GOVERNANCE GOVERNANCE EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT Public Sector Employment and Compensation: An Assessment Framework EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 1 © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved. This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, 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. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO), http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: 2021. Public Sector Employment and Compensation: An Assessment Framework. EFI Insight-Governance. Washington, DC: World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. Graphic Designer: Maria Lopez / lopez.ten@gmail.com >>> Contents Acknowledgements iv Abbreviations v 1. Introduction 1 2. The Assessment Framework 3 3. Assessing Wage Bill Dimensions 6 3.1 Wage Bill Planning 6 3.2 Wage Bill Controls 11 3.3 Employment Levels and Distribution 13 3.4 Wage Competitiveness 19 3.5 Wage Equity 27 3.6 Wage Incentives 32 4. Measuring Impact 35 4.1 Fiscal Sustainability 35 4.2 Public Sector Productivity 38 4.3 Labor Allocation Between the Public and Private Sectors 43 5. Conclusion 48 Annex 1. Definitions and Data Sources 50 Annex 2. Streamlined Assessment for Poor Data Country Contexts 53 References 58 Figures Figure 1. Assessment Framework for Public Sector Employment and Compensation 4 Figure 2. Variations in the Wage Bill by Country Income 7 Figure 3. Volatility in Annual Recruitments, an Example from Brazil 9 Figure 4. Public Sector Is a Large Employer Globally 13 Figure 5. Public Sector Employment Varies Considerably Cross-Nationally 14 Figure 6. Most Public Sector Employees Work in Public Administration, Education, and Health 15 Figure 7. Proportion of Public Sector Workers with No or Primary Education 16 Figure 8. The Proportion of Public Sector Employees Who Are Female 17 Figure 9. Public Sector Wage Premium (Compared to All Private Paid Employees, Excluding Benefits) 21 Figure 10. Access to Benefits in the Public and Private Sectors 22 Figure 11. Public Sector Generally Pays a Premium for Education and Health Professionals 23 Figure 12. Pay Compression Ratio: Public vs Private Sectors 25 Figure 13. Wages of Hospital Doctors Relative to Global Median 26 Figure 14. Relative Wages of Key Service Delivery Staff Can Vary Significantly 28 Figure 15. Pay Inequity in the Brazilian Public Sector 29 Figure 16. Gender Inequity in Wages in the Public Sector 30 Figure 17. Summarizing the Evidence on Performance-Related Pay 33 Figure 18. There Is No Strong Correlation between the Size of the Wage Bill and Fiscal Balances 36 Figure 19. The Public Sector Wage Premium Is Not Correlated with Measures of Institutional Quality 41 Figure 20. Younger Public Sector Workers in Low and Lower- Middle Income Countries Have a Higher Wage Premium 44 Tables Table 1. Approaches to Measuring Public Sector Productivity 39 Boxes Box 1: Accounting for Pensions in Public-Private Wage Differential Estimates 22 Box 2: Brazil’s Wage Bill Micro-Data Analytics 37 Box 3: Pay and Employment and Public Sector Productivity in the European Union 40 Box 4: Public Sector Wages and the Labor Market in Ethiopia 45 >>> Acknowledgments This note was produced by a team consisting of Zahid Hasnain (lead author), Faisal Baig, Ravi Somani, Wouter van Acker, Anca Butnaru, Robert Palacios, and Daniel Ortega. The team is grateful to comments and guidance from Jim Brumby, Tony Verheijen, Shiho Nagaki, Moritz Piatti, and Srinivas Gurazada. It was prepared under the overall guidance of Tracey Lane and Ed Olowo-Okere. For the World Bank staff interested in this topic and wishing to explore relevant World Bank cases and relevant literature, please contact govgplearns@worldbank.org. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< iv >>> Abbreviations CPIA Country Policy and Institutional Assessment GDP Gross Domestic Product IMF International Monetary Fund OECD Organisation for Economic Co-operation and Development PEFA Public Expenditure and Financial Accountability PISA Programme for International Student Assessment WWBI Worldwide Bureaucracy Indicators EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< v 1. >>> Introduction Effective management of public sector employment and compensation is a vital activity of governments, with broad implications for fiscal sustainability, public sector productivity, and the competitiveness of the overall labor market. Governments’ expenditures on employee compensation represent approximately 30 percent of their expenditures, and wage bill management therefore has obvious fiscal and expenditure efficiency implications. Public sector wages affect the selection, retention, and motivation of public sector workers, which in turn, impacts productivity, or the amount and quality of government outputs such as infrastructure, regulations, and public services produced per public sector worker. The public sector is a large employer, accounting for 37 percent of global formal employment, and the wage bill represents livelihoods as well as budgetary costs, and changes in government wages are likely to produce significant effects on the whole labor market and the overall economy, including potentially crowding out private sector employment. In many low- and middle-income countries, especially those experiencing fragility, public sector employment is the core ingredient of the political settlement and wage bill reforms have immediate and often severe implications for political stability, peace, and security. The objective of employment and wage policies should be to maximize public sector productivity in a fiscally sustainable manner and without distorting the overall labor market. This normative goal does not deny that governments may have other important social and political objectives, such as ensuring adequate representation of important constituencies and marginalized groups, and providing formal jobs where the private sector is underdeveloped. Nevertheless, these policies require governments to manage difficult technical and political tradeoffs and all governments need to consider their impacts on service delivery, fiscal sustainability, and the labor market. Fiscal crises have often been an impetus for reforms, necessitating short-term fixes like across-the-board wage freezes or cuts, and downsizing through voluntary or involuntary schemes to cut expenditures and meet fiscal targets. However, EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 1 many of these short-term measures have adverse impacts on are usually divided across jurisdictions, occupational groups, long-term growth and welfare, and political viability, and often and organizations. This divide is obvious in federal systems create distortions and perverse incentives.1 For example, where public sector personnel are employed and managed freezing basic wages has often resulted in a mushrooming separately by central and subnational governments. Even of less transparent allowances and salary supplements that in unitary systems, there may be separate legal regimes for reduce wage bill transparency, harm pay equity, and hurt different categories of public servants, and often delegated or productivity. Public sector employment and wage practices, fragmented management responsibilities across administrative therefore, need to give equal primacy to their effects on units. Depending on local context the assessment, therefore, recruitment, retention, and motivation of workers and ultimately will need to be applied to a sufficient number of administrative productivity even during a fiscal crisis. units to provide a reasonably complete picture of public sector employment and compensation practices and their impacts. This paper aims to provide a framework for conducting public sector employment and compensation Another challenge is where to demarcate the boundary assessments that can help develop evidence-based between wage bill management and organizational and reforms. Such a framework is necessary given growing debt human resource management. Most of the impacts of distress and the need for greater expenditure efficiency in employment and compensation policies are mediated by many of the World Bank’s client countries, and also due to personnel management practices. For example, whether the urgency for addressing global challenges like pandemics, public sector employees are productive depends on merit- climate change, building human capital, and reducing based recruitment and performance management as well inequality, all of which require a strong role for the public as appropriate staff numbers and competitive pay. This sector. Inspired by the success of the Public Expenditure and assessment framework recognizes these overlaps but, for Financial Accountability (PEFA) framework to diagnose the pragmatism and to keep the scope manageable, strictly limits state of public financial management policies and outcomes, the framework to employment and compensation practices. A the objective of this paper is to provide a consistent and diagnostic framework for personnel management is a potential robust approach for conducting public sector employment area for future work. Similarly, the paper does not cover the and compensation assessments. Like PEFA, this approach important, but complex, issue of public sector pensions as that is anchored in a conceptual framework, a set of guidance too would greatly expand its scope, though it does discuss the questions, and quantitative indicators on the wage bill, implications of pensions on public-private wage gaps, and on employment, and wages that provide the core evidence base of fiscal sustainability. such assessments and which can be tracked over time. Unlike PEFA, however, the goal is not to develop an assessment The paper is organized as follows: The next section score of the various aspects of wage bill management, or a presents the assessment framework. Section 3 then goes detailed toolkit, but rather to provide more general evidence- through each of the dimensions presenting cross-national based guidance that can be adapted to country contexts and stylized facts to highlight important elements that should can help to improve the quality of Public Expenditure Reviews, be assessed, and proposes a set of guidance questions for Human Capital Expenditure and Institutional Reviews, and the assessment and key indicators as the evidence base overall policy advice. for the assessment. Section 4 discusses how to assess and measure the impact of pay and employment on the three One challenge for the diagnostic is that most governments outcomes of fiscal sustainability, public sector productivity, do not have a centralized or unified system for managing and labor allocation between public and private sectors. its workforce. While analytically the concepts of “public The final section concludes with some suggestions on sector” or “general government” (as defined in Annex 1) are framing recommendations. useful, the actual responsibilities for wage bill management 1. IMF (2016). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 2 2. >>> The Assessment Framework The proposed diagnostic framework consists of two parts (Figure 1). The first part assesses six dimensions of public sector employment and compensation practices: • • The robustness of the wage bill planning process such that annual planning and budgeting is anchored in a fiscal sustainability framework, is informed by comprehensive and accurate data, and is transparent. The strength of wage bill controls to ensure that the budget is executed as planned and that only legally employed staff on the payroll are paid the wages that are due to them. • The employment levels and distribution of the public sector workforce so that there is neither over- or under-staffing, and that the right staff are employed in the right location and in the right positions. • The wage competitiveness of the public sector to ensure that qualified staff are attracted to public sector employment without creating skills shortages in the private sector. • Wage equity so that staff in similar jobs with similar skills and similar performance are paid equivalently. • Wage incentives in the form of appropriate performance-based career growth and performance bonuses aimed at maximizing worker productivity. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 3 > > > F I G U R E 1 - Assessment Framework for Public Sector Employment and Compensation 1 Wage Bill Planning Fiscal sustainability 2 Wage Bill Controls 3 Employment Levels and Distribution Public sector productivity 4 Wage Competitiveness 5 Wage Equity Labor allocation between public and private sectors 6 Wage Incentives Source: Authors. These dimensions cover both the institutional and • Labor allocation between the public and private behavioral aspects of employment and compensation sectors: Whether the public sector is crowding out the practices. The first two dimensions explore the “macro” labor market, resulting in skills shortages in the private institutional aspects of the strength of public financial sector and impeding private-sector job growth. management and human resource management systems for wage bill expenditure planning and execution; the next four Given that employment and compensation diagnostics are more “micro” elements likely to influence the behaviors should be “problem-driven,” the linkages between these and motivations of individual public employees. practices and outcomes are necessary for identifying priority reforms. While the main outcomes of interest will The second part of the assessment explores the impact of vary depending on country context and greater emphasis these practices on: can be given to one outcome over another in an assessment, recognizing the importance of all three is necessary for • Fiscal sustainability: Whether the wage bill is putting a balancing tradeoffs and informing policies. Some linkages stress on public finances, resulting in either rising fiscal between practices and outcomes are theoretically reasonably deficits or reductions in other important government clear—for example, weak wage bill controls and presence of expenditures, such as capital expenditures. large numbers of ghost workers hurt fiscal sustainability and productivity. But others, such as whether competitive wages • Public sector productivity: Whether employment and improve service delivery, are open empirical questions. compensation practices are capacitating and incentivizing the public sector workforce to effectively deliver core government outputs such as infrastructure, services, and regulations. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 4 Implicit in the framework are both complementarities and The Worldwide Bureaucracy Indicators (WWBI), used tradeoffs across the dimensions. These are most apparent here in cross-national comparisons, while not addressing across wage competitiveness, wage equity, and wage these differences in the scope of the public sector, incentives. In many countries, personnel of some agencies does minimize the risk of inconsistent definitions are paid higher than equivalently skilled and experienced and measurement errors. The WWBI are derived from personnel of other agencies to compete for high demand harmonized household surveys using a common taxonomy skills, which compromises wage equity. Similarly, performance which enables cross-country comparisons based on similar bonuses and other incentive payments can also be perceived sources and common definitions of employment and wages. as inequitable. Assessments will need to recognize these Annex 1 provides more details on some of the key definitions tradeoffs and make judgements about whether they are and data sources used in this paper. appropriate in the local context. This report recognizes that these data requirements are Rigorous evidence needs to underpin the assessment, challenging, particularly in low-income countries. The and this report will highlight cross-national stylized framework provided in this report utilizes data already existing facts drawing on comparable data sources. Any global within the public sector—labor force survey data collected by benchmarking needs to be used with caution as definitions of national statistical agencies and administrative data collected the wage bill, public sector employment, and compensation by ministries of finance and public administration. However, can vary from country to country. While it is common to do in many developing country contexts, statistical agencies cross-national comparisons of the wage bill – either as a share may lack the capacity to conduct the large, nationally of GDP or government expenditures – based on government representative surveys needed, and administrative systems fiscal data, differences in which employees are included or may be too weak to allow for this micro-founded approach to excluded from the payroll, and differences in the elements wage bill diagnostics. However, there are also risks with more of pay that are accounted for in employee compensation qualitative, or expert assessment approaches as these can makes these comparisons inaccurate. There is also, as will be be subjective and discretionary. Cognizant of these two facts, detailed in this report, no “right size” of the wage bill. Countries the report also provides an alternative, limited assessment have different approaches on what proportion of services instrument in Annex 2 that utilizes the same conceptual should be publicly delivered with public sector personnel as framework and guidance questions. This relies on a more opposed to publicly financed but outsourced to private sector limited minimum set of quantitative indicators derived from providers for delivery. Other country characteristics, such administrative data that most governments do collect as part as size and whether it is a unitary or federal state, can also of their daily operations. impact the wage bill and public employment, particularly for service delivery personnel. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 5 3. >>> Assessing Wage Bill Dimensions 3.1 Wage Bill Planning The wage bill represents a large and less flexible component of government expenditures with significant future liabilities. Globally, and noting the difficulties with cross-country comparisons, the wage bill represents approximately 30 percent of government expenditures, with significant variation around this average. In many low and middle-income countries, the wage bill can take up almost half of all government expenditures, and is an even larger share of expenditures for labor-intensive services like teaching and healthcare. For example, teacher salaries represent more than 80 percent of public education expenditures in developing countries.2 The wage bill, as a share of GDP, is larger in higher income countries, reflecting the bigger scope of governments as incomes rise; but is higher as a share of expenditures in lower income countries (Figure 2). The wage bill is also “sticky” as it is politically difficult to remove public sector workers or to cut their wages. For example, in higher income countries wage bill increases during economic upturns are not matched by similar reductions during downturns.3 Evaluations of public sector voluntary retirement schemes and downsizing programs reveal that these are generally not sustainable.4 These wage bill numbers underestimate the full fiscal costs of public sector workers given the generous pensions benefits that they enjoy. In Brazil, for example, the wage bill is 13 percent of GDP, and public sector pensions expenditures are another 4 percent of GDP.5 2. UNESCO (2017). 3. IMF (2016). 4. IEG (2008). 5. World Bank (2017). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 6 Estimating the size of the wage bill is of paramount paid through cash outside of the payroll, were estimated to importance, but is often difficult due to non-transparent be almost a third of the overall wage bill.6 Honoraria and per budgeting and accounting, weak data systems, and diems similarly serve as a large component of compensation informalities in employment and compensation. The in Cameroon, Zambia, and Zimbabwe.7 wage bill should at a minimum cover all permanent and contract workers employed by both central and subnational Estimating the true wage bill also necessitates an governments, and include all elements of compensation assessment of the quality of payroll and human resources (basic wages and allowances) and benefits and payments that data systems. Ideally, government personnel should be de facto serve as compensation (such as honoraria and per managed and paid through either a centralized and integrated diems). In many low- and middle-income countries, the official payroll and human resource management system or through wage bill in budgetary reports underestimates the actual wage a process of consolidation and reconciliation of the different bill as it does not include all workers or all these elements payrolls covering different jurisdictions and the various of compensation. For example, the government payroll may categories of personnel. Often, however, these systems cover exclude employees paid from extra-budgetary funds, or contract only a subset of employees, and are inaccurate, as human workers that are effectively full-time employees and perform resource data is not linked to the payroll which weakens similar functions to permanent staff. Budget documents may incentives for agencies to regularly update personnel records also classify de facto salary payments, like honoraria or per given that there are no costs in terms of missed salaries diems, under goods and services. In Indonesia, for example, associated with inaccurate data. honoraria for attending meetings and workshops, which were > > > F I G U R E 2 - Variations in the Wage Bill by Country Income PUBLIC SECTOR WAGE BILL (SHARE OF) 10 30 8 26 Percent of expenditures Percent of GDP 6 26 4 24 2 22 0 20 Low Income Low-Middle Income Upper-Middle Income High Income Percent of GDP (LHS) Percent of general movement expenditure (RHS) Source: IMF Government Compensation and Employment Dataset. 6. World Bank (2013). 7. World Bank staff assessments in Country Policy and Institutional Assessment (CPIA) write-ups. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 7 Wage bill decisions should ideally be part of a medium- • In federal systems, whether the inter-governmental term budget framework to ensure fiscal sustainability. An fiscal transfer system can create perverse incentives International Monetary Fund (IMF) review of practices in 42 for overstaffing by local governments and challenges countries revealed that half of the countries did not integrate to comprehensive planning resulting from the shared wage bill decisions in their budget planning procedures, personnel management responsibilities between central and that most low- and middle-income countries did not and subnational governments. undertake medium-term forecasting of the wage bill as part of their budgetary planning. Many countries do have fiscal • If the pay regime is characterized by collective bargaining, rules with wage bill expenditure ceilings, though it’s unclear then the union density, or percentage of public sector whether these rules constrain decisions. For example, in workers that are unionized, and whether there is a single Brazil, the Fiscal Responsibility Act stipulates ceilings on comprehensive collective agreement for all public sector personnel expenditures as a share of revenue, with states that employees or several, decentralized negotiations can surpass this limit not having access to federal guarantees for impact wage bill planning. Decentralized negotiations can international financing. However, steady wage bill growth over be organizational (with separate agreements per agency the past decade has meant that over half of the states have and ministry), occupational (with separate agreements per exceeded this threshold.8 Many countries also have automatic job group), geographical (with separate agreements for indexation of wages with inflation which can result in wage bill central, regional and local governments), or a combination growth through wage-price spirals and reduce governments’ of these. discretion in adjusting policies to respond to crises.9 In some low income and fragile and conflict affected Understanding the institutional mechanisms for setting countries donors create parallel civil services and service public sector wages is an essential part of the wage bill delivery arrangements that compromise integrated wage planning process. Pay setting regimes refers to the laws, bill planning. While these arrangements are motivated by regulations, and norms that govern the structure of public the understandable need to quickly scale up capacity and sector wages, and the processes by which these are set. At a help deliver donor-funded programs, they risk compromising minimum, a review of the pay setting regime should analyze: nascent planning processes and longer-term sustainability, especially when the payroll for these parallel arrangements • The predominant regime of wage determination, does not flow through the budget. In Afghanistan for example, whether it is a unilateral decision by the government or thousands of donor-funded—both on and off-budget— through collective bargaining between government and contracted staff under the National Technical Assistance trade unions. program undertake core policy and administrative functions and enjoy significantly higher salaries than civil servants.10 • If this is a unilateral decision by government, then a review Such programs also create wage bill growth pressures of the salary structure, in particular the pay and grading when these contractors need to be absorbed by the public system. The scope of the review will vary depending on administration once donor funding ends. the degree of centralization versus decentralization in pay determination—these range from completely centralized Hiring decisions should ideally be informed by strategic systems where pay for all federal public sector employees staffing plans that assess needs but are often driven by is set in a unified pay scale, as for example in Germany, political considerations. There should be a clear line of sight India, and the Philippines, to completely decentralized between an organization’s medium-term strategy, staffing systems where individual agencies are free to set their needs, and available resources yet political cycles in hiring, own pay scales within a salary envelope, as in Sweden. or ad hoc changes in policy, are very common. For example, One important element is the extent to which there are studies reveal that even in European Union member countries, different laws and regulations for different categories of in which personnel decisions are presumably more insulated public sector employees. from harmful political considerations, there are political cycles 8. World Bank (2020). 9. Holm-Hadulla et al. (2010). 10. World Bank staff assessments in CPIA write-ups. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 8 in hiring with increases in public sector employment around One approach that can increase transparency and election years.11 Such a political cycle represents decisions accountability around these policy decisions is the public that can occur despite the robustness of fiscal planning and disclosure of wage bill information. Some countries produce generally data-informed policymaking. Wage bill assessments annual reports on public sector employment and wages to can analyze employment trends from administrative data to disclose progress on various reforms and for informing future highlight these inefficiencies. Volatility in hiring is particularly policy developments, and involve academics, think tanks, disruptive to sound fiscal management but is common, as and institutions like fiscal councils, for fact-checking and to illustrated in Figure 3 in a subnational government in Brazil. lend credibility to their analysis.12 Involving academia and think tanks can create more evidence-based policies, build a culture of rigorous evaluation, and help insulate employment and compensation from detrimental politicization. > > > F I G U R E 3 - Volatility in Annual Recruitments, an Example from Brazil 10,000 8,773 8,359 8,096 8,000 Number of new employees 6,275 (values >= 2,000) 6,000 3,972 4,000 3,608 2,866 3,010 2,891 2,603 2,313 2,000 0 2001 2002 2003 2008 2010 2011 2012 2013 2014 2016 2017 Source: WB staff calculations based on administrative data. 11. See Martins (2010) for evidence on Portugal; and Dahlberg and Mork (2008) for evidence on Sweden and Finland. 12. Examples include the Australian Public Service Remuneration Report and the UK: Annual Report on Senior Salaries. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 9 GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. What are the main stylized facts about the wage bill: its actual size, evolution, and decomposition into employment and wages? 2. Is wage bill management anchored in comprehensive, accurate, and timely data? a. Does the data cover all government employees paid from the budget and from extra-budgetary funds? b. Is the data derived from integrated payroll and human resource systems that are updated every pay cycle? c. Are budget classifications transparent so that all payroll expenditures are accurately recorded? 3. Is wage bill planning integrated in a broader medium-term fiscal framework? Are forward budget estimates derived from models of future wage bill commitments? 4. How does the institutional regime for salary setting impact wage bill planning? Are there particular challenges for subnational wage bill planning due to mismatches in staffing decisions and funding between central and subnational governments? 5. Are staffing increases informed by needs and medium-term strategic planning, or primarily by short-term political considerations? 6. Is there any public debate or public disclosure around wage bill policies? K E Y I N D I C AT O R S Indicator Description Data Sources Wage bill as share of Actual wage bill estimates that include all government IMF Government GDP, expenditures, and workers that are paid from the payroll and from employment and revenues, ideally over the extra-budgetary funds; and all elements of de facto compensation dataset; past 10 years compensation such as basic wages, allowances, Boost; administrative data honoraria, and per diems. Wage bill as a share of An indicator of the how sensitive the fiscal impact of Administrative data domestic revenues, that is, the wage bill is to external aid flows, especially in low- excluding foreign grants income countries. Wage bill decomposed by Disaggregates the wage bill by the main organizations Administrative data administrative and industry – ministries and agencies – and the main occupational classification groups, including public administration, education, health, and security that contribute to the wage bill. Annual recruitment Provides an indication of whether annual recruitment Administrative data patterns over time follows strategic needs or is ad hoc and driven by political considerations. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 10 3.2 Wage Bill Controls A key question for the effectiveness of wage bill workers who can get paid from the government payroll without controls is whether wage bill ceilings are binding for being employed by the government– are a major problem in line ministries and agencies. The sanctity of these ceilings many low and middle-income countries. A related problem is of depends on whether they are based on authorized positions legitimate workers getting illegal payments, such as teachers or anticipated personnel numbers; whether they are set by and health workers having multiple part-time contracts that economic classification (i.e., there is a salary envelope within add up to more than the allowed full-time working hours. The the overall budget ceiling) and therefore forbid reallocation main mechanism to ensure effective wage bill execution, between wage expenditures and other expenditures or only by as spelled out in PEFA indicator PI-23, is integrated payroll administrative organization and provide agencies flexibility in and human resource management information systems that reallocation; and on the power of the ministry of finance. One have comprehensive databases of authorized staff positions, common problem is wage bill budgeting based on positions personnel, and payroll with all elements of compensation that instead of personnel which can lead to cash management is regularly updated with appropriate safeguards and audit problems and perverse incentives as ministries and agencies trails. For example, Nigeria’s extensive ghost worker problem may deliberately maintain vacancies in order use the savings was revealed when the government implemented a digital from vacant positions to fund salary top-ups. Ministries and ID system for civil servants which enabled it to remove over agencies also often violate ceilings by hiring contract workers 60,000 ghost workers, saving $1 billion annually.14 who are then later converted into staff, or by ignoring the ceilings altogether. In Gabon for example, employment ceilings Absent these integrated systems, censuses of public set in the budget law each year are not adequately adhered sector employees, or risk-based payroll audits, are also an to, resulting in recruitments above the ceiling established for effective control mechanism. Several Sub-Saharan African each ministry.13 Some budgeting systems, like performance countries – Central African Republic, Chad, Democratic budgeting as compared to line-item budgeting, give spending Republic of Congo, Ghana, Guinea, and Malawi, for example agencies more flexibility in reappropriating funds from the – have recently conducted censuses and audits to remove payroll to other expenditures. The question, then, is whether thousands of ghost workers and retirees from payrolls, and this delegated authority is appropriately used or abused. to suspend illegal payments, thereby generating significant fiscal savings. Deviations between budgeted and actual wage bill expenditures for the general government or the central A related problem is delays in payments to legitimate government, and differences between budgeted and employees. Inefficiencies in public financial management filled positions are good indicators of the effectiveness systems can result in delays in budget authorization and of wage bill ceilings. These are akin to the PEFA dimension payments, and salary arrears are a common problem in many on budget credibility which examines deviations between the low-income and even middle-income countries. For example, overall budget and actual expenditures. in the past, community teachers in Chad had payment arrears of 10 months, and half of newly hired kindergarten teachers in Another key issue is effective enforcement and control so the Philippines received their salaries late.15 In both countries, that only legally employed public servants receive wages electronic payment of salaries and use of mobile payments, and benefits to which they are entitled. Ghost employees – has greatly ameliorated this problem. 13. World Bank staff assessment. 14. Gelb and Clark (2013). 15. Evans and Yuan (2018). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 11 GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Are wage bill ceilings set by the Ministry of Finance adhered to by line ministries? a. Are ceilings set based on positions or actual personnel? b. Do ministries agencies have discretion in spending the savings from vacant positions? c. If agencies have flexibility in reappropriations from payroll to other expenditures, are these being done transparently? 2. How effective are payroll controls as measured by PEFA indicator PI-23? a. Integration of personnel and payroll records b. Management of payroll changes c. Internal control of payroll d. Use of payroll audits 3. Is there a high incidence of ghost workers? What are the reasons for these workers being on the payroll (fake workers, retirees still on the payroll, or legitimate employees getting illegal pay)? Has a payroll audit been conducted to identify issues? 4. Is there are problem of salary arrears and delays in the payment of salaries for public servants? K E Y I N D I C AT O R S Indicator Description Data Sources Deviations between This indicator measures the extent to which aggregate Government budget and budgeted and actual actual expenditure on the compensation of government outturn data general government wage workers reflects the amount originally approved, as bill expenditures in the defined in government budget documentation and previous three fiscal years fiscal reports. Filled positions as a This indicator measures the sanctity of wage Administrative data proportion of budgeted expenditure ceilings as high percentage of vacant positions positions can create perverse incentives of using savings from vacant positions to finance compensation. Ghost workers as a Measures the effectiveness of payroll controls. Public employee census percentage of total workers and payroll audits Wage bill arrears as a Measures delays in salary payments. Administrative data percentage of total wage bill expenditures. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 12 3.3 Employment Levels and Distribution A key question that these assessments need to analyze 37 percent of formal sector wage employment (Figure 4).16 is whether the public sector is appropriately staffed. The size of the public sector as a share of total employment Important metrics in assessing aggregate staffing levels increases with a country’s level of economic development— are public sector employment as a share of total paid (i.e., reflecting the increasing role of the state in providing social those working for wage labor, which excludes self-employed services as incomes rise (Figure 5 , top panel). However, workers), and formal sector paid employment (those who there is no discernible relationship between country income have a formal contract and receive benefits like pensions). levels and public sector employment as a share of salaried The first metric measures the overall labor market footprint employment, which suggests that the public sector grows of the public sector, while the latter two are better measures along with private formal sector wage employment (Figure 5, of the public sector’s size in the subset of the labor market, bottom panel). There is considerable cross-country variation namely wage and formal sector employment, which is around these global averages with public sector shares of more comparable. total employment ranging from less than two percent to over 40 percent, and of paid employment from 10 percent to 70 Globally, the public sector is responsible for 16 percent of percent, with four- to five-fold variations in these shares at any total employment, 30 percent of wage employment, and given income level. > > > F I G U R E 4 - Public Sector Is a Large Employer Globally PUBLIC SECTOR EMPLOYMENT As a % of Formal 37.1% Employment As a % of Paid 29.8% Employment As a % of Total 16.2% Employment Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). 16. Total employed individuals are defined as those workers, aged 15 and above, who in the household surveys responded that they had a job in prior week. Wage employees are those whose basic remuneration is not directly dependent on the revenue of the unit they work for and are instead paid in wages and salaries, piece work, or in-kind, and therefore, exclude self-employed workers. Formal sector wage employees are those who have an employment contract, have health insurance, belong to a union, or who are inscribed in a pension program. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 13 > > > F I G U R E 5 - Public Sector Employment Varies Considerably Cross-Nationally a. PUBLIC SECTOR EMPLOYMENT, AS A SHARE OF TOTAL EMPLOYMENT 50% (share of total employment) 40% Public sector 30% 20% 10% 0% 2.5 3 3.5 4 4.5 5 Log GDP per capita (Constant 2010 USD) b. PUBLIC SECTOR EMPLOYMENT, AS A SHARE OF PAID EMPLOYMENT 60% 50% (share of paid employment) 40% Public sector 30% 20% 10% 0% 2.5 3 3.5 4 4.5 5 Log GDP per capita (Constant 2010 USD) Source: Worldwide Bureaucracy Indicators, World Development Indicators, latest observations per country (multiple years). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 14 Equally important is to determine the occupational classification systems also differ across countries. Labor distribution of public sector employment. Countries have force surveys that use internationally accepted standard unique legal and occupational classifications of public sector industry and occupational classifications can be a basis for employees that make cross-national comparisons difficult. cross-national comparisons. These reveal that the public In many countries, all employees are classified as “civil administration (including police), education, and health servants,” meaning they enjoy distinct legal protections; in industries employ most public sector employees (Figure 6). others, only management and policy staff are civil servants The public sector is the predominant provider of education and other workers, particularly service delivery staff, have and healthcare services with three-fourths and two-thirds of fewer privileges and are governed by the labor code similar the education and healthcare paid workforce employed in the to formal private sector employees. The public sector job public sector, respectively. > > > F I G U R E 6 - Most Public Sector Employees Work in Public Administration, Education, and Health PUBLIC SECTOR PAID EMPLOYMENT BY INDUSTRY 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% East Asia & Europe & Latin America Middle East & South Asia Sub-Saharan Pacific Central Asia & Caribbean North Africa Africa Education Healthcare Public Administration Others Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). Cross-national comparisons of the “right” numbers of education and health professionals are possible because there are norms, albeit approximate, for benchmarking. Such estimates are more difficult for public administration as it is a broad category that contains a variety of occupations. In education, based on decades of research, a ratio of one teacher for every ten children is considered adequate for pre-school and early childhood programs.17 Student-teacher ratios can be higher in primary and secondary schools; in the OECD countries, for example, the average primary student-teacher ratio is 16.18 The World Health Organization specifies a norm of 2.3 health workers per 1,000 inhabitants. Cross-national comparisons should be complemented with within-country analysis, with appropriate nuances for differences in localities (rural versus urban, remote 17. Bowne et al. (2017). 18. OECD (2014). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 15 versus well connected) and the populations to be served, given that the distribution of these public sector workers in public administration, health, education, and security across localities can be unequal. One study found that a quarter of the variation in teacher allocation to schools could not be explained by the number of students.19 While no such standards exist for policy or administrative staff, interregional variations can still point to staffing inefficiencies. > > > F I G U R E 7 - Proportion of Public Sector Workers with No or Primary Education 50% Share of public employment 40% 30% 20% 10% 0% 2.5 3 3.5 4 4.5 5 Log GDP per capita (Constant 2010 USD) Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). The assessment should also explore the skill distribution high as 40 percent in some cases. A corollary to this high of public sector employees. While on average public proportion of low-skilled workers is a high proportion of clerical sector workers are more educated than their private sector or support jobs. A functional review of Serbia’s executive counterparts, there is variation in the proportion of public branch found that approximately one-third of jobs were for sector workers with tertiary education across countries. A high internal administrative support functions in ministries and proportion of low-skilled workers points to the public sector agencies which could potentially be consolidated into shared serving a social welfare function and points to potential fiscal services centers.20 savings without compromising public sector productivity through outsourcing of some elementary functions. As Figure The age, grade, and seniority profile of public sector 7 shows, the proportion of public sector workers with no workers can point to skills gaps. For example, prolonged or only primary education declines with country incomes. hiring freeze or disruptions to recruitment because of conflict, Many low-income countries have on average 20 percent of can result in missing cadres, as was the case in Cameroon and such low skilled employees, with the proportions rising to as Sierra Leone. Another problem is having a large proportion of 19. Majgaard and Mingat (2012). 20. World Bank (2016). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 16 older workers, as in Democratic Republic of Congo (DRC), are female is higher than the proportion of private sector where the inability to finance pensions has meant that many workers who are female. Women, however, are significantly retirees stay on the payroll. An aging public sector workforce underrepresented in the public sector in some low- and is also a problem in high income countries; in Romania, for middle-income countries (Figure 8). There is also considerable example, 30 percent of public employees are approaching horizontal and vertical occupational segregation with women retirement in the next 10 years, which can have implications concentrated in certain industries and positions. Globally, 74 for staff motivation and productivity, and fiscal sustainability and 69 percent of the public sector education and healthcare given growth in pensions expenditures.21 workforce, respectively, is female, but women occupy only 30 percent of senior official positions. The reasons for this Improving gender equality in public sector employment inequality in public sector employment have not been studied, should be an important policy objective of governments. but drawing on the academic literature on the private sector, The public sector’s large labor market footprint means that it likely entail: differential caring responsibilities that limit can be a strategic leader in changing norms and behaviors women’s career progression; social norms and attitudes about and promote greater gender equality in employment. Globally, what type of work women are more suited to; and biases in 46 percent of all public sector workers are women, and in task assignments so that women are less likely to receive most countries the proportion of public sector workers who more visible and career-enhancing responsibilities.22 > > > F I G U R E 8 - The Proportion of Public Sector Employees Who Are Female 100% Females (share of public paid employees) 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2.5 3 3.5 4 4.5 5 Log GDP per capita (Constant 2010 USD) Source: Worldwide Bureaucracy Indicators, World Development Indicators, latest observations per country (multiple years). 21. World Bank (2019). 22. Czibor (2021). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 17 In many countries, there is a dual labor regime for public have an incentive to work harder in order to get civil servant sector employees, with a large contingent of contract status, there are question marks about the sustainability of workers in service delivery sectors performing similar jobs this policy.23 Most contract staff, with the backing of employee as permanent employees. Contract teachers and community unions, successfully transition to permanent status and there health workers are very common in developing countries and are significant risks that the lower screening criteria risk more are different from their civil servant counterparts in that they patronage-based hiring. In Indonesia, for example, hiring are paid less, often significantly so, enter employment through of contract teachers by subnational governments is highly less rigorous recruitment criteria, and have time-bound politically motivated, spiking in election years and increasing appointments. While some rigorous impact evaluations have in numbers after the introduction of direct elections of shown that in the short-term contract teachers can outperform district heads.24 regular teachers as they are ostensibly more accountable and GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Is the public sector over- or understaffed in aggregate and what have been the employment trends over time? 2. Is the public sector over- or understaffed for particular skills and occupations, both in aggregate and in specific organizational and territorial jurisdictions? 3. Are there a large proportion of public sector workers with low educational qualifications or in low-skilled jobs? 4. Is the public sector a gender-equal employer? a. What are the proportion of women in aggregate? b. Is there horizontal and vertical segregation in employment? 5. Is there a dual labor regime with a large segment of contract workers, and what are its implications for public sector employment? K E Y I N D I C AT O R S Indicator Description Data Sources Public sector employment Public sector employment, as defined by the WWBI; labor forces as a share of total International Classification of Status in Employment surveys employment (ICSE),25 as a share of all currently employed individuals. Public sector employment Public sector employment, as defined by ICSE, as a WWBI; labor forces as a share of paid share of all individuals in paid employment, excluding surveys employment non-paid, self-employed, and employers. Public sector employment Public sector employment, as defined by ICSE, as a WWBI; labor forces as a share of formal paid share of all individuals with formal contracts, social surveys employment security, health insurance or union membership. Industry and occupational Decomposition of public sector employment by industry WWBI; labor forces distribution of public sector classification or occupational group. surveys employment 23. Bruns et al. (2011). 24. Pierskala and Sacks (2017). 25. https://ilostat.ilo.org/resources/concepts-and-definitions/classification-status-at-work/. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 18 Indicator Description Data Sources Industry and occupational Decomposition of public sector employment by industry WWBI; labor forces distribution of public sector classification or occupational group. surveys employment Student-teacher ratios; Provides a measure of the adequacy of key service Administrative data per-capital health workers delivery staff compared to international norms. Educational profile of Decomposition of public sector workers by educational WWBI; labor forces public sector workers qualifications. surveys Distribution of public sector Decomposition of public sector workers by skill level Administrative data workers by pay grade and jobs. Females as a share of Female employment within the public sector as a share WWBI; labor forces public sector employees of public sector paid employment. surveys Females as a share of Distribution of female employment in the public sector WWBI; labor forces public sector employees by disaggregated by industry classification. surveys industry Females as a share of Distribution of female employment in the public sector WWBI; labor forces public sector employees by disaggregated by occupational group. surveys occupation 3.4 Wage Competitiveness Public sector wage levels are crucial for attracting and as the European Union); or where significant migration – for retaining competent workers for expenditure efficiency, example, of doctors and nurses – occurs.26 Third, what is the fiscal sustainability, and for equilibrium outcomes in distribution of public sector pay—that is, is there too much or the entire labor market. Answering the question, “Does the too little pay for certain jobs relative to others within the public public sector pay too much or too little?” naturally requires sector— which will be considered in the next section under an appropriate comparator. There are three potential wage inequity? comparisons. First, how much does the public sector pay relative to the country’s private sector, given that the likely Theoretically, public sector wages should be set at a level alternative work opportunities are in the domestic private relative to private sector wages so that there is a small sector? Estimating public-private wage differentials is the most public sector wage penalty, and this penalty should be natural comparison that has been explored in a large academic stable over time.27 Under the theory of “compensating wage and policy literature. Second, how much does the public sector differentials,” a particular job should pay more or less than in that country pay for certain occupations, compared to the another if there are other non-wage undesirable or desirable public sectors of other similar countries? This comparison is characteristics of that job that need to be compensated for. useful for public sector occupations that may not have a clear Given the pecuniary and non-pecuniary benefits of the public private sector comparator (such as the police); where there sector, the most important of which is job security, monetary is a natural regional labor market for public sector jobs (such compensation should be lower than in the private sector, 26. In countries with high levels of emigration, like the Philippines and Croatia, the appropriate comparator for certain occupations, like nurses and doctors, will also need to include the destination country private sector labor markets. 