Abstract
Selected Issues
Wage Bill Dynamics and Reform Options1
A. Introduction
1. Mozambique’s economy faces serious challenges. The economy slowed down starting in 2015 due to several factors, including lower commodity prices, adverse weather conditions, and loose macroeconomic policies. The economic situation deteriorated further following the disclosure in 2016 of undisclosed borrowing by two public companies amounting to about 11 percent of GDP and the ensuing freeze in donor budget support. While the external sector is showing signs of stabilization, the fiscal stance remains excessively loose.
2. Significant fiscal adjustment is required over the medium term to restore macroeconomic balance. Prior to the crisis, strong revenue performance, especially sizeable capital gain taxes from the mining sector, provided amply fiscal space for spending, while the government tapped into non-concessional borrowing to scale up public investment in anticipation of future revenues from natural gas production. However, the lack of fiscal buffers implies that fiscal policy became unsustainable as the crisis hit, while the depreciation and the disclosure in April 2016 of previously hidden debt caused the stock of public debt to surge to a distressed level.
3. Going forward, containing wage bill spending should be an important component of fiscal consolidation. Mozambique’s central government wage bill has experienced a rapid increase in recent years, rising from 8 percent of GDP in 2010 to 11.3 percent in 2016, which is significantly above the median across a sample of countries with similar level of development. The observed increase in wage bill spending could potentially reflect several factors, including ad-hoc wage and salary adjustments, rapid increases in hiring, and issues related to wage bill planning and budgeting, and payroll execution and control. Containing wage spending will be important not only for achieving fiscal targets, but also to avoid a continuation of the crowding out of priority social and infrastructure spending (Figure 1).
Mozambique: Central Government Wage and Capital Spending
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Note: Capital expenditures excludes those financed by project grants.Mozambique: Central Government Wage and Capital Spending
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Note: Capital expenditures excludes those financed by project grants.Mozambique: Central Government Wage and Capital Spending
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Note: Capital expenditures excludes those financed by project grants.4. This study reviews compensation and employment policies and developments, and provides options for reform. The appropriate mix of policy options depends on the objectives, priorities, and preferences of the government. Section II discusses recent trends in compensation and employment and places these considerations in an international context. Section III discusses issues related to wage bill management and identifies challenges in implementing reforms. The last section provides options for reforming the wage bill.
B. Trends in Compensation and Employment
5. Wage bill spending by the central government reached 11.3 percent of GDP in 2016. The central government’s wage bill is set by the legislature as part of the overall budget process and comprises wage bill spending by the budgetary central government, the provinces and districts, the autonomous agencies, and other public bodies that are fully or partially funded by the state’s budget. The wage bill expenditure item recorded in the fiscal accounts excludes compensation paid to 3,755 doctors financed by the Common funds (0.15 percent of GDP in 2016), and social security contributions (0.5 percent. Central government employment is estimated at around 393.7 thousand employees in 2016 (1.4 percent of the total population; 3.4 percent of the working age population).2 This figure covers all sectors except defense, police and security, on which information is not made available.3 Overall, regular full-time employees represent around 80 percent of the total workforce in the central government.
6. Education and other (non-health) sectors absorb the largest shares of the wage bill (Figure 2). Education is the largest employer, with a share of 52.2 percent of total employment in 2016, absorbing around 40.4 percent of total wage spending. The health sector’s share is rather small, at only 12.8 percent of total employment, and absorbs only 9.6 percent of total compensation. The other sectors (grouped together under one category), and for which a breakdown is not made available, constitutes the second largest group of employers in the government at 35.0 percent in 2016, but absorbs the largest share of total compensation (at around 50.0 percent).4
Mozambique: Central Government Wage Bill Spending by Sector
(Percent of total spending)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill Spending by Sector
(Percent of total spending)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill Spending by Sector
(Percent of total spending)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.7. The central government wage bill in Mozambique is very high compared to peers. At 11.3 percent of GDP in 2016, the wage bill exceeds the median for a group of 42 low-income and Developing countries (6.6 percent; Figure 3). It is the fourth highest after Zimbabwe, Lesotho, and Liberia. Moreover, it is very high as a share of primary expenditures (40.3 percent) and as a share of domestic revenues (47.11 percent). The former indicator suggests that the control of government spending dynamics can be a very challenging task, while the latter indicator points to sustainability issues of wage outlays in Mozambique.
