## wpiea2026092-source-pdf

## Source details

**Canonical URL:** [wpiea2026092-source-pdf](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026092-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026092-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026092-source-pdf.pdf.json)

---

### Introduction — scope and objectives
- Research questions:
  - Whether regional governments have increased their fiscal importance over time in Latin America.
  - How changes in revenue and expenditure composition contributed to decentralization.
  - Whether the growing fiscal role of regional governments affects fiscal cyclicality and the role played by transfer revenues.
- Dataset characteristics:
  - A new long-term, highly granular database on subnational public finances in Latin America.
  - Coverage: nearly four decades, from the early 1980s to 2024.
  - Includes regional-level revenue and expenditure data for major countries in the region, and municipal-level fiscal variables and debt for some countries.
  - Provides longer time coverage and finer geographic breakdown at the regional level than existing sources.

### Data construction, coverage, and harmonization
- Geographic and temporal coverage:
  - Database covers the largest economies in the region: Argentina (only regions), Brazil, Colombia, Mexico, Peru (only regions) and Uruguay (only regions).
  - The database starts as early as 1989, and most countries have information up to 2023.
  - Very small economies were excluded; Bolivia lacks consistent time series; Chile’s regional governments have very limited fiscal autonomy and municipalities are the primary subnational fiscal entities.
- Fiscal variable definitions (harmonized to IMF GFS principles):
  - Revenues consolidated into three categories: own-source revenue; transfers revenue; total revenue.
  - Expenditures grouped into: current; capital; debt-servicing.
  - Definitions preserved as in source: own-source revenues, transfer revenues, current expenditures, capital expenditures, debt-servicing expenditures.
- Extension of IMF Fiscal Decentralization dataset:
  - Incorporates historical revisions and newly covered data for the years of 2021-2023.
  - Expands the number of economies to 97; new LAC additions include Argentina, Dominican Republic, Ecuador, Panama and Uruguay.
  - Close alignment observed between granular regional sums and national GFS aggregates (Figure 1, 1989–2023).

### Key empirical findings — fiscal importance, composition, and transfers
- Fiscal importance and trends:
  - Regional government revenues as a share of GDP have risen in all countries in the sample, roughly doubling on average.
  - Revenues in countries such as Argentina or Brazil reaching a level close to 15 percentage points of GDP.
  - Mexico’s regional revenues are about 10 percentage points of GDP in 2023.
  - Regional revenues as a share of general government revenues show a similar, though less pronounced, increase.
- Role and composition of transfers:
  - In all countries considered in the sample, transfers account for more than 30 percent of regional government revenues.
  - Transfers account for more than 65 percent in Argentina and Mexico, about 50 percent in Colombia, roughly 30 percent in Brazil and Uruguay.
  - Mexico remains highly dependent on federal transfers, which account for up to 80–90 percent of state revenues.
  - Transfers are higher to poorer regions rather than systematically lower own-source revenues in those regions—transfer systems operate primarily as redistribution mechanisms across regions.
  - Own-source revenues, normalized by regional GDP per capita, are relatively flat across the income distribution.
  - Transfers (as percent of regional GDP) show a clear negative association with regional GDP per capita.

### Fiscal cyclicality — elasticities and asymmetries (headline results)
- National-GDP elasticities (short-run):
  - A 1 percent increase in nominal national GDP is associated with a 0.87 percent increase in regional government revenues in the same year.
  - A 1 percent increase in nominal national GDP is associated with a 0.71 percent increase in regional government expenditures.
  - Central government short-term revenue response is described as "near one-to-one" with national GDP.
- Long-run elasticities:
  - Long-run revenue elasticities for both regional and central governments are not statistically different from one.
- Decomposition of revenue elasticities:
  - Regional own-source revenues display elasticities close to or above unity with respect to both national and regional GDP.
  - Regional transfer revenues exhibit weaker elasticity to national GDP, typically below 0.7.
  - The elasticity of transfers to regional GDP is close to 0.4.
- Decomposition of expenditure elasticities:
  - Current expenditures at the regional level have short-run elasticities to national and regional GDP in the range of 0.4–0.5.
  - Capital expenditures exhibit substantially higher short-run responsiveness, with elasticities well above one.
  - The combination yields total expenditure elasticity in the range of 0.5-0.6 for both central and regional perimeters.
- Interpretation:
  - Low elasticity of transfers and limited revenue-raising capacity of regions suggest weak cyclical responsiveness among subnational governments in Latin America and relatively shallow fiscal integration.
  - Lower expenditure elasticity is primarily explained by the rigidity of current spending; the null hypothesis that capital expenditures respond proportionally to GDP is not rejected.

### Within-country heterogeneity, time stability, and robustness
- Heterogeneity:
  - Regions with above-median transfer dependency within each country exhibit lower short-term revenue elasticities than regions with below-median transfer dependency.
  - Richer regions tend to exhibit lower dependence on transfers as a share of total revenues.
- Time stability:
  - Elasticity patterns are mostly constant over time: similar revenue and expenditure elasticities are found across subperiods 1995-2010 and 2011-2021.
- Robustness:
  - Estimation uses panel Error Correction Models (ECM) combined with Common Correlated Effects (CCE) estimators to control for cross-sectional dependence.
  - Results are robust across alternative estimators: Mean Group (MG), Clustered Common Correlated Effects (CCEC), Dynamic CCE (DCCE), and CCET with country-specific linear trends.
  - MG and CCET yield results very similar to benchmark CCE; CCEC yields slightly smaller elasticities; dynamic models yield marginally higher elasticities.

