Final Report Summary - LOCAL PARTISAN TIES (Partisan Ties that Matter? Exploring the Advantages of National Incumbency for Local Governments)
The objective of the project is to understand electoral incumbency advantage that accrues to political parties from intergovernmental partisan ties. This question is analyzed in two unitary countries – Chile and Turkey – that fall into the upper middle income category for most of the period of analysis in this project. The main hypothesis of the project is that political parties use municipal resources to build effective local organizations when they control the local government and that the party’s local organization, in turn, benefits the political party electorally in national elections.
Data on municipal level local and national elections are gathered for both of countries, Chile and Turkey, in the post-transition period. These data are merged with socio-economic indicators, average income, for the case of Chile and, illiteracy rate, for the case of Turkey. Through OLS regression elections are analyzed separately for each main political party in both cases and for also two main electoral coalitions, Concertacion (leftist) and Alianza (rightist), in Chile. Compared to their performance in the previous elections, both Chilean electoral coalitions increased their vote shares in the municipalities that were controlled by the coalition parties in the three competitive elections following the regime transition in 1990, except for Alianza in the 1997 elections. Neither of the coalitions seem to have benefitted from their co-partisans’ control over the local government in the 2005 elections. In Turkey, when the elections are analyzed separately in a similar manner, in the 1999 legislative elections the pro-Islamic party of the period, FP (Virtue Party), and one of the center-right parties, DYP (True Path Party), and in the 2007 and 2002 legislative elections the center-left party, CHP (Republican People’s Party), benefitted from their partisan ties at the local level. Since financial resources are allocated in a nondiscretionary manner to local governments in Chile the national incumbent coalition was not expected to benefit disproportionately from its control over the local government. Even though Concertacion was in power for twenty years at the national level until 2010, as expected, such national incumbency did not benefit the Concertacion disproportionately over its main competitor, Alianza. In Turkey, in contrast, there are opportunities for discretionary allocation of resources to local governments. Therefore, it is possible for the national incumbent party to channel more resources to its co-partisans at the local level. Since more financial resources would help the party to build a stronger local organization, the national incumbent political parties would be expected to benefit more from their ties with the local governments. However, the results of the project from the district municipalities reveal that it is rather the main opposition parties, the FP in 1999 and the CHP in 2002 and 2007, that seem to enjoy incumbency advantage from intergovernmental partisan ties.
Although systematic budgetary data are not available for district municipalities in Turkey, municipal financial data exist for the Turkish metropolitan municipalities – the municipalities of large cities. In a co-authored paper with the former research assistant of this project, Elif Ozdemir, we examine metropolitan municipalities’ budgetary items in the period between 1997 and 2013. In this paper we focus more on the current national incumbent party, the AKP (Justice and Development Party) and whether intergovernmental partisan ties helped it to establish its dominance in the Turkish party system. As expected, we find that the municipalities that are controlled by the AKP received more discretionary resources. When we analyze municipalities’ spending choices we find that health expenditures played a crucial role. The AKP seemed to have effectively guided its co-partisans in the metropolitan municipalities to increase health expenditures in competitive districts. Higher municipal health expenditures, in turn, seem to have helped the AKP to increase its vote shares in the legislative elections. When we compare these results with the 1997-2002 period, prior to the AKP’s national incumbency, we find that former incumbent parties could not effectively enforce their local co-partisans to target competitive districts.
In another co-authored paper with Resat Bayer and Ece Kural, we try to understand whether stronger local party organizations that are built with municipal resources help to alleviate some of the negative consequences of national incumbency. Terrorist attacks that pose a challenge for security are clearly incidences that are expected to have such negative consequences. We compare the early post-transition period in Chile when anarchist and radical leftist groups carried out terrorist activities with the recent period in Turkey when Kurdish terrorist attacks formed a major challenge for the incumbent parties. We indeed find that municipal government control, most probably through strong local party organizations, helps the incumbent parties not to lose votes as a consequence of terrorism. As the first step to this research, Resat Bayer and I wrote a paper that analyzes the 2011 Turkish legislative elections in order to understand how terrorist attacks affect right-wing parties’ vote shares. Although previous literature argues that terrorist attacks benefit right-wing parties, we find that a right-wing party is hurt by terrorist attacks when it is the national incumbent party.
