These negotiations tend to follow after negotiations between the more powerful private sector unions and the associations of private employers. As a 8 As the basic law on wages in the budget sector requires all wages to be indexed, Parliament had to enact special one-time laws in 2010 and 2011 freezing the wages of civil servants, judges, and non-multiplier staff at their 2009 levels. 13 result, agreements affecting private sector workers tend to establish bench-marks for the Angestelle. These, in turn, influence negotiations over the wages of Beamte. In recent years, the Government and the unions have had difficulty reaching agreements. As a stop-gap, Government has opted for one-time increases—valid for one year—in the absence of a consensus on permanent increase. Until 2012, such stop-gaps had been used in seven of the last ten years. In 2012, the Government finally reached an agreement with the unions on a three year schedule of wage increases for both civil servants and public employees. It was approved by Parliament in late June, 2012. 22. Serbia, as another example, has explicitly used indexation to gradually thaw wages after an extended freeze. Serbia imposed a comprehensive wage freeze in 2009 and retained it through 2010. In 2011, in response to growing pressures for wage increases, the Government (in conjunction with an IMF Standby Arrangement) initiated a schedule of gradual increases. In April, all public sector wages were increased by the rate of inflation in the previous six months plus one-half the rate of GDP growth in 2010. In October, they were increased by the rate of inflation in the subsequent six months (i.e., April-September, 2011). At the time, the Government intended to continue with this schedule of twice yearly increases, linked to inflation and GDP growth, until the aggregate wage bill fell below a target percentage of GDP.9 23. For staff in the Polish civil service, the equivalent measure would be to index the base reference wage. The Government’s fiscal adjustment plan, for example, limits the growth in central government expenditures to the trend rate of GDP. The Government might want to apply the same rule explicitly to the wage bill component of each ministry’s budget allocation. Or adopt a more interventionist policy (as in Serbia) and index not only ministerial budget allocations but individual wages. For non-multiplier staff, the equivalent measure would be to index the zloty-denominated salary schedules of this group of staff. Poland, of course, might want to adopt a more conservative index. Rather than tying wages increases to GDP growth, the Government may want to index them only to inflation. 24. There is ample international precedent for this approach. The pay levels for government employees are indexed to inflation in Belgium, Hungary, Italy, and Slovakia—although in some cases the index acts as a floor for subsequent negotiations, rather than an absolute determinant. Tying wages to inflation does not, however, guarantee that public sector wages will remain competitive. Wages for equivalent positions in the private sector may rise, due to increased productivity. To ensure competitiveness, some countries base public sector wage increases on surveys of wages in the private sector. Changes in the wage scales of US federal government employees, for example, are based on surveys of private sector wage levels. While much maligned, this system has succeeded in maintaining a rough parity between private and public sector wages. Over the longer term, the Government could consider a similar approach, indexing public sector wages to wage growth in the private sector. At the same time, the Government should periodically review individual ministries’ experience in recruitment and retention— 9 In the event, the Serbian Government has departed from this policy. In response to political pressures, the Government increased all salaries by two percent in January 2011. Subsequent increases in April and October 2011 and April 2012 were consistent with the indexation formula. The October 2012 increase was not. In response to increasing fiscal pressures, wages were increased only two percent, despite an inflation rate of 7.2 percent over the preceding six months. The Government again intends to increase public sector by two percent in April 2013, regardless of the rate of inflation. 14 particularly in high skilled more competitive jobs—to ensure that wage levels remain sufficiently high to attract and retain qualified staff. Substantial and sustained problems in recruitment and retention would justify an eventual adjustment in the coefficients for such positions. See Box 2. Box 2. Improving the Quality of Public Sector Employment and Wage Data At present, the MOF does not monitor staffing levels in the public sector. Nor does it monitor the level of individual wages. This information is monitored, instead, by individual ministries. It is, however, accessible to the MOF. Information on staffing levels and wages at the subnational level (except for teachers) is not so readily obtained. Local governments in Poland are independent employers. Subject only to national legislation and collective bargaining agreements, they determine their own staffing needs and wage levels and are not required to report them to the central government. Should the MOF attempt to improve the quality or coverage of public sector wage and employment data? Data collection can be costly and time-consuming. Major efforts should be undertaken only when they are essential for policy making. As noted in the main text, periodic surveys of recruitment and retention experience in the public sector, as well as surveys of comparable wages in the private sector, may be useful in deciding how much to increase base wages. But it is not so clear that the Ministry needs to collect such information on an ongoing basis. Nor is it obvious that the Ministry has any immediate use for data on subnational wages and staffing. Are Salaries High Enough to Attract and Retain Qualified Staff? 25. Now we turn to the micro issue: the question of whether wages are sufficient to attract and retain qualified staff.10 (One way to address this question is to compare wages in public sector positions with wages for comparable positions in the private sector. This presumes that if wages in the public sector are too far below private sector wages, the public sector will have difficulty attracting and retaining the sort of staff it requires. 