### Regional debt and fiscal risks
- Regional debt:
  - Regional debt is low relative to regions' GDP in most cases, with notable outliers in some regions of Brazil and Argentina.
  - When debt is measured relative to own-source revenues, several regions show higher debt burdens, suggesting potential dependence on intergovernmental transfers to service obligations.
  - Debt patterns do not appear to differ systematically across income levels in the smaller subsample with debt data.

### Quantitative regression excerpt (selected coefficients and statistics)
- Coefficient: Normalized GDP per capita = -0.134 ** (standard error 0.058) for Transfer Dependency (dependent variable).
- Coefficient: Normalized GDP per capita = 0.001 (standard error 0.003) for Revenue Own-Source (%GDP).
- Coefficient: Normalized GDP per capita = -0.058 *** (standard error 0.019) for Revenue Transfer (%GDP).
- Country × Year fixed effects: Yes.
- Observations: 2794.000 for each regression reported.
- R-squared: 0.711 for Transfer Dependency; 0.664 for Revenue Own-Source (%GDP); 0.457 for Revenue Transfer (%GDP).
- Significance notation: * p < 0.10, ** p < 0.05, *** p < 0.01.

### Policy implications and recommendations
- Strengthen intergovernmental fiscal frameworks to balance redistribution and incentives for local revenue mobilization.
- Increase revenue autonomy to enhance fiscal flexibility at the subnational level; options include expanding regional tax bases or creating shared income tax arrangements.
- Redesign transfer systems to incorporate performance-linked elements to reduce rigidity and promote more own-source financing.
- Better integrate subnational governments into medium-term fiscal planning and implement appropriate regional-level fiscal rules.
- Address spending rigidities (pensions, wages, interest, institutional constraints) through structural reforms and efficiency-enhancing reprioritization.
- Protect high-quality public investment given the reliance on capital spending as the primary short-run adjustment margin and the risks of procyclicality for long-term growth.
- Integrate subnational governments into comprehensive fiscal reporting, credible borrowing and fiscal rules to identify and mitigate macro-fiscal risks from subnational finances.

_Italic source: IMF estimations using subnational-level budget information for subnational governments compiled from individual countries (Content unit: wpiea2026092-source-pdf - sections 3.1–5.2)._

### References .............................................................................................................

### References

### Figures and Tables (as listed)
- Figures include:
  - Figure 1: Comparison of Subnational Databases (1989–2023)
  - Figure 2: Revenues in Subnational Governments Across Regions
  - Figure 3: Expenditures in Subnational Governments Across Regions
  - Figure 4: Regional Government Revenues
  - Figure 5: Regional Government Expenses
  - Figure 6: Regional Government Revenues Breakdown
  - Figure 7: Regional Government Expenditures Breakdown
  - Figure 8: Dispersion of Regional Government Revenues Across Regions, as Percent of Regional GDP
  - Figure 9: Regional Government Transfers Across Years, as Percent of Total Revenues
  - Figure 10: Dispersion of Regional Government Transfers Across Regions, as Percent of Total Revenues
  - Figure 11: Regional Government Total Revenues (As Percent of Regional GDP) and GDP Per Capita
  - Figure 12: Regional Government Transfers (As Percent of Regional Revenues) and GDP Per Capita
  - Figure 13: Regional Government Own-Source Revenues (As Percent of Regional GDP) and GDP Per Capita
  - Figure 14: Regional Government Transfers (As Percent of Regional GDP) and GDP Per Capita
  - Figure 15: Short-Run and Long-Run Fiscal Elasticities of Regional and Central Governments to National GDP
  - Figure 16: Short-Run Revenue and Expenditure Elasticities of Regional Governments to Regional
  - Figure 17: Short-Run Revenue and Expenditure Elasticities of Regional Governments
  - Figure 18: Short-Run Revenue and Expenditure Elasticities of Regional Governments to National GDP, by Time Period and by Transfer Dependency of Regions within Each Country
  - Figure 19: Robustness Tests Across the Different Estimation Models
  - Figure 20: Regional Government Revenues as Percent of General Government Revenues
  - Figure 21: Regional Governments Debt
- Tables include:
  - Table 1: Summary of Coverage of Subnational Fiscal Data
  - Table 2: Fiscal and Administrative Decentralization Arrangements by Country
  - Table 3: Descriptive Statistics for the Subnational Database at the Regional Level
  - Table 4: Common Sample Period for Regional and National Fiscal Series and GDP Series
  - Table 5: Transfer Mechanisms by Country
  - Table 6: Regression Results of Transfer Dependency and Revenues in Regions across Countries

### Introduction — scope and objectives
- Research questions:
  - Whether regional governments have increased their fiscal importance over time in Latin America.
  - How changes in revenue and expenditure composition contributed to decentralization.
  - Whether the growing fiscal role of regional governments affects fiscal cyclicality and the role played by transfer revenues.
- Dataset characteristics:
  - A new long-term, highly granular database on subnational public finances in Latin America.
  - Coverage: nearly four decades, from the early 1980s to 2024.
  - Includes regional-level revenue and expenditure data for major countries in the region, and municipal-level fiscal variables and debt for some countries.
  - Provides longer time coverage and finer geographic breakdown at the regional level than existing sources.

