Essays on Voting and Government Inefficiency

Sammanfattning: This thesis consists of three self-contained essays on political economics.The first essay, The Effects on Political Rent-seeking of Allowing Campaign Contributions in Politics, presents a theoretical analysis of the effects on political rent-seeking of allowing campaign contributions in politics. I define rents as the difference between taxes and government spending. In a two-income group framework, I find that rents are higher when neither of the groups give campaign contributions, as compared to the case when both groups do. The mechanism behind this result is that candidates need campaign contributions in order to increase their probability of election success and voters will give campaign contributions to the candidate committed to the lowest level of rents. Rents are also found to be decreasing (increasing) in income inequality, when only the rich (poor) income group gives campaign contributions.The second essay, Government Inefficiency and Campaign Contributions, deals with campaign contributions and inefficiency in policy making. Intuitively, the interaction rests on a trade-off for political candidates between the need for campaign contributions (to win the election) and the effort associated with decreasing government inefficiency. Hence, voters will decrease the contributions to candidates likely to run a relatively less efficient policy. The main theoretical findings are that the inefficiency in the candidates' platforms is increasing in income inequality among voters, while their received campaign contributions are decreasing in income inequality. These predictions are empirically tested using US state level data for the period 1977-1995, and found to be statistically significant.The third essay, Explaining the Spending on Elderly Care in Swedish Municipalities Using Economic Theory, is, as indicated by the title, an empirical examination of local public spending in Sweden. This essay examines if the spending on elderly care in Swedish municipalities can be explained by the identity of the median voter using the voter characteristics (i) age, (ii) sex and (iii) income. The findings yield support for the median voter theory in the sense that the share of women in the population is found to negatively affect municipalities' spending on elderly care. On average, spending increases by 3-7 %, if the share of women in the population decreases by one standard deviation.