26. There are two principal sources of such information in Poland. One is a biennial survey by the Central Statistical Office. The survey for 2010 found that on average, wages in the public sector are 17 percent higher than in private sector.11 But this is partly because of the nature of jobs in the public sector. While private sector jobs run the gamut from low skilled blue collar positions to high skilled managerial ones, public sector positions tend to be concentrated in white collar professions. What is therefore more revealing are differences in wages for comparable positions. Analysis by economic activity shows that, in the case of the ‘public administration’ sector, wages in the private sector are twice those in public sector.12 Somewhat smaller discrepancies are evident in ‘professional, scientific and technical activities’ and ‘information and communication’ (both 25 percent). Analysis by profession shows similar discrepancies. Wages of ‘higher level officials and managers’ are 16 percent higher in the private sector than in 10 Note that this section of the report deals only with a subset of staff in the budgetary sphere: civil servants and non- multiplier staff. The decision to focus on this group was dictated by the resource constraints of this study and does not imply that problems in this group of staff are more severe than elsewhere in the public sector. Within this group, much of the statistical evidence is confined to the civil service, due to the absence of data on non-multiplier staff. 11 Public sector includes state owned enterprises and subnational units of administration. 12 Presumably, the number of private sector jobs in this sector is small; diminishing the validity of the comparison. 15 the public sector. Wages of ‘specialists’ are 19 percent higher. In less skilled professions, the opposite is true. ‘Office employees’ earn 13 percent more in the public sector than in the private sector. Employees in so-called ‘simple jobs’ earn roughly the same amount in the private and public sectors. This suggests that for higher level management and technical positions, wages in the public sector may be too low. And that for less skilled positions, they may be too high. Figure 4. Comparison of Private Sector and Civil 27. This conclusion is reinforced by Service Wages, 2011 second source, a report by an international consulting firm retained by the Government to look into the competiveness of public sector compensation.13 The consultant classified a sample of existing positions into fourteen groups (ranging from ‘Technical 1’ to ‘Managerial 5’) and compared them with private sector equivalents. As shown in Figure 4, wages in technical (T1-T4) and lower- level professional and managerial positions (P1-P2, M1-M2) in Source: Report prepared by HRM Partners for the Chancellery the civil service in 2011 were roughly of the Prime Minister equivalent to their private sector Figure 5. Trends in Staff Turnover Rates, Percent comparators. This equivalence disappears at higher-level professional and 25 managerial positions (P3-P5, M3-M5), 20 however. The report found that salaries 15 for the highest-level professionals and 10 managers (P5, M5), for example, are 5 three times higher in the private sector 0 than in the civil service. Tax Control Civil Service Total Ministeries Tax Offices Other Offices Administration Administration Administration Tax Chambers Joint Voivodship Chambers Administration Joint Powiat Voivodship Central 28. It is not yet clear that this discrepancy actually impedes the public sector’s ability to recruit and retain 2007 2008 2009 2010 higher level professional and managerial Source: KPRM, World Bank staff estimates. staff at the present time. Overall14 the number of applications per advertised position is fairly high. In 2010 and the first half of 2011, it averaged 25. It is true that there are far more applicants for lower level positions than for higher levels ones. In 2011, there were 50 applicants per vacancy for support level positions, 25 per vacancy for lower level expert positions, 15 per vacancy for higher level expert positions and less than ten per vacancy for managers. This may be a reflection of the relatively low level of wages in higher level positions. But it may also reflect the relative abundance of job seekers who consider themselves qualified only for lower level positions. 13 HRM Partners, 2010. Note that this task was entirely separate from the Government’s efforts to consolidate positions into grades. 14 Note that while the evidence suggests that, overall, wages are sufficient to attract and retain qualified staff, difficulties have reportedly arisen in professional and technical positions in the regional offices of some ministries. 16 29. A more convincing indicator is the turnover or vacancy rate. A report by the Prime Minister’s office found that the average turnover rate in the civil service as a whole was 12.8 percent in 2007. The turnover rate in ministries and voivodships was particularly high, presumably due to the ‘restructurings’ that occurred after the 2007 elections. By 2010, the turnover rate had fallen to 9.1 percent in the ministries and 6.7 percent in the voivodships. (See Figure 5, above.) For the civil service as a whole is has fallen to seven percent. While the latter figure is low by international standards, it may reflect the propensity of civil servants to hang onto their jobs during the economic slowdown.15 Whether they will continue to cling to them when the economy picks up remains to be seen. Is Wage Setting Fair? Does it Motivate Staff? 30. So far, the answer to the two questions posed at the outset of this report is a qualified ‘yes’. Is the system able to control the aggregate wage bill? Yes—at least while the Ministry of Finance is able to impose its fiscal priorities on the sectoral ministries. Are wages high enough to attract and retain qualified staff? Yes—at least during the present economic slowdown. But when the economy booms again, the discrepancy between private and public sector wages for higher level professional and managerial positions may make recruitment and retention for such positions more difficult. 31. Even so, our sense is that all is not well. We are particularly concerned that the absence of clearly defined job descriptions and high degree of discretion that unit managers have to determine the multipliers of specific individuals may adversely affect staff morale and the ability to attract highly skilled people who hope to pursue a career in the public sector. 