Other co-authored work with Resat Bayer examines whether political parties benefit from the allocation of central government resources more in the municipalities that they control. We find that higher number of green cards (free health care benefits) allocated to a municipality increases the vote share of the AKP, but only when the party also controls the local government. Again the mechanism through which this advantage works seems to be the strong local party organizations that the AKP builds through municipal resources. Since the finding that national incumbent party benefits from the distribution of green cards more in the local municipalities that it controls suggests that national incumbent parties should, then, allocate more resources to the municipalities that they control, we also examine the allocation of green cards and find support for this expectation in an analysis of green cards per capita through regression discontinuity design (RDD). These results reveal a challenge for the distributive politics literature where the core versus swing voter debate is dominant. Since the districts where core supporters of party are densely located tend to elect the party’s candidate for the mayor, disentangling the effect of core supporters from local government control poses a challenge for the empirical analysis. RDD helps solve some of this problem by analyzing the impact of local incumbency in competitive elections and treating the results of elections in competitive districts as random.
The final paper, co-authored with Selim Erdem Aytac and Ali Carkoglu, merges municipal electoral data with survey data from the most recent three legislative elections in Turkey and analyzes the electoral campaign activities of locally incumbent parties. We find that respondents who attend the party’s rallies and who live in districts where the party’s vote share is high are contacted more by local incumbent parties during campaigns. These results suggest that local party organizations target constituencies that are already close to them when they contact voters with personal visits.
All of these results from works that analyze the effects of local and national incumbency in Chile and Turkey together suggest that we can find examples of incumbency advantage through intergovernmental partisan ties. Incumbency advantage, in turn, is a concern for democratic systems since it can lead to an uneven competition in the elections. Therefore, not only academics but policy makers and civil society organizations that are concerned about the quality of democratic systems should be cognizant of these findings.
Data on municipal level local and national elections are gathered for both of countries, Chile and Turkey, in the post-transition period. These data are merged with socio-economic indicators, average income, for the case of Chile and, illiteracy rate, for the case of Turkey. Through OLS regression elections are analyzed separately for each main political party in both cases and for also two main electoral coalitions, Concertacion (leftist) and Alianza (rightist), in Chile. Compared to their performance in the previous elections, both Chilean electoral coalitions increased their vote shares in the municipalities that were controlled by the coalition parties in the three competitive elections following the regime transition in 1990, except for Alianza in the 1997 elections. Neither of the coalitions seem to have benefitted from their co-partisans’ control over the local government in the 2005 elections. In Turkey, when the elections are analyzed separately in a similar manner, in the 1999 legislative elections the pro-Islamic party of the period, FP (Virtue Party), and one of the center-right parties, DYP (True Path Party), and in the 2007 and 2002 legislative elections the center-left party, CHP (Republican People’s Party), benefitted from their partisan ties at the local level. Since financial resources are allocated in a nondiscretionary manner to local governments in Chile the national incumbent coalition was not expected to benefit disproportionately from its control over the local government. Even though Concertacion was in power for twenty years at the national level until 2010, as expected, such national incumbency did not benefit the Concertacion disproportionately over its main competitor, Alianza. In Turkey, in contrast, there are opportunities for discretionary allocation of resources to local governments. Therefore, it is possible for the national incumbent party to channel more resources to its co-partisans at the local level. Since more financial resources would help the party to build a stronger local organization, the national incumbent political parties would be expected to benefit more from their ties with the local governments. However, the results of the project from the district municipalities reveal that it is rather the main opposition parties, the FP in 1999 and the CHP in 2002 and 2007, that seem to enjoy incumbency advantage from intergovernmental partisan ties.
