The Incentive Effect of Equalization Grants on Tax Collection

If fiscal capacity were measured accurately, such transfers would not create a disincentive for local governments to raise revenues, because at the margin local governments still bear full fiscal responsibility for expenditure and tax decisions, essentially because transfers are lump sum in nature (Smart 2007).

The FCM, in turn, is self-financing: municipalities must contribute a share of their main sources of local revenue to the FCM every year.

The theoretical literature on the impact of equalization grants takes into account that consumer and producer mobility in response to fiscal incentives is a distinctive characteristic, especially when jurisdictions have fiscal autonomy and the tax rate varies among them; hence, this literature is closely related to tax competition literature (Wilson 1999; Koethenbuerger 2002; Smart 1998)

If I assume an additively separable utility function for private and public goods, the results do not change.

I assume that the budget constraint is satisfied with equality because in Chile municipalities may not take on debt, although some Chilean municipalities run large deficits. Even if I include debt as a third source of revenue, the incentive effect still exists.

Note that due to the dynamic structure of the model, it is possible to separate the implicit tax, which must be anticipated by the local government, from the contribution rate () which is contemporaneous, unlike what has been done previously. In a static model, this is not possible (see Appendix A).

Moreover, in 2002 and 2006 the update was delayed by one year. The municipalities were only informed of this decision at the end of 2001 and 2005, respectively. This kind of reform changes the number of years between revenue collection and the update of information for the remaining years.

Since 2007, the information used in the distribution formula has been updated every year and the revenue component considers the IPP per capita measured two years before the updating year. So if the updating year is t, the grant in year t considers the IPP per capita of the year t-2. Thus, since 2007 every year has been relevant for the distribution formula and the number of years between revenue collection and the update of information is always two years. Finally, the weight of the equalizing component is 35%. I do not consider 2007 because a sophisticated compensation mechanism was established that year for the municipalities that received a lower grant than in the previous year.

For instance, during most of 2001, municipalities were aware that the information would be updated in 2002 and that therefore, collected revenue in 2001 would have an impact from 2005 to 2007 (that is, in four years’ time). Since the update was delayed by one year, which was known at the end of 2001, the collected revenue in 2002 also affected the grant distribution after four years. The collected revenue in 2001 ultimately affected the grant distribution in 2003, which was known ex-post, and that is why it is not considered in the diagram.

For instance, for municipalities with IPP per capita closer to the national average, the part of the FCM grant associated with the IPP component is quite small in comparison to municipal revenue. However, because the per capita IPP of those municipalities is near the national average, the rate at which their future grants decrease due to increased collection is also high.

To calculate the relative importance of the part of the FCM grant associated with the IPP component with respect to municipal revenue, I use the average data for the 1990–2006 period.

For instance, for 1996, I consider the collected revenue that year and the implicit tax is calculated with the average IPP per capita of 1992, 1993 and 1994. There is another difference between per capita local revenue and the IPP per capita used in the distribution formula: the way the population is measured. For per capita local revenue, the projected population over the 1990–2006 period was considered, based on censuses through 2002, whereas the IPPP per capita includes the projected population based on the latest censuses, which changes depending on the updating year.

Five municipalities were not considered in the panel because their data present serious measurement errors.

FNDR is an important vertical grant source used to finance investment projects

Business equity is the tax base for business taxes.

For the years in which this information is not available, the value of the preceding year was assigned. Municipalities for which this information was not available were assigned the regional poverty data.

This threshold was defined on the basis of the average proportion of municipalities in which the political coalition to which the mayor belongs was reelected and the distribution of the predicted likelihood of paying tax in every year. That is, on average, the proportion of municipalities in which the political coalition to which the incumbent mayor belongs was reelected is close to 60%. Therefore then I set the threshold at the predicted likelihood that corresponds to the centile 40 which is, on average, close to 70%.

The effect measured in standard deviations is calculated using the descriptive statistics of Table 3, i.e., considering municipalities that pay a positive implicit tax.

If I consider that this kind of revenue does not depend on fiscal effort because it is determined by subsidy amount per pupil (defined at the national level) and the average attendance of students enrolled in municipal schools (although lately municipal authorities have made efforts to encourage student attendance and thus increase education resources), the estimated coefficients will represent the compensated (Marshallian) incentive effect.

Since there are different sources of revenue (property tax, business tax, vehicle tax) I could use different measures of fiscal capacity, but the only data available was the proportion of exempt properties and business equity for each municipality.

However, there could be a substitution effect between collected revenue and education resources when the cost of efforts to collect in relation to the effort to increase attendance is higher.