Mozambique: Central Government Wage Bill, Low Income and Developing Countries, 2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill, Low Income and Developing Countries, 2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill, Low Income and Developing Countries, 2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.8. The high wage bill level compared to peers seems to reflect high average compensation, rather than high employment levels (Figure 4). The central government employment to population ratio in Mozambique is around the median ratio across peers (a sample of low income and developing countries). This suggests that high wage spending levels in Mozambique most likely reflect higher compensation levels.
Mozambique: Central Government Wage Bill and Employment
(Low income and developing countries, latest year)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Source: IMF staff calculations.Mozambique: Central Government Wage Bill and Employment
(Low income and developing countries, latest year)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Source: IMF staff calculations.Mozambique: Central Government Wage Bill and Employment
(Low income and developing countries, latest year)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Source: IMF staff calculations.9. Not only is the wage bill relatively high compared to peers, it has also been rising rapidly since 2010, reflecting compensation dynamics in some sectors and employment dynamics in other sectors. The wage bill increased from around 8 percent of GDP in 2010 to around 11.3 percent of GDP in 2016 (Figure 5), raising concerns about fiscal sustainability going forward. This is partly explained by in large differential wage dynamics across different sectors in the government since 2013. Government decisions to grant wage increases to different sectors have initially favored mostly doctors and medical staff who organized strikes in 2013 demanding higher wages. Teachers and police workers have also received preferential wage increases since 2013. Overall, the cumulative, inflation adjusted, granted wage increases over 2013–2015 amounted to around 32, 18, and 11 percent for doctors, teachers and police, and other civil servants, respectively. By 2016, the cumulative real wage increases that were granted by the government were still relatively high despite the large increase in inflation from 2.4 percent in 2015 to 19.2 in 2016. On the other hand, the rapid increase in wage spending seems to be also driven by an increase in employment outside the health and education sector, over time (Figure 6).5
Mozambique: Central Government Wage Bill Spending, 2010–2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill Spending, 2010–2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Wage Bill Spending, 2010–2016
(Percent of GDP)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.
Mozambique: Central Government Employment, 2010–16
(Percent of working age population)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Employment, 2010–16
(Percent of working age population)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Central Government Employment, 2010–16
(Percent of working age population)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.10. Increases in wage bill spending have been discretional, driven by both compensation and employment dynamics over time (Figure 7). We decompose the increase in the wage bill into the contributions of compensation and employment over time for the overall central government.6 The analysis suggests that compensation has been the main driver behind the recent increase in wage bill spending, with about 67 percent (median over 2014–2016) of the growth being due to rising compensation. Such relative importance of compensation as a driver of wage bill growth has increased over time from about 43 percent (median over 2011–2013). Further decomposition of wage bill spending changes by sector over the period from 2014 to 2016 suggests that while compensation has also been the main driver behind the increase in wage bill spending in the health and “Other” sectors, employment dynamics have played a more prominent role in the education sector.
