Three essays in applied economics


Request a copy
Serval ID
PhD thesis: a PhD thesis.
Three essays in applied economics
Jametti M.
Von Ungern-Sternberg T.
Institution details
Université de Lausanne, Faculté des hautes études commerciales
Publication state
Issued date
Number of pages
REROID:R003699544; 30 cm; Old school value: Université de Lausanne
Introduction and Summary
This thesis is the output of four years of doctoral studies. The final task is to write a few lines to introduce summarize and present the chapters to follow.
The unifying theme of the thesis is applied economics. The idea behind each chapter was to focus attention to one (although maybe small) economic problem that exists somewhere in the "real world", use economic and econometric theory to understand the issue at hand and to provide, if possible, some ideas to improve the situation.
The thesis consists of three chapters. Chapter one looks at housing insurance in Switzerland, chapter two at natural disaster insurance in France, and chapter three at tax externalities in Switzerland.
In chapter one, I present an empirical application to housing insurance prices in Switzerland. In each canton (region) one of the two following systems is applied. Either housing insurance is provided by a publicly owned monopoly, or private insurance companies offer the insurance. A first-hand observation is that the private insurers work with substantially higher premiums than their monopoly counterparts. The question is to explain how much of this difference can be attributed to the differences in the systems (monopoly versus private). Several issues arise when dealing with this question. First, a commonly used measure, the claims - premium ratio, which measures the percentage of premium income that is spend on claims payments, is not suited for the analysis, because it does not take into account that some costs for the insurers, such as administrative costs and commissions, are independent of the level of claims. Second, insurance companies are likely to fix their premium rates according to some definition of "normal" claims, of which annual observations are a noisy measure. In the chapter a solution to this issue is proposed using techniques of measurement error. Taking into account these issues, one observes that the public insurance companies are about 20% more cost efficient than their private counterparts. This is mainly due to the fact that the public insurers work with significantly lower administrative costs. These cost savings amount to approximately 300 million Francs per year. Considering the important amounts of prevention expenditures of the public insurers, which could explain partly the important differences in claims levels between public and private regions, it is argued that this cost efficiency might actually be even more important.
The second chapter builds a theoretical model of the French natural disaster insurance, "cat. nat.". In recent years the discussion about how to adequately insure against natural disasters has become a much debated issue on the policy agenda internationally. One could think that it might be interesting to consider the French solution for reforms in other countries. The goal of the chapter is to show that this is not the case. The model takes up some institutional feature of the French system, such as the uniform (and independent of the risk of damage!) premium rate across the country, and the existence of a public reinsurance company, offering reinsurance to the private insurers at particularly favorable conditions. It is shown that this particular institutional setup is likely to lead to a "specialization" among the insurers, i.e. insurers serve either high or low risk regions, but not both together. This implies that the reinsurance company, offering service at a unique price, suffers from risk selection, ending up with a portfolio constituted of "bad" risks. The result of this outcome was that the reinsurance company had to be refinanced in 1999, together with a significant increase in premium rates. I show that increasing the premium rate also increases the tendency to risk selection and thus, improving the financial situation of the reinsurer comes at an important cost to the final customer, who also happens to be the tax payer.
The final chapter addresses the issue of tax externalities in an empirical application to Switzerland. The general view of tax externalities relies on the concept of "horizontal" tax externalities. These denote the effect of tax setting in a specific jurisdiction on its neighbors. In general, such externalities lead to what is known as a "race to the bottom", resulting in tax rates that are below the social optimum. More recently, it has been shown that within a federation, where jurisdictions of different hierarchical levels tax the same (mobile) base, the existence of "vertical" tax externalities can lead to a situation where tax rates are higher than the social optimum. Given that the two externalities work in different directions, it is interesting to investigate which one dominates in a real setting. In the chapter it is shown that an increase in fragmentation leads to lower tax rates if horizontal externalities dominate and to higher tax rates if vertical externalities dominate. This result constitutes a readily applicable test on the dominance of externalities. This test is applied to a sample of municipalities in Switzerland. The overall results are that horizontal externalities appear to dominate. However, there is a caveat one should consider. The model and the test it implies, is based on benevolent governments setting taxes. If one were to introduce different government objectives, such as revenue maximization (Leviathan), the test is biased towards the dominance of horizontal externalities. Assuming that governments with more direct-democratic implication in the tax setting process are more likely to behave benevolently, I find that the result of a dominance of horizontal externalities is driven mainly by (potential) Leviathan government objectives. Hence, fragmentation is a means to "tame the Leviathan".
Create date
19/11/2007 10:23
Last modification date
29/05/2020 10:33
Usage data