Political Decentralization and Fiscal Management of Local Governments

Political Decentralization and Fiscal Management of Local Governments: Evidence and Lessons from Korea No-Wook Park Center for Fiscal Analysis Korea Institute of Public Finance

Purpose of the Presentation n Share Korean experiences by examining the effects of political decentralization (resumption of local elections) on the fiscal management of local governments – Resumption of Korean local elections in 1995 – Identify differences in fiscal managements before and after the resumption of local elections – Derive lessons and implications from Korean experiences

Decentralization Process n Decentralization is a globally popular since 1990 s – Many developing countries have experimented with decentralization reforms – Decentralization reforms are in full swing with the initiatives of the current administration in Korea

What is Decentralization? n Devolution of various authorities (powers) to the lower level of governments – Fiscal decentralization • More fiscal decision making authority and economic resources in local governments – Political decentralization • Local governments are formed by local residents – Administrative decentralization • More administrative decision making authority in local governments

Why Decentralization? n Expected benefits of decentralization – Increase in efficiency • Local governments have better local information • Competition among local governments improves the performance of local governments n Expected benefits of centralization – Realization of economies of scale – Internalize externalities across local governments n Recent trend reflects the perception that movement towards decentralization may bring in benefits.

How to Decentralize? n Proper assignment of various authorities between central and local governments n Capacity consideration is important. – Capacity for decentralization needs to be developed. – Local governments tend to lack capacity to cope with various issues. • Administrative capacity • Restraining corruption

Supporting Institutions for Decentralization n Intergovernmental Fiscal Institutions – Intergovernmental Transfers – Transfer Dependency and Local Revenue Raising – Expenditure Assignment and Local Autonomy Political Institutions n Capital Markets n Factor Mobility and Land Market n Hierarchical Mechanisms n

Political Decentralization n How to select local officials – Election: Politicians – Appointment: Bureaucrats n Resumption of local elections – Appointment Election

Why Do Local Elections Matter? n Different Incentives – Politicians • Winning Elections – Bureaucrats • Career Concerns n Local Elections as an Issue-Unbundling Mechanism – Elected Local Officials • Mainly Care for Local Issues – Appointed Local Officials • Care for both of Central and Local Issues

The Korean Case n Resumption of Local Elections in 1995 – Opportunity to examine the effects of political decentralization – No significant devolution of economic resources to local governments – Appointed officials Elected Officials

Research Questions n Do elected officials manage differently in their fiscal issues? n In addition to elections as a way of selecting local officials, do other political factors matter? – Political competition in each district – Alignment with the incumbent president

Summary of Results 1 n More fiscal independence – Exert efforts in increasing internal revenues n Raising revenues through non-tax revenues as opposed to through local taxes – Minimize political costs associated with raising internal revenues

Summary of Results 2 n Do not face harder budget constraints – Rely more on intergovernmental transfers than local borrowings n Political competitive districts receive more intergovernmental grants – Political decentralization may activate distributive politics

Institutional Backgrounds n Political decentralization n Regional identity based voting and imperial president

Political decentralization 1 n Resumption of local elections for the head of local governments in 1995 – Suspended since 1961 – 3 year term and term-limit is 3 times – Higher level local governments • 7 Metro-cities and 6 provinces – Lower level local governments • 232 cities, counties and districts in 2000

Political decentralization 2 n No significant fiscal decentralization during 1991 -1999 n Local tax revenue – 20 percent of total tax revenue n Local government’s budget share – 35 -43 percent of total government revenue

Voting Behavior and Imperial Presidency n Regional identity-based voting n Imperial presidency n Importance of Swing Voters (Districts) – Expected to be important given strong regionalism n Importance of Alignment with president – May not be important

Empirical Strategy n Devise Fiscal Indices n Test significance of decentralization dummy which is 1 after decentralization n Test significance of political variables in explaining differences of indices pre and post decentralization

Indices of Fiscal Management 1 n Internal-revenue vs. external-revenue – Fiscal independence index – FI 1 = (local tax+non-tax)/total expenditure – FI 2 = (local tax+non-tax)/total revenue

Indices of fiscal management 2 n Revenue raising through non-tax vs. local tax – Source of raising internal revenue – RI 1 = current non-tax/(local tax+current non-tax) – RI 2 = non-tax/(local tax+non-tax) – Non-tax revenues tend to carry less political costs.

Indices of fiscal management 3 n External financing through subsidy vs. local borrowing – Source of external funding – SI 1 = National subsidy/(national subsidy+local borrowing) – SI 2 = All intergovernmental transfers/(intergovernmental transfers + local borrowing) – Transfers tend to be soft money.

Data 1 n Local government fiscal year book during 1991 -1999 – No significant change in fiscal rules related to local governments n Cities, counties, and district level data – 232 cities, counties, and districts in 2000 – Due to administrative boundary changes in 1995, 123 local governments are analyzed.

Results - Break of Trend 1 n Significant break of trend after the resumption of local elections n Unconditional average – Fiscal independence indices increase by 1 -2% – Subsidy indices decrease by 1 -2% – Non tax revenue indices increase by 6 -9%

Table 1 Average Indices of Pre and Post Decentralization Pre-Decentralization (1991 -1995) Index Post-Decentralization (1996 -1999) Mean Standard Deviation Number of Observations Independence Index 1 0. 6231 0. 4304 615 0. 6547 0. 2779 492 Independence Index 2 0. 4817 0. 3178 615 0. 4925 0. 1734 492 Non-tax Revenue Index 1 0. 2946 0. 0697 615 0. 3529 0. 0665 492 Non-tax Revenue Index 2 0. 6098 0. 1019 615 0. 6988 0. 1001 492 Subsidy Index 1 0. 9351 0. 1072 615 0. 9213 0. 1092 492 Subsidy Index 2 0. 9660 0. 0750 612 0. 9554 0. 0721 492

Results - Break of Trend 2 n Controlling for district fixed effect and year effect n Significantly positive decentralization dummy (=1 after local elections) for every category of indices – Independence indices increase 3 -5% – Subsidy indices increase 0. 56 -0. 35% – Non tax indices increase 5 -8. 4%

Table 2 Comparison of Indices between Pre and Post Decentralization (Weighted Least Squares Logit Estimates) Index Decentralizati on Dummy Estimated Predecentralization Fiscal Index Estimated Postdecentralization Fiscal Index R 2 Number of Observations Independence Index 1 0. 2524 (0. 040) 0. 4657 (0. 1710) 0. 4954 (0. 1671) 0. 9031 1228 Independence Index 2 0. 2724 (0. 055) 0. 5525 (0. 1652) 0. 6023 (0. 1577) 0. 7416 1107 Non-tax Revenue Index 1 0. 3074 (0. 032) 0. 2998 (0. 0552) 0. 3499 (0. 0575) 0. 7000 1107 Non-tax Revenue Index 2 0. 5410 (0. 043) 0. 6178 (0. 0932) 0. 7014 (0. 0822) 0. 7532 1228 Subsidy Index 1 0. 2886 (0. 147) 0. 8400 (0. 1117) 0. 8456 (0. 1067) 0. 539 761 Subsidy Index 2 0. 3650 (0. 145) 0. 9117 (0. 0920) 0. 9152 (0. 0893) 0. 626 760

Results – Political determinants 1 n Robust results – Political competition have a negative impact on fiscal independence index 2 – Political variables do not matter for non-tax revenue indices – Political competition has a positive correlation with subsidy index 1

Table 3 Average Political Factors Affecting Changes in Fiscal Management Independe nce Index 1 Independe nce Index 2 Revenue Index 1 Revenue Index 2 Subsidy Index 1 Subsidy Index 2 Alignment -0. 001 (0. 021) -0. 066 (0. 072) 0. 002 (0. 024) -0. 006 (0. 029) -0. 017 (0. 038) -0. 037** (0. 019) Competition 0. 001 (0. 064) -0. 505** (0. 208) -0. 096 (0. 065) -0. 063 (0. 077) 0. 199** (0. 076) 0. 064* (0. 037) Alignment*Competition -0. 029 (0. 012) 0. 028 (0. 052) -0. 020 (0. 19) -0. 020 (0. 022) -0. 083*** (0. 025) -0. 022* (0. 013) Population Density -0. 003*** (0. 000) -0. 010*** (0. 002) -0. 000 (0. 001) -0. 004*** (0. 001) 0. 009*** (0. 002) 0. 004*** (0. 001) Number of Observations 89 89 89 R 2 0. 3866 0. 264 0. 049 0. 2374 0. 2968 0. 2388

Summary 1 n Fiscal management changed after the resumption of local elections – More fiscal independence – More non-tax revenue raising efforts – More dependence on subsidy for external funding

Summary 2 n Political characteristics of districts matter for attracting subsidy from the central government after political decentralization. n Internal revenue raising efforts through nontax revenues are significant regardless of political characteristics.

Efficiency implications n Increased revenue raising through non-tax – Efficiency implication depends on how local governments use increased revenues n Active political factors – May have negative implications n Decentralization efficiency gains. itself does not guarantee – Supporting institutions are important.

Other Implications n Political institutions are important in realizing benefits of decentralization. n Political environments shaping incentives of local officials are crucial. – Can local residents (voters) discipline local officials? – Availability of Information on the fiscal management of local governments – How to avoid a possibility of collusion between local officials and powerful local interest groups

Thank you!