Mozambique: Contributions of Wages and Employment to the Wage Bill Growth, 2011–16
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance and IMF staff calculations.Mozambique: Contributions of Wages and Employment to the Wage Bill Growth, 2011–16
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance and IMF staff calculations.Mozambique: Contributions of Wages and Employment to the Wage Bill Growth, 2011–16
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance and IMF staff calculations.11. Allowances, bonuses, and other supplements account for a large share of total compensation (Figure 8). Allowances, bonuses and other supplements currently account for about 43 percent of the total wage bill. This is well above the level typically found in OECD countries, where the base wage component averages nearly 90 percent of the total wages of civil servants. Furthermore, the share of these salary supplements in the wage bill has been increasing over time, up from around 40 percent in 2012. Other supplements contain several categories of remuneration, including a “13th month” salary.7
Mozambique: The Composition of Wage Bill Spending, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry Economy and Finance; and IMF staff calculations.Mozambique: The Composition of Wage Bill Spending, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry Economy and Finance; and IMF staff calculations.Mozambique: The Composition of Wage Bill Spending, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry Economy and Finance; and IMF staff calculations.12. The contribution of seniority-related increments to the wage drift in Mozambique is estimated to be about 1.3 percent per year. The wage drift refers to aspects of government employment policy—such as seniority-related increases in compensation, promotion policies, and the reclassification of employment positions of their salary grade—that often automatically drive public wage growth independently of wage policy and employment decisions. The contribution of seniority-related increments to wage drift in the Mozambique is estimated to be about 1.3 percent per year for civil servants who enter the workforce in their early 20s and retire at the statutory retirement ages (55 for women and 60 for men). While often unnoticed, the wage drift—which is not directly linked to either performance or productivity—is inevitably compounding upward pressure on wage costs.
13. Recent slippages in the wage bill in 2016 are concerning and raise questions about compliance issues (Figure 9). Recent slippages of the wage bill in 2016 amounted to around 1.2 percent of GDP relative to the budgeted amount.8 Data provided by the authorities suggest that the health sector only accounts for 14 percent of the slippage, while the bulk of the slippage occurred in the education sector (around 72 percent).9 Poor forecasting and unrealistic budgeting seems to have played a major role. For instance, salary increases granted by the government in April 2016 were somehow not incorporated in the revised supplementary budget of August 2016. The deactivation or overriding of the controls embedded in the systems and legal procedures at the central or provincial and district levels may have also played a role. A buildup of wage arrears from previous years, other salary increases beyond those granted in April 2016, and a significant increase in the hiring of new staff in other sectors are additional possible factors. Without detailed data, it is not possible to uncover the factors behind the observed slippage and quantify their relative contribution.
Mozambique: Deviation of Wage Outlays Relative to the Budget, by Sector, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Deviation of Wage Outlays Relative to the Budget, by Sector, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.Mozambique: Deviation of Wage Outlays Relative to the Budget, by Sector, 2016
(Percent)
Citation: IMF Staff Country Reports 2018, 066; 10.5089/9781484345634.002.A004
Sources: Ministry of Economy and Finance; and IMF staff calculations.14. In the absence of reforms, wage spending is expected to rise from 11.3 percent of GDP in 2016 to 12.0 by 2021. Wage bill projections—which include the impact of measures that have already been taken in 2017 as well as in the context of the 2018 budget—presume that: (i) basic salaries will rise at the rate of inflation plus one-half the rate of real GDP growth;10 (ii) the “13th month” salary will be paid in full to all employees from 2018 onwards; (iii) the cost of allowances, overtime, bonuses, and other salary supplements (as a share of basic salaries) will remain unchanged over the projection horizon; (iv) the number of employees in the Education sector will grow in line with the population of persons under the age of 20; and (v) the number of employees in the Health sector and the “Other” sector category will rise in line with overall population growth.
C. Key Issues in Wage Bill Spending
Compensation
15. Public sector wages in Mozambique are set according to a formula that introduces a form of indexation, effectively reducing the flexibility of the government in addressing inefficiencies in wage levels and structures and in managing macroeconomic imbalances. Within this framework, the growth of public wages considers (i) the rate of inflation, (ii) half the rate of real GDP growth (as a proxy for productivity), and (iii) a factor “delta” that presumably reflects negotiations between the government and professional employee associations.11 In practice, the factor “delta” is unpredictable and introduces a subjective and rather ad-hoc element to the wage setting mechanism. This leads to unpredictable dynamics in public wages over time. For instance, the existing wage rule has resulted in basic salaries rising substantially faster than inflation in previous years. In combination with the seniority-based wage increases, most public sector employees can receive significant real increases in the value of their basic wages every year, regardless of their productivity or performance.
