## wp1910 ## Source details **Canonical URL:** [wp1910](https://www.imf.org/-/media/files/publications/wp/2019/wp1910.pdf) ## Other formats - [Markdown version](/-/media/files/publications/wp/2019/wp1910.pdf.md) - [Structured JSON version](/-/media/files/publications/wp/2019/wp1910.pdf.json) --- ### Research objective and context - Objective: provide empirical evidence on the financing of wage bill increases and on how non-wage spending and revenue aggregates behave in response to wage bill increases. - Key questions: - How do countries finance wage bill increases—increasing revenues, adjusting non-wage expenditures, or worsening the fiscal balance? - Do wage bill increases revert shortly after or persist? - Do dynamics differ across country groups? - Rationale: spending on the wage bill absorbs around one-fifth of total spending on average in advanced economies and nearly 30 percent in emerging markets and low-income and developing countries (IMF, 2016a). ### Data, sample, grouping and definitions - Dataset and time span: - Annual panel, 1992–2015. - Wage bill from IMF FAD’s Government Compensation and Employment Dataset (IMF, 2016a); revenues, other expenditures and GDP from IMF World Economic Outlook. - Final sample size: 137 countries (original 148 reduced to 137 after dropping outliers and countries with insufficient consecutive observations). - Country group counts: - Advanced economies: 27 countries. - Non-advanced economies: 110 countries. - High-debt economies: 71 countries. - Low-debt economies: 66 countries. - Resource-rich economies: 25 countries (of which only one is an advanced economy). - Grouping criteria: - Advanced vs non-advanced (non-advanced comprises emerging markets and low-income and developing economies). - Public debt classification: advanced economies with public debt above 72 percent of GDP are high-debt; emerging/low-income/developing countries with public debt above 43 percent of GDP are high-debt. - Resource-rich if revenues from natural resources represent at least twenty percent of total revenues. - Variable definitions: - Wage bill: total compensation (cash or in kind) payable to a government employee, including wages and salaries, allowances, and social security contributions made on behalf of employees to social insurance schemes. - Revenue: taxes, revenues from transfers, income from ownership of assets, and sales of goods and services. - Other expenditure: all expenditures other than the wage bill (e.g., use of goods and services, consumption of fixed capital, interest payments, subsidies, grants, social benefits and other expenses). - Primary other expenditure: non-wage expenditures minus interest payments (used for robustness checks). - Additional data: election-year dummy from Database of Political Institutions. ### Methodology overview - Empirical approach: - Panel vector autoregressive (PVAR) methodology estimated in first differences. - PVAR of order 3 estimated using a large annual panel dataset of 137 countries, 1992–2015. - Lag length selected: three years (Akaike and Bayesian information criteria). - Estimation: GMM-style estimator following Abrigo and Love (2015). - Identification: orthogonalized impulse responses with standard Cholesky decomposition; wage bill ordered as most exogenous, revenue as least exogenous. - Rationale for first differences: - Panel unit root tests did not reject a stochastic trend in levels and panel cointegration tests did not indicate cointegration among revenue, wage bill and other expenditures. - Presentation: - Results shown as responses of individual variables and cumulative impulse response functions; heterogeneity analysis across country groups. ### Stylized facts and descriptive statistics (selected) - Sample-level trends (levels and first differences): - For the whole sample, the average annual change of the wage bill between 1995–15 was 0.05 percentage point of GDP. - For non-advanced economies the average annual change of the wage bill was 0.07 percentage point of GDP over that period. - For the average non-advanced economy, revenues increased by about 0.41 percentage point of GDP and other expenditures increased by about 0.36 percentage point of GDP over 1995–15. - In advanced economies, the wage bill share of GDP has remained broadly stable; increases in revenues and other expenditures have been negligible. - Cross-country dispersion (examples preserved from source): - Revenues exceeded 70 percent of GDP in Kuwait, Libya and Marshall Islands at some points; revenues fell below 5 percent of GDP in Democratic Republic of Congo, Zimbabwe