6591
Views
3390
Downloads |
The political economy of local government in Croatia: winning coalitions, corruption, and taxes
Vuk Vuković* 
Article | Year: 2017 | Pages: 387 - 420 | Volume: 41 | Issue: 4 Received: June 1, 2017 | Accepted: October 20, 2017 | Published online: December 11, 2017
|
FULL ARTICLE
FIGURES & DATA
REFERENCES
CROSSMARK POLICY
METRICS
LICENCING
PDF
Variable | Observations | Mean | Std. Dev. | Min | Max | W/S ratio | 546 | 0.270 | 0.102 | 0.034 | 0.681 | W/S_binary | 546 | 0.469 | 0.499 | 0 | 1 | Corruption Index | 546 | 0.083 | 0.127 | 0 | 0.75 | Tax rate (prirez) | 546 | 0.042 | 0.043 | 0 | 0.18 | Re-election | 546 | 0.731 | 0.444 | 0 | 1 | Years in power | 546 | 8.668 | 3.532 | 2 | 12 | Turnout | 546 | 0.478 | 0.100 | 0.231 | 0.886 | HDZ vote share | 546 | 0.206 | 0.115 | 0.003 | 0.675 | SDP vote share | 546 | 0.244 | 0.098 | 0.019 | 0.523 | Mayor gov | 546 | 0.676 | 0.468 | 0 | 1 | Size of gov coalition | 546 | 2.170 | 1.136 | 1 | 7 | Mayor majority | 545 | 0.585 | 0.112 | 0.231 | 1 | EPI | 546 | 0.818 | 0.177 | 0 | 0.996 | Unemployment | 546 | 0.193 | 0.093 | 0.04 | 0.567 | Income per capita | 546 | 1696.8 | 438.5 | 519.9 | 2973.6 | Grants per capita | 546 | 507.3 | 998.5 | 0 | 11575.9 | Croats | 546 | 0.889 | 0.172 | 0.018 | 1 | Average age | 546 | 42.642 | 3.330 | 33.1 | 63.3 | Years of education | 546 | 9.842 | 0.875 | 5.925 | 12.132 | War disabled per 1000 | 546 | 8.539 | 8.082 | 0 | 64.98 | Settlement size | 546 | 3.006 | 0.545 | 1.65 | 5.78 | Assembly size | 546 | 13.385 | 3.897 | 7 | 51 |
Note: The data was collected for all 556 municipalities but 10 of them had no procurements in the given years so they were excluded from the dataset. Including them as a zero value does not influence the findings. Sources: See appendix.
Dependent variable:
Years in power (Ti) | (1) | (2) | (3) | (4) | W/S ratio | -4.135 (1.55)*** | -3.779 (1.56)** | -4.898 (1.61)*** | -3.913 (1.71)** | Corruption index | | 2.581 (1.2)** | 2.446 (1.24)** | 2.36 (1.18)** | Size of gov coalition | -0.613 (0.15)*** | -0.641 (0.145)*** | -0.661 (0.143)*** | -0.681 (0.143)*** | Mayor majority | 8.889 (1.13)*** | 8.851 (1.14)*** | 9.078 (1.11)*** | 9.332 (1.107)*** | EPI | -1.281 (0.97) | -1.454 (0.98) | -1.725 (0.97) | -1.172 (1.07) | Mayor gov | 0.259 (0.41) | 0.25 (0.40) | 0.343 (0.40) | 0.275 (0.41) | HDZ vote share | 2.294 (1.57) | 2.196 (1.59) | 2.843 (1.54)* | 2.818 (1.93) | SDP vote share | 6.424 (2.15)*** | 5.915 (2.18)*** | 3.339 (2.42) | 3.738 (2.61) | Unemployment | | | -5.77 (2.28)** | -7.29 (2.55)*** | Income p/c | | | -0.0007 (0.0004) | -0.0007 (0.0006) | Grants p/c | | | 0.0002 (0.0001) | 0.0003 (0.0001)** | Controls | No | No | No | Yes | N | 546 | 546 | 546 | 546 | R2 | 0.1869 | 0.1951 | 0.2102 | 0.2194 | F-test (p-value) | 27.53 (0) | 24.29 (0) | 21.42 (0) | 15.70 (0) |
Note: OLS regressions performed throughout. Control variables include the municipality and city specific controls: Croats, Average age, Years of education, War disabled, Settlement size, and Assembly size. Standard errors are robust to heteroskedasticity. *** denotes significance at 1%, ** at 5%, and * at 10%.