16. While rule-based frameworks for wage setting may appear to offer a remedy against repeated and large public sector wage increases, they can have several disadvantages. These include (i) limiting the ability of the government to use fiscal policy for macroeconomic management;12 and (ii) potentially locking into place undesirable public sector wage levels and wage structures. The latter reduces the capacity of the public sector to address inefficiencies by automatically granting wage increases to a wide range of workers, regardless of any determination of whether their salaries are at an appropriate level.
17. Regression analysis on individual-level data from the Integrated Household and Labor Force Survey of 2015 suggests that some central government employees receive a wage premium over their private sector counterparts. After accounting for differences in socio-economic characteristics of workers, including age, gender, education, experience, and regional effects, the magnitude of the wage premium varies significantly by education level. The premium, in absolute terms, for lower-skilled employees with primary and secondary education is around 63 and 91 percent, respectively, relative to private sector counterparts with similar socio-economic characteristics. On the other hand, high-skilled government employees with university level education tend to have a wage discount, effectively earning significantly less than their private sector counterparts. This tends to reflect the observed gap between lower-skilled and high-skilled wage premiums exists across countries, which is especially pronounced in low income and developing economies. Nevertheless, wage premiums analyses do not take account the fact that government workers have access to numerous generous benefits (including generous public sector pension schemes, and in-kind and non-monetary benefits) and are provided with greater job security. Thus, the government does not always necessarily need to offer salaries that are exactly on par with the private sector to retain high-quality employees. For this reason, a modest wage discount for public sector workers is sometimes expected. Appendix I describes the empirical analysis for wage premiums.
18. The compensation structure is very complex and entails large fiscal costs. Micro data from the registry of public employees (e-CAF) and the payroll management database system (e-Folha),13 suggests that there are at least 46 different salary supplements. These supplements comprise (i) 19 allowances; (ii) 4 types of bonuses; and (iii) 23 different other salary supplements (including compensation for overtime). Some of these supplements are well entrenched in the salary structure. For example, the profitability bonus, which is awarded to civil servants or state officials who are deemed to have a very good performance assessment, is equal to 100% of the corresponding base salary. Another example is the reintegration allowance, which consists of the inclusion of 75 percent of the base salary, for each year spent in a governing position.14 For instance, in the case where a civil servant occupies a governing position (e.g., as Head of Administrative Post) for a period of four years, he or she is entitled to receiving a reintegration allowance of 300 percent (75 percent, cumulative over four years) upon returning to his or her original position.
Employment
19. There are issued related to compliance in the hiring process and the application of rules governing contracting employees. The procedures of appointment are subject to double-control by the Administrative Tribunal (AT)—first at the time of an employee’s provisional appointment, and second at the permanent appointment.15 The AT is responsible for ensuring compliance with established procedures, most importantly with regard to the authorization of appointments, which should reflect the need for the appointment, the presence of vacancies, and resource availability. Nevertheless, despite these procedures, there appear to be problems with procedural compliance. The AT has identified challenges – relating to the adoption of manual procedures for ensuring compliance with hiring – which have led to hiring of employees above the authorized figures.16
20. Issues related to the identification of core functions within ministries merit examination, as they may be creating hiring pressure. There are issues that may be related to the duplication of roles and responsibilities in some government ministries, departments, and agencies. For instance, in the education sector, a sizeable share of employees who are teachers by training are effectively working as public administrators. Effectively, there are teachers in the education sector who do not teach in classrooms. This raises serious concerns about the definition of functions, roles and responsibilities in the sector, particularly given that pressures on the hiring of teachers are mounting and are effectively driving an increase in overtime pay. Overtime pay is costly, and can be reduced by redefining the functions and responsibilities of teachers working as public administrators in the sector.