### Key empirical findings (preserve numeric values exactly)
- Fiscal importance and trends:
  - Regional government revenues as a share of GDP have risen in all countries in the sample, roughly doubling on average.
  - Revenues in countries such as Argentina or Brazil reaching a level close to 15 percentage points of GDP.
  - Regional revenues as a share of general government revenues show a similar, though less pronounced, increase.
- Role of transfers:
  - In all countries considered in the sample, transfers account for more than 30 percent of regional government revenues.
  - Reliance on transfers varies widely across regions, with richer regions consistently exhibiting lower transfer dependence.
  - Disaggregation indicates transfers are higher to poorer regions rather than systematically lower own-source revenues in those regions—transfer systems operate primarily as redistribution mechanisms across regions.
- Fiscal cyclicality (elasticities and asymmetries):
  - A 1 percent increase in nominal national GDP is associated with a 0.87 percent increase in regional government revenues in the same year.
  - Own-source revenues respond almost one-for-one to national GDP, while transfer revenues are significantly less flexible.
  - Elasticity of regional revenues with respect to regional GDP is also below one, driven by weak responsiveness of transfers.
  - A 1 percent increase in nominal national GDP is associated with a 0.71 percent increase in regional government expenditures.
  - Lower expenditure elasticity is primarily explained by the rigidity of current spending; the null hypothesis that capital expenditures respond proportionally to GDP is not rejected.
- Interpretation:
  - Low elasticity of transfers and limited revenue-raising capacity of regions suggest weak cyclical responsiveness among subnational governments in Latin America and relatively shallow fiscal integration.

### Methodology and data sources
- Empirical approach:
  - Estimates of elasticity use error-correction models and common correlated effects estimators.
  - Analysis distinguishes short-term versus long-term responses and compares responsiveness to national and regional GDP.
- Databases used:
  - A highly granular dataset of subnational fiscal information (regional or municipal levels) compiled from administrative data obtained directly from fiscal authorities in Latin America.
  - An extended version of the IMF’s Fiscal Decentralization dataset with comprehensive and internationally comparable decentralization indicators at the country level.
  - Complementary country- and regional-level demographic and economic information, including regional GDP or population.
- Coverage and harmonization:
  - The compilation and harmonization choices are described in detail in Section 3 (Data Sources).

### Literature context and contributions
- Builds on first-generation (Tiebout, Musgrave, Olson, Oates) and second-generation theories of decentralization (public-choice perspective, incentive problems).
- Places findings within literatures on:
  - Determinants of government size (Wagner’s Law, insurance, redistribution, institutional factors).
  - Historical patterns of fiscal decentralization in Latin America (CAF (2025), Inter-American Development Bank (2022)).
  - Fiscal cyclicality and revenue elasticities (Lane, 2003; Frankel et al., 2013; Heimberger, 2023; Jalles et al., 2024; Cornevin et al., 2020; Lagravinese et al., 2020; Gupta et al., 2022).
  - Subnational cyclicality studies (Dougherty and Biase, 2020; Arena and Revilla, 2009; Meloni, 2018; Xing and Fuest, 2018; Groves and Kahn, 1952; Sørensen et al., 2001).
- Contribution:
  - Provides a long-term, multi-country assessment of fiscal cyclicality at the regional level in Latin America using newly constructed and harmonized subnational fiscal data and frontier panel econometric methods.
  - Investigates mechanisms underlying heterogeneity, focusing on revenue composition—particularly intergovernmental transfers—and expenditure categories.

### Policy implications and conclusions
- Efficiency and redistribution:
  - Expansion of regional governments may have generated efficiency gains by better tailoring public goods to local preferences.
  - Intergovernmental transfer systems may have contributed to reducing inequality, partially compensating for income differences across regions.
- Macroeconomic stabilization and fiscal risk:
  - Limited cyclical responsiveness of regions’ revenue structures—dominated by intergovernmental transfers—could have weakened their role as automatic stabilizers.
  - Adjustment burden may shift to expenditures, particularly public investment, during downturns.
- Policy recommendations highlighted:
  - Strengthen intergovernmental fiscal frameworks.
  - Better integrate subnational governments into medium-term fiscal planning.
  - Implement appropriate regional-level fiscal rules.
  - Ensure transfer and expenditure rules support both macroeconomic stabilization and sustainable public investment.