and Rwanda at some points. - Wage bill as share of GDP exceeded 20 percent in Djibouti, Lesotho, Libya and Marshall Islands in some observations; wage bill below 1 percent of GDP in Democratic Republic of Congo, Equatorial Guinea and Zimbabwe in some observations. - Selected descriptive table rows (values preserved exactly as in source for wage bill/revenue/other expenditure blocks; example rows shown): - "All" (Min Med Ave Max): 4.9 27.9 30.3 72.1 0.8 8.5 8.4 23.0 3.9 22.6 23.5 62.5 - "Resource Rich" (first block): 4.9 30.4 32.0 72.1 0.8 5.8 7.1 17.4 3.9 19.8 21.4 62.5 - "Non-Resource Rich" (first block): 8.3 27.7 30.0 65.5 1.6 8.9 8.7 23.0 5.0 23.0 23.9 46.7 - "Low Debt": 8.4 34.9 33.5 65.5 1.6 9.2 9.0 23.0 6.9 26.4 25.4 51.9 - "High Debt": 4.9 24.9 27.3 72.1 0.8 7.9 7.9 21.4 3.9 20.4 21.7 62.5 - "Advanced" (first block): 17.4 41.8 42.4 57.4 3.5 11.0 10.9 17.6 10.4 34.1 32.9 46.7 - Changes over 1995–2013: - The wage bill as a share of GDP increased in 80 out of 137 countries over 1995–2013. - Negative correlation between wage bill changes and overall fiscal balance changes observed in about 65 percent of countries. - Wage bill increases associated with deteriorating overall fiscal balance in 52 countries; wage bill decreases associated with improving overall fiscal balance in 36 countries. - Group patterns: - Wage bill-to-GDP ratio in resource-rich economies is lower than in non-resource-rich economies by about 2 percentage points of GDP. - Average annual change in the wage bill-to-GDP ratio is smaller in resource-rich than in non-resource-rich economies. - Convergence in wage bill-to-GDP ratios between low-debt and high-debt economies by 2015 due to higher average wage bill annual changes in high-debt economies. ### Panel unit root and cointegration test outcomes (Section IV) - Variables expressed as GDP shares: R_it (government revenue), W_it (wage bill), O_it (other non-wage expenditure); fiscal balance = R_it − W_it − O_it. - Panel unit root tests on levels (p-values reported): - Revenue: Dickey Fuller 0.63; Phillips Perron 0.00; Levin-Lin-Chu 0.82; Im-Pesaran-Shin 0.39; Hadri 0.00. - Wage Bill: Dickey Fuller 0.26; Phillips Perron 0.23; Levin-Lin-Chu 0.16; Im-Pesaran-Shin 0.34; Hadri 0.00. - Other Expenditure: Dickey Fuller 0.62; Phillips Perron 0.02; Levin-Lin-Chu 0.49; Im-Pesaran-Shin 0.41; Hadri 0.00. - Null hypotheses: H0 (Dickey/Levin/Im tests) = All panels contain unit root; H0 (Hadri) = All panels are stationary. - First-differences stationarity tests (p-values reported): - Revenue: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.00; Im-Pesaran-Shin 0.00; Hadri 1.00. - Wage Bill: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.09; Im-Pesaran-Shin 0.00; Hadri 0.25. - Other Expenditure: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.01; Im-Pesaran-Shin 0.00; Hadri 1.00. - Interpretation: first differences are stationary for the three series by the majority of tests. - Cointegration: - Four Westerlund (2007) error-correction based panel cointegration tests applied. - None reject the null hypothesis of absence of cointegration among revenue, wage bill and other expenditure. - Implication: no cointegration relationship found; VAR estimated in first differences. ### Main empirical findings from PVAR and Granger causality - Exogeneity and causality: - Changes in revenue and other expenditure have very little predictive content for the wage bill (Granger causality tests). - The wage bill does not follow changes in revenue or other non-wage expenditures in the short or medium term. - Persistence and financing (sample average): - Wage bill increases do not revert in the medium term on average. - Wage bill increases are roughly equally financed by both larger deficits and additional revenues on average. - Non-wage spending remains broadly unaffected on average. - Granger causality test p-values (Table 6): - Equation: Revenue — Excluded Variable: Wage Bill 0.02; Other Expenditure 0.31; Wage Bill and Other Expenditure 0.01. - Equation: Wage Bill — Excluded Variable: Revenue 0.09; Other Expenditure 0.86; Revenue and Other Expenditure 0.18. - Equation: Other Expenditure — Excluded Variable: Revenue 0.00; Wage Bill 0.99; Revenue and Wage Bill 0.00. - Null hypothesis H0: Excluded variable does not Granger-cause equation variable. - Univariate panel fixed-effect estimates (IMF 2016a): - A revenue surge of one percentage point of GDP is on average associated with: - other non-wage expenditures being higher by about 0.3 percentage point of GDP, and - the overall fiscal balance improving on average by about 