21. The challenges discussed above also reflect potential weaknesses in PFM. (i) The budgeting and forecasting framework does not always make use of all available information or rely on data with poor quality, resulting in underestimation in the budget. For instance, medical staff who are funded temporarily by donor organizations and whose wage costs are accounted for under investment spending, are not adequately incorporated in the wage bill forecasting and planning process in anticipation of their transfer to the budget. (ii) There is an inadequate degree of monitoring and oversight in budget execution following the decentralized control over hiring at levels beneath the central government; (iii) There are capacity limits in the information management (IT) system, which relies on many manual processes and is unable to create useable and timely reports for use in expenditure control. The Administrative Tribunal lacks an adequate IT system that is linked to personnel management and payroll systems at all levels of government (central, provinces and districts).
D. Options for Reform
22. There is an urgent need to develop a comprehensive reform plan that can help ensure the sustainability of wage bill spending over the medium term. The government has already undertaken measures to contain wage bill spending. For instance, the government took a decision in December 2016 to suspend the 13th salary owed from fiscal year 2016 for all managerial level positions in the government, and granted all other employees in the government only half of it.17 On another front, the government took a decision in April 2017 to disproportionally increase the salaries of government employees. The salaries of low paid works (earning the minimum wage of around 3800 Meticais per month, as defined in the government sector) were increased by 21 percent. The nominal amount of that increase was then granted to all other employees in the government (which is small relative to their wage levels). Finally, in the context of the 2018 budget, the government proposed to reduce several allowances including the special bonus, the localization allowance, and the adaptation allowance.
23. While recent reform measures are encouraging, they are not by themselves sufficient and adequate to bend the trajectory of the wage bill going forward to a more sound and sustainable path. For instance, the measures taken in the context of the 2018 budget, unless reversed, will have a permanent impact on the wage bill going forward. However, the fiscal savings that they generate is at best modest (at around 0.5 percent of GDP in 2018). Going forward, it is essential to start devising a wage policy plan that encompasses structural measures that will help in putting the wage bill on a sustainable path, and unlock fiscal savings, some of which can be used to increase the much needed social and infrastructure spending.
24. Public wage spending is most efficient—and mostly likely to be sustainable—when it reflects sound compensation and hiring policies. To a limited degree and for the short-term, the growth of wage spending can be contained using crude measures that address wage levels or employment. Such measures, however, typically have limited fiscal reach. Nominal wage levels can always be fixed—or prevented from rising faster than the rate of inflation—but typically cannot be reduced. Measures to reduce public sector employment are, in practice, mostly limited to attrition-based policies—given the job protections enjoyed by most public employees—and, consequently, can provide only limited immediate fiscal relief. Over the medium-term, therefore, public wage spending is best managed by developing and implementing compensation and hiring policies that reduce the need for short-term containment measures.
25. International experience shows that crude short-term wage bill reform measures introduced in support of fiscal consolidation are often unsustainable. Short-term wage policies that greatly degrade the real (i.e., inflation-adjusted) value of employee wages often generate upward pressure for wage increases once consolidation targets have been met. Attrition-based hiring policies are difficult to target, which risks the loss of key personnel and the degradation of essential public services. Prolonged use of such short-term may lead to pressures that can eventually lead to a reversal of reforms. It can also undermine the attractiveness of government employment and inducing disruptive departures of skilled government employees.
26. The driving goal of any wage bill reform should be to introduce structural measures that can induce permanent savings in the wage bill. Structural reforms of the compensation structure can facilitate stronger wage bill control while ensuring wages are competitive, equitable and transparent. However, structural measures may require time and careful planning and implementation. In the meantime, while these measures are being designed and implemented, rationalizing the wage bill through short-run consolidation measures—consistent with the authority’s immediate fiscal objectives—including restraints on wages and employment in non-priority sectors, may provide fiscal saving.
E. Reforms Related to Compensation
27. In the short-term, two policy levers are available for constraining the growth of compensation. First, the existing wage rule can be replaced with a flexible wage policy that enables basic salaries to be increased more slowly. Wages can be fixed at their current nominal values or allowed to rise at a rate less than the rate of inflation. Increasing basic salaries in such a manner would generate meaningful short-term fiscal relief but could generate upward wage pressure if the real value of wages is too greatly degraded. Second, some elements of compensation can be reduced or eliminated. For instance, there are a large number of benefits and supplements (around 40 percent of compensation) that, collectively, reduce the transparency and equity of public sector compensation, engender wide disparities across sectors, and aggravate the task of fiscal management. In addition, the provision of overtime pay can gradually be curtailed through improved workforce planning.