32. There are, in principle, certain advantages to Poland’s highly flexible system of wage setting. Each of the unit managers, in concert with his DG, can adapt salary levels to local labor market conditions and can reward high-performing staff, subject only to the extremely broad range of multipliers assigned to each grade and his ability to negotiate corresponding wage bill allocation from her DG. But this arrangement places a high degree of reliance on the ability of the unit managers and DGs to make these judgments—to decide who gets an increase in multiplier (or salary increase in the case of non-multiplier staff) and who does not. Directions for Reform 33. Other countries in the region have grappled with this issue (i.e.,. how much leeway unit managers should have over the salaries of their individual staff.) They have come up with a wide range of practices. These run the gamut from near total autonomy (Sweden) to no autonomy at all (Germany). There may be lessons in their experience for Poland. Four such cases were analyzed in detail for this report. These are summarized in Table 6 at the end of this section. 15 The consultant’s report found anecdotal evidence of recruitment difficulties in certain highly specialized fields in which there is private sector demand. These include meteorology, cartography, IT, and civil engineering. The consultants reported that only one qualified candidate applied for the position of chief specialist in the Department of IT projects in the Department for Computerization and Court Registries. 17 International Comparisons 34. In comparing countries, it should be noted that the organizational structure of the public sector varies among the cases. All of them, of course, have a central policy making body, e.g., a prime minister, a cabinet of ministers and a Parliament. All of them have ministries charged with line and staff functions. These ministries are divided into subordinate units administration (departments, divisions, etc.) each with its own manager. But the number of subordinate levels and the size of subordinate units varies considerably. Scandinavian countries (e.g. Sweden) also make a sharp distinction between the realm of the ministry (which is limited) and the realm of individual agencies, which operate at arm’s length from their respective ministries and have broad autonomy.16 For our purposes, we will refer to a single level of organization below the level of ministry. We will call it the agency. Sweden 35. Among the cases analyzed for this report, Sweden is the most liberal—in the sense of giving agency managers the most discretion to set wages. Subject only to an overall administrative budget constraint, agency managers have the authority to set the wages not only of positions but of individual staff. 36. Calculating the Salary Envelope. Each agency’s administrative budget envelope is determined through the annual budget process. Parliament begins the process by setting a global expenditure ceiling for the public sector as a whole. Detailed budget preparation proceeds through negotiations between the Ministry of Finance (MOF) and the various line ministries. The ceiling set by Parliament is subdivided into policy areas and then into policy ‘activities’ which include administrative authorizations. In the budget proposal that the Government submits to Parliament, there is a single administrative authorization for each agency. Salaries must be paid from the administrative authorization. But the budget law does not dictate how much of the administrative allocation must be spent on salaries. This decision is left to the agency manager. The agency manager is also free to decide how much of the administrative budget allocation will be used to hire more staff or to raise salaries of existing employees. An agency manager can also dismiss staff, subject to the provisions in the general Employment Protection Act, if available funds are insufficient. 37. Calculating Individual Wages. Agency managers have the authority to set the salaries of individual staff. Thus a manager may offer different salaries to two staff occupying equivalent positions. Agencies are free to design their own internal salary systems and may adopt uniform salary schedules for certain positions or career stages. This is common in agencies with many employees performing similar tasks. But they are not required to do so. 38. There are, of course, constraints on the agency manager’s freedom of action. In addition to the overall administrative budget constraint facing each agency, there are unions to contend 16 Scandinavian agencies may also have subordinate organizational units within them, but for purposes of this discussion, the agency is regarded as the smallest organization with control over wage setting 18 with.17 Sweden, like other Nordic countries, has two-level system for collective bargaining. Central agreements provide the framework for wage setting. These typically contain guarantees for minimum individual and collective salary increases. In the central negotiations for the state sector, employees are represented by the Union of Civil Servants (within the Federation of Professional and Salaried Employees—TCO) and various professional unions affiliated with SACO, the Confederation of Professional Associations. The employers are represented by SAGE-- the Agency for Government Employers.18 39. At the agency level, individual salaries are set through negotiations between the agency manager and the trade union representing each staff. (Salaries for non-unionized employees are set unilaterally by the agency.) If the agency and the relevant union are unable to reach an agreement, the union may leave the agency to act unilaterally or it may request central negotiations. (In the latter case, the agency’s proposal would lapse.) But unions cannot expect much from central negotiations. The central union has no leverage in such cases. It can only ensure that the central agreement is respected and that negotiations at the agency level have been pursued fairly, with no apparent cases of patronage or discrimination. 40. Agency directors are also constrained--albeit only slightly--by central guidelines. SAGE, in additional to negotiating central agreements, provides guidelines for individual wage setting. These specify, for example, that salaries should be based on objective factors such as level of responsibility, difficulty and complexity of task, skills, and performance. But SAGE does not have the authority to monitor compliance with the guidelines or to turn down agency-level agreements that appear to be inconsistent with them. SAGE’s role is purely advisory. 