0.7 percentage point of GDP. - Other expenditures accommodate quickly to positive surprises in revenues; the wage bill remains unchanged in response to revenue surges. - Non-wage expenditure increases are about 60 percent deficit financed in the medium term. ### Financing patterns and heterogeneity by country characteristics - By level of development: - Advanced economies: - Initially, higher wage bill associated with fiscal balance deterioration. - Two years after the increase, impact on the balance disappears as additional wage spending is fully compensated by higher revenue and lower other expenditures. - Revenue adjustment in advanced economies exceeds that in non-advanced countries. - Non-advanced economies: - Fiscal balance deterioration persists in the medium term because wage bill surge is only partly financed with revenues while other non-wage expenditures remain broadly unchanged. - By public-debt level: - High-debt economies: - Increase in the wage bill is not associated with revenue adjustment. - About half of the wage bill increase is financed by gradually crowding out other expenditures and the remaining half by a deteriorating fiscal balance. - Low-debt economies: - Revenues tend to respond more to increases in the wage bill; other expenditures also seem to increase to some extent. - The higher wage bill is strongly associated with a larger deficit at the time of the increase and it shrinks to about 75 percent of its initial size after about three years. - By resource wealth: - Resource-rich economies: - Rise in the wage bill does not seem to be accompanied by adjustments in either revenues or other expenditures, resulting in a deterioration of the balance of the same magnitude. - Non-resource-rich economies: - Revenues marginally increase over time, other expenditures increase slightly, and wage bill increases tend to be nearly fully financed with larger deficits. ### Robustness checks and empirical scope - Sample: 137 countries, 1992–2015. - Robustness checks performed: - Different orderings of variables in the PVAR (Appendix 2). - Alternative measure for non-wage expenditures excluding interest payments (primary non-wage expenditure) — did not significantly affect impulse responses (Appendix 3). - Inclusion of real GDP growth and alternatively the output gap — neither had a significant impact on the wage bill and vice versa. - Results remain broadly unchanged when reducing the sample to post 2009 data. - Caveat: small precision in estimates for resource-rich group likely due to only 27 resource-rich countries out of 137 in the sample; 26 out of 27 resource-rich countries are non-advanced. ### Policy implications and recommendations - Wage increases are difficult to scale back; countries should have a strong institutional framework to adequately manage the wage bill. - Given the potential large fiscal impact of wage bill changes, and expected pressures to increase wage bills across countries, efforts should focus on strengthening institutions that provide for adequate management of the wage bill (as described in IMF 2016a). - Policymakers should be aware that: - Increases in the wage bill could lead to a worsening fiscal balance, higher levels of taxation, and in some cases cuts in other expenditures. - The composition of financing (revenues adjustment versus crowding out versus deficit financing) varies with country characteristics (level of development, debt level, resource wealth), so tailored institutional arrangements and medium-term fiscal frameworks are important. *Source: IMF staff calculations; analysis based on "I. INTRODUCTION", "II. DATA AND STYLIZED FACTS", and "IV. Results from Panel Unit Root and Panel Cointegration Tests" of the provided content unit.* ### 1. Size of the Sample and Sub-Samples by Income Level, Access to Natural Resources ### 1. Size of the Sample and Sub-Samples by Income Level, Access to Natural Resources and Public Debt ### Research objective and context - Objective: provide empirical evidence on the financing of wage bill increases and on how non-wage spending and revenue aggregates behave in response to wage bill increases. - Key questions: - How do countries finance wage bill increases—increasing revenues, adjusting non-wage expenditures, or worsening the fiscal balance? - Do wage bill increases revert shortly after or persist? - Do dynamics differ across country groups? - Rationale: spending on the wage bill absorbs around one-fifth of total spending on average in advanced economies and nearly 30 percent in emerging markets and low-income and developing countries (IMF, 2016a). Small increases in compensation or employment can have large implications for the fiscal balance. ### Data, sample and grouping - Dataset: annual panel of government wage bill, revenues, and non-wage expenditures for the general government; wage bill from IMF FAD’s Government Compensation and Employment Dataset (IMF, 2016a); revenues, other expenditures and GDP from IMF World Economic Outlook. - Time span: 1992–2015. - Final sample size: 137 countries (original sample of 148 reduced to 137 after dropping outliers and countries with insufficient consecutive observations). - Country group counts (sample composition): - Advanced economies: 27 countries. - Non-advanced economies: 110 countries. - High-debt economies: 71 countries. - Low-debt economies: 66 countries. - Resource-rich economies: 25 countries (of which only one is an advanced economy). - Grouping criteria: - Advanced vs non-advanced (non-advanced comprises both emerging markets and low-income and developing economies). - Public debt classification: advanced economies with public debt above 72 percent of GDP are high-debt; emerging/low-income/developing countries with public debt above 43 percent of GDP are high-debt (IMF (2014a) criteria applied). - Resource rich if revenues from natural resources represent at least twenty percent of total revenues. - Additional data: election-year dummy from Database of Political Institutions to capture election effects. ### Definitions and variable construction - Wage bill: total compensation (cash or in kind) payable to a government employee, including wages and salaries, allowances, and social security contributions made on behalf of employees to social insurance schemes (IMF, 2014b). - Revenue: taxes, revenues from transfers, income from ownership of assets, and sales of goods and services. - Other expenditure: all expenditures other than the wage bill (e.g., use of goods and services, consumption of fixed capital, interest payments, subsidies, grants, social benefits and other expenses). - Primary other expenditure: non-wage expenditures minus interest payments (used for robustness checks). ### Methodology overview - Empirical approach: panel vector autoregressive (PVAR) methodology estimated in first differences (PVAR of order 3 estimated using a large annual panel dataset of 137 countries, 1992–2015). - Variables transformed to first differences because panel unit root tests did not reject a stochastic trend in levels and panel cointegration tests did not indicate cointegration among revenue, wage bill and other expenditures. - Results presented as responses of individual variables to increases in the other endogenous variables; impact on overall budget balance is derived and analyzed. - Heterogeneity analysis: results compared across country groups defined by level of development, public indebtedness and access to natural resources. ### Stylized facts and descriptive statistics - Sample-level trends (levels and first differences): - For the whole sample, the average annual change of the wage bill between 1995–15 was 0.05 percentage point of GDP. - For non-advanced economies the average annual change of the wage bill was 0.07 percentage point of GDP over that period. - For the average non-advanced economy, revenues increased by about 0.41 percentage point of GDP and other expenditures increased by about 0.36 percentage point of GDP over 1995–15. - In advanced economies, the wage bill share of GDP has remained broadly stable; increases in revenues and other expenditures have been negligible. - Cross-country dispersion: - Government revenues, wage bill and other non-wage expenditures (all in percent of GDP) are higher in advanced than in non-advanced economies by average and median values. - Non-advanced economies show large heterogeneity: examples cited in the source text include revenues exceeding 70 percent of GDP in Kuwait, Libya and Marshall Islands at some points, while revenues fall below 5 percent of GDP in Democratic Republic of Congo, Zimbabwe and Rwanda at some points. - Wage bill as share of GDP exceeds 20 percent in Djibouti, Lesotho, Libya and Marshall Islands in some observations; wage bill below 1 percent of GDP in Democratic Republic of Congo, Equatorial Guinea and Zimbabwe in some observations. ### Key empirical findings - Persistence and financing of wage bill increases (sample average): - Wage bill increases do not revert in the medium term on average. - On average, wage bill increases are roughly equally financed by both larger deficits and additional revenues. - Non-wage spending remains