27. Gomes (2016). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 19 for equivalent workers in equivalent jobs, so that the total on occupations like administration, healthcare, security, and compensation accounting for all attributes of jobs is roughly teaching that are not well represented in the private sector with equal in the two sectors. Under this optimal compensation its much higher proportion of agricultural and manufacturing policy, the public sector will be competitive without being jobs. Given that many public sector jobs are unique and may distortionary, and there will not be any shortage of skills in not have an obvious private sector comparator complicates either sector. The same principle implies that the wage this analysis. Another challenge is that the elements of total premium should be annually monitored to ensure that no compensation, such as the proportion paid in wages versus gap emerges between the public and private sectors that can bonuses or benefits, can be quite different between the cause a departure from this theoretical optimum. two sectors and may not be accurately measured in labor force surveys. Estimating public sector wage competitiveness, compared to the private sector, is methodologically Cross-national wage regressions from labor force surveys complicated. The standard approach in the academic reveal that public sector workers on average earn higher literature is to measure differences in total compensation – wages than observably similar private sector workers. basic wages, allowances, bonuses, and monetized in-kind Using a wage regression, where wages are a function of payments like housing and transportation – between the public observable worker characteristics, such as education, age and private sector for statistically similar workers in similar (a proxy for work experience), gender, location, and the jobs. Given the demographic differences of workers between sector of employment (public sector or private sector), public the two sectors, this approach ideally requires controlling for sector workers have approximately 19 percent higher basic observable worker characteristics, such as age, education, wages (excluding allowances and bonus payments) across work experience, and gender that impact human capital and the 111 countries for which the World Bank has data. Eighty therefore earnings; accounting for unobserved characteristics of these countries have a positive premium, and there is no such as ability, risk aversion, and public service motivation; clear pattern in the size of the premium with country incomes and controlling for occupations given that similar workers can (Figure 9). This finding also holds for gross wages that include have very different responsibilities in different occupations. A employer social insurance contribution and allowances for simple, raw comparison of average wages in the private and the 27 European Union member states for which data is public sectors is misleading as public sector workers are older available.28 Several academic studies, either single-country and more educated than their private sector counterparts, or cross-country, similarly find public sector wage premiums.29 have different career objectives and motivations, and work 28. WB staff calculations based on Eurostat EU SILC data. 29. Finan et al. (2017) EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 20 > > > F I G U R E 9 - Public Sector Wage Premium (Compared to All Private Paid Employees, Excluding Benefits) 120% 100% 80% Public sector wage premium 60% 40% 20% 80 out of 111 countries > 0 0% -20% -40% -60% 2 2.5 3 3.5 4 4.5 5 5.5 Log GDP per capita (Constant 2010 USD) Source: Worldwide Bureaucracy Indicators, World Development Indicators, latest observations per country (multiple years). Public sector premia are likely to be higher globally case of Thailand, it rose from 15 percent to 42 percent.30 when allowances and benefits, particularly pensions, Box 1 provides some examples of how including pensions are accounted for. A much higher proportion of public can have a large impact on total compensation differentials sector workers receive a job contract, health insurance, and between public and private sector workers. These differentials pensions (social security) than private sector workers (Figure are likely to be even larger when non-pecuniary benefits, like 10). For example, in Indonesia, Pakistan, and Thailand the job security, are factored in as public sector workers have inclusion of expected pensions benefits, monetized annually, much more stable jobs than their private sector counterparts. significantly increased the public sector wage premia; in the 30. Whitehouse (2006); Palacios and Jain (2020). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 21 > > > F I G U R E 1 0 - Access to Benefits in the Public and Private Sectors 60% 50% Public sector wage premium 40% 30% 20% 10% 0% Contract Social Security Health Insurance Union Membership Public Sector Healthcare Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). > > > B O X 1 - Accounting for Pensions in Public-Private Wage Differential Estimates A complete estimation of public-private compensation differentials using econometric methods needs to include pensions which are considered as deferred compensation and may exacerbate or offset earnings differentials. However, this is rarely done. An early example is the analysis of compensation differentials between public and private sector workers in Pakistan produced by Whitehouse (2006). Pension wealth is calculated using the defined benefit formula from the separate schemes that cover each type of worker and assuming life expectancy at retirement. The way pensions are indexed after retirement is also factored. The pension wealth is discounted to present terms and divided by expected years of service to estimate pension rights accrued for each year of service. This present value of (marginal) pension entitlements is added to current earnings of workers in each sector to calculate the ‘compensation’ variable. Including this pension entitlement significantly offset the negative wage differential for workers with more education at higher grade levels. Instead of a wage penalty, these workers received a compensation premium. More recently, a World Bank study of Indonesia (Jain and Palacios, forthcoming) revealed that the defined benefit scheme for public sector employees yielded higher replacement rates than the scheme that covers private sector workers, but allowances were not included in the public sector pensionable wage and allowances varied across the wage distribution. Adjusting for these differences, the wage premium for less educated workers in the public sector doubled from 15 to 30 percent after including pensions. In contrast to the case of Pakistan, the significant public sector wage premium EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 22 for university graduates, whose total remuneration includes a higher share of allowances, was reduced slightly by including pensions. Including pensions in public sector compensation diagnostics becomes even more important in the context of reforms. Changes to the wage bill often have long run repercussions on pensions; including allowances in the pensionable wage base in Indonesia for example, would greatly increase public-private differentials. Meanwhile, sensible parametric reforms to civil service pension schemes like moving from final earnings to a longer earnings period can reduce these public- private compensation differentials while also improving equity across civil servants with different age-earnings profiles. The size of the wage premium is sensitive to the choice wage premium globally decreases to seven percent, if public of the private sector comparator. What matters for workers sector workers are compared only to formal sector workers, are their alternative employment prospects when applying for with 52 out of 84 countries having a positive premium. Similarly, public sector jobs or when considering leaving public sector for specialized occupations, the comparison should not be to employment. The relatively few studies that have analyzed the overall private sector but for those specific occupations movements of workers between the two sectors suggest that in the private sector. For example, there is a wage premium the alternative to public sector employment is more likely to be for public sector education and health professionals compared the formal sector for relatively skilled workers, and the informal to their private sector counterparts, though interestingly, the sector for relatively unskilled workers.31 The average basic premium decreases with country income levels (Figure 11). > > > F I G U R E 1 1 - Public Sector Generally Pays a Premium for Education and Health Professionals a. PUBLIC SECTOR WAGE PREMIUM, BY INDUSTRY: HEALTH (COMPARED TO ALL PRIVATE EMPLOYEES) 2.0 1.5 Public sector wage premium 1.0 0.5 0.0 -0.5 -1.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Log GDP per capita (Constant 2010 USD) Low Income Lower Middle Income Upper Middle Income High Income 31. Yassin and Langot (2018) EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 23 b. PUBLIC SECTOR WAGE PREMIUM, BY INDUSTRY: EDUCATION (COMPARED TO ALL PRIVATE EMPLOYEES) 2.0 1.5 Public sector wage premium 1.0 0.5 0.0 -0.5 -1.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 Log GDP per capita (Constant 2010 USD) Low Income Lower Middle Income Upper Middle Income High Income Source: Worldwide Bureaucracy Indicators, World Development Indicators, latest observations per country (multiple years). How the wage premium varies across different categories demand for public sector jobs, with hundreds of applicants for of public sector workers can reveal potential inefficiencies each vacancy. in the wage bill. A common finding in several academic studies, confirmed in the WWBI data, is that the premium is higher for Even if the public sector pays higher wages on average lower skilled workers and for lower wage occupations. Public there may be rigidities in pay policies that prevent it from sector workers with tertiary education have a wage premium paying competitive wages for high demand occupations of two percent as compared to a nine percent public sector or for the most senior jobs. As has been noted in many wage premium for workers with primary education. Clerical studies, pay dispersion in the private sector has increased occupations enjoy higher premia in the public sector than greatly over the past four decades and some occupations senior managerial occupations. In many countries, the public like finance and information technology can pay multiples for sector pays a wage penalty for the most skilled employees.32 similarly educated workers than other occupations. Private Studies have also noted that premiums are higher for entry- sector organizations also typically have steeper hierarchies level civil servants than senior civil servants, which may be than public sector organizations, with pay rising more sharply particularly disruptive to the overall labor market. For example, with seniority. One common metric for wage dispersion is the in Brazil, the public sector wage premium is 45 percent for wage compression ratio which is the ratio of the 90th percentile new entrants in the federal civil service, and in Indonesia, wage to the 10th percentile wage in the salary distribution. entry level civil servants earn approximately 28 percent more This ratio is lower in the public sector for 70 out of 99 countries than similarly situated peers in the private sector.33 Given for which there is data in the WWBI (Figure 12). Studies show these high premia, it is not surprising that there is massive that when the wage structure in the public sector becomes 32. Hausman, Nedelkoska, and Noor (2020). 33. Kuipers (2020). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 24 relatively more compressed, it is more difficult to attract and maintain high-skilled workers or skills that are in high demand in the labor market, such as for digital skills or other highly technical jobs.34 The public sector will also likely pay less, often significantly so, for the most senior positions as compared to chief executives and senior management of private sector companies, and these differences are unlikely to be captured in labor force surveys as they are in the “tails” of the salary distribution. > > > F I G U R E 1 2 - Pay Compression Ratio: Public vs Private Sectors 20 18 16 14 12 Public sector 10 8 6 4 2 0 0 2 4 6 8 10 12 14 16 18 20 Private sector Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). Note: The 45-degree line represents equal values for the two axes. These changes in the structure of the labor market to attract and retain unique skills that are in high demand in prompted moves towards greater pay flexibility in the the labor market, it also creates risks of pay fragmentation and public sector in the OECD countries, including greater inequity and underlines the tradeoff between equity and pay discretion to ministries and agencies in setting pay scales competitiveness that governments need to manage. and more individualization of pay. Pay flexibility measures include having broad pay grade bands so that managers have For certain occupations, there may be no obvious private more discretion in placing jobs in a pay range based on staff sector comparator and instead the appropriate benchmark skills and demand; individual contracts determined by agency may be the public sectors of other countries. Police and heads that enables pay to be set at any point of the pay band security occupations have limited private sector alternatives. for that job; and higher pay scales for finance and regulatory Similarly, globally, over 77 percent of all workers in the agencies.35 While such flexibility can better enable agencies education sector work in the public sector, with higher ratios in 34. Borjas (2002). 35. World Bank (2014) EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 25 low- and middle-income countries. Doctors and nurses have reveal, for example, that South Africa pays the highest wages a high incidence of migration and retaining these workers of hospital doctors relative to the global median among the requires tracking the wages for these occupations in destination countries in Sub-Saharan Africa (Figure 13). Such international countries.36 The International Comparison Program’s wage public sector wage benchmarking for high demand jobs with data enables cross-national wage comparisons for specific relevant countries can be a valuable complement to public- occupations adjusted for purchasing power parity (PPP). They private wage comparisons. > > > F I G U R E 1 3 - Wages of Hospital Doctors Relative to Global Median USA 6 HKG 5 KWT 4 SGP ITA SAU 3 CAN ABW BHR GRC 2 AUS IND ZAF BMU GRC ISR 1 FJI KGZ NIC EGY LKA SDN 0 East Asia & Europe & Latin America Middle East & North South Asia Sub-Saharan Pacific Central Asia & Caribbean North Africa America Africa Source: Worldwide Bureaucracy Indicators based on ICP data, 2017. GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Does the public sector over or under-pay (taking into account all elements of compensation) compared to the private sector in aggregate and for different industries, occupations, locations, and demographic groups? 2. How does pay for occupations with limited domestic private sector comparators compare to that in the public sectors of other relevant countries? 3. Is there sufficient flexibility in wage practices to better align public sector wages to the needs of specific labor markets to attract specialized and high demand talent? How does this flexibility impact wage inequity and is the tradeoff well managed? 4. Does the government use data to model fiscal impact and identify salary mismatches and to inform wage policy? 36. Antwi and Phillips (2013). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 26 K E Y I N D I C AT O R S Indicator Description Data Sources Public vs. private sector Most widely used indicator of overall wage WWBI; labor forces wage premia (controlling competitiveness. surveys for worker and job characteristics) Public-private wage premia Measures the difference between public and private WWBI; labor forces by occupation sector pay for specific occupations. surveys; administrative data Public-private wage premia Measures the difference between public and private WWBI; labor forces by education level sector pay for different skill levels. surveys Public-private benefits Measures the difference between public and private WWBI; labor forces premia sector benefits (for example, health insurance, surveys pensions, access to training). Pay compression ratios Measures pay dispersion and is an indicator for public- WWBI; labor forces in the public and private private wage differentials at different points in pay surveys sectors. distribution. Pay comparisons for Measures pay of a particular public sector occupation ICP data; administrative occupations across public compared to global, regional, or benchmark country data sectors pay for that occupation. 3.5 Wage Equity Wage equity – whether workers in similar jobs, with can also have significant implications for pension liabilities similar skills, similar years of service, and similar as many defined benefit schemes link pension payments to performance are paid equally, and differences in salary highest earned salaries. between different occupations are deemed legitimate – has strong psychological roots and impacts worker There are many sources of wage inequity, a common one motivation and productivity. Studies have shown that being a fragmented wage setting regime with different wage equity is ingrained in peoples’ sense of fairness and salary laws for different occupational groups, territorial self-worth, that workers compare their wages to their peers jurisdictions, and even ministries and agencies. This in an organization, just as they do to the private sector, and pay differentiation may be a deliberate policy choice, as wage differentials that are not perceived to be justifiable can in several Western European and OECD countries, where be demotivating.37 Surveys of public employees conducted responsibilities for pay setting have been devolved to ministries by the World Bank reveal that this pay inequity is a major and agencies to better enable managers to individualize pay source of work dissatisfaction—for example, 65 percent to the competencies of staff and to attract high-demand skills of civil servants surveyed in the Liberia forestry agency to the public sector. Or it may be an accumulation of ad hoc stated that unequal and unfair pay was a source of conflict sector-specific changes reflecting bargaining by different between staff.38 Unjustifiably high wages for certain jobs interest groups. In Brazil for example, there are hundreds of 37. Rosenfeld (2021). 38. World Bank (2020b). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 27 different pay scales, and associated employee unions, for scales for distinct occupational groups—for example, public different occupational groups in the federal government alone. administration, education, health, and security. Often, job Such fragmentation can cause considerable pay inequity evaluations are not systematically done or regularly updated. between similar workers and can also create pressure for As public sector employment grows and new positions are wage increases as it encourages competitive bargaining created, relative pay between jobs can become ad hoc and between the unions to increase salaries for their respective not grounded in objective and credible criteria. For example, occupational groups. university teachers are much better paid relative to primary and secondary school teachers in low and lower-middle income Wage inequity can also arise with unified pay scales countries, earning almost 4.5 times the wages of government if these are not grounded in a robust job evaluation clerks compared to 1.5 times for primary teachers (Figure 14). methodology. Ideally, different jobs should be “weighted” These disparities are much lower in high income countries. based on standard criteria like complexity of tasks entailed Similarly, nurses are relatively underpaid compared to doctors and other relevant traits, and positioned relative to each other in low-income countries. either by grade on a single pay scale or on separate pay > > > F I G U R E 1 4 - Relative Wages of Key Service Delivery Staff Can Vary Significantly a. WAGES OF TEACHERS RELATIVE TO CLERKS 4 3 2 1 0 Low Income Lower Middle Income Upper Middle Income High Income Primary school Secondary school University b. WAGES OF MEDICAL WORKERS RELATIVE TO CLERKS 3.5 3 2.5 2 1.5 1 0.5 0 Low Income Lower Middle Income Upper Middle Income High Income Nurse Doctor Source: Worldwide Bureaucracy Indicators based on ICP data, 2017. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 28 A high proportion of allowances can cause pay dispersion For example, it is common to pay public sector workers even between similar workers in similar jobs. Brazil is a honoraria for attending board meetings and workshops, often case in point. For example, in the Federal Social Security, paid in cash and charged to “goods and services” instead Health, and Labor occupational group, gross pay can vary of the payroll, which can become a significant source of tenfold for workers with similar levels of experience, largely rents to those handing out the honoraria and create obvious a result of non-performance related payments and not basic perverse behavior. In Cameroon for example, these per diems pay (Figure 15).39 Many countries also have a high number for attending meetings were equivalent to civil servants’ of allowances or salary supplements. Often, these are at the base wages.40 discretion of a minister or agency head, and are often abused. > > > F I G U R E 1 5 - Pay Inequity in the Brazilian Public Sector DISPERSION: TOTAL COMPENSATION Intermediate positions in the social security, health and labor career - NI 30,000.00 20,000.00 2017 reais 10,000.00 0.00 0 10 20 30 40 50 Total compensation Regression line Source: World Bank 2020. Note: Each dot is an employee; the horizontal axis is years of service; the vertical axis is wages. Pay inequity can be justifiable if the labor market requires it for the public sector to attract and retain talent. Even if jobs across occupations have the same attributes and are graded similarly in the government pay scale, there may be higher demand for some jobs that require them to receive a higher wage than other jobs for which there is less demand in the private sector. For example, there is a high demand for doctors and nurses in the European Union member states which implies that these medical staff may need to be paid higher than, say, equivalently graded police officers in order to ensure that they remain in government service. There should be a clear empirical justification for this job specific premium so that the inequity it creates is not demotivating for other staff. 39. World Bank (2020a). 40. World Bank (2018). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 29 > > > F I G U R E 1 6 - Gender Inequity in Wages in the Public Sector FEMALE TO MALE WAGE RATIO IN THE PUBLIC SECTOR (USING MEAN) 160% 140% 120% Female to male wage ratio 100% 80% 60% 40% 20% 0% 2.5 3.0 3.5 4.0 4.5 5.0 Log GDP per capita (Constant 2010 USD) 20% 0% Gender wage premium -20% -40% -60% 2.5 3.0 3.5 4.0 4.5 5.0 Log GDP per capita (Constant 2010 USD) Public administration Education Healthcare Source: Worldwide Bureaucracy Indicators, World Development Indicators, latest observations per country (multiple years). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 30 Women earn less than men for doing the same work. qualified men (Figure 16, bottom panel). This gender pay While the public sector has more gender-equal pay than the gap for similar workers in similar jobs may be the outcome private sector, there remains a gender pay gap. Globally, of differential access to allowances and salary supplements, women’s average wages are 88 percent of male wages in the different opportunities for promotion within an occupation, and public sector, as compared to 74 percent of male wages in the impact of differential norms on family care responsibilities. the private sector. There is, predictably, considerable variation across countries on female-mean average wage ratios, though Quantifying these aspects of wage inequity requires an interestingly wage inequity persists in the higher income assessment of the public sector compensation structure. countries (Figure 16, top panel). One reason, as discussed, A review of the salary legislation and pay grading structure, for the gender pay gap is occupational differences with women supplemented with the analysis of micro-level payroll data under-represented in senior positions. For example, in the (i.e., anonymized individual salary records) can provide a state of Parana in Brazil, while women account for more than complete picture of whether there is generally “equal pay for 60 percent of tenured employees across government, they equal work” and coherence in compensation across the public occupy only 39 percent of leadership positions.41 The wage sector. It can also clarify whether wage inequity is justifiable gap however, persists even after controlling for occupations in certain cases and a necessary tradeoff to improve wage and worker characteristics like age and education. For competitiveness for skills that are in high demand in the example, women working in public administration, education, labor market. and health sectors have a wage penalty compared to similarly GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Is there “equal pay for equal work”? Do workers in similar jobs with similar skills get similar wages? Answering this question requires an assessment of the public sector salary structure. 2. Is there gender equality in pay in aggregate and across major occupations? Does government have a policy to address gender inequity? 3. What are the sources of wage inequity? Are pay relativities in salary scales derived from robust job evaluations? Are there discretionary elements of pay, such as cash payments or allowances? 4. Is there a justifiable tradeoff between wage inequity and wage competitiveness in order to attract high-demand talent while maintaining general coherence of pay across the public sector? K E Y I N D I C AT O R S Indicator Description Data Sources Public sector pay Ratio of the 90th to 10th percentile of wages, which WWBI; labor forces compression ratio provides an estimate of wage dispersion. surveys; administrative data; review of salary legislation Relative wages between Provides an estimate of pay equity given the different WWBI; administrative data; different occupations responsibilities and demands of different jobs. review of salary legislation Within-occupation public Provides estimates of wage differences controlling for Administrative data; review sector pay dispersion occupational groups and worker characteristics. of salary legislation 41. World Bank staff assessment based on government administrative data. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 31 Indicator Description Data Sources Female-male wage ratio Measures the raw wage difference between men and WWBI; labor forces (public vs private sector) women. surveys Gender wage premium Measures the wage differences between men and WWBI; labor forces women controlling for occupations and observable surveys worker characteristics. 3.6 Wage Incentives The extent to which wage increases are explicitly linked past year; and in Romania 95 percent of staff self-reported to performance is important for employee motivation receiving the highest rating in their annual evaluation.43 These and public sector productivity. There is increasing high performance ratings are ubiquitous in bureaucracies, rigorous evidence that unconditional salary increases have, though they do not rule out that promotion decisions may still unsurprisingly, no impact on staff performance. In Indonesia, reflect merit and be based on other sources of information than for example, a doubling of teacher salaries had no effects the annual performance evaluation. Many civil services pay on improving student learning outcomes.42 The two main scales have substantial automatic annual pay increases which channels for wage growth that can potentially be conditioned are fiscally costly, limit governments’ flexibility to respond to on performance are pay progression linked to promotions, and fiscal constraints, and have no incentives for performance. In annual bonuses or pay increments tied to the achievement Brazil, for example, generous automatic annual pay increases of yearly goals. Both are dependent on the extent to imply that most staff can more than double their real wages which competencies and performance inform promotion after 10 years of service.44 decisions and performance pay, which in turn, depends on how objectively performance can be measured and the An opposite and similarly pervasive problem arises from effectiveness of the broader human resource management flat hierarchies and limited promotion opportunities. In processes. While these personnel management aspects are some countries, public servants quickly progress to higher likely beyond the scope of employment and compensation grades and then exhaust any opportunities for further assessments, what can be explored are some key stylized promotion. In Romania, for example, 70 percent of civil facts that can provide suggestive evidence of the productivity servants had reached the highest technical grades and were and fiscal implications of pay incentives. These include: the ineligible for promotion.45 Limited promotion is common in magnitude of promotion-based and seniority-based annual teaching and healthcare, in part because of the specialized wage increases; the expected wage increases over a nature of the job. For example, in Bangladesh less than 10 representative employee’s career; and the coverage and key percent of teachers were promoted to a higher position in their design features of performance bonuses. careers, and almost no teachers were able to double their salaries over a 30-year career in many African countries.46 Seniority-based promotions and significant annual within- grade pay increases in many public sector bureaucracies Performance bonuses can be an important incentive weaken pay incentive effects. Useful indicators of this problem for jobs that have more easily measurable outputs. are the percentage of staff undergoing annual performance Performance bonuses are quite rare in low and middle-income appraisals and the percentage of staff that receive the highest countries and face a variety of design and implementation performance ratings. In Liberia, for example, only 30 percent problems when applied. By contrast, two-thirds of OECD of civil servants surveyed by the World Bank stated that their countries have some form of performance pay for their public manager had conducted their performance evaluation in the sector. Performance-related pay in the public sector is a 42. De Rhee et al (2017) 43. From World Bank surveys of public sector employees. 44. World Bank (2020a). 45. World Bank (2019). 46. Evans and Yuan (2018). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 32 controversial topic given the difficulties in measuring outputs, are the proportion of public sector staff eligible for the multi-dimensional nature of work where measuring some receiving the incentive, the probability of receiving activities can incentivize workers to ignore the unmeasured the incentive, and the size of the incentive. Universal tasks, and risks of favoritism and pay inequity that can result performance-pay schemes that do not factor the different in the absence of objective performance measures.47 It is attributes of jobs, especially the measurability of outputs, also an area of considerable academic research and the are likely to be costlier with unclear productivity impacts than available evidence shows that performance incentives can targeted schemes. If all staff receive the performance bonus, improve productivity for tasks with standardized delivery as has been common in many countries, then the incentive processes and relatively easily measurable outputs. These disappears, and the bonus becomes just another salary include processing of welfare payments, licensing and supplement that staff are guaranteed to receive irrespective of registration, tax and customs administration and, more performance. In the Brazilian federal government, for example, controversially, education and health. Figure 17 below two-thirds of the occupational groups have a performance summarizes the evidence from rigorous studies. Performance bonus scheme, and over 90 percent of staff in each of these pay for senior civil servants can also complement measures groups received the incentive.49 In some countries, such as to improve organizational performance management and the Philippines, there is a mandated distribution of rankings results-based budgeting.48 so that only a subset of staff are eligible for the reward. Also, if the performance incentive is too small, it will likely have no The design features of performance pay schemes that have impact, but if it is too big it can create incentives for cheating direct wage bill and productivity implications irrespective or tensions between staff because of the high stakes involved. > > > F I G U R E 1 7 - Summarizing the Evidence on Performance-Related Pay Jobs with repetitive Jobs with varied tasks Jobs with varied tasks tasks and easy-to- and difficult-to-measure but measurable outputs measure outputs outputs Administrative Teaching, Civil service services healthcare Evidence is positive Evidence is mixed Evidence is negative Source: Hasnain et al (2014). 47. Hasnain et al. (2014). 48. World Bank (2014) 49. Ibid. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 33 Another important type of wage incentive is additional large and financial incentives alone are usually not enough of payments to encourage health workers, teachers, an incentive and need to be combined with non-financial and and administrators to serve in rural areas or hardship career incentives to work.51 From a wage bill management locations, particularly in large and sparsely populated perspective what is important is that financial incentives be countries. Living conditions are often difficult in rural locations targeted to essential workers that are scarce in particular and a small, but growing, body of experimental literature localities, and to demographic groups, like younger doctors shows that higher salaries can help address these constraints and nurses, who may be more willing to serve in hardship and improve the recruitment and retention of employees in locations and would require lower incentive payments to remote areas.50 The size of the incentive though, needs to be compensate for opportunity costs. GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. What wage growth is typical for public sector employees in different occupational groups over the course of their career? How of much of this wage growth is dependent on promotions and how much can occur “within grade”? 2. Does the annual performance management system distinguish between high and low performers? Are promotion decisions largely seniority based and/or automatic? 3. Are there performance bonus schemes and if so what occupation groups and percentage of the staff receive them; and how large is the bonus that staff are expected to receive? 4. Is there targeted, significant additional pay for essential workers to serve in hard-to-staff locations such as rural areas? K E Y I N D I C AT O R S Indicator Description Data Sources Percentage of staff Indicator of the robustness of performance-based Administrative data receiving the highest promotions and performance bonuses. performance evaluations The size of annual, within- Measures the importance of seniority vis-à-vis Administrative data; Salary grade pay increases performance for wage increases. legislation Percentage of staff An indicator of whether performance bonuses are Administrative data receiving a performance regular salary supplements or distinguish between bonus good and poor performers. The size of the An indicator of the magnitude of the incentive. Administrative data; Salary performance bonus as a legislation percentage of basic pay The size of rural or An indicator of the magnitude of the incentive. Administrative data; Salary hardship allowance legislation and percentage of staff receiving it 50. Chelwa et al (2018). 51. Bhatti and McDonald (2019). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 34 4. >>> Measuring Impact 4.1 Fiscal Sustainability The wage bill can potentially have a major effect on fiscal balances, but there are no simple benchmarks for the “right” size of the wage bill. The most used metric for estimating the size of the wage bill, the wage bill as a share of GDP, is not a good indicator of fiscal impact given the cross-national heterogeneity in government functions, scope, and size. While the global wage bill is approximately nine percent of GDP, it is incorrect to conclude that countries with wage bills below this number, or below some other average for comparable countries, have more fiscally sustainable wage bills than countries with higher averages. Cross-nationally, there is no correlation between the size of the wage bill, either as a share of GDP or expenditures, and fiscal balances (Figure 18). For example, Denmark has one of the highest wage bills in the world at over 17 percent of GDP, but has generally achieved budgetary surpluses. A better measure is the wage bill as a share of expenditures and revenues, but even here there is weak correlation with fiscal deficits. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 35 > > > F I G U R E 1 8 - There Is No Strong Correlation between the Size of the Wage Bill and Fiscal Balances WAGE BILL (% OF PUBLIC EXPENDITURES) AND FISCAL DEFICITS 70 60 Wage bill (% of pub. exp.) 50 40 30 20 10 0 -10 -5 0 5 10 Fiscal deficits Source: Worldwide Bureaucracy Indicators, IMF World Economic Outlook, latest observations per country (multiple years). Wage bill dynamics within countries are a better contrast to low- and middle-income countries that tend to predictor of fiscal unsustainability, though the evidence finance it by increasing deficits. is limited. There are only a few studies that have explored the relationships between the wage bill, fiscal balances, and The sources of wage bill growth vary across countries. non-wage expenditures.52 These find that increases in wage An IMF review of 20 case studies revealed that the drivers bill tend to worsen fiscal balances, and wage expenditures are can be increases in wages (Latvia, Moldova, Romania, and procyclical, rising during periods of economic growth, but not South Africa), expansion in government employment (El falling as much during downturns due to structural rigidities. Salvador and Portugal), or a combination of the two (Kenya One study of 137 countries found that a one percentage and Tunisia).55 Sometimes, the growth is driven by increases point increase in wage-bill-to-GDP was associated with a 0.5 in allowances and non-transparent salary supplements as percentage point deterioration in fiscal balances.53 Another a back-door means to increase salaries after a period of study on the European countries found a similar pattern, fiscal consolidation, as was the case in Cameroon. Wage where a one percentage point increase in the wage bill as a expenditures also have a built-in momentum since public share of GDP correlates with an increase in the fiscal deficit sector salary scales have seniority-based pay increments that of half a percentage point.54 These studies also find that fiscal result in a natural rate of wage bill growth, even if employment deterioration is less in high income countries, which that is held steady and there are no across-the-board salary are more likely to finance wage bill increases by increasing increases. The example of Brazil is indicative, where micro- revenues and reducing other categories of expenditures, in level data allows for modeling scenarios that can decompose 52. IMF (2016); Eckhardt and Mills (2014); Dybczak and Garcia-Escribano (2019). 53. Dybczak, K. and M. Garcia-Escribano (2019). 54. Eckhardt and Mills (2014). 55. IMF (2016). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 36 the effects of wage bill increases due to additional hiring from agencies, as opposed to centrally negotiated collective those due to staff moving up the pay scale with increasing bargains with “peak unions” (representing unified smaller years of service. In some Brazilian states, for example, the unions), likely have higher transaction costs and compromise wage bill increases by over two percent in real terms annually, wage bill planning and the ability of governments to reduce even when staffing levels are fixed and there is a one-to-one wage bills in the face of crises. replacement of relatively higher paid retirees with relatively lower paid new hires, and annual nominal increases to the Wage bill modelling is useful for estimating the risk of salary scale are limited to inflation.56 fiscal unsustainability and for quantifying the savings from different reform options. Many governments have The type of wage setting arrangements can have integrated human resource and payroll systems that they use implications for fiscal sustainability. Although the studies for undertaking transactions, but which are also a vital, but are limited, the tentative findings suggest that the risk of often under-used, resource for wage bill planning. The micro- “fiscal drift” – an increase in wage bill and an increase in data from these systems can form the basis for simulations budget deficit – is positively correlated with union density, the that estimate the fiscal impact of different scenarios, such as percentage of public servants who are members of a union. freezing salaries or decreasing the rates at which salaries Also, the chances of fiscal tightening – reductions in the wage are readjusted; reducing the size of annual pay progressions bill and deficits – during an economic recession as well as or increasing the time interval between them; reducing during election years decrease with higher union density. staff replacement rates, a hiring freeze; and reducing pay Decentralized collective bargaining agreements, where dispersion. Brazil is a good example of such collaboration sectoral unions negotiate with their respective counterpart between World Bank and the authorities (Box 2). > > > B O X 2 - Brazil’s Wage Bill Micro-Data Analytics The World Bank gained access to vast amounts of individual level administrative data to analyze the Brazilian federal and state wage bills, showing its immediate impact on the country’s fiscal outlook. For the federal government, the analysis revealed that: • Lowering starting salaries of new hires by 10 percent (the federal civil servants receive almost double the wages of private workers with similar characteristics) could lead to savings of approximately R$26.4 billion and reduce the annual wage bill growth from 2.9 percent to 1.02 percent of GDP. • Adjusting the current replacement rate of 1.29 new employees for every retiree to 1:1 would generate savings of approximately R$44 billion. • Limiting salary increases to promotions for three years could have cumulative savings of approximately R$187 billion by 2030; and limiting salary increases to promotions for three years and thereafter adjusting the value of other salary increases not related to promotions to inflation could yield savings of over R$230 billion by 2030. The World Bank’s analysis received significant media attention because most of the debate used to revolve around anecdotal evidence as opposed to data. This attention focused on certain elements of the findings, in particular the average retirement age of public servants, the wage premium, and the automatic wage increases in the public sector. The Ministry of Economy is finalizing an administrative reform, which would require a constitutional amendment, that is informed by the World Bank report. Source: World Bank (2020a). 56. WB staff calculations based on micro-level payroll data. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 37 GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. What are the wage bill dynamics and how are changes in the wage bill being financed— i.e., through deficits, raising revenues, or reducing other expenditures? 2. What are the main drivers of wage bill growth? 3. What are the wage bill projections for baseline and different policy option scenarios based on wage bill modelling? K E Y I N D I C AT O R S Indicator Description Data Sources Correlation between wage To measure wage bill dynamics and fiscal sustainability. Budget data bill growth and fiscal balances Correlation between wage Provides an indication of whether the wage bill crowds Budget data bill growth and changes out other spending. in capital and non-wage recurrent expenditures Decomposition of wage To understand the sources of wage bill growth. Administrative data growth: employment and wages 4.2 Public Sector Productivity A key question is whether the public sector workforce is of outputs to the currency-value of all inputs. Productivity is performing well and delivering high quality infrastructure, difficult to estimate for the public sector, mainly because of services, and regulations, which is really a question of the difficulty in defining and measuring government outputs, public sector labor productivity. Productivity measures particularly for functions such as defense, foreign relations, the efficiency with which inputs, like labor, are converted into and environmental protection, and for calculating prices for outputs, and is a more precise and economically meaningful services for which there are no market transactions. concept than “performance,” since presumably performance can be improved by spending more while productivity Outputs Total Factor Productivity = measures whether more is produced and delivered for a All Inputs given wage bill. For public employment and compensation assessments, the main measure should be labor productivity, Outputs Labor Productivity = though a more commonly used metric in academic studies is Expenditure on Labor total-factor productivity, which is the ratio of the currency-value EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 38 Given these difficulties, a diversity of approaches that go for International Student Assessment (PISA), divided by beyond the textbook definitions of outputs necessarily sectoral wage expenditures, or the number of teachers for must be used, but which nevertheless provide useful a measure of per-teacher productivity, mortality rates or proxies for productivity. These alternative measures can vaccines administered per health worker; and health outputs be distinguished into “macro” approaches, which provide like number of consultations or patients discharged per doctor information at the level of an organization, sector, or service or wage spending in the health sector. The use of outcome as a whole; and “micro” approaches, which can be applied measures is not without its flaws as it typically includes factors to the individual employee, task, project, and process (Table that are beyond the control of the public organization or public 1).57 A common macro approach is to use service delivery official—for example, quality of support outside of school for outcome indicators per service delivery staff, which, given test results, and lifestyle choices for mortality rates, family and the high proportion of education and health personnel in housing conditions and other demand-side factors. Budget the public sector, provides a reasonably comprehensive execution rates can be another useful indicator to measure measure for a large segment of the public sector. These the productivity of public administration employees across a measures of productivity include standardized test scores variety of sectors and organizations, with the caveat that fast in the education sector, such as the OECD’s Programme spending does not necessarily imply good spending. > > > T A B L E 1 - Approaches to Measuring Public Sector Productivity Macro (Organization, Sector, Service) Cost-weighted output (Atkinson, 2005) Service delivery indicators Budget execution rates Micro (Employee, Task, Process) Revenue collection Procurement outcomes Staff and user satisfaction Subjective assessments (by employees, by stakeholders) Independent observers (`mystery shoppers’) and process productivity The knowledge/capacity of public officials Source: Somani, 2021. Micro approaches can be a good complement by providing productivity estimates for large and important organizations like revenue and customs agencies, infrastructure agencies, and social security administrations. Examples of outputs include revenue collection and tax audits conducted per revenue agency staff; project-completion rates for local public-infrastructure projects; organization-level road-construction-completion information; and individual employee-level case completion times in the social-security administration.58 Box 3 provides an example of applying some of these macro and micro approaches in the European Union member states. 57. Somani (2021). 58. Rasul and Rogger (2018). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 39 > > > B O X 3 - Pay and Employment and Public Sector Productivity in the European Union Below, we present trends in proxies of public-sector productivity for the EU as a whole, based on direct measures of public-sector output. We index the 2010 value at 100 and present the relative changes in the indicator over time. a. DOING BUSINESS: LABOR PRODUCTIVITY (PER WORKER) b. DOING BUSINESS: TOTAL PRODUCTIVITY 140 140 130 130 120 120 110 110 100 100 90 90 80 80 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 c. PISA READING: TOTAL PRODUCTIVITY d. LIFE EXPENTANCY: TOTAL PRODUCTIVITY 140 140 130 130 120 120 110 110 100 100 90 90 80 80 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 e. HOSPITAL DISCHARGES: LABOR PRODUCTIVITY f. REVENUE COLLECTION: TOTAL PRODUCTIVITY (PER WORKER) 140 140 130 130 120 120 110 110 100 100 90 90 80 80 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017 The data is sourced from the World Bank’s Doing Business Database, the OECD’s PISA Database, and Eurostat. The top-left figure shows the trend in the Doing Business Score (DBS) divided by the total number of civil servants working in the central administration, as a proxy for labor productivity in the area of business regulation. The top-middle figure shows the DBS divided by government expenditure on general public services, as a proxy for total factor productivity in business regulation. The top-right figure shows PISA reading scores divided by expenditure on education; the bottom-left the life expectancy at birth expenditure on health; the bottom-middle hospital discharges over the number of hospital staff; and the bottom-right the ratio of government revenue to the number of civil servants in central public administration. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 40 Bearing the caveats around such measures (presented above) in mind, these trends are still informative of important trends in the quality of public services in the EU, especially when the trends in outputs (numerators) and inputs (denominators) are analyzed separately. The evidence suggests a gain in productivity in business regulation; a decrease in total-factor productivity in social sectors, but evidence of a constant rate of labor productivity over the same period; and evidence of a substantial increase in revenue-collection productivity. Source: WB staff calculations based on Eurostat, Doing Business, and PISA data. In the absence of productivity measures, public the Worldwide Governance Indicators (WGI), reveal no employment and compensation assessments can explore clear relationship between the public sector wage premium correlations between the wage bill and expert-based and these measures of the quality of governance (Figure measures of institutional quality. While we may expect 19). In fact, some studies find that nepotism and corruption better-paid public employees to be more motivated and is higher in countries with large public sector wage premia, less corrupt, or to have less of an incentive to supplement suggesting that the higher value of public sector jobs can their salaries with bribes, cross-national estimates, using create opportunities for rent-seeking in recruitment.59 > > > F I G U R E 1 9 - The Public Sector Wage Premium Is Not Correlated with Measures of Institutional Quality a. WAGE PREMIUM AND GOVERNMENT EFFECTIVENESS 100% 80% Public sector wage premium 60% 40% 20% 0% -2 -1 0 1 2 -20% -40% Government Effectiveness Index 59. Rijckeghem and Weder (2001). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 41 b. WAGE PREMIUM AND CORRUPTION 80% 60% Public sector wage premium 40% 20% 0% -2 -1 0 1 2 -20% -40% Control of corruption index Source: Worldwide Bureaucracy Indicators, World Governance Indicators, latest observations per country (multiple years). GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Is total labor expenditure in the sector, including staff wages, outsourced labor costs, and external consultancy fees, correlated with sector-specific improvements in public-sector outputs and outcomes, so that productivity measures are stable or increasing within sector over time? 2. Using subnational data, if available, are there any outliers in the current distribution of total labor expenditure, public- sector output measures, outcome measures, or measures of productivity? Are there any outliers in the trends of these subnational measures—for example, recent rapid growth or declines in a particular subnational entity)? What might be the underlying features explaining the existence of such outliers? 3. Are total labor expenditures and public sector wages correlated with measures of institutional quality over time? Measures of institutional quality include the Worldwide Governance Indicators, perceptions of the quality of governance through household surveys – such as Eurobarometer, Afrobarometer, or World Values Surveys – the OECD’s Government at a Glance Indicators, the Quality of Government Index, and the World Bank’s Doing Business Indicators. With household level data and any other subnational data on measures of institutional quality, are there any outliers across subnational entities in the current distribution of institutional quality or outliers in the trends of measures of institutional quality? EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 42 K E Y I N D I C AT O R S Indicator Description Data Sources Correlation between To measure trends in public-sector productivity and the Budget data (wage-bill wage-bill growth and impacts of wage-bill expenditure on desired outputs/ growth). public-sector outputs and outcomes. outcomes. Administrative data; for example, agency annual reports, and statistics agencies for outputs and outcomes. 4.3 Labor Allocation between the Public and Private Sectors Public sector employment and compensation policies is much anecdotal evidence from even high-income countries can have major impacts on the entire labor market, of thousands of people applying for a few public sector particularly in low- and middle-income countries where vacancies.60 In many contexts affected by fragility, conflict, and the public sector is the dominant formal sector employer. violence (FCV), there may be few formal sector alternatives to As noted earlier, globally, the public sector accounts for 38 public sector employment, and a major challenge is how to percent of formal employment, reaching almost 50 percent for gradually relax this social role of the state in employment to low-income and lower-middle and countries. This large labor enable the private sector to grow. Box 4 provides an example market presence, combined with public sector wage premia of these labor market effects of public sector employment and job security, can lead to unemployment, particularly for and compensation in Ethiopia; similar conclusions have been fresh university graduates who queue for public sector entry- drawn based on research on the Colombian and West African level job openings and reject private sector job offers. There labor markets.61 60. Garibaldi and Gomes (Forthcoming). 61. See Albrecht et al. (2018) for Colombia; and Girsberger and Meango (2018) for West Africa. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 43 > > > F I G U R E 2 0 - Younger Public Sector Workers in Low and Lower-Middle Income Countries Have a Higher Wage Premium PUBLIC SECTOR WAGE PREMIUM, BY AGE OF WORKERS 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% -2% 15-24 yrs 25-45 yrs 46-64 yrs Low income countries Lower-middle income countries Upper-middle income countries High income countries Source: Worldwide Bureaucracy Indicators, latest observations per country (multiple years). Analyzing these labor market effects is technically older cohorts, which suggests that the distortionary effects of complicated, but assessments can explore a few stylized relatively high public sector wages are more pronounced for facts that can give an indication of these impacts. Given the youth. Figure 20 shows the data compared to formal sector that public sector turnover is low, and vacancies are more private workers. Government administrative data can also be likely to occur for entry-level positions, the public-private wage used to analyze these impacts by measuring the number of gap by age can provide evidence as to whether public sector applications per job opening, with inordinately high numbers wages are driving youth unemployment. Cross-nationally, also suggestive of these distortions. Complementary data we find that the youngest cohort of public sector workers in from investment climate assessments or enterprise surveys low- and lower-middle income countries – those in the 15 to can be used to assess whether there are skills shortages in 24 age group – have a higher wage premium compared to the private sector. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 44 > > > B O X 4 - Public Sector Wages and the Labor Market in Ethiopia The public sector is a highly attractive employment prospect in the Ethiopian labor market and favored towards those with a tertiary level of education. Wages are 50 percent higher in public administration than in services, the second best-paid industry, and individuals working in public administration are three times more likely to have a tertiary education than those working in services, the second-highest. As a result of this large wage premium, as opportunities to access tertiary education grew (through an expansion in local public universities), many individuals left the private sector, mainly from service-based jobs, to search for public employment (see top figure among those shown below). Furthermore, it was only in localities where the public sector was a relatively larger employer, improving the chances of landing a public-sector job, and where the public-sector wage premium was larger (making it more attractive to join the public sector, and hence to undertake a tertiary education in the first place) where individuals chose to enter higher education after a local university was established (bottom figures, left and right, respectively). For example, in localities where the public sector accounted for 50 percent of local employment, tertiary education attainment increased by almost 25 percentage points when a new public university was established. Tertiary education attainment increased much less in districts where the public sector was a smaller employer. Similarly, in districts where the public sector paid double the private sector, local tertiary education attainment increased by 13 percentage points, but much less in areas where the public sector was less lucrative. Together, this provides evidence that the number and wages of public-sector jobs impact: (i) education decisions—the types of skills and human capital in an economy; and (ii) private-sector labor supply. a. ESTIMATED EFFECTS ON EMPLOYMENT 4% 3.5% 3% 2% 1% Services 0% Public Administration -1% -2% -3% -4% 3.7% -5% EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 45 b. PERCENTAGE INCREASE IN DISTRICT’S TERTIARY EDUCATION ATTAINMENT BY SIZE OF LOCAL PUBLIC SECTOR 30 25 20 15 10 5 0 -5 -10 0 10 20 30 40 50 Size of district’s public sector (Proportion of local employment) Public wage premium: -1 = wages are twice as high in private sector; 1= wages are twice as high in public sector c. PERCENTAGE INCREASE IN DISTRICT’S TERTIARY EDUCATION ATTAINMENT BY PUBLIC WAGE PREMIUM 30 25 20 15 10 5 0 -5 -10 -1 -.5 0 .5 1 District’s Public Wage Premium Public wage premium: -1 = wages are twice as high in private sector; 1= wages are twice as high in public sector Source: Somani (Forthcoming). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 46 GUIDANCE QUESTIONS FOR THE ASSESSMENT 1. Is the public sector crowding out the labor market, impeding private-sector job growth for specific skills, occupation, demographic, and geographical groups? 2. Is the public sector having a distortionary impact on younger cohorts, particularly fresh university graduates, causing them to queue for public sector jobs and reject private sector jobs? K E Y I N D I C AT O R S Indicator Description Data Sources Public sector wage Provides an indicator of relative supply of young WWBI; labor force surveys premium by age cohorts workers to the public and private sectors. Number of applications per Provides an indicator of queuing for public sector jobs. Administrative data position (for a sample of positions) Correlation between public- To identify whether public-sector employment or labor WWBI; labor force surveys; sector employment growth demand is crowding out private-sector labor demand or administrative data or recruitment patterns and job growth (and the most-affected groups). private-sector employment growth or recruitment patterns (by skill, demographic, occupation, and geographical groups) Trends in educational To understand trends in the skills and skill demands WWBI; labor force surveys profiles of public-sector of the public sector relative to the private sector (the workers relative to public-sector education premium) and whether specific private-sector workers (by skill groups are increasingly drawn to or away from the demographic, occupation, public sector. and geographical groups) Correlation between To understand how the public-sector wage premium is WWBI; labor force surveys public-sector wage impacting labor supply to the public and private sectors. premium and relative employment growth in the public sector (by skill, demographic, occupation, and geographical groups) Skills shortages in the To understand if skills shortages are a constraint for the Enterprise surveys and private sector (from firm private sector. Investment Climate surveys) Assessments EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 47 5 >>> Conclusion This paper has aimed to provide a comprehensive framework for conducting public sector employment and compensation assessments to help develop evidence-based answers to three policy questions that governments care about: how can employment and compensation policies (i) contribute to sound fiscal management, (ii) increase public sector productivity, (iii) and improve the competitiveness of labor markets and help grow jobs? The importance of these policy questions will vary by country context. Most governments pay more attention to their wage and employment policies in times of fiscal distress and, therefore, the fiscal sustainability impacts are likely to remain paramount in such assessments. What this paper has emphasized is the importance of a holistic approach since the wage bill is qualitatively different from other government expenditures and other inputs in the government production function, and assessments need to also explore the productivity and labor market impacts. Understanding all these dimensions are necessary for a rigorous diagnostic and for identifying feasible yet meaningful reforms. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 48 Political economy factors are a major reason why pay and data and to better use the data that they do collect. employment reforms are difficult, but these have been Sometimes, robust data are not available but more often the discussed only briefly in this paper. With some exceptions, data that is available is not used. The WWBI provides robust such as on the political cycle in hiring, the assessment cross-national data, but there are gaps in country coverage framework is largely technical, mainly because political due either to the absence of labor force surveys or restrictions economy factors are country specific and difficult to identify in their use for research and policy purposes. Assessments and standardize in a set of diagnostic questions. Political can be used to advocate for improving national statistical economy factors though are paramount, either explicitly so, capacity or for making the data collected publicly available as in the role of trade unions, or implicitly as in most countries with appropriate safeguards to protect sensitive individual public sector employees are a powerful stakeholder and have data and abide by international data privacy standards. Many a significant voice in what reforms are on or off the table. governments have invested heavily in payroll and human The country-specific applications of the framework will need resource management information systems, often with to enrich the technical analysis with the underlying political World Bank and development partner support, but use these economy factors that result in the status quo practices, and systems largely for conducting transactions rather than as what reforms are politically feasible or require a change in the a basis for informing policies. These administrative systems political equilibrium to be implemented. provide a rich source of data and can help governments devise the flexible and targeted reforms that are needed, and The data needs are challenging, but these assessments to monitor the impact of these reforms. can be an impetus for governments to invest in better EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 49 >>> Annex 1. Definitions and Data Sources This report uses the standard definitions of the “general While limiting the analysis to the general government is likely government” and “public sector” in the analysis of the six appropriate and sufficient for most diagnostic assessments, employment and compensation practices and their impacts. in some countries the broader consideration of the public As the figure below depicts, the general government are all sector may be necessary, if there is a large state-owned institutional units controlled directly, or indirectly, by central enterprise sector that has significant claims on the general and subnational governments, and includes the wage bill of government budget, and these public corporations can have all workers employed by these units—for example, public claims on the central, state, or local governments. Moreover, administrators, service delivery personnel, military and many household and labor force surveys do not allow for security personnel). The public sector is general government an accurate disaggregation between general government plus public- or state-owned corporations. This conceptual and public enterprises and, therefore, much of the analysis, distinction between the public sector and general government particularly public-private comparisons, requires the broader is purely to clarify concepts and in no way implies that definition of the public sector. responsibilities for employment and compensation policies so neatly divided in countries. Public Sector Public General Corporations Government Central State Local Government Governments Governments Source: Authors. The main data source used for cross-national analysis is the Worldwide Bureaucracy Indicators (WWBI). The objective of the WWBI is to provide comprehensive, cross-national data on public sector employment and compensation to better enable researchers, development practitioners, and policymakers identify and implement evidenced-based reforms. The WWBI country- level indicators are constructed from nationally representative household surveys that are designed and implemented by national statistical agencies. The WWBI are based on microdata from 909 labor force and household welfare surveys, which translates to 53 million unique survey observations, and consist of 112,919 estimations across 192 indicators for 202 countries and territories between 2000 and 2018. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 51 The WWBI encompass five categories of variables: occupational categories in the public and private sector, the relative wages of key occupations within the public sector, and 1. The size and demographics of the public and private the cross-country comparisons of the compensations of public sector workforces (107 indicators). sector workers by occupations. Indicators on gender pay 2. Public sector wage premiums (39 indicators). gap compare the wages of females to their male colleagues 3. Relative wages within the public sector (35 indicators). in the public and private sectors as well as by industry of 4. Gender pay gaps (9 indicators). employment. Indicators on the relative size of the wage bill 5. The public sector wage bill (2 indicators). offer a glimpse into the structure and affordability of the public sector within the larger economy. The demographics of public and private employment track key characteristics, including the size of the public sector The wage data in the WWBI denote the income associated with workforce in absolute and relative numbers, their age, the occupation of employment used in the analysis (which the and distributions across gender, rural and urban locations, individual dedicated most of their time in the week preceding academic qualifications, wage quintiles, industry categories the survey) and excludes both bonuses, allowances, and and occupational groups. The indicators on public sector other in-cash/-kind payments from the same job as well as all wage premiums capture the overall competitiveness of public additional sources of income (from other jobs) or investments sector wages (compared to the private sector) as well as the and transfers. Due to the almost complete lack of information decomposed public-private wage differentials by gender, on taxes, the wage from primary job is not net of taxes. For academic qualifications, industry category and occupation all those with self-employment or owners of businesses, this groups. Indicators on pay compression ratios present the corresponds to net revenues (net of all costs excluding taxes) relative wages of the top and bottom earners in the public or the amount of salary withdrawn from the business. and private sectors, the ratios of wages for employees of EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 52 >>> Annex 2. Streamlined Assessment for Poor Data Country Contexts This report has emphasized the advantages of using well- detailed in the report. The streamlined approach requires designed, nationally representative surveys of individuals less data for the key indicators to help answer the guidance in the labor force to assess public sector employment and questions, including only for a few, critical organizations, and compensation policies and their impact on productivity, jurisdictions in cases where a complete picture of the general labor allocation, and fiscal sustainability. However, the government is not available. It also limits the data sources heterogeneous distribution of capacity within national to budget and administrative data that most governments statistical offices to collect this information may impact its collect as part of their daily operations, such as registering usability in a wage bill assessment. Therefore, this section employees and processing salary payments. This reliance lays out a more limited approach for use in data-constraint on only administrative data does significantly constrain the environments, utilizing the same micro-foundations as the assessment, particularly on the dimensions of employment framework presented in the report. and wage competitiveness. Nevertheless, the following core sets of indicators combined with the guidance questions The structure of the assessment remains unchanged, focusing should provide an acceptably empirical assessment of public on six core dimensions of pay and employment practices and sector employment and compensation in such contexts. their three impacts, with the same set of guidance questions A S S E S S I N G WA G E B I L L D I M E N S I O N S - C O R E I N D I C AT O R S Indicator Description Data Sources 1. Wage Bill Planning Public expenditures To measure expenditure dynamics in the wage bill Budget data; IMF Article IV across main expenditure vis-à-vis other expenditure categories like pensions, data annexes categories transfer, and good and services. Reported wage bill To measure wage bill dynamics and fiscal sustainability, Budget data; IMF as share of GDP, recognizing that the reported wage bill may not capture Government employment expenditures, and all expenditures. and compensation dataset revenues Wage bill trends for To understand the trends in usually the two largest Budget data education and health public employment sectors. sectors Public sector employment To understand the sources of wage bill growth for the Administrative data trends in key sectors largest public employment sectors. 2. Wage Bill Controls Deviations between This indicator measures the extent to which aggregate Budget and outturn data budgeted and actual wage actual expenditure on the compensation of government bill expenditures in the workers reflects the amount originally approved, as previous three fiscal years: defined in government budget documentation and fiscal For central government reports. The data can be limited to a few, high spending (or a sample of ministries) ministries and local governments. and a sample of local governments EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 54 Indicator Description Data Sources Filled positions as a This indicator measures the sanctity of wage Administrative data proportion of budgeted expenditure ceilings as high percentage of vacant positions (for a sample positions can create perverse incentives of using of ministries and local savings from vacant positions to finance compensation. governments) Ghost workers as a Measures the effectiveness of payroll controls. Payroll audits percentage of total workers in education and health sectors Wage bill arrears as a Measures delays in salary payments. Administrative data percentage of total wage bill expenditures (for select ministries) 3. Employment Levels and Distribution Educational profile of Decomposition of public sector workers by educational Administrative data public sector workers (for qualifications allows for an assessment of the level of select ministries and local skills in the public workforce. governments) Distribution of public sector The distribution of public sector employment across Administrative data workers by pay grade (for clerical and ancillary staff, implementing staff, and select ministries and local managerial staff measures the quality of government governments) workforce planning. Student-teacher ratios; Provides a measure of the adequacy of key service Administrative data per-capital health workers delivery staff compared to international norms. Share of public sector Measures the representativeness of the public sector Administrative data employees (across key workforce. personal demographics) for select ministries and local governments 4. Wage Competitiveness Public-private wage gap Measures the wage difference between public and Administrative data; Data for select occupations private sector pay for specific occupations. The premia from a sample of private (teachers, medical are calculated for jobs without any demographic providers workers) based on raw controls on the assumptions that this detailed micro- (without controls) wage data is not available for the private service providers. comparisons of jobs EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 55 Indicator Description Data Sources Turnover rates for public Indicator of the relative competitiveness between the Administrative data sector employees (within public and private markets for key occupations where occupations with private individuals can choose between the public and private sector alternatives like sector workforces. teachers and health workers) Pay comparisons for select Measures pay of a particular public sector occupation ICP data; administrative occupations (e.g., teachers compared to that in a few regional comparators for that data and medical workers) occupation. across public sectors 5. Wage Equity Public sector pay The ratio of salaries between the highest and lowest Administrative data; review compression ratio (select deciles in the pay scale. of salary legislation ministries and local governments) Relative wages between Provides an estimate of pay equity given the different Administrative data; review different occupations (for a responsibilities and demands of different jobs. of salary legislation sample of occupations) Within-occupation public Provides estimates of wage differences controlling for Administrative data; review sector pay dispersion (for a occupational groups and worker characteristics. of salary legislation sample of occupations) 6. Wage Incentives Percentage of staff Indicator of the robustness of performance-based Administrative data receiving the highest promotions and performance bonuses. performance evaluations (sample of ministries or local governments) The size of annual, within- Measures the importance of seniority vis-à-vis Administrative data; Salary grade pay increases performance for wage increases. legislation Percentage of staff An indicator of whether performance bonuses are Administrative data receiving a performance regular salary supplements or distinguish between bonus (sample of ministries good and poor performers. or local governments) The size of the An indicator of the magnitude of the incentive. Administrative data; Salary performance bonus as a legislation. percentage of basic pay EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 56 Indicator Description Data Sources The size of rural or An indicator of the magnitude of the incentive. Administrative data; Salary hardship allowance legislation and percentage of staff receiving it A S S E S S I N G I M PA C T - C O R E I N D I C AT O R S Indicator Description Data Sources 1. Fiscal Sustainability Trends in wage bill and To measure wage bill dynamics and fiscal sustainability. Budget data fiscal balances Trends in wage bill and Provides an indication of whether the wage bill crowds Budget data capital expenditures out other spending. 2. Public Sector Productivity Correlation between To measure trends in public-sector productivity and the Budget data and service wage-bill growth and impacts of wage-bill expenditure on desired outputs/ delivery indicators socio-economic outcomes outcomes. (for key public facing ministries/departments 3. Labor Allocation between Public and Private Sectors Number of applications per Provides an indicator of queuing for public sector jobs. Administrative data vacancy in public sector (for a sample of ministries and positions) Turnover of public sector An indicator of demand for private sector jobs. Administrative data employees (for a sample of ministries and positions). EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 57 >>> References Albrecht, J., M. Robayo-Abril, and S. Vroman. 2018. “Public-Sector Employment in an Equilibrium Search and Matching Model.” Economic Journal 129(617): 35–61. Antwi, J., and G. Phillips. 2013. “Wages and health worker retention: Evidence from public sector wage reforms in Ghana.” Journal of Development Economics 102: 101-115. Araujo, Felipe A. 2020. “Selection Bias and the Returns to Public-Sector Employment.” Mimeo, University of Pittsburgh. Australian Public Service Remuneration Report. https://www.apsc.gov.au/sites/default/ files/18330_apsc_-_remuneration_report-accessible-v3.pdf). Borjas, G. 2002. The Wage Structure and the Sorting of Workers into the Public Sector. Working Paper 9313, National Bureau of Economic Research. https://www.nber.org/papers/w9313. Bhatti, Z. and L. McDonald. 2019. Overcoming shortages of skilled civil servants in remote and hardship areas. Governance Notes. Washington, DC: World Bank. Bowne, J. B., K. Magnuson, H. Schindler, G. Duncan, and H. Yoshikawa. 2017. “A Meta-Analysis of Class Sizes and Ratios in Early Childhood Education Programs: Are Thresholds of Quality Associated with Greater Impacts on Cognitive, Achievement, and Socioemotional Outcomes?” Educational Evaluation and Policy Analysis 39(3): 407–428. Bruns, B., D. Filmer and H. A. Patrinos. 2011. Making Schools Work: New Evidence on Accountability Reforms. Washington, DC: World Bank. Chelwa, G., M. Pellicer, and M. Maboshe. 2018. Teacher Pay and Educational Outcomes: Evidence from the Rural Hardship Allowance in Zambia. ERSA Working Paper. https://www. econrsa.org/system/files/publications/working_papers/working_paper_764.pdf Czibor, E. 2021. “Gender Equality in the Public Sector: Bridging the Gap Between Experimental Evidence and Practice.” Presentation given to the World Bank, March 2021. Dahlberg, M., and E. Mork. 2008. “Is There an Election Cycle in Public Employment? Separating Time Effects from Election Year Effects.” CESifo Economic Studies 57(3): 480-498. De Rhee, J., K. Muralidharan, M. Pradhan, and H. Rogers. 2017. “Double for nothing? Experimental Evidence on an Unconditional Teacher Salary Increase in Indonesia.” Quarterly Journal of Economics 133(2): 993-1039. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 58 Dybczak, K., and M. Garcia-Escribano. 2019. Fiscal Implications of Government Wage Bill Spending. IMF Working Paper. Washington, DC: International Monetary Fund. Eckhardt, S., and Z. Mills, 2014. What Goes Up Must Come Down. Cyclicality in Public Wage Bill Spending. World Bank Policy Research Working Paper. Washington, DC: World Bank. Evans, D., and F. Yuan. 2018. The Working Conditions of Teachers in Low and Middle-Income Countries. Washington, DC: World Bank. Finan, F., B. Olken, and R. Pande. 2017. “The Personnel Economics of the Developing State.” In Handbook of Economic Field Experiments. Vol. 2 (Abhijit Vinayak Banerjee, Esther Duflo, eds): pages 467-514. https://doi.org/10.1016/bs.hefe.2016.08.001. Garibaldi, P., and P. Gomes. Forthcoming. The Economic of Public Employment: An Overview for Policymakers. Gelb, A., and J. Clark. 2013. “Identification for Development: The Biometrics Revolution.” Working Paper 315. Washington, DC: Center for Global Development. Girsberger, E., and R. Meango. 2018. “The Puzzle of Educated Unemployed in West Africa.” IZA Conference paper. https://conference.iza.org/conference_files/2017_Labor/girsberger_ e9967.pdf. Gomes, P. 2016. https://www.youtube.com/watch?v=Uar2IvyLxLI. Hasnain, Z., N. Manning, and J. Pierskalla. 2014. “The Promise of Performance Pay? Reasons for Caution in Policy Prescriptions for the Core Civil Service” World Bank Research Observer 29(2): 235-264. Hausman, R., L. Nedelkoska, and S. Noor. 2020. “You Get What You Pay for: Sources and Consequences of the Public Sector Premium in Albania and Sri Lanka.” CID Faculty Working Paper No. 376. https://growthlab.cid.harvard.edu/files/growthlab/files/2020-02-cid-wp-376- public-sector-premium.pdf Holm-Hadulla, F., K. Kamath, A. Lamo, J. Pérez, and L. Schuknecht. 2010. “Public Wages in the Euro Area: Towards Securing Stability and Competitiveness.” Occasional Paper Series. Frankfurt: European Central Bank. Independent Evaluation Group. 2008. Public Sector Reform: What Works and Why? Washington, DC: World Bank. International Monetary Fund. 2016. Managing Government Compensation and Employment: Institutions, Policies, and Reform Challenges. Washington, DC: World Bank. Jain, H., and R. Palacios. Forthcoming. The Challenge of Civil Service Pension Reform in Asia. Washington, DC: World Bank. Kuipers, N. 2020. “The Indonesian Public Sector Wage Premium.” Mimeo, University of California, Berkeley. Martins, P. S. (2010): “Cronyism.” IZA Discussion Papers 5349. https://ftp.iza.org/dp5349.pdf. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 59 Majgaard, K., and A. Mingat. 2012. Education in Sub-Saharan Africa: A Comparative Analysis. Washington, DC: World Bank. OECD. 2014. Education at a Glance: OECD Economic Indicators. Palacios, R., and H. Jain. 2020. Pension Provision in Thailand: Rethinking old age income support in a rapidly aging society. Washington, DC: World Bank. Pierskala, J., and A. Sacks. 2017. “Personnel Politics: Elections, Clientelistic Competition and Teacher Hiring in Indonesia.” British Journal of Political Science 50(4): 1283-1305. Rasul, I., and D. Rogger. 2018. “Management of Bureaucrats and Public Service Delivery: Evidence from the Nigerian Civil Service.” The Economic Journal 128(608): 413-446. Rijckeghem, C. V., and B. Weder. 2001. “Bureaucratic corruption and the rate of temptation: do wages in the civil service affect corruption, and by how much?” Journal of Development Economics, 65(2): 307–331. Rosenfeld, J. 2021. You’re Paid what you are Worth: And Other Myths of the Modern Economy. Cambridge, Massachusetts: Belknap Press. Somani, R. 2021. Public Sector Productivity: Why is it Important and how can we Measure it? Equitable Growth, Finance, and Institutions Insight. Washington, DC: World Bank. Somani, R. Forthcoming. The Returns to Higher Education and Public Employment. World Development. UNESCO. 2017. UIS Database on Education. United Kingdom. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/ attachment_data/file/740064/Fortieth_Annual_Report_on_Senior_Salaries_2018.pdf. Whitehouse, E. 2006. Pakistan: Reforming Civil Service Pensions. Washington, DC: World Bank. World Bank. 2013. Pay Reform in the Indonesian Civil Service. Washington, DC: World Bank. World Bank. 2014. Pay Flexibility and Government Performance. Washington, DC: World Bank. World Bank. 2016. Serbia: Horizontal Functional Review of Central Government. Washington, DC: World Bank. World Bank. 2017. A Fair Adjustment: Efficiency and Equity of Public Spending in Brazil. Washington, DC: World Bank. World Bank. 2018. Cameroon Public Expenditure Review. Washington, DC: World Bank. World Bank. 2019. Romania: Baseline Review of the National Framework for HRM and its Institutionalization. Washington, DC: World Bank. World Bank. 2020a. Brazil: Public Sector Wage Bill and Personnel Management. Washington, DC: World Bank. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 60 World Bank. 2020b. Liberia Forestry Development Authority: An Institutional Capacity Assessment. Washington, DC: World Bank. Yassin, S., and F. Langot. 2018. “Informality, public employment and employment protection in developing countries.” Journal of Comparative Economics 46(1): 326–348. EQUITABLE GROWTH, FINANCE & INSTITUTIONS INSIGHT <<< 61