Dependent variable: Corruption index (ki) | (1) | (2) | (3) | (4) | W/S ratio | -0.138 (0.06)** | -0.122 (0.06)** | -0.121 (0.06)* | -0.113 (0.06)* | Tax rate (prirez) | | | | 0.267 (0.129)** | Size of gov coalition | 0.01 (0.005)* | 0.006 (0.005) | 0.006 (0.005) | 0.005 (0.143) | Mayor majority | 0.014 (0.039) | 0.02 (0.038) | 0.031 (0.039) | 0.033 (0.04) | EPI | 0.067 (0.032)** | 0.083 (0.031)** | 0.078 (0.032)** | 0.074 (0.032)** | Mayor gov | 0.003 (0.01) | 0.001 (0.01) | 0.005 (0.01) | 0.006 (0.01) | HDZ vote share | 0.038 (0.068) | 0.008 (0.069) | 0.018 (0.088) | 0.014 (0.089) | SDP vote share | 0.197 (0.10)* | 0.05 (0.10) | 0.058 (0.12) | 0.063 (0.12) | Unemployment | | 0.117 (0.09) | 0.092 (0.101) | 0.079 (0.10) | Income p/c | | 8.38x10-5 (1.48x10-5)*** | 4.59x10-5 (2.43x10-5)* | 4.06x10-5 (2.4x10-5)* | Grants p/c | | -8.14x10-6 (3.17x10-6)** | -9.42x10-6 (4.2x10-6)** | -9.44x10-6 (4.1x10-6)** | Controls | No | No | Yes | Yes | N | 546 | 546 | 546 | 546 | R2 | 0.0498 | 0.0989 | 0.1088 | 0.1158 | F-test (p-value) | 4.56 (0) | 8.54 (0) | 6.90 (0) | 6.99 (0) |
Dependent variable: Re-election (Ri) | Small winning coalition
(W/S < 0.26) | Large winning coalition (W/S > 0.26) | | (1) Probit | (2) LPM | (3) Probit | (4) LPM | Corruption ki | 0.948 (0.508)* | 0.954 (0.524)* | 0.229 (0.353) | 0.318 (0.447) | Corruption squared ki2 | -1.868 (0.948)** | -1.891 (0.927)** | -1.182 (0.795) | -1.576 (1.401) | Controls | Yes | Yes | Yes | Yes | N | 295 | 295 | 250 | 250 | Pseudo R2 | 0.1276 | 0.1522 | 0.2717 | 0.1821 | Wald/F-test (p-value) | 43.67 (0) | 4.36 (0) | 40.93 (0) | 1.26 (0.19) |
Dependent variable: Tax rate (ti) | (1) | (2) | (3) | (4) | W/S ratio | -0.051 (0.016)*** | -0.039 (0.016)** | -0.041 (0.018)** | -0.023 (0.019) | Corruption index | 0.053 (0.016)*** | 0.034 (0.015)** | 0.030 (0.015)** | 0.029 (0.014)** | Size of gov coalition | | | 0.005 (0.002)*** | 0.004 (0.002)** | Mayor majority | | | -0.012 (0.015) | -0.007 (0.016) | EPI | | | 0.003 (0.012) | 0.011 (0.012) | Mayor gov | | | -0.004 (0.005) | -0.005 (0.005) | HDZ vote share | | | 0.022 (0.023) | 0.013 (0.025) | SDP vote share | | | -0.017 (0.03) | -0.021 (0.03) | Unemployment | | 0.084 (0.024)*** | 0.063 (0.027)** | 0.046 (0.031) | Income p/c | | 3.05x10-5 (5.87x10-6)*** | 3.11x10-5 (6.13x10-6)*** | 1.88x10-5 (8.2x10-6)** | Grants p/c | | -2.37x10-6 (1.7x10-6) | -2.59x10-6 (1.7x10-6) | 3.72x10-7 (1.8x10-6) | Controls | No | No | No | Yes | N | 546 | 546 | 546 | 546 | R2 | 0.0442 | 0.0958 | 0.1208 | 0.1513 | F-test (p-value) | 11.62 (0) | 10.37 (0) | 6.19 (0) | 5.75 (0) |
Note: OLS regressions performed throughout. Control variables include the municipality and city specific controls: Croats, Average age, Years of education, War disabled, Settlement size, and Assembly size. Standard errors are robust to heteroskedasticity. *** denotes significance at 1%, ** at 5%, and * at 10%.
