We empirically investigate effects of political economy factors of local governments on their public debt employing a panel dataset for Japanese cities. We mainly examine weak government hypothesis that government fragmentation leads to higher public deficits and debt. But, we also examine whether political business cycle exists in local governments or not. By using datasets of all cities and cities with population of a hundred thousand, we also analyze if size of city influences impact of these political factors. Our results are as follows. First, we find that party fragmentation leads to the greater fiscal deficit of Japanese cities. If the seat share of ruling party (coalition) in the council is small or the number of parties in ruling coalition is great, fiscal deficit increases. Especially, they have a robust effect on the deficit in cities with more than a hundred thousand people. In addition, if a city had a mayor who had received recommendation from more than two parties in the last election, fiscal deficit of it would expand. Second, we find that strong political leadership might contribute to cutting fiscal deficit of local governments. If a mayor in a city with a hundred thousand people got higher percentage of votes, public debt would decrease more and bond would be issued less in a year. If the mayor in the city won more times, bond would be issued less by the city. We also have findings on political business cycle and effects of city size. We do not have evidence on political business cycle for all cities but have for cities with a hundred thousand people. Judging from our results on this and weak government hypothesis, population size of city is very likely to influence effect of political factors on local pubic fiscal debt and policy.

抄録全体を表示

This paper focuses on candidates’ policy positioning mechanism and its impact on vote fluctuation in the Japanese Lower House elections. In recent years, a serious fiscal crisis sometimes hits some democratic countries like Japan, the United States, and Greece. With regard to the crisis, Buchanan and Tullock have argued that contra-cyclical policies and rentseeking behaviors cause such fiscal problem in democratic countries. Then this paper analyzes the electoral relationship between whole voters and the candidates’ behaviors aiming at winning their seats and argues about the background of expansion and inefficiency of the Japanese government spending. Concretely, this paper has two kinds of analyses. One is to demonstrate the causal relationship between candidates’ position shifts (budget increase decrease on 13 policy areas) between the current and previous elections and its determinants the result of previous election (this is based on the Dynamic Responsiveness Theory / Fowler 2005), the gap of candidates’ numbers between the elections (Powell 1982), the number of successful elections (Instead of Incumbency / Powell 1982 Burden 2004), the candidates’ attributes, the urbanization level of each SMD, and party affiliation. The purpose of the second analysis is to investigate the effect of candidates’ positioning on their vote fluctuation. According to the results, the candidates who won in the previous election and have experienced party elites and local politicians tend to shift their positions to budget increase for the purpose of reelection. Moreover, candidates are inclined to take position on budget increase in case that they compete with more candidates in their districts than ever. DPJ candidates are more likely to shift than LDP. On the other hand, most voters select candidates not being affected by the candidates’ repositioning oriented toward budget increase. To summarize, voters do not always seek their self-interests when they vote, and to be blame may be the candidates who have shifted their positions toward an increase in order to get more votes. That is, representative democracy does not always cause the fiscal crisis in Japan. In addition, this paper also suggests that the Dynamic Responsiveness Theory works in Japan.

抄録全体を表示