41. Another constraint—again mild—is the Swedish system of state sector pay statistics, termed BESTA. BESTA reports pay levels throughout the state sector, using a standardized nine- digit position classification system. The first two digits indicate the type of tasks associated with position. The third indicates the extent, complexity, and responsibility of the position. The fourth indicates whether or not the position is managerial (The others are used for fine tuning.) While BESTA enables agency managers to compare the level of salaries in their respective agencies with those of equivalent positions elsewhere in the state sector, it is not binding. It makes discrepancies obvious and serves as a reference, but does not directly impinge on the wage setting autonomy of agency managers. Germany 42. Germany is an example of the opposite extreme. Unlike Sweden, Germany is a federal country. Federal-level regulations nevertheless govern the wage setting process in all three tiers (Federal, provincial (Land) and local) of government. 43. Calculating the Salary Envelope. In Germany, each agency’s salary envelope for civil servants is determined, in effect, as a function of number of positions the agency is authorized to fill and the salary assigned to each position. Each agency has an establishment plan setting out 17 Public sector trade union density in Sweden is 80 percent . 18 SAGE is a confederation of government employers. It has an executive board elected by the managers of the agencies and is responsible for negotiating and implementing collective agreements for the national public service 19 the number of authorized positions, by title. Each title is linked to a grade. The salary for each grade is set out in Federal legislation. Each agency’s budget must fund each position at its respective wage level. No more. And no less. 44. There are several pay scales. Each takes the form of a grid, consisting of grades and steps. Table 5, below, shows the 2012 pay grid for the majority of Federal civil servants.19 As shown, there are fifteen grades (A2-A16) reflecting different job requirements and responsibilities. Each grade has series of steps. These reflect the individual employee’s length of service.20 As shown, there are eight length-of-service steps for each grade. The wage for each cell in the grid is a single amount, expressed Euros. The range among grades is narrow. A person entering the federal civil service at the top grade (A16) earns only 2.8 times as a person entering at the lowest grade (A2). The range within grades (based on length of service) is also fairly narrow. Without a promotion to a higher grade, the salary of an A2 would be only 14 percent higher at the end of a career than at the start of it (not counting across the board increases awarded to all staff). 45. All adjustment in wages scales are made centrally. Adjustment in the wages of Angestelle are normally settled through collective bargaining agreements with public employees unions. Any percentage increase in wages that is agreed upon applies uniformly to all pay scales and all career groups. The outcomes of these negotiations are in turn reflected in wage adjustments for the Beamte. Table 5. Germany: 2012 Pay Scales for Federal Civil Servants Step (Length of Service) € 1 2 3 4 5 6 7 8 A2 1,802 1,845 1,888 1,920 1,954 1,987 2,021 2,054 A3 1,875 1,919 1,963 1,999 2,035 2,070 2,106 2,142 A4 1,916 1,969 2,022 2,064 2,106 2,148 2,190 2,229 A5 1,931 1,997 2,050 2,102 2,154 2,207 2,258 2,309 A6 1,974 2,051 2,129 2,188 2,250 2,309 2,375 2,432 A7 2,077 2,145 2,235 2,326 2,416 2,507 2,575 2,643 A8 2,202 2,284 2,400 2,517 2,633 2,714 2,796 2,878 A9 2,384 2,465 2,592 2,722 2,849 2,936 3,023 3,109 A 10 2,558 2,669 2,830 2,990 3,150 3,261 3,372 3,484 A 11 2,936 3,101 3,265 3,431 3,544 3,658 3,771 3,885 A 12 3,148 3,343 3,540 3,735 3,872 4,006 4,141 4,278 A 13 3,691 3,875 4,058 4,241 4,368 4,495 4,622 4,746 A 14 3,796 4,033 4,270 4,507 4,670 4,834 4,998 5,162 Grade A 15 4,640 4,854 5,017 5,180 5,343 5,505 5,668 5,829 A 16 5,119 5,367 5,555 5,743 5,930 6,119 6,307 6,493 19 A separate schedule, for example, applies to high level positions, such as directors and director generals. It has eleven salary grades but no steps. There are also separate tables for provincial (Land) employees in each Land. 20 Progression from one step to another is not automatic, however. The employee must demonstrate ‘satisfactory performance’ before advancing to the next step. 20 46. Calculating Individual Wages. Agency managers have no discretion whatsoever over the wages of individual staff, or even over the wages of groups of staff within a given grade. Managers do have the authority to grant one-time bonuses for outstanding performance and special allowances to facilitate recruitment of a civil servant in case of acute staff shortage. (The latter may not exceed ten percent of the starting salary for the position and must be approved at the ministry level.) For an individual staff, the only way to obtain an increase in salary (outside of the normal adjustments in the grid) is through a promotion. Finland 47. Between the Swedish and German extremes, there are many intermediate cases. Finland is one of them. As in Sweden, Finnish agency managers receive a salary budget envelope and have some discretion over how to use it. But they have less discretion than their Swedish counterparts. 48. Calculating the Salary Envelope. In Finland, the salary envelope of each agency is determined through budget negotiations at the central level. In principle, the salary envelope is based on the agency’s allocation in the preceding year, adjusted for any general salary increases granted by the Government, and further adjusted up or down based on government priorities. Within that envelope, the agency manager may choose to add positions or adjust the salaries of his staff. 49. Calculating Individual Wages. But, unlike in Sweden, there is a structured system for salary setting with a position-based component, and an assessment-based individual component. 21 The scheme may vary among agencies (unlike in Germany). Each agency must have a defined salary scheme, but has the authority to devise its own salary scheme, subject only to agreement with the relevant unions. The position based component links positions to grades and grades to specific salaries. Some are very elaborate. The Ministry of Finance’s scheme, for example, has nine levels, each divided into five sublevels, for a total of 45. The individual-based component links individual assessment outcomes to salary steps. The Ministry of Finance’s scheme has seven steps ranging from 3 to 50 percent of the position based component. The largest possible individual component varies across agencies, but may not be larger than 55 percent. The total ensuing wage bill, must fit within the overall salary envelope for the agency. 50. As in Sweden, agency managers are subject to additional constraints. Like Sweden, Finland has two-tier collective bargaining system. The central collective agreement typically specifies an average salary increase for all public employees. Negotiations at the level of individual agencies determine the increase for each employee within the agency. And like Sweden, Finland has a system that allows agency directors to observe wages in comparable positions outside the agency. In the Finnish case, the comparisons are made to private sector employees. The comparison is performed by the Agency for the Public Sector as Employer (SAMV) and serves as an input to the central collective bargaining process but are not binding. 21 Individual performance awards have become less common. Five years ago, 20 percent of agencies had such reward systems. Now only six percent do, covering ten percent of staff and representing only 2-3 percent of employees’ annual salary 21 Lithuania 51. And then there is the case of Lithuania. 52. Calculating the Salary Envelope. In Lithuania, the government sets a salary envelope for each agency. As in Germany, it is based on the number of authorized positions. But unlike in Germany, the salary envelope is not a function of the number of authorized positions in each grade multiplied by the salary for each grade. Instead it is calculated as the total number of positions in each agency (regardless of grade) multiplied by the average salary for those positions. Within that envelope, the agency manager has some flexibility. 53. Calculating Individual Wages. All positions must be classified into a salary grid that applies uniformly throughout the public sector. (The grid consists of 20 grades, with steps for length of service.) While each grade is assigned a single coefficient, agency managers have some discretion in deciding which grade a specific position belongs to. . (In a sense, this is similar to the current Polish system, although in Lithuania, it is the grade, not the coefficient within a grade, that the agency manager can influence.) Similarly, the agency manager cannot add to the total number of positions in his agency. But he can add positions at one grade and subtract a corresponding number from another grade. As a result, he can adjust both the profile of positions in his agency and the level of individual salaries, as long as he observes the constraints on overall staffing levels, the coefficient or ranges of allowable coefficients in each grade and his overall personnel budget constraint. Table 6. Summary of Agency-Level Wage Setting Practices in Case Study Countries Sweden Finland Lithuania Germany Wage envelope Ministerial Ministerial Avg wage x total Wage of each based on: negotiations negotiations authorized staffing in authorized position in agency agency Authority of Agency Director to: Increase Yes Yes, by only by Only by reclassifying No individual amending the locally position into a higher salaries designed salary grade . system. Add positions Yes Yes No, but can substitute No higher level positions for lower level positions and vice versa Options for Poland 54. In theory, the Swedish approach has much to recommend it. It allows agency managers to determine staffing and salary levels subject only to an overall administrative budget constraint (and the compliance of relevant unions). As a result, an agency manager can adjust the number and nature of positions to match current needs and adjust salaries to reflect market conditions. But if Poland’s aim is to improve pay equity--to ensure equal pay for equal work—the Swedish model may be a leap too far--or at least too early. 22 Box 3. Performance Based Pay: International Experience Should the proposed system, with its narrow salary ranges, be supplemented by individual performance bonuses? The idea is tempting. In theory, such performance-related pay (PRP) offers a means to motivate staff. The empirical record is discouraging, however. While PRP is widely used*, international experience has generally been disappointing. The World Bank’s Governance and Social Development Resource Center reports that “there is no strong evidence that PRP improves worker or organizational performance—at least not in staff (as opposed to line) functions. In the latter case, performance can, to varying degrees, be measured objectively, for example, according to the number of patients seen, and can be rewarded accordingly. But in management and policy making roles, it is very difficult to objectively measure and quantify performance.� Similarly, OECD studies play down the effectiveness of PRP, noting that “performance assessment is inherently difficult in the public sector. [It] requires a large element of managerial judgment. The notion of performance itself is complex, owing to the difficulty of finding suitable quantitative indicators and because performance objectives often change with government policy.“ Other studies have concluded that the impact of PRP on performance is limited and may even be negative. PRP systems are particularly difficult to implement when management is weak, because managers tend to assign ratings on the basis of nonperformance factors, such as personal friendships. Attempts to counteract this by adopting more quantitative performance measures can create difficulties. Focusing too much on easy-to-measure quantitative targets can drive out harder-to-measure but more important qualitative aspects of performance. Over the last 10 years OECD countries have been moving away from quantitative performance measures. Their performance evaluations now rely less on numbers and more on previously set objectives and on dialogue with line management.. Countries where these appraisals have a significant direct effect on salaries have also instituted procedures for appeals and reviews of appraisals. These are standardized in Lithuania, based on explicit expectations in Finland and Germany and on generally followed recommendations in Sweden. Such measures are important for assuring the effectiveness and legitimacy of appraisals and individualized salary setting. For Poland, the best option may be to rely on promotions to reward staff performance. Annual performance evaluations can be used as an opportunity for managers to discuss performance issues and career prospects with individual staff, but not as a basis for financial rewards. * In the first wave at the end of the 1980s, the central governments of Denmark, the Netherlands, New Zealand, Spain, Sweden, the United Kingdom, and the United States all adopted some form of perfom PRP. Australia, Finland, Ireland, and Italy followed in the early 1990s. In 2004, France started experimenting with PRP for top civil servants in six pilot ministries (Finance, Defense, Interior, Equipment, Agriculture, and Civil Service). Most recently, in the late 1990s and early 2000s, Germany, Korea, Switzerland, Hungary, Slovakia, and the Czech Republic began to put PRP mechanisms in place. 55. In many respects, wage setting in Poland is already very similar to the Swedish model. Agency managers can set the salary coefficients of individual employees, subject only to very broad limits on the range for each grade. But the Swedish model works, in part, because agency managers can be relied upon to responsibly exercise their autonomy over individual salaries— something that has taken the Swedes thirty years to achieve. This is reinforced by the presence of unions (which limit opportunities for managers to exercise favoritism and patronage), the BESTA system (which would make any glaring wage discrepancies immediately obvious) and 23 the SAGE guidelines. At present, Poland has none of these supportive conditions. No unions. No BESTA. And while Poland does have guidelines, they are not specific enough.22 . 56. While Sweden may be a model for the long term, in the short term a less liberal approach, along the lines of Finland or Lithuania, may be appropriate. As described earlier, this would classify all positions into grades and assign each grade a single coefficient.. Ideally, such an approach would begin with a rigorous job classification exercise for all positions. This would classify positions in each agency according to its specific characteristics and responsibilities. (As noted earlier, Poland’s earlier grading exercise did not succeed in doing so.) Based on this classification system, positions would be grouped into grades. Each grade would be assigned a single wage coefficient (as in Finland and Lithuania.). Given regional variations in the cost of living within Poland and their effects on local labor markets, this single coefficient (or narrow band) could be supplemented by a regional wage supplement. This would enable ministries to offer competitive salaries to staff located in Warsaw, without overpaying staff located in areas where prevailing wages are lower. 57. A rigorous grading exercise would improve pay equity in two respects. By assigning staff within each grade to a single salary coefficient (or a narrow range), current discrepancies in pay within each agency would be reduced. And by imposing the common classification and salary scales on the entire civil service, discrepancies among agencies will be reduced. 58. In the course of this exercise, the number of grades should be increased from the current five. As noted above, Germany has fifteen. Lithuania has 20. Finland’s Ministry of Finance has 45. This would allow grades to be more precisely differentiated and offer staff a clearly defined career ladder. It would also allow staff to be evaluated against objective promotion criteria more frequently in their careers, rather than depending on the ad hoc decisions of their managers. Within their overall salary budget constraint, agency managers could be left free to alter the composition of positions within an overall staffing constraint (as in Lithuania.) They would merely have less arbitrary discretion over salaries than they do now. Implementation 59. The first step in a grading (or regrading) exercise is to define the criteria that will be used to classify positions. A key supposition is that it is the content of the position that is classified. The personal properties of the present holder of the position should not affect the classification of the position. More specifically, the salary of the present holder should be disregarded. 60. The work of devising classification criteria is normally carried out by a central personnel management agency (e.g., the Office of Personnel Management in the US, the SPF Personnel et Organisation in Belgium, or the SAMV in Finland) often with the aid of specialized consulting firms. There are a variety of standard sets of criteria used in position classification exercises. As 22 For the civil service, Ordinance no. 3 (Standards of the Civil Service) provides general guidelines on salary determination. For employees as a whole (both inside and outside the public sector) the Labor Code provides that employees have a right to equal pay for equal work (or work of equal value) and defines the rights of employees who have been unequally remunerated. The State Labor Inspectorate and labor courts have the authority to verify that the provisions of the Labor Code are being observed, both by employers and employees. 24 detailed in Annex 3, Hay Associates (one of the major consulting firms in this industry) classifies positions according to three broad criteria:  know how: the knowledge required to perform in the position;  problem solving: the degree to which the position requires judgment as opposed to merely the ability to follow instructions); and  accountability: the impact the position has on the company or agency. 61. Mercer, another major firm, uses the same three criteria (with slightly different names) and adds another two:  communication: the extent to which the position requires the ability to communicate and with whom; and  risk: the degree to which the position entails exposure to risk of mental or physical injury. 62. In the US, positions in the federal government are Table 7. US Federal classified on the basis of a so-called factor evaluation system Government: Matching Points to (FES). 23 The FES sets out nine factors which are to be used Grades and Salaries in classifying positions: (1) the knowledge required by Point Range GS 2012 position, (2) the extent of supervisory controls and (3) Grade Salary* guidelines; (4) the complexity of the work; (5) its scope and 190-250 1 $17,803 effect; (6) the extent and (7) nature of contacts with the 255-450 2 20,017 public (and other government staff); (8) its physical 455-650 3 21,840 655-850 4 24,518 demands and (9) the quality of the work environment. 855-1100 5 27,431 (Further details are provided in Annex 3.) As will be noted, 1105-1350 6 30,577 these are quite similar to the Hay and Mercer criteria. 1355-1600 7 33,979 1605-1850 8 37,631 63. For each factor, the FES designates a range of 1855-2100 9 41,563 ‘levels’ with a corresponding point score. For example, 2105-2350 10 45,771 under Factor 1, (knowledge required by the position) a 2355-2750 11 50,287 position can be classified into one of nine levels. These 2755-3150 12 60,274 range from level 1 ‘knowledge of simple routine or 3155-3600 13 71,674 repetitive tasks that typically include step-by-step 3605-4050 14 84,697 4055-up 15 99,628 instructions and require little or no previous training or *Starting salary experience’ (50 points) to level 9 ‘mastery of a professional field sufficient to generate and develop new hypotheses and theories (850 points). Similarly, Factor 3 (guidelines) is divided into five levels, ranging from level 1 ‘specific guidelines covering all important aspects of the assignment are provided to the employee, the employee works in strict adherence to the guidelines with any deviations to be authorized by supervisor (25 points) to level 5 (guidelines are broadly stated and non-specific, requiring employee to use judgment and ingenuity in interpreting the intent of any guidelines that do exist (650 points). 23 The FES applies only to white collar employees of the federal government, including g professional, technical, administrative, and clerical positions. 25 Based on total number of points, each position is classified into one of fifteen grades (GS-1, GS- 2, etc. up to GS-15). Each grade is assigned a single salary, expressed in dollars. 24 In 2012, salaries ranged from $17,803 (GS1) to $99,628 (GS15. Table 7 shown the current match between points, grades, and salaries. 64. In Belgium’s recently reformed system of position classification, detailed criteria are specified only for higher level (A-level) positions. These consist of administrators and managers. A-level positions are classified on the basis of twelve criteria. The manual describing the classification system includes a form with 12 corresponding sections that has to be filled in, and a list of 80 verbs that can be used in describing the content of the position. There are extensive supporting materials and arrangements to ensure that the classification of A-level positions is coherent across the federal administration. On the basis of this classification system, A-level positions are divided into five sub-levels (A1-A5) along with 17 professional categories (filières de métiers).25 Lower level administrative and support positions (levels B, C and D) are classified largely on the basis of educational requirements. (Less attention has been given to the classification of B, C, and D level positions, perhaps because there are no hierarchical subdivisions within each of these groups.) 65. In Finland, there is no common scheme for the classification of positions. This is instead left to each state agency, although guided by centrally agreed criteria and a classification manual. The schemes drawn up by the State institutions have to be submitted to SAMV, which sometimes advises against initial proposals. SAMV also translates the position levels used by the different line organisations into seven standard levels for use as the base for salary statistics. 66. Once the classification criteria are determined, they have to be applied to the actual positions in the organization. This is no small task, particularly where (as in Poland) the existing position classification system does not provide much to go on. While the overall postion classification system is defined by the central personnel management office, individual position classifications are normally undertaken by the individual state agencies, who are assumed to have superior knowledge of the actual content of the existing positions. A large state agency might even need to delegate the initial classification to subunits. 67. But state agencies cannot be left on their own. Because positions are ultimately linked to salaries, job reclassification is likely to encounter resistance, particularly from staff whose positions would be downgraded. To mollify such staff, managers are likely to over-grade—i.e., to classify positions into higher grades than are actually warranted. To thwart this tendency, the classification process is normally subject to a review by an oversight body. In the US, OPM has the legal authority to determine whether agencies are classifying positions in a manner ‘consistent with published standards’ and can revoke the classification authority of any agency that falls short. In Belgium, the first proposal for the classification of a position is drawn up by responsible line manager. There are then three higher authorities that have to approve the 24 The GS schedule allows for increases based on length of service. Each grade has 10 steps. Normally, staff receive one step increase each year in first three years, then one every two years for next three years and then one every three years until step ten.) 25 Although some professional categories are sector specific, many of them correspond to specific ministerial responsibilities Among the 17 are ‘human and animal health’ and ‘population and security’ but also ‘budget and public finances’, ‘employment’, and ‘taxation’. 26 classification: the state institution concerned, the federal administration and a classification committee. Each of these can fail a proposal and send it back for revision. In Finland, schemes drawn up by state institutions also have to be submitted to SAMV, which sometimes advises against initial proposals. Finland has two other mechanisms to guard against over-grading. First, all remuneration costs have to be financed within each state institution’s annual administrative allocation. Thus managers must ensure that any revisions to position classifications do not result in an increase in overall personnel costs. Second, classification decisions have to be made through collective agreements with the representative trade unions. They have an interest in ensuring a coherent salary structure across the State administration. 68. At the same time, governments can counter employee resistance with money. When a position reclassification exercise is undertaken, positions to be downgraded are normally grandfathered—that is to say, the employee occupying the position is protected against any nominal decrease in salary. The salary for the position is reduced only when the employee leaves and the position falls vacant or when inflation reduces the real value of the salary to its appropriate level. But grandfathering can be expensive. Any savings resulting from downgrading positions can take years to have an impact. Meanwhile, employees often expect that any increase in salary arising from re-grading will go into effect immediately. One solution to the latter problem is to disappoint them—to phase in the increase over a period of several years. 