*International Monetary Fund — IMF Working Papers: Fiscal Decentralization and Cyclicality in Latin America*

### 3.1 Subnational-Level Budget Information for Subnational Governments

### 3.1 Subnational-Level Budget Information for Subnational Governments

### Data construction and scope
- A highly granular database was constructed with fiscal information at the subnational level using a variety of sources from fiscal authorities in Latin America.
- Systematic collection and harmonization of data from official national sources involved manual downloading of budget execution and fiscal reports of regions and municipalities available from finance ministries or national statistical offices.
- The dataset covers the largest economies in the region: Argentina (only regions), Brazil, Colombia, Mexico, Peru (only regions) and Uruguay (only regions).
- The database starts as early as 1989, and most countries have information up to 2023, offering a long-term view of subnational fiscal dynamics.
- Very small economies in the region were excluded due to near absence of subnational structures or publicly available data; Bolivia lacks consistent time series; Chile’s regional governments have very limited fiscal autonomy and municipalities are the primary subnational fiscal entities.

### Focus and future extension
- Primary empirical analysis focuses on regional governments (states, provinces, or regions) because, on average, regional governments account for a larger share of GDP than local governments (municipalities) and typically enjoy greater fiscal autonomy.
- The authors note that extending analysis to the municipal level—especially in countries where subnational activity is mainly concentrated at the local level—would provide a more comprehensive assessment of fiscal decentralization.

### Harmonization and fiscal variable definitions
- Fiscal variables were standardized following IMF Government Finance Statistics (GFS) principles.
- Revenue consolidated into three categories:
  - own-source revenue (including tax and non-tax collections),
  - transfers revenue for the subnational government,
  - total revenue recorded.
- Expenditures grouped into:
  - current,
  - capital,
  - debt-servicing spending.
- Definitions:
  - Own-source revenues: revenues directly controlled and collected by subnational governments (local taxes, fees, charges for public services).
  - Transfer revenues: resources received from other levels of government (central government grants or shared revenues).
  - Current expenditures: day-to-day operating spending including wage payments, purchases of goods and services, routine administrative costs.
  - Capital expenditures: investments in physical assets and infrastructure (schools, hospitals, transport facilities).
  - Debt-servicing expenditures: interest payments of public debt.

### Country coverage decisions
- Coverage guided by economic relevance and data availability; excluded small countries with limited subnational structures.
- Specific exclusions/limitations noted: Bolivia (inconsistent series), Chile (regional fiscal autonomy limited), and selective inclusion of regions vs. municipalities in some countries.

---

### 3.2 Country-Level Budget Information for Subnational Governments

### Extension of IMF Fiscal Decentralization dataset
- The IMF Fiscal Decentralization dataset primarily uses indicators from the Government Finance Statistics (GFS) database for central, regional (aggregated across states), and local (aggregated across municipal) levels and spans several decades for more than 165 economies.
- The authors extend the IMF Fiscal Decentralization dataset:
  - incorporate historical revisions and newly covered data for the years of 2021-2023,
  - expand the number of economies available to 97.
- New countries added in LAC coverage include Argentina, Dominican Republic, Ecuador, Panama and Uruguay where subnational fiscal data have become increasingly accessible and comparable.

### Consistency assessment
- Comparison between the granular subnational database and the extended IMF Fiscal Decentralization dataset reveals a high degree of consistency.
- Figure 1 (1989–2023) plots sum of regional revenues for Argentina, Brazil, Colombia, and Mexico (y-axis) against corresponding aggregate values reported by national authorities to the IMF (x-axis), showing close alignment around the 45-degree line.
- Minor discrepancies likely reflect differences in reporting standards between GFS submissions and more detailed regional data, measurement errors, and classification differences.

---

### 3.3 Additional Country-Level and Subnational-Level Demographic and Economic Information

- Demographic and economic indicators incorporated at national and subnational level include population and national or regional GDP obtained from countries’ statistical agencies and ministries of finance to ensure consistency.
- These indicators enable estimation of per capita regional GDP and construction of fiscal ratios expressed as a percent of subnational GDP, providing standardized bases for cross-country comparison and assessment of subnational fiscal capacity.
- Table 4 in the Annex summarizes all years available for each country and database used in the paper.

---

### Key empirical findings and stylized facts (from Sections 4.1–4.3)

### Global and regional trends (high-level)
- Total fiscal revenues across all government levels as a share of GDP have increased across all regions over time; levels in LAC remain below those observed in advanced economies and EMDEs.
- The importance of local governments remains comparatively limited in LAC versus advanced economies; regions in LAC reached in 2023 a high level of importance relative to advanced and emerging countries.
- Expenditures broadly mirror revenue patterns, with LAC exhibiting a more consistent upward trend at all government levels.
- Local government revenues and expenditures tend to move closely together across regions, suggesting tighter constraints on local governments’ ability to finance fiscal deficits.

### Long-term trends in Latin America (1989–2023)
- Regional revenues, as a share of national GDP, have generally increased across the six economies in the sample over 1989–2023.
- Argentina and Brazil’s regional budgets are the largest in the period, with revenues close to 15 percentage points of GDP in 2023.
- Mexico’s regional revenues are about 10 percentage points of GDP in 2023 and have experienced pronounced growth over the period.
- Uruguay’s regional government size remains the smallest in the period, followed by Colombia and Peru.
- Expenditures as a share of GDP follow a broadly similar pattern to revenues.

