Taxation and Political Models

Abstract

This thesis examines the viability of financial transaction taxes (FTT) to raise large amounts of revenue for the government and if these taxes would have helped prevent the financial crisis. The history of transaction taxes are examined along with some plausibility analysis of transaction taxes in general. FTTs are found not to be a general panacea able to increase government revenue without significant costs. The revenue generated from transaction taxes are expected to drastically under-perform the amounts predicted by its advocates and lead to incentives for tax evasion.

The thesis looks at how including structural characteristics such as skewness and kurtosis of revenue may assist policy makers when choosing between different tax-policy regimes. Traditional economic analysis commonly concerns itself with minimisation of any dead-weight loss associated with any taxation regime. Commonly, this is done through analysis of the tax base and rate. Revenue shape characteristics (such as variance, skewness, and kurtosis) coupled with basic political realities also matter when selecting a tax policy regime. This chapter demonstrates how revenue characteristics can be translated into a policy relevant tool for policy makers.

Finally, the thesis investigates the different ways of modelling how wars are treated in political forecasting models. Traditionally, the involvement of a country in a war was either ignored entirely or simply noted through the use of a binary dummy variable. The analysis asserts that capturing war costs and looking at how long the war progresses over time capture relevant information for use in models. To illustrate this fact, dummy variable models are compared with time progression and war cost models. Dummy variable models are shown to be inferior to models which either capture a time trend or capture direct war costs.

Keywords: FTT, Revenue Estimation

JEL Classification: C13, C14, C33, C69, C82, F36, G28, G18, H29, H61, P26