69. A fallback option to an overall pay and grading reform would be to simply narrow the range of coefficients assigned to each existing grade. Under this approach, only the outliers— existing positions at the top and bottom of each of the existing salary ranges—would have to be re-classified. Positions at the top of each range could be reclassified into a higher grade—or suffer a decrease in coefficient. Positions at the low end of each range, by the same token, would be reclassified into a lower grade or have their coefficients increased. The charts in Annex 2 suggest that the number of such cases would not be particularly large. Only seven percent of clerical positions and 18 percent of secretarial positions would have to be reclassified in order to reduce the salary coefficient for these positions to a range of 1.7-2.3. 70. But there are shortcomings to this approach. First, it would not address the existing misclassification of positions in the middle range. More importantly, it would not provide for a career ladder within the current grading system. Staff in the current clerical grade, for instance, would remain at that grade throughout their careers unless and until they meet the qualifications for promotion to the specialist grade. Increasing the number of grades would create more plausible and therefore more motivating career paths. 27 Annex 1. Glossary Glossary Public sector – For purposes of international comparisons, the term ‘public sector’ is defined according to the European System of National and Regional Accounts (ESA 95) and includes all institutional units that are non-market producers. It includes three sub-sectors: (1) central government, including all administrative departments of the state and other central agencies, (2) local governments, including those types of public administration whose competence extends to only a local part of the economic territory, and (3 social security funds, including all central, state and local institutional units whose principal activity is to provide social benefits. Public administration – is defined according to the European Classification of Economic Activities (NACE2) and consists of persons employed in ‘public administration, defense and compulsory social security’ (Section O). Because NACE (unlike ESA) does not distinguish between private and public sector employees, some employees of private firms may be included in this category. By the same token, many public sector employees are excluded: including those (such as teachers and health workers) who are not employed in ‘public administration, defense or compulsory social security.’ State budgetary sphere – in Poland, includes all units of the public sector that cover their expenses directly from the state budget and/or transfer their revenues to the state budget (Public Finance Act from 2009, Article 10-13). In this report, only the central budgetary sphere is covered. 28 Annex 2: Distribution of Wage Coefficients by Title, Polish Civil Service Director of Department Deputy Director Head of Division 25 60 200 Frequency Frequency 20 Frequency 150 15 40 10 100 5 20 50 0 0 0.80 2.00 3.20 4.40 5.60 6.80 8.00 0 0.80 2.30 3.80 5.30 6.80 More 5.60 0.80 2.00 3.20 4.40 6.80 8.00 Bin Bin Bin Advisor to Minister Chief Specialist Senior Specialist 60 500 400 400 Frequency Frequency Frequency 300 40 300 200 200 20 100 100 0 0 0 3.80 0.80 2.30 5.30 6.80 0.80 2.30 3.80 5.30 6.80 More More 0.80 2.00 3.20 4.40 5.60 6.80 8.00 Bin Bin Bin Specialist Secretary Inspector 500 40 100 400 80 Frequency Frequency Frequency 30 300 60 20 40 200 10 20 100 0 0 0.80 2.30 3.80 5.30 6.80 More 0 5.60 0.80 2.00 3.20 4.40 6.80 8.00 3.20 0.80 2.00 4.40 5.60 6.80 8.00 Bin Bin Bin 200 Clerk 150 Frequency 100 50 0 0.80 2.00 3.20 4.40 5.60 6.80 8.00 Bin 29 Annex 3: Examples of Position Classification Systems Hay Associates Know how Technical Knowledge: depth and breadth of technical or specialized knowledge needed to achieve desired results Management Breadth: ability to undertake managerial functions, such as planning and organizing staff or directing or controlling resources to achieve business results over time Human Relations Skills: interpersonal skills required for successful interaction with individuals and groups, inside and outside the organization Problem solving Thinking Environment: job’s context and degree to which problems and solutions are defined Thinking Challenge: the nature of addressable problems and the difficulty of identifying solutions Accountability Freedom to Act: whether action is defined by regulations guidelines, versus broadly defined Magnitude: breadth of impact on company operations Impact: the degree to which position affects outcomes; e.g. whether advisory or executive Mercer Impact Size of organization measures by sales, activities or employees Nature of impact position has on organization Contribution that position holder makes in context of impact Communication Required communication ability Nature of communication Innovation Required ability to identify or improve procedures, services, or products Required ability to deal with complexity Knowledge Required knowledge to accomplish objective How knowledge is to be applied in team Geographic context in which knowledge is to be applied Risk Exposure to risk of mental or physical injury Quality of environment 30 US Federal Factor Evaluation System Factor 1 - Knowledge Required by the Position (50-1850 points) • Kind or nature of knowledge and skills needed. • How the knowledge and skills are used in doing work. Factor 2 - Supervisory Controls (25-650 points) • How the work is assigned. • Employee's responsibility for carrying out the work. • How the work is reviewed. Factor 3 - Guidelines (25-650 points) • Nature of guidelines for performing the work. • Judgment needed to apply the guidelines Factor 4 - Complexity (25-450 points) • Nature of the assignment. • Difficulty in identifying what needs to be done. • Difficulty and originality involved in performing work. Factor 5 - Scope and Effect (25-450 points) • Purpose of the work. • Impact of the work product or service. Factor 6 - Personal Contacts (10-110 points) • People and conditions under which contacts are made. Factor 7 - Purpose of Contacts (20-220 points) • Reasons for contacts in Factor 6. Factor 8 - Physical Demands (5-50 points) • Nature, frequency, and intensity of physical activity. Factor 9 - Work Environment (5-50 points) • Risks and discomforts caused by physical surroundings 31