### Revenue and expenditure composition across countries
- Transfers from the central government play a dominant role in regional government revenues:
  - accounting for more than 65 percent in Argentina and Mexico,
  - about 50 percent in Colombia,
  - roughly 30 percent in Brazil and Uruguay,
  - transfers account for only about 10 percent of total regional government revenues in advanced economies.
- Capital spending as a share of total regional government expenditures:
  - more than 80 percent in Colombia,
  - 50 percent in Peru,
  - below 20 percent in the rest of the region (similar to advanced economies).
- Subnational debt service expenditure remains minimal in the region given relatively low debt levels and limited access to financing.

### Transfer dependence and dispersion
- Transfer systems vary by country: federal countries (Argentina, Brazil, Mexico) use constitutionally established revenue-sharing; unitary countries (Colombia, Peru, Bolivia) use sector-specific formulae and often natural-resource–based transfers.
- Mexico remains highly dependent on federal transfers, which account for up to 80–90 percent of state revenues.
- Over time transfers have remained relatively constant overall, but with substantial variation in transfer dependence across regional governments:
  - Uruguay shows a gradually rising dependence on transfers,
  - Argentina and Colombia display very wide dispersion in transfers as a share of total revenues with an interquartile (25th–75th percentile) range above 60 percent for most years,
  - Brazil and Uruguay display more moderate dispersion,
  - Mexico and Peru exhibit very high transfer dependence—exceeding 80 percent of total revenues—with very limited dispersion especially in Peru.
- Data for Bolivia is only available for 2023 in the comparison of regional transfer dispersion.

### Regional disparities and fiscal capacity
- Large dispersion of regional revenues as percent of regional GDP indicates wide intra-country variation in revenue autonomy and capacity, especially in Argentina and Brazil.
- Colombia and Mexico show more compact distributions of revenue-to-GDP ratios.
- Uruguay shows low and tightly clustered ratios, indicating limited regional revenue autonomy.

### Relationship between regional income and fiscal measures
- Richer regions (higher GDP per capita relative to national GDP per capita) tend to exhibit lower revenue ratios measured as a share of regional GDP:
  - total regional revenues can reach up to 50 percent of regional GDP in poorer regions,
  - regions with per capita GDP exceeding twice the national average display revenue shares closer to 10 percent of regional GDP.
- Richer regions tend to exhibit lower dependence on transfers as a share of total revenues (inverse relationship between transfer dependence and regional GDP per capita).
- High transfer dependency of poorer regions is driven by larger transfers, not by less own-source revenues:
  - own-source revenues, normalized by regional GDP per capita, are relatively flat across the income distribution,
  - transfers (as percent of regional GDP) show a clear negative association with regional GDP per capita—poorer regions receive larger transfer inflows.

### Interpretation
- Regional differences in transfer dependency in Latin America are mainly driven by redistributive transfer design rather than differences in regional fiscal effort.
- Intergovernmental transfers appear to play an equalizing role by channeling more resources to lower-income regions, but persistent high transfer dependency and wide dispersion imply significant regional fiscal disparities remain.
- Policy implications include the need to balance redistribution objectives with incentives for local revenue mobilization and efficient expenditure management.

*Source: IMF Working Paper — Fiscal Decentralization and Cyclicality in Latin America (sections 3.1–4.3 as provided).*

### 5.1 Empirical Strategy

### 5.1 Empirical Strategy

### Measurement of fiscal cyclicality
- Fiscal cyclicality is measured as the responsiveness of revenues and expenditures to changes in economic activity, proxied by national or regional GDP.
- The analysis compares cyclical behavior across regional and central governments and investigates drivers of heterogeneity across levels of government and countries.
- The study considers elasticities of regional governments with respect to national GDP as well as to regional GDP.
- The sensitivity of total revenues and their subcomponents, and of total expenditures and their subcomponents, to changes in economic activity is examined.

### Econometric framework
- The estimation of fiscal elasticities is based on a panel Error Correction Model (ECM) following Engle and Granger (1987), which captures both short-run and long-run relationships between fiscal variables and economic activity.
  - The ECM framework is suited for macro-fiscal data where variables tend to be non-stationary but cointegrated.
  - The error-correction term ensures deviations from long-run equilibrium are corrected over time, distinguishing temporary cyclical movements from persistent structural adjustments.
  - The framework yields both short-run and long-run elasticities in a coherent model.

- To address common shocks affecting all countries or regions simultaneously, the Common Correlated Effects (CCE) estimator (Pesaran, 2006) is used.
  - The CCE estimator introduces cross-sectional averages of the dependent variable and GDP (cross-sectional average term) to control for unobserved global or regional factors (e.g., global commodity price swings, synchronized business cycles).
  - Combining ECM and CCE allows accounting for country-specific dynamics and broader macroeconomic interdependencies, producing robust and comparable estimates of:
    - short-run elasticity to GDP (response to contemporaneous changes in GDP), and
    - long-run elasticity to GDP (adjustment to sustained shifts in output).

