Please summary this article and  follow the requirements below：At least two pages，Need to answer all the questions provided：Critical Review Summary TemplateCRITICAL REVIEW1. Full APA citation o

r2009 Public Financial Publications, Inc. A Reconnaissance of Alternative Measures of Effective Property Tax Rates MICHAEL E. BELL and CHARLOTTE KIRSCHNER Informed policy discussions on local property tax reform initiatives require accurate and reliable information about the tax’s current use. This paper reviews two popular measures of property tax burden and their limitations. The paper then reports the results of a reconnaissance of measures of effective property tax rates and how they vary across the 50 states. The paper also reports the findings of a 50-state survey that indicates how states calculated and reported effective property tax rates. These comparisons lead to a suggested framework for certain methodologies for calculating effective property tax rates. INTRODUCTION The property tax generated nearly US$360 billion for state and local governments in 2006, making it the largest single revenue source for state and local governments. In- Michael E. Bell, Research Professor, George Washington Institute of Public Policy, George Washington University, P.O. Box 869, McHenry, MD 21541. Dr. Michael Bell is a Research Professor at the George Washington University Institute for Public Policy, State and Local Fiscal Policy Research Program. He is also Executive Director of the Coalition for Effective Local Democracy, a nonprofit working to strengthen local democratic governance. Dr. Bell’s background is in public finance, