Tiebout and Crime. How crime affect the income tax capacity.

Despite the extensive literature on the relation between crime and migration, not much is known about how crime affects the tax capacity of local communities. This paper empirically investigates whether the Flemish local income tax base yield is sensitive to changes in the local crime level. The underlying assumptions are twofold. In a Tiebout world rational voters holding the local government accountable for the safety of its citizens move out when the local level of security gets too much alienated from what they want it to be (first assumption). If migration is due to crime, the more wealthy citizens are expected to move first. Looking for a place elsewhere implies transaction costs, which the more wealthy citizens are more likely to be able to afford. As a consequence, the average income per capita and so the income distribution will be affected, which in turn, will influence the local income tax base yield (second assumption). Ceteris paribus, the decreasing average income per capita, results in a lower local income tax base yield. In absence of a higher level governments' compensation, decreasing local tax revenues could prove to be disastrous for a crime-ridden municipality. When communities do not succeed in forcing back the number of offences, this can be the onset of a cumulative process of urban deterioration. A spatial panel data model containing several proxies for the local level of crime in 306 Flemish municipalities covering the period 2000-2014 is used to test the relation between crime and the local income tax base yield. When using aggregate crime rate statistics, the total number of offenses per capita, no significant results were found. Yet when using the type specific crime statistics, significant results do were found. By regressing the specific types of crime and the control variable onto household mobility (per capita efflux- en influx of citizens) the first assumption was confirmed. Respectively, a negative and a positive link was found between the per capita efflux- and influx of citizens, and the change in the relative number of violent crime per capita. Results regarding the efflux of citizens were persistent up until a 1 year lag. To investigate the second assumption, a 2SLS panel data analysis was used. These results show that crime induced household mobility does indeed lower the income tax base yield and therefore confirms the second assumption. An increase in per capita violent crimes, will stimulate the per capita efflux of people, in turn lowering the local income tax base yield. A negative result was also found for the crime induced per capita efflux of citizens, although it was barely insignificant. By confirming both assumptions we present evidence for a relation between changes in local crime rates and the income tax base yield of a municipality, coming about thru household mobility

http://ift.tt/2nJi4Db