| (1) PS matching | (2) NN matching | (3) Radius matching | (4) Kernel matching | Impact of W/S on tenure (Ti) | -0.987 (0.440) ** | -0.543 (0.32)* | -0.644 (0.338)* | -0.669 (0.302)** | R2 | 0.3059 | 0.2221 | 0.2144 | 0.2430 | F-test (p-value) | 12.9 (0) | 13.09 (0) | 10.29 (0) | 13.89 (0) | Impact of W/S on corruption (ki) | -0.0558 (0.038) | -0.0526 (0.026)** | -0.0367 (0.028) | -0.046 (0.026)* | R2 | 0.0846 | 0.088 | 0.0713 | 0.0912 | F-test (p-value) | 1.85 (0.04) | 2.56 (0) | 2.59 (0) | 2.67 (0) | Impact of W/S on tax rates (ti) | -0.0102 (0.005) * | -0.0086 (0.004)** | -0.0097 (0.003)** | -0.0099 (0.003)*** | R2 | 0.1855 | 0.1290 | 0.0962 | 0.1168 | F-test (p-value) | 4.85 (0) | 3.81 (0) | 4.25 (0) | 4.29 (0) | Controls | Yes | Yes | Yes | Yes | N | 257 | 484 | 546 | 537 |
| (1) Entropy balancing | (2) Propensity score matching | (3) Unmatched sample | Impact of W/S on tenure (Ti) | -0.857 (0.275)*** | -0.987 (0.440)** | -0.62 (0.30)** | R2 | 0.2519 | 0.3059 | 0.2018 | F-test (p-value) | 23.88 (0) | 12.9 (0) | 20.49 (0) | Impact of W/S on corruption (ki) | -0.0602 (0.030)** | -0.0558 (0.038) | -0.054 (0.026)** | R2 | 0.0856 | 0.0846 | 0.0752 | F-test (p-value) | 3.09 (0) | 1.85 (0.04) | 5.21 (0) | Impact of W/S on tax rates (ti) | -0.012 (0.004)*** | -0.0102 (0.005)* | -0.0097 (0.037)** | R2 | 0.184 | 0.1855 | 0.1245 | F-test (p-value) | 3.51 (0) | 4.85 (0) | 6.45 (0) | Controls | Yes | Yes | Yes | N | 546 | 257 | 546 |
Note: The first column reports the estimates of the three given relationships after entropy balancing, the second replicates the same results for propensity score matching from column 1 in table 6, and the final are OLS estimates from an unmatched sample reported in column 4 of tables 2, 3 and 5. The full list of control variables is the same as in column 4 of tables 2 and 3. Standard errors are robust to heteroskedasticity. *** denotes significance at 1%, ** at 5%, and * at 10
Variable | Description | Source | W/S ratio | Winning coalition divided by the entire selectorate (votes for incumbent divided by all eligible voters) | State Electoral Commission of the Republic of Croatia | Corruption index | Weighted average of all the given proxies of corruption defined through various cases of potentially fraudulent public procurements | Vukovic (2017), based on interviews with key stakeholders in the procurement process. Source of procurements: Official Gazette of the Republic of Croatia | Tax rate (prirez) | Local tax surcharge rate on income tax | Ministry of Finance, Tax Administration | Re-election | =1 when incumbent mayor gets re-elected, =0 otherwise | State Electoral Commission of the Republic of Croatia | Years in power | Total years in power until 2013 | State Electoral Commission of the Republic of Croatia | Turnout | Municipality-level turnout in the 2013 local election | State Electoral Commission of the Republic of Croatia | HDZ vote share | Proportion of municipal vote for the HDZ in the 2011national election | State Electoral Commission of the Republic of Croatia | SDP vote share | Proportion of municipal vote for the SDP in the 2011national election | State Electoral Commission of the Republic of Croatia | Mayor for government | = 1 when mayor from the same party as the national government; = 0 otherwise | Glaurdić and Vuković (2017); source of original data: State Electoral Commission of the Republic of Croatia | EPI | Electoral polarization index, based on municipality-level results in the directly preceding national election | Glaurdić and Vuković (2017); source of original data: State Electoral Commission of the Republic of Croatia | Size of governing coalition | Number of seats the governing coalition has in the local assembly | Glaurdić and Vuković (2017); source of original data: State Electoral Commission of the Republic of Croatia | Mayor majority | Proportion of deputies in the local council supporting the mayor | Glaurdić and Vuković (2017); source of original data: State Electoral Commission of the Republic of Croatia | Unemployment | Average monthly unemployment rate | Croatian Unemployment Bureau | Income | Income per capita (deflated to 2010 levels), natural log transformed | Croatian Bureau of Statistics | Grants | Per capita grants (deflated to 2010 levels), natural log transformed | Ministry of Finance | Croats | Proportion of population ethnically Croatian | Croatian Bureau of Statistics | Average age | Average age for entire population in municipality | Croatian Bureau of Statistics | Years of education | Average years of education for population older than 15 years of age | Croatian Bureau of Statistics | War disabled per 1000 | Number of disabled persons whose disability was caused by war | Glaurdić and Vuković (2016); source of original data: Croatian Bureau of Statistics | Settlement size | Weighted average of settlement size in a municipality as a measure of urban-rural cleavage | Glaurdić and Vuković (2016); source of original data: Croatian Bureau of Statistics | Assembly size | Total number of representatives in the local assembly | State Electoral Commission of the Republic of Croatia |
Table 1Summary statistics DISPLAY Table
Table 2Impact of winning coalition size on tenure in power DISPLAY Table
Table 3Impact of winning coalition size on corruption DISPLAY Table
Graph 1Smaller winning coalition size environments on average exhibit greater levels of corruption DISPLAY Graph
Table 4Impact of corruption on re-election in small and large winning coalitions DISPLAY Table
Graph 2A small winning coalition environment has on average higher local tax rates DISPLAY Graph
Table 5Impact of winning coalition size on local tax rates DISPLAY Table
Graph 3A comparison of sample balancing using four different matching algorithms DISPLAY Graph
Table 6Comparing the results for different matching algorithms DISPLAY Table
Graph 4A balanced sample using entropy balancing (triangles), compared with an unbalanced sample (diamonds), and a sample matched using propensity score matching (circles) DISPLAY Graph
Table 7Entropy balancing DISPLAY Table
Table A1Variables used in analysis, their description and sources DISPLAY Table
* The author would like to thank to three anonymous referees for their valuable comments and suggestions.
The article was judged the best paper in the regular category in the annual competition for the Prof. Dr. Marijan Hanžeković Prize for 2017.
1 In Croatia this is the income tax surcharge rate ( prirez) that is used to finance the local government.
2 In reality, with multiple party races it takes much less than 50% of the electoral vote to win. A random parameter could have been used instead of the ½ threshold, however this would not have changed the theoretical prediction.
3 The index is calculated using the following formula, adapted from Glaurdić and Vuković (  2017) and originally proposed by Reynal-Querol’s ( 2002) ethnic polarization index:  πi, where πi represents the proportion of votes given to three political blocs: the conservative bloc led by the HDZ containing all parties that classify as right-wing, the social-democrat bloc led by SDP containing all parties that classify as left-wing, and a few unaffiliated parties and independent candidates. This variable measures to which extent party competition within a city or municipality was polarized between the two main left and right blocs.
4 This is the official definition of the Total index corruption proxy, according to Vuković ( 2017).
|
|
December, 2017
IV/2017
|
