|
Article Excerpt
Property tax assessment ratios--a property's assessed value divided by its market value--determine the effective tax rates paid by owners. When ratios in the same assessment district vary, the property tax will not be applied uniformly and the burden of financing local public services will be unevenly distributed. When ratios for similarly priced houses are not uniform, "horizontal inequities" exist. "Vertical inequities" exist when assessment ratios are not consistent across a range of values. The tax can be considered "regressive" if lower--valued houses are assessed at a higher proportion of their market value, or "progressive" in the opposite case.
Researchers have found much evidence of both kinds of inequities. However they have not provided adequate explanations for their findings.
We propose that one source of variation within price strata--namely the relatively sparse number of comparable sales in certain neighborhoods--may also explain some of the tendency toward relatively high assessment ratios for lower-priced properties. Our goal is to test the notion that assessment ratios will be not only less uniform but also higher in areas where prices are relatively unpredictable and lower.
We believe this might be the case because in markets with few sales, relevant information cannot be incorporated quickly into either a property's market price or, by association, its assessed valuation. Therefore, assessments in inactive markets may vary for nearly identical properties in the same price strata and may well be biased (i.e., horizontal inequities). Because sales of both very low- and very high-value properties are rare, assessment ratios for such properties are likely to be more error-prone and more variable. If only owners of the high-value properties appeal their tax bills, however, a taxing jurisdiction may end up with a more regressive distribution of assessment ratios. In other words, owners of lower-valued homes would pay a relatively higher price for public services.
We test our hypotheses using sales and assessment data for small residential properties from the City of Chicago. The results of our models provide evidence that sales density has a significant effect on horizontal equity. Having fewer sales in a census tract increases the probability of having both very high and very low assessment ratios. We find much evidence that Chicago's distribution of assessment ratios is regressive. Owners of lower-valued homes pay a higher effective tax rate because their assessments constitute a higher share of their homes' market values. However, sales frequency is not the primary cause of this kind of vertical inequity. When we include the sales frequency variable, we find that the magnitude of regressivity is reduced but not eliminated.
INTRODUCTION
Property tax assessment ratios--a property's assessed value divided by its market value--determine the effective tax rates paid by owners and, therefore, the burden of financing local public services. When ratios in the same assessment district vary, the property tax will not be applied uniformly. Some amount of variation is to be expected and is, on occasion, the result of an intentional policy, such as assessing different property types (e.g., industrial, residential) at different rates. Other types of non-uniformity are potentially more problematic. In particular, researchers have explored whether the property tax burden is shared fairly by taxpayers within and across different housing price strata. During a decade of steadily increasing tax bills, the distribution of the property tax has significant implications for homeowners' standard of living.
When ratios for similarly priced houses are not uniform, horizontal inequities exist. For example, assessment ratios may vary systematically with a property's location (Haurin, 1988; Allen and Dare, 2002), size (Berry and Bednarz, 1975), or age (Goolsby, 1997). In the last case, for example, owners of houses that are almost identical with the exception of the building's age were found to pay different effective tax rates for the same bundle of services.
Vertical inequities occur when assessment ratios are not consistent across a range of values. The tax can be considered "regressive" if lower-valued houses are assessed at a higher proportion of their market value, or "progressive" in the opposite case. The tendency for higher-priced properties to receive lower assessment ratios has been documented by Baar (1982); Bell (1984); Black (1977); Clapp (1990); Cornia and Slade (2005); Engle (1975); Haurin (1988); Ihlanfeldt (1982); Paglin and Fogarty (1972); and Sirmans, Diskin and Friday (1995). Such findings are taken as evidence of the regressivity of the property tax system since higher assessment ratios lead to higher effective tax rates when other factors are held constant. (1)
Assessment studies provide extensive documentation of the tendency for assessment ratios to vary within and between different price segments of the same market, but they have provided little explanation of such findings. A lack of horizontal uniformity has been attributed to the limited information about relevant property and neighborhood factors available to buyers, sellers, and assessors (Allen and Dare, 2002; Goolsby, 1997). In such cases, more sophisticated appraisal techniques, such as interior inspections and mapping, appear to help assessors overcome such information deficiencies and reduce errors (Bowman and Mikesell, 1978; Bowman and Butcher, 1986). The primary explanation put forth for the potentially more serious problem of vertical inequity is that higher-priced properties may appreciate more quickly relative to the natural lag in assessments, which only change at fixed intervals (Mikesell, 1980; Clapp, 1990; Ihlanfeldt, 2004).
While we concur with these technical explanations for assessment inequities, we propose that one source of variation within price strata--namely the relatively sparse number of comparable sales in certain neighborhoods--may also explain some of the tendency toward relatively high assessment ratios for lower-priced properties. In other words, we ask whether the root cause of horizontal and vertical assessment inequities is one and the same. Our goal is to test the notion that assessment ratios will be not only less uniform but also higher in areas where prices are relatively unpredictable. In a thin market with few sales, relevant information cannot be incorporated quickly into either a property's market price or, by association, its assessed valuation. Therefore, assessments of land and improvements in inactive markets may vary significantly for nearly identical properties in the same price strata and may well be biased systematically (i.e., horizontal inequities). Because sales of both very low- and very high-value properties are rare, assessment ratios for such properties are likely to be more error-prone and more variable. If only owners of the high-value properties appeal their tax bills, however, a more regressive distribution of assessment ratios may result.
Using assessment data from Chicago for small (six units or fewer) residential properties in 2003 and sale prices for 2004-2005, we test whether assessment uniformity and regressivity are influenced by the number of nearby sales. Chicago (as well as the rest of Cook County) has an unusually complicated classified property tax system with a statutory assessment ratio of 16 percent of market value for residential properties. In practice, the average assessment ratio is less than ten percent. Moreover, we demonstrate that the city's assessments are more variable and regressive than is considered acceptable by the International Association of Assessing Officers. The city is home to a variety of housing markets, some of which are very active while others are considerably more dormant. In this environment of routine under-assessment and high variability of assessment ratios, homeowners in Chicago's thin markets may find it difficult to gather the information required to appeal and reduce an assessment that is higher than expected yet still lower than the statutory ratio.
We replicate the traditional model to test for vertical equity, which regresses assessment ratios on sales prices. In addition, we include a variable that measures market activity by the number of residential sales in each Chicago census tract in which a property is located. Adding our new sales frequency variable to the model does not significantly reduce the degree of regressivity indicated by the coefficient for sales price, which changes very little when the number of residential sales is added to the regression. However we do find some evidence of horizontal inequities: assessment ratios tend to be higher in areas with few sales.
Our emphasis on assessment variability is not adequately captured by the traditional regression of assessment ratios on sales prices. A regression estimates a conditional expectation: what is the increase in the expected assessment ratio as sales prices increase? In contrast, our concern is with the entire distribution: as sales prices and sales frequencies increase, are assessment ratios more tightly clustered around a central value? As such, we propose a multinomial logit model to characterize the relationship between the assessment ratios and both sales prices and sales frequencies. The three-choice multinomial logit model estimates the probability that an assessment ratio is in either the bottom or top 25 percent of the distribution relative to the middle 25 percent-75 percent. We find that having more sales in a census tract decreases the probability of having both high and low assessment ratios, which provides additional evidence of horizontal inequities. However, adding the sales frequency variable to the logit model, again, reduces but does not eliminate the finding of regressivity.
These results suggest that the traditional assessment ratio study...
|