28. Over the medium-term, these short-term fiscal consolidation measures should be replaced by structural measures. The most important structural measures relating to compensation include steps to revise the wage grids for public sector employees and develop new laws and regulations that govern the other elements of compensation.
29. In the first phase, the government can enact structural pay reforms to align job-specific requirements with compensation and to restore parity—in a systematic way—between public and private sector remuneration for workers of comparable skill and experience. The objective of structural pay reform is to identify the proper level of remuneration—to include basic salaries and other elements of compensation—for each type of position within the public sector for workers with different qualifications and experience. In this regard, a structural pay reform should be the first step in a two-step process of compensation reform.
30. In the second phase, a systematic review of all elements of compensation can be undertaken. The objective is to simplify remuneration, by eliminating nonessential allowances and other supplements, and streamlining the remaining ones by incorporating them into the base pay. Increasing the share of basic salaries in total compensation improve transparency and fairness in pay. This will allocate the levels of remuneration developed during the first phase (i.e., the structural pay reform phase) across the various components of compensation. However, consolidating allowances into the base pay may be fiscally costly if it leads to higher future pension costs.
F. Reforms Related to Employment
31. In the short-term, the size of the public-sector workforce cannot be easily adjusted in direct proportion to a shrinking fiscal envelope. To some degree, attrition-based measures for cutting employment can be used in pursuit of fiscal consolidation, but their fiscal impact is constrained by the number of persons leaving public employment. Moreover, attrition-based measures are not without risk as they can result in the loss of critical skills—ultimately leading to their subsequent reversal—or in the deterioration in the quality of public services. For this reason, when designing such measures, it is crucial to use—to the extent possible—a targeted approach (e.g., by focusing on non-essential personnel and exempting critical sectors, such as education and healthcare). Other policy measures to reduce the size of the workforce—such as retrenchment and voluntary severance—entail similar risks and, because they take time to design and implement, are unlikely to generate meaningful fiscal relief in the short term.
32. Over the medium-term, attrition-based policies should be replaced by structural measures to “right-size” public employment. The most important structural measures for employment include the following:
33. Functional reviews of ministries, departments, and agencies to identify areas of overlap or duplication and to clarify and codify organizational goals and responsibilities. The objective of functional reviews is to look strategically at how the public sector is organized at the institutional level with the goal of eliminating duplicative governmental functions and cutting (or curtailing) nonessential public services.
34. Institution-level restructuring and process reengineering. The objective is to look surgically at individual ministries, departments, and agencies with the aim of (i) aligning their personnel resources with their responsibilities and spending priorities, and (ii) improving public sector efficiency by leveraging technology and streamlining processes and administrative systems. Ultimately, process reengineering should then be used as the basis for streamlining institutional staffing plans (i.e., the number of authorized positions for every category of employment) with the goal of “right-sizing” overall employment.
G. Reforms Related to Public Financial Management
35. The challenge of managing wage spending also extends beyond compensation and employment policy to issues pertaining to financial management. Managing wage spending requires that public financial management processes and procedures be sufficiently developed to enable effective wage bill budgeting, execution, and control. Of importance are the following:
PFM Reform Strategy: develop a medium-term strategy for PFM reform, based on a strategic plan and an appropriate action plan, taking into consideration the advice already provided by development partners.
Forecasting and Budgeting: expand the quantity and quality of data made available for fiscal analysis, budgeting, and the monitoring of public sector employment and compensation; prepare budgets based on staffing priorities and limit hiring to fit within budgetary resources; this is particularly important with respect to workers funded by donor organizations whose wage costs will eventually be transferred to the budget.
Budget Execution: strengthen procedures governing payroll and address the problems of decentralized control over hiring at levels beneath the central government; develop an operational plan for strengthening oversight over budget execution.