### Empirical specification
- The empirical specification is expressed as:

  (1)   훥푙푛푌
  푖,푡
   = 휃
  푖
   훥푙푛퐺퐷푃
  푖,푡
   + 훾
  푖
   (푙푛푌
  푖,푡−1
   − 훽
  푖
   푙푛퐺퐷푃
  푖,푡−1
  ) + 휇
  푖
   + 훿
  1푖
  푚푒푎푛
  푖
  (푙푛푌
  푖,푡
  ) + 훿
  2푖
  푚푒푎푛
  푖
  (푙푛퐺퐷푃
  푖,푡
  ) + 휀
  푖,푡

- Notation and interpretation:
  - 푌
  푖,푡
    is a fiscal variable (e.g., revenues or expenditures) of geographic unit 푖 in year 푡 (e.g., a region in a given country or the entire country).
  - 휃푖 is the short-run elasticity to GDP.
  - 훽푖 is the long-run elasticity to GDP.
  - 훾푖 is the coefficient on the error-correction term.
  - 휇푖 denotes unit fixed effects.
  - 훿1푖 and 훿2푖 multiply cross-sectional mean terms to implement the CCE corrections.
  - The model is estimated for alternative fiscal indicators and GDP measures to identify both long- and short-run elasticities at national and regional levels.

### Robustness and alternative estimators
- Results are verified using four alternative estimators:
  - Mean Group (MG) estimator: baseline that does not account for cross-sectional dependence.
  - Clustered Common Correlated Effects (CCEC) estimator: accounts for within-country cross-sectional dependence by including country-level cross-sectional means.
  - Dynamic Common Correlated Effects (DCCE) estimator: addresses lagged cross-sectional dependence by incorporating the cross-sectional mean at time t-1.
  - CCET model: adds country-specific linear trends to control for the long-run growth of subnational governments.

*Source: IMF estimations using subnational-level budget information for subnational governments compiled from individual countries.*

### 5.2 Empirical Results

### 5.2 Empirical Results

### Short-run and long-run fiscal elasticities: headline findings
- On average, a 1 percent increase in nominal national GDP is associated with a 0.87 percent rise in regional revenues in the short-term.
- Central government short-term revenue response is near one-to-one with national GDP (described as "near one-to-one" in the source).
- Long-run revenue elasticities for both regional and central governments are not statistically different from one, implying revenues move broadly in line with national GDP over time.
- Short-term elasticities for expenditures to national GDP are broadly similar for regional and central governments, but spending adjusts less than revenues in the short term.
- Short-term expenditure elasticities tend to be lower than revenue elasticities for both central and regional governments; however, the expenditure elasticity for central governments to national GDP has a wide confidence interval.

### Revenue decomposition: own-source vs transfers
- Regional own-source revenues display elasticities close to or above unity with respect to both national and regional GDP.
- Regional transfer revenues exhibit weaker elasticity to national GDP, typically below 0.7.
- The elasticity of transfers to regional GDP is close to 0.4.
- The lower responsiveness of regional transfer revenues is attributed to rigid transfer rules from central to regional governments in many Latin American countries.
- The countercyclical aspect of transfers provides a backstop for regions but increases rigidity for central government expenditures.

### Expenditure decomposition: current vs capital
- Current expenditures at the regional level have short-run elasticities to national and regional GDP in the range of 0.4–0.5.
- Capital expenditures exhibit substantially higher short-run responsiveness, with elasticities well above one for both fiscal perimeters.
- The combination of current and capital elasticities leads to a total expenditure elasticity in the range of 0.5-0.6 for both central and regional perimeters.
- The pattern—greater cyclical sensitivity of capital spending and rigidity of current spending—suggests investment spending is the key short-run adjustment margin during upturns and downturns.

### Within-country heterogeneity and time stability
- Within-country comparisons show that greater dependence on intergovernmental transfers dampens regional revenue elasticity.
- Regions with above-median transfer dependency within each country exhibit lower short-term revenue elasticities than regions with below-median transfer dependency.
- Elasticity patterns are mostly constant over time: similar revenue and expenditure elasticities are found across two subperiods, 1995-2010 and 2011-2021.
- Results are robust to adding controls for electoral years.

### Robustness checks and alternative estimators
- Robustness tests across estimators (MG, CCET with country-specific linear trends, benchmark CCE, CCEC, dynamic models) reveal a broadly consistent picture.
- MG and CCET yield results very similar to the benchmark CCE specification.
- Slightly smaller elasticities are observed when using the CCEC estimator; by including country-level cross-sectional means, the estimator isolates within-country variation in state-level GDP and indicates that pure state-level shocks are even more countercyclical.
- Dynamic models yield marginally higher elasticities, consistent with macroeconomic shocks at t-1 partially driving the response of current revenues and expenditures because budgets are drafted and approved in t-1.

### Regional debt findings (extension)
- Regional debt is low relative to regions' GDP in most cases, but there are notable outliers, particularly in some regions of Brazil and Argentina.
- When debt is measured relative to own-source revenues rather than regional GDP, several regions show higher debt burdens, suggesting greater potential dependence on intergovernmental transfers to service obligations.
- These debt patterns do not appear to differ systematically across income levels; richer and poorer regions exhibit broadly similar debt levels (based on the smaller subsample with debt data).