Although systematic budgetary data are not available for district municipalities in Turkey, municipal financial data exist for the Turkish metropolitan municipalities – the municipalities of large cities. In a co-authored paper with the former research assistant of this project, Elif Ozdemir, we examine metropolitan municipalities’ budgetary items in the period between 1997 and 2013. In this paper we focus more on the current national incumbent party, the AKP (Justice and Development Party) and whether intergovernmental partisan ties helped it to establish its dominance in the Turkish party system. As expected, we find that the municipalities that are controlled by the AKP received more discretionary resources. When we analyze municipalities’ spending choices we find that health expenditures played a crucial role. The AKP seemed to have effectively guided its co-partisans in the metropolitan municipalities to increase health expenditures in competitive districts. Higher municipal health expenditures, in turn, seem to have helped the AKP to increase its vote shares in the legislative elections. When we compare these results with the 1997-2002 period, prior to the AKP’s national incumbency, we find that former incumbent parties could not effectively enforce their local co-partisans to target competitive districts.
In another co-authored paper with Resat Bayer and Ece Kural, we try to understand whether stronger local party organizations that are built with municipal resources help to alleviate some of the negative consequences of national incumbency. Terrorist attacks that pose a challenge for security are clearly incidences that are expected to have such negative consequences. We compare the early post-transition period in Chile when anarchist and radical leftist groups carried out terrorist activities with the recent period in Turkey when Kurdish terrorist attacks formed a major challenge for the incumbent parties. We indeed find that municipal government control, most probably through strong local party organizations, helps the incumbent parties not to lose votes as a consequence of terrorism. As the first step to this research, Resat Bayer and I wrote a paper that analyzes the 2011 Turkish legislative elections in order to understand how terrorist attacks affect right-wing parties’ vote shares. Although previous literature argues that terrorist attacks benefit right-wing parties, we find that a right-wing party is hurt by terrorist attacks when it is the national incumbent party.
Other co-authored work with Resat Bayer examines whether political parties benefit from the allocation of central government resources more in the municipalities that they control. We find that higher number of green cards (free health care benefits) allocated to a municipality increases the vote share of the AKP, but only when the party also controls the local government. Again the mechanism through which this advantage works seems to be the strong local party organizations that the AKP builds through municipal resources. Since the finding that national incumbent party benefits from the distribution of green cards more in the local municipalities that it controls suggests that national incumbent parties should, then, allocate more resources to the municipalities that they control, we also examine the allocation of green cards and find support for this expectation in an analysis of green cards per capita through regression discontinuity design (RDD). These results reveal a challenge for the distributive politics literature where the core versus swing voter debate is dominant. Since the districts where core supporters of party are densely located tend to elect the party’s candidate for the mayor, disentangling the effect of core supporters from local government control poses a challenge for the empirical analysis. RDD helps solve some of this problem by analyzing the impact of local incumbency in competitive elections and treating the results of elections in competitive districts as random.
The final paper, co-authored with Selim Erdem Aytac and Ali Carkoglu, merges municipal electoral data with survey data from the most recent three legislative elections in Turkey and analyzes the electoral campaign activities of locally incumbent parties. We find that respondents who attend the party’s rallies and who live in districts where the party’s vote share is high are contacted more by local incumbent parties during campaigns. These results suggest that local party organizations target constituencies that are already close to them when they contact voters with personal visits.
All of these results from works that analyze the effects of local and national incumbency in Chile and Turkey together suggest that we can find examples of incumbency advantage through intergovernmental partisan ties. Incumbency advantage, in turn, is a concern for democratic systems since it can lead to an uneven competition in the elections. Therefore, not only academics but policy makers and civil society organizations that are concerned about the quality of democratic systems should be cognizant of these findings.