The analysis of real estate taxes across states is an important source of information for estimating economic impact of housing on local economies, comparing affordability of housing and cost of living in general, and making decisions about relocating or retiring in a different state. What complicates such cross-country comparisons is the fact that property taxes are imposed by multiple local jurisdictions that follow different assessment, administration, and reporting procedures. In addition, 37 states also collect property taxes on the state level. As a result it is often difficult to compare effective tax rates on residential real estate based on state and local government data.
The decennial Census is one source of data on residential real estate taxes across the country since the Census long form asks homeowners to report their overall annual real estate taxes. At NAHB’s request, the Census Bureau expanded the coverage of the aggregate property taxes paid by owners to include more detailed level of geography for the 2000 Census. This made it possible to use the decennial Census to study property taxes across counties and metropolitan areas. The American Community Survey (ACS) is designed to replace the Census long form to provide more timely data. Prior to 2005, the ACS did not report property tax information for detailed geography. However, much of the geographic detail was restored in 2005 (again, at the request of NAHB).
NAHB previously reported the 2000 real estate tax rates in the Housing Economics article “
Property Taxes in the 2000 Census” (PDF) —Paul Emrath, December 2002. Since that time substantial changes have taken place in housing markets in many parts of the country. For example, some markets experienced double digit home value increases. It may have been difficult for tax assessments to keep pace with these increases. Some jurisdictions also passed property tax limits. For these reasons, residential property tax rates may have changed considerably since 2000. This article presents new estimates of property taxes across US states and counties based on the 2005 ACS data.
State Residential Property Taxes per Home
Table 1 presents the median home values, the median real estate taxes per home, tax rates per $1,000 of property value for all states and the District of Columbia in 2005, as well as the state tax rankings and property tax revenue in the state as a percent of all revenue from state and local sources. The state with the highest median real estate taxes is New Jersey where more than fifty percent of all households pay in excess of $5,352 in property taxes per home. This top position cannot be simply attributed to high home values in New Jersey, since states like California, Hawaii, District of Columbia, and Massachusetts have higher home values but on average collect significantly smaller real estate taxes per property. High residential property taxes in New Jersey are even more striking considering that a state with the second highest property taxes, New Hampshire, on average collects $1,432 less per home.
At the other end of the spectrum states in the South Census Region, such as Arkansas, Mississippi, West Virginia, Alabama, Louisiana, where median real estate taxes do not even reach $500. In Louisiana, half of all households pay less than $175 in taxes per home.
The Census Bureau’s Annual Survey of State and Local (S&L) Government Finances helps explain some of these sizeable cross-country differences in the amount of real estate taxes collected by S&L governments per home. The last column of table 1 shows statewide property taxes as a share of S&L government own-source revenue (own-source revenue excludes revenue from government owned water, gas, electric, transit utilities, and liquor stores), based on data from the most recently available 2004 survey.
Overall, property taxes in the US accounted for about 22 percent of S&L government revenue. By contrast, state and local governments in New Jersey and New Hampshire rely on property taxes more than governments in any other state. New Hampshire, a state that does not tax wages and salaries, derives almost 43 percent of statewide government own-source revenue from property taxes. New Jersey derives more than 35 percent. All six New England states are among the ten states most dependent on property taxes as a source of government revenue. On the other hand, Delaware, New Mexico and southern states such as Alabama, Arkansas, Louisiana, derive no more that 9-11 percent of statewide revenue from property taxes.
A simple correlation analysis confirms that the extent to which state and local governments rely on property taxes to fund local services greatly affects the amount of real estate taxes they collect per home. The correlation between median prope
rty taxes per home and statewide property taxes as a share of S&L government revenue is 0.81. At the same time, the correlation between state median RET and home values is only 0.52.
[1]
Property Tax Rates across States
Since state property taxes per home to a certain extent reflect median home values, it is often more meaningful to compare state tax rates per $1,000 of property value. These numbers are presented in the fourth column of Table 1. Louisiana with its generous homestead exemption has the lowest real estate tax rates in the nation, $1.72 per $1,000 of home value. Two states with the highest effective rates are Wisconsin and Texas, where rates exceed $18 per $1,000 of property value. These numbers and the color-coded map (figure 1) once again highlight the long established tradition of southern states, with the exception of Texas, to rely less heavily on real estate taxes.
Figure1
It is worth noting that median home values in the highest-rate states (Wisconsin and Texas) are well below the national average. At the same time, the three states with the most expensive homes (California, Hawaii and the District of Columbia) have some of the lowest property tax rates in the nation, $4.77, $2.04, $3.76 per $1,000 of value respectively. A simple correlation analysis confirms a negative relationship between house values and property tax rates (the correlation is -0.14). Academic literature offers several compelling arguments to explain this relationship: low property v
alues drive state and local government to raise tax rates to boost government revenues; poorer neighborhoods require more public services that in turn dictate higher tax rates; when capitalized, lower tax rates boost home values.
[2]
Differences within States
Table 2 presents counties with the highest and lowest median property tax rates in each state, as well as the ratio between the highest and lowest rates, the median home values in these counties, and the correlation between county tax rates and property values.
The county with the highest effective median tax rate is Niagara, New York with a rate of $28.12 per $1,000 of value. New York also stands out by having the most diverse tax rates across the state. Kings County, the lowest-rate NY county, has a tax rate of $4.79 per $1,000 of value that is almost six times lower than in Niagara, the highest-rate county. By comparison, among states where the ACS covers more than two counties, Nebraska has the most consistent tax rates across its counties. St. Bernard Parish, Louisiana, Maui, Hawaii and Apache County, Arizona have the lowest median tax rates among all counties, $1.35, $1.62, $1.64 per $1,000 of value respectively. Hawaii has consistently low tax rates across its three counties; its high-low ratio is only 1.38. In contrast, Arizona is noticeable for high variability; its highest rate is almost five times higher than its lowest county rate.
The negative correlation between median property tax rates and median home values (-0.19) is stronger across counties than across states, but there are numerous states within which the correlation is actually positive, probably reflecting presence of metro counties in these states that traditionally require more state and local services and have higher home values.
For interested readers, NAHB also provides a complete list of real estate tax rates for all counties and metropolitan statistical areas on its website.
_______________________
Fotnotes:
[1] Corr
elation indicates the strength and direction of a linear relationship between two variables and ranges from -1 to 1. Larger correlation numbers (i.e., further away from zero) mean that variables tend to move closer together in the same (positive correlation) or opposite direction (negative correlation). Zero correlation means that changes in two variables are completely independent or unrelated to each other.