### Quantitative regression excerpt (selected coefficients from Table 6)
- Coefficient: Normalized GDP per capita = -0.134 ** (standard error 0.058) for Transfer Dependency (dependent variable).
- Coefficient: Normalized GDP per capita = 0.001 (standard error 0.003) for Revenue Own-Source (%GDP).
- Coefficient: Normalized GDP per capita = -0.058 *** (standard error 0.019) for Revenue Transfer (%GDP).
- Country × Year fixed effects: Yes for the regressions reported.
- Observations: 2794.000 for each regression reported.
- R-squared: 0.711 for Transfer Dependency; 0.664 for Revenue Own-Source (%GDP); 0.457 for Revenue Transfer (%GDP).
- Significance notation in source: * p < 0.10, ** p < 0.05, *** p < 0.01.

### Policy implications and recommendations
- Increase revenue autonomy to enhance fiscal flexibility at the subnational level; expanding regional tax bases or creating shared income tax arrangements could strengthen own-revenue countercyclical capacity.
- Redesign transfer systems to incorporate performance-linked elements to reduce rigidity and promote more own-source financing.
- Strengthen coordination between central and regional governments in budget planning and implement fiscal rules at the regional level to mitigate asymmetric shocks while preserving aggregate fiscal discipline.
- Address spending rigidities (pensions, wages, interest, institutional constraints) through structural reforms and efficiency-enhancing reprioritization to reduce rigidity and improve adjustment capacity.
- Integrate subnational governments into comprehensive fiscal reporting, medium-term fiscal frameworks, and credible borrowing and fiscal rules to identify and mitigate macro-fiscal risks from subnational finances.
- Protect high-quality public investment given the reliance on capital spending as the primary short-run adjustment margin and the risks of procyclicality for long-term growth.

_Italic source: IMF estimations using subnational-level budget information for subnational governments compiled from individual countries (Content unit: wpiea2026092-source-pdf - 5.2 Empirical Results)._

### References

### References

### Fiscal decentralization, theory, and public finance classics
- Musgrave, R.A., 1959, The Theory of Public Finance: A Study in Public Economy. New York: McGraw-Hill.  
- Tiebout, C.M., 1956. A pure theory of local expenditures. Journal of Political Economy, 64(5), pp. 416–424.  
- Wagner, A., 1883. Finanzwissenschaft. Leipzig: C.F. Winter.  
- Olson, M., 1969. The principle of “fiscal equivalence: The division of responsibilities among different levels of government. American Economic Review, 59(2), pp.479–487.  
- Oates, W.E., 1972. Fiscal Federalism. New York: Harcourt Brace Jovanovich.  
- Oates, W.E., 1999. An essay on fiscal federalism, Journal of Economic Literature, 37(3), pp.1120–1149.  
- Oates, W.E., 2005. Toward a second-generation theory of fiscal federalism. International Tax and Public Finance, 12(4), pp.349–373.  
- Persson, T. and Tabellini, G., 1999. The size and scope of government: Comparative politics with rational politicians, European Economic Review, 43(4–6), pp.699–735.  
- Meltzer, A.H. and Richard, S.F., 1981. A rational theory of the size of government, Journal of Political Economy, 89(5), pp.914–927.  
- Benabou, R., 1996. Equity and Efficiency in Human Capital Investment: The Local Connection. The Review of Economic Studies, 63(2), 237.  
- Groves, H.M. and Kahn, C., 1952. The stability of state and local tax yields. American Economic Association Studies in Income and Wealth, Vol. 14. Princeton, NJ: Princeton University Press.  
- Kenen, P.B., 1969. The Theory of Optimum Currency Areas: An Eclectic View, in Mundell, R.A. and Swoboda, A.K. (eds.) Monetary Problems of the International Economy. Chicago: University of Chicago Press.

### Decentralization, subnational politics, and regional evidence (Latin America & elsewhere)
- Falleti, T.G., 2010. Decentralization and Subnational Politics in Latin America. Cambridge: Cambridge University Press.  
- Brosio, G. and Jiménez, J.P. (eds.), 2012. Decentralization and Reform in Latin America. Cheltenham: Edward Elgar.  
- Faguet, J.P., 2004. Does decentralization increase government responsiveness to local needs?: Evidence from Bolivia. Journal of public economics, 88(3-4), pp.867-893.  
- Stein, E., 1999. Fiscal decentralization and government size in Latin America. Journal of applied Economics, 2(2), pp.357-391.  
- Meloni, O., 2018. Fiscal policy cyclicality at the subnational level: Evidence from Argentine provinces. Economic Modelling, 68, pp. 164–180.  
- Martínez-Vázquez, J. and McNab, R.M., 2003. Fiscal decentralization and economic growth, World Development, 31(9), pp. 1597–1616.  
- CAF – Development Bank of Latin America, 2025. Decentralization and subnational public finances in Latin America. Caracas: CAF.  
- Inter-American Development Bank, 2022. Outlook of Fiscal Relations among Government Levels in Latin America and the Caribbean. Washington, DC: IDB.  
- Inter-American Development Bank, 2024. Centralized Information on Territorial Expenditures (CITE) Database. Washington, DC: IDB.  
- Arena, M. and Revilla, J.E., 2009. Pro-cyclical fiscal policy in Brazil: evidence from the states. World Bank Policy Research Working Paper, (5144).