broadly unaffected on average. - Heterogeneity across country groups: - Advanced economies: wage bill increases are not associated with a worsening fiscal balance; these countries fully finance wage spending surges by both increasing revenues and cutting non-wage expenditures. - Non-advanced economies: on average, wage bill increases are associated with higher deficits and additional revenues rather than cuts in non-wage spending. - Low-debt and non-resource-rich economies: on average, additional wage bill spending is associated with higher revenue more than in high-debt and resource-rich economies. - Robustness: - Results are robust to alternative variable orderings in impulse response estimation. - Results are robust to using primary other expenditure (non-wage expenditures excluding interest payments) as an alternative measure. ### Implications highlighted in the source - Small increases in compensation or employment could have large unintended adverse implications for the fiscal balance given the wage bill’s sizable share of total spending. - Financing modalities have different macro-fiscal implications: - Financing wage increases via persistent deterioration in the fiscal balance will worsen public debt. - Financing via cuts in non-wage spending could crowd out crucial spending for growth and poverty reduction (public infrastructure, social protection). - Financing via higher revenues could erode private sector competitiveness and harm growth. - Financing via revenue surges during economic upswings risks amplifying output fluctuations and undermining fiscal stabilization roles. *Source: IMF staff calculations, from "I. INTRODUCTION" and "II. DATA AND STYLIZED FACTS" sections of the provided content unit.* ### Section IV discusses results from the panel unit root and panel cointegration tests. ### Section IV — Results from Panel Unit Root and Panel Cointegration Tests ### Sample and empirical setup - Sample size: 137 countries. - Variables (expressed as GDP shares): R_it (government revenue), W_it (wage bill), O_it (other non-wage expenditure). The overall fiscal balance is R_it − W_it − O_it. - PVAR specification: - Reduced-form three-dimensional panel VAR of order p. - Lag length selected: three years (supported by Akaike and Bayesian information criteria). - Estimation: GMM-style estimator following Abrigo and Love (2015). - Identification: orthogonalized impulse responses with standard Cholesky decomposition; wage bill ordered as most exogenous, revenue as least exogenous (wage bill responds to contemporaneous changes with a lag). - Robustness: alternative ordering and alternative measure “primary non-wage expenditures” (excludes interest payments) used as checks. ### Stationarity and cointegration test outcomes - Panel unit root tests on levels (p-values reported in Table 4): - Revenue: Dickey Fuller 0.63; Phillips Perron 0.00; Levin-Lin-Chu 0.82; Im-Pesaran-Shin 0.39; Hadri 0.00. - Wage Bill: Dickey Fuller 0.26; Phillips Perron 0.23; Levin-Lin-Chu 0.16; Im-Pesaran-Shin 0.34; Hadri 0.00. - Other Expenditure: Dickey Fuller 0.62; Phillips Perron 0.02; Levin-Lin-Chu 0.49; Im-Pesaran-Shin 0.41; Hadri 0.00. - Null hypotheses stated: H0 (Dickey/Levin/Im tests) = All panels contain unit root; H0 (Hadri) = All panels are stationary. - First-differences stationarity tests (p-values reported in Table 5): - Revenue: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.00; Im-Pesaran-Shin 0.00; Hadri 1.00. - Wage Bill: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.09; Im-Pesaran-Shin 0.00; Hadri 0.25. - Other Expenditure: Dickey Fuller 0.00; Phillips Perron 0.00; Levin-Lin-Chu 0.01; Im-Pesaran-Shin 0.00; Hadri 1.00. - Interpretation: first differences are stationary for the three series by the majority of tests. - Cointegration: - Four Westerlund (2007) error-correction based panel cointegration tests applied. - None of these tests reject the null hypothesis of absence of cointegration among revenue, wage bill and other expenditure. - Implication: no cointegration relationship found; VAR estimated in first differences. ### Descriptive and subgroup evidence (selected statistics and patterns) - Wage bill developments and group comparisons (selected rows from descriptive statistics table): - "All" (Min Med Ave Max): 4.9 27.9 30.3 72.1 0.8 8.5 8.4 23.0 3.9 22.6 23.5 62.5 - "Resource Rich" (first block): 4.9 30.4 32.0 72.1 0.8 5.8 7.1 17.4 3.9 19.8 21.4 62.5 - "Non-Resource Rich" (first block): 8.3 27.7 30.0 65.5 1.6 8.9 8.7 23.0 5.0 23.0 23.9 46.7 - "Low Debt": 8.4 34.9 33.5 65.5 1.6 9.2 9.0 23.0 6.9 26.4 25.4 51.9 - "High Debt": 4.9 24.9 27.3 72.1 0.8 7.9 7.9 21.4 3.9 20.4 21.7 62.5 - "Advanced" (first block): 17.4 41.8 42.4 57.4 3.5 11.0 10.9 17.6 10.4 34.1 32.9 46.7 - (Note: table contains multiple blocks and repeated groupings for Revenue, Wage Bill, Other Expenditure; values preserved exactly as in source.) - Changes over 1995–2013: - The wage bill as a share of GDP increased in 80 out of 137 countries over 1995–2013. - Negative correlation between wage bill changes and overall fiscal balance changes observed in about 65 percent of countries. - Wage bill increases associated with deteriorating overall fiscal balance in 52 countries; wage bill decreases associated with improving overall fiscal balance in 36 countries. - Group dynamics: - Splitting sample into high- and low-debt economies is relevant: despite initially higher wage bill-to-GDP ratio in low-debt economies, ratios converged by 2015 (Figure 1b). - Convergence resulted from higher average wage bill annual changes in high-debt economies compared to average low-debt economies (Table 3). - Controlling for level of development: low-indebted economies exhibit larger increases in revenues than high-debt economies. - Natural resources: - Wage bill-to-GDP ratio in resource-rich economies is lower than in non-resource-rich economies by about 2 percentage points of GDP. - Average annual change in the wage bill-to-GDP ratio is smaller in resource-rich than in non-resource-rich economies. - Non-resource economies display positive increases in revenues and other expenditures of similar magnitude. - For resource-rich economies, the observed increase in revenues exceeds the average annual change of non-wage expenditures. ### Main empirical findings from the PVAR results - Exogeneity and causality: - Changes in revenue and other expenditure have very little predictive content for the wage bill (supported by Granger causality tests). - The wage bill does not follow changes in revenue or other non-wage expenditures in the short or medium term. - Finding contrasts with literature that argued for cyclicality of the wage bill. - Financing of wage bill increases: - Adjustments to the wage bill are usually associated with a deterioration in the fiscal balance that persists into the medium term. - Wage bill increases are only partially financed by increases in revenues; other expenditure items remain broadly unchanged (Figure 3). - The report presents cumulative impulse response functions of revenues, wage bill and other non-wage expenditures to analyze how increases to the wage bill have been financed. - Presentation of results: - Results shown for full sample and for six subsamples: advanced economies; non-advanced economies; high-debt; low-debt; resource-rich; non-resource-rich. - PVAR estimated in first differences; cumulative impulse response functions reported to show cumulative effects on levels of GDP ratios over time. *Source: Section IV, "wp1910 - Section IV discusses results from the panel unit root and panel cointegration tests."* ### 0.5 percentage points and deficit increases of almost equal magnitude. In other words, increases ### wp1910 - 0.5 percentage points and deficit increases of almost equal magnitude. In other words, increases ### Key findings on wage bill, revenues, other expenditures, and fiscal balance - An increase in the wage bill is associated with a deterioration of the overall balance in the medium term because these spending increases are only partially compensated with additional revenues. - Granger causality tests indicate past values of the wage bill have predictive power for revenue but no predictive power for other expenditures (see Table 6 p-values). - Applying univariate panel fixed effect models (IMF 2016a) suggests that increases in the wage bill have on average been associated with increases in other government spending and with a deterioration of the overall balance as these spending increases are only partially compensated with additional revenues. - A revenue surge of one percentage point of GDP is on average associated with: - other non-wage expenditures being higher by about 0.3 percentage point of GDP, and - the overall fiscal balance improving on average by about 0.7 percentage point of GDP. - Other expenditures seem to accommodate quickly to positive surprises in revenues, whereas the wage bill remains unchanged in response to revenue surges. - Non-wage expenditure increases are associated with some worsening of the fiscal balance in the medium term: about 60 percent of the increase in other expenditure outlays is deficit financed, with the