The Tiebout Model
The bedrock of this paper is a model put forward by Charles Tiebout in 1956, in his seminal paper "A Pure Theory of Local Expenditures", that has been known since then as the Tiebout Model. This model suggests that, given costless mobility and perfect information on the tax rates and the quality of services provided by localities, residents will allocate themselves in communities that best suit their preferences. This will result with an optimal market type solution for the provision of local public goods with competition between localities offering different bundles of services.
From a normative perspective, the Tiebout Model is often regarded as positing a serious problem of growing inequalities. Since localities are incentivized to attract wealthier residents, that can broaden the local tax base, as well as to deter lower-income residents, a competition between communities will result in an endless process of the poor chasing the rich. Other criticisms of the Tiebout Model, from a positivist perspective, are raising doubts over its relevance to the actual reality of local government, since the model calls for a rigorous set of assumptions in order to apply.
This research aims to empirically test the claim that the Tiebout Model causes this normatively undesirable process of the poor chasing the rich. The research question of this paper is as follows: How does greater municipal autonomy influence local expenditures on welfare? Given greater municipal autonomy (that is equivalent to a greater degree of adherence to the Tiebout Model), it is interesting to test whether localities are ignoring the needs of their low-income residents by minimizing their expenditures on welfare. Another important factor that will be taken into account is the demand for local welfare services. Even more important, the interaction effect between municipal autonomy and demand for local welfare services will also be tested. The answer to the research question of this paper will be provided through quantitative research of panel data showcasing 74 cities in Israel between the years 2010-2014.
Municipalities in Israel are categorized by the Central Bureau of Statistics according to their socio-economic level on a scale of 1-10 (the lowest socio-economic level is marked 1 and the highest socio-economic level is marked 10). This scale reflects the wealth of residents, taking into account the local unemployment rate, average salaries, car ownership and other economic factors.
Evident from Figure 1 is the varying levels of auto-investment rate (AIR) which is used here as a proxy of municipal autonomy. Wealthier cities are unsurprisingly more independent in the sense that they have more money that is allocated at their discretion. The exception to this claim is that cities in the lowest socio-economic level also boast a high auto-investment rate. This situation is explained by the fact that allocations from the Ministry of Welfare are supplied under a matching agreement, that the local authority is obligated to provide for 25% of the welfare expenditures. The poorest cities cannot even afford this partial participation in welfare expenditures and therefore cannot receive the earmarked allocations from the Ministry of Welfare.
The Interaction Effect
The results of this research show that as localities enjoy more autonomy, they are more sensitive to the needs of their residents. This finding is in alignment with the Tiebout Model, since the competition between localities incentives local leaders to be sensitive to the needs of the citizenry. The ability to move freely between localities is the driving force of the competition between localities. This does not allow local leaders to choose bundles of local services that are not in demand.
Figure 2 is a visualization of the pooled OLS regression model, showing the interaction between the auto-investment rate and 4 subsets of the demand for welfare services (MW variable). When the demand for welfare services is low (MW closer to 0, representing less people who are earning up to the minimum wage) the line slope is negative, but as the demand for welfare services increases, the slope changes direction and becomes positive. This visualizes the interaction between the demand for welfare services and municipal autonomy causing a different impact on the share of welfare in the local expense budget.
Why do the main effects, when tested independently, have a negative impact on the welfare rate? It is likely that wealthier localities exhibit higher degrees of municipal autonomy while on the other hand they have fewer residents who require social services. It is also likely that poorer localities are the exact opposite - their tax base is smaller and therefore they allocate fewer funds at their discretion while they also have more low-income residents in need of social services. From a normative perspective, this situation seems inherently flawed, since wealthier municipalities can offer better welfare services, but there is no demand for such services, while residents in poorer municipalities who have great desire for welfare services cannot receive them because the tax base is not wide enough to have funds that can be autonomously allocated. The interaction between the two independent variables contradicts this undesirable impression. Wealthier localities with greater municipal autonomy and greater demand for welfare services allocate a greater share of the expenditures on services to these services. In a given municipality, one unit increase of the interaction variable is projected to add 1.6-2.3 percentage points to the municipality welfare rate.
A Path for Urban Welfare
In conclusion - firstly, we found no empirical support for the poor chasing the rich hypothesis. Secondly, local politicians can use these findings to advocate for greater decentralization of spending powers. Lastly, this research showed that lower-income residents are better served in cities that exhibit greater municipal autonomy and greater demand for welfare services.
My goal is to continue to measure cities' behavior with quantitative tools and investigate urban problems to offer data-driven solutions.