Payroll Systems: automate existing manual systems to improve accuracy and strengthen controls; update and verify employee-related service records on a more routine and timely basis; strengthen the regulations, guidelines, and responsibilities for the periodic reconciliation of payroll expenditures.
Appendix I. Estimating the Public-Private Wage Premium
This appendix presents the details of the analysis on the relative competitiveness of compensation levels for government workers in Mozambique, as presented in the report. Specifically, it discusses the empirical approach used in estimating the wage premium of government workers relative to private sector workers using micro-level data from the 2015 Integrated Household and Labor Force Survey.
Data from the labor force survey suggests that in 2015, the average wage was higher in the public sector by about 30 percent. Nevertheless, in Mozambique, the share of better educated workers is much higher in the public sector than in the private sector (Table AI.1). Given that compensation is highly associated with levels of education (and other socio-economic factors), it is important to consider them when assessing wage differentials across individuals working in the public and the private sectors. We turn next to a more careful empirical analysis that allows us to control for such factors when estimating the wage premium.
Average Wage and Worker Characteristics, 2015
Average Wage and Worker Characteristics, 2015
| Educational attainment (share of total) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Number of workers | Average monthly wage (in Meticais) | Primary | Secondary | Tertiary | Professional | Average age | Female (percent) | ||
| Public sector | 3059 | 11189.3 | 0.075 | 0.464 | 0.352 | 0.109 | 38 | 0.375 | |
| Private sector | 6685 | 8603.1 | 0.400 | 0.478 | 0.091 | 0.031 | 35 | 0.194 | |
Average Wage and Worker Characteristics, 2015
| Educational attainment (share of total) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Number of workers | Average monthly wage (in Meticais) | Primary | Secondary | Tertiary | Professional | Average age | Female (percent) | ||
| Public sector | 3059 | 11189.3 | 0.075 | 0.464 | 0.352 | 0.109 | 38 | 0.375 | |
| Private sector | 6685 | 8603.1 | 0.400 | 0.478 | 0.091 | 0.031 | 35 | 0.194 | |
We use the following Mincer-type wage equation to estimate the public-private wage premium
where y is the logarithm of the monthly compensation, and x is a vector comprising variables that are found to be relevant for explaining wage differentials across individuals, including a set of binary variables indicating educational attainment, age and age squared (to proxy for experience), gender, and area of residence (to control for price effects). The binary variable public takes the value of 1 if an individual works in the public sector. We control for potential heterogeneity in the effects, essentially allowing the wage differential estimate to differ across skill (or educational) levels. To do so, we include interaction terms between the public-sector binary variable and the educational binary variables, where xj is a subset of x that comprises four binary variables indicating if an individual worker has tertiary, secondary, primary, or professional education.1 The difference between the coefficients in λ and the coefficient on the corresponding category of education represents the percent difference of public sector workers’ average wage compared to the average wage of private sector counterparts with similar socio-economic characteristics.
The wage equation is estimated using ordinary least squares regression techniques, using data from the 2015 Mozambique Integrated Household and Labor Force Survey. The focus of the analysis is limited to salaried private sector workers who report monthly earnings, and thus excludes self-employed workers. Furthermore, in an attempt to exclude workers in the informal private sector, we drop from the sample (i) workers earning below the minimum wage, (ii) domestic workers, farmers, and agriculture workers, and (iii) workers with no education. For hypothesis testing, we use standard errors that are robust to heteroscedasticity. The estimation results are shown in Table A2.2, and can be summarized as follows:2
There is a wage premium, in absolute terms, for lower-skilled employees with primary education of around 91 percent (Table AI.2).
There is a wage premium, in absolute terms, for lower-skilled employees with secondary education of around 63 percent (Table AI.2).
On the other hand, high-skilled government employees with university level education tend to have, on average, a wage discount of around 100 percent (Table AI.2).
Finally, government employees who hold a professional education (e.g., technical education and teacher training) earn a wage discount of around 11 percent relative to their private sector counterparts. However, the estimate is not significantly different from zero at the 5 percent level (Table AI.2).