remainder financed with higher revenues. ### Financing patterns by country characteristics - By level of development: - Initially, a higher wage bill is strongly associated with fiscal balance deterioration in both advanced and non-advanced economies. - Two years after the increase: - Advanced economies: the impact on the balance seems to disappear as additional wage spending is fully compensated by higher revenue and lower other expenditures. - Non-advanced economies: fiscal balance deterioration persists in the medium term because the wage bill surge is only partly financed with revenues while other non-wage expenditures remain broadly unchanged. - Revenue adjustment in advanced economies exceeds that observed in non-advanced countries. - By public-debt level: - The association between higher wage bill spending and a worsening fiscal balance is comparable in high and low-debt economies. - High-debt economies: the increase in the wage bill is not associated with revenue adjustment; about half of the wage bill increase is financed by gradually crowding out other expenditures and the remaining half by a deteriorating fiscal balance. - Low-debt economies: revenues tend to respond more to increases in the wage bill, while other expenditures also seem to increase to some extent. The higher wage bill is strongly associated with a larger deficit at the time of the increase and it shrinks to about 75 percent of its initial size after about three years. - By resource wealth: - Resource-rich economies: rise in the wage bill does not seem to be accompanied by adjustments in either revenues or other expenditures, resulting in a deterioration of the balance of the same magnitude. - Non-resource-rich economies: revenues marginally increase over time, other expenditures increase slightly as well, and wage bill increases tend to be nearly fully financed with larger deficits. ### Empirical scope and robustness - Sample: 137 countries between 1992 and 2015. - Method: panel VAR estimation with cumulative impulse response functions and Granger causality tests on three variables: revenues, wage bill, other expenditures. - Robustness checks performed: - Different orderings of variables in the PVAR (see Appendix 2). - Alternative measure for non-wage expenditures excluding interest payments (primary non-wage expenditure) — did not significantly affect impulse responses (Appendix 3). - Inclusion of real GDP growth and alternatively the output gap to address potential endogeneity from omitting the economic cycle — neither variable had a significant impact on the wage bill and vice versa. - Results remain broadly unchanged when reducing the sample to post 2009 data. - Caveat: small precision in estimates for resource-rich group likely due to only 27 resource-rich countries out of 137 in the sample; 26 out of 27 resource-rich countries are non-advanced. ### Granger causality test (Table 6 p-values) - Equation: Revenue - Excluded Variable: Wage Bill — 0.02 - Excluded Variable: Other Expenditure — 0.31 - Excluded Variable: Wage Bill and Other Expenditure — 0.01 - Equation: Wage Bill - Excluded Variable: Revenue — 0.09 - Excluded Variable: Other Expenditure — 0.86 - Excluded Variable: Revenue and Other Expenditure — 0.18 - Equation: Other Expenditure - Excluded Variable: Revenue — 0.00 - Excluded Variable: Wage Bill — 0.99 - Excluded Variable: Revenue and Wage Bill — 0.00 - Null hypothesis H0: Excluded variable does not Granger-cause equation variable. ### Policy implications and recommendations - Wage increases are difficult to scale back; countries should have a strong institutional framework to adequately manage the wage bill. - Given the potential large fiscal impact of wage bill changes, and expected pressures to increase wage bills across countries, efforts should focus on strengthening institutions that provide for adequate management of the wage bill (as described in IMF 2016a). - Policymakers should be aware that: - Increases in the wage bill could lead to a worsening fiscal balance, higher levels of taxation, and in some cases cuts in other expenditures. - The composition of financing (revenues adjustment versus crowding out versus deficit financing) varies with country characteristics (level of development, debt level, resource wealth), so tailored institutional arrangements and medium-term fiscal frameworks are important. *Source: IMF staff calculations; analysis based on a sample of 137 countries between 1992 and 2015 as presented in the source document.* --- _Source: https://www.imf.org/-/media/files/publications/wp/2019/wp1910.pdf_