### Fiscal cyclicality, tax revenue dynamics, and fiscal policy over the cycle
- Lane, P.R., 2003. The cyclical behaviour of fiscal policy: Evidence from the OECD. Journal of Public Economics, 87(12), pp. 2661–2675.  
- Gavin, M. and Perotti, R., 1997. Fiscal policy in Latin America, NBER Macroeconomics Annual, 12, pp.11–61.  
- Frankel, J.A., Végh, C.A. and Vuletin, G., 2013. On graduation from fiscal procyclicality. Journal of Development Economics, 100(1), pp. 32–47.  
- Heimberger, P., 2023. Fiscal policy over the business cycle: A meta-analysis. Oxford Economic Papers, 75(2), pp.373–404.  
- Martorano, B., 2018. Cyclicality of fiscal policy in Latin America over the period 1990–2015, Review of Development Economics, 22(1), pp.67–86.  
- Jalles, J.T., Kiendrebeogo, Y., Lam, R. and Piazza, R., 2024. Revisiting the countercyclicality of fiscal policy. Empirical Economics, 67(3), pp.877-914.  
- Fricke, C. and Süssmuth, B., 2014. Growth and volatility of tax revenues: Evidence from a dynamic panel ECM. International Tax and Public Finance, 21(5), pp.874–908.  
- Lagravinese, R., Liberati, P. and Sacchi, A., 2020. Tax buoyancy in OECD countries: New evidence from a multivariate approach. Public Finance Review, 48(4), pp. 483–517.  
- Cornevin, A., Corrales, J.S. and Mojica, J.P.A., 2024. Do tax revenues track economic growth? Comparing panel data estimators. Economic Modelling, 140, p.106867.  
- Klemm, A., 2014. Fiscal policy in Latin America over the cycle, IMF Working Paper, No. 14/59. Washington, DC: International Monetary Fund.  
- Meloni, O., 2018. Fiscal policy cyclicality at the subnational level: Evidence from Argentine provinces. Economic Modelling, 68, pp. 164–180.  
- Dougherty, S. and Biase, M., 2020. Who bears the burden of economic adjustment? A comparative analysis of subnational fiscal cyclicality, OECD Working Papers on Fiscal Federalism No. 27. Paris: OECD Publishing.

### Subnational fiscal risk sharing, rules, and management
- Asdrubali, P., Sørensen, B.E. and Yosha, O., 1996. Channels of interstate risk sharing: United States 1963–1990, The Quarterly Journal of Economics, 111(4), pp.1081–1110.  
- Furceri, D. and Zdzienicka, A., 2015. The euro area crisis: need for a supranational fiscal risk sharing mechanism?. Open Economies Review, 26(4), pp.683-710.  
- Saxena, S., 2022. How to Manage Fiscal Risks from Subnational Governments. International Monetary Fund How To Notes, 003.  
- Eyraud, L., Hodge, A., Ralyea, J. and Reynaud, J., 2020. How to Design Subnational Fiscal Rules, A Primer. International Monetary Fund How To Notes, 01.  
- De Biase, P. and Dougherty, S., 2021. Federalism and public health decentralisation in the time of COVID-19. OECD Working Papers on fiscal federalism, (33), pp.1-43.

### Empirical methods and econometrics referenced
- Engle, R.F. and Granger, C.W., 1987. Co-integration and error correction: representation, estimation, and testing. Econometrica, pp.251-276.  
- Pesaran, M.H., 2006. Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4), pp.967-1012.  
- Sørensen, B.E., Wu, L., Yosha, O. and Zhu, Y., 2001. Output fluctuations and fiscal policy: U.S. state and local governments 1978–1994. European Economic Review, 45(7), pp.1271–1310.

### Government size, composition, and cross-country perspectives
- Shelton, C.A., 2007 The size and composition of government expenditure, Journal of Public Economics, 91(11–12), pp.2230–2260.  
- Rodrik, D., 1998. Why do more open economies have bigger governments?, Journal of Political Economy, 106(5), pp. 997–1032.  
- Rodríguez-Pose, A. and Ezcurra, R., 2010. Does decentralization matter for regional disparities? A cross-country analysis, Journal of Economic Geography, 10(5), pp.619–644.  
- Xing, J. and Fuest, C., 2018. Central-local government fiscal relations and cyclicality of public spending: evidence from China. International Tax and Public Finance, 25(4), pp.946-980.  
- Gupta, S., Jalles, J.T., Melina, G. and Schena, F., 2022. Fiscal policy, growth, and income inequality in developing countries. IMF Working Paper No. 22/40. Washington, DC: International Monetary Fund.

*Fiscal Decentralization and Cyclicality in Latin America: Evidence from a New Regional Budget Database Working Paper No. WP/2026/092*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026092-source-pdf.pdf_