Estimate of the Public-Private Wage Gap
(By skill or educational level)
Estimate of the Public-Private Wage Gap
(By skill or educational level)
| Coefficient | t-statistic | |
|---|---|---|
| Tertiary x Public | −0.337 | −7.33 |
| Professional x Public | −0.112 | −1.76 |
| Secondary x Public | 0.147 | 7.14 |
| Primary x Public | 0.128 | 3.08 |
| Tertiary education | 0.672 | 9.73 |
| Secondary education | −0.485 | −8.46 |
| Primary education | −0.783 | −13.42 |
| Age | 0.062 | 15.85 |
| Age squared | −0.001 | −12.64 |
| Male | 0.180 | 10.89 |
| Urban | 0.042 | 1.79 |
| Constant | 7.574 | 78.76 |
| N | 9,744 | |
| R-squared | 0.3064 |
Estimate of the Public-Private Wage Gap
(By skill or educational level)
| Coefficient | t-statistic | |
|---|---|---|
| Tertiary x Public | −0.337 | −7.33 |
| Professional x Public | −0.112 | −1.76 |
| Secondary x Public | 0.147 | 7.14 |
| Primary x Public | 0.128 | 3.08 |
| Tertiary education | 0.672 | 9.73 |
| Secondary education | −0.485 | −8.46 |
| Primary education | −0.783 | −13.42 |
| Age | 0.062 | 15.85 |
| Age squared | −0.001 | −12.64 |
| Male | 0.180 | 10.89 |
| Urban | 0.042 | 1.79 |
| Constant | 7.574 | 78.76 |
| N | 9,744 | |
| R-squared | 0.3064 |
By Chadi Abdallah.
The numbers of workers in other sectors (for which information were not available) is estimated to be around 45,000.
The wage bill that is recorded in the fiscal accounts and is publicly available includes spending on wage and salaries of the defense, security and police workers. While the authorities provided information on total wage spending for these three sectors, they did not provide data on their employment levels.
This category is referred to in the report as the “Other” sector.
And while these other sectors received relatively more modest wage increases, the rapid increase in employment that they experienced magnified the impact on the wage bill.
For consistency, the decomposition analysis is performed using wage bill spending that excludes the compensation of defense, police and security personnel, since, as pointed earlier, the employment figures for these sectors are not readily available.
The 13th month salary is an entitlement for civil servants in Mozambique, and is mandated by the government. It is equal to the base salary of the current fiscal year, but is paid in the following fiscal year (typically in January).
The authorities argue that substantial amount of hiring occurred in the health sector, which explains most of the slippage. Data provided by the authorities suggests that there were no additional hiring above planned budgeted amounts in the education sector.
The category of all other sectors combined explains the remaining 14 percent of the slippage.
This is consistent with the wage rule that is used by the government to adjust wages. Section III discusses the wage rule in more detail.
No labor unions currently exist for government employees in Mozambique. There are professional associations that engage in social dialogue with the government on behalf of civil servants, with respect to the wage adjustment mechanism.
For instance, under inflation indexation, wage bill spending increases at a time when fiscal restraint may be needed to address macroeconomic imbalances. In that sense, it can contribute to inflationary pressures, and further jeopardize fiscal sustainability going forward.
The data from the e-CAF and e-Folha systems pertain to the June 2017 vintage.
Law 7/98 establishes the rights and duties of governing positions (article 12(5)). Reintegration occurs when duties are terminated and the employee effectively returns to his/her previous position.
The awarding of positions is deemed provisional for the first two years, at the end of which employees are formally entered into the permanent civil service if the requirements specified by the General Statutes for Civil Servants have been met.
This seems to reflect ongoing technical difficulties with existing IT systems, which have, at least, partly compounded compliance issues regarding the enforcement of procedural requirements.
The suspension of the 13th salary may potentially imply unpaid obligations to be paid in the future, especially since it was not the result of a change in the law.
The category of workers with professional education includes those with technical education and teacher training degrees.
As a robustness check, the regression analysis has also been applied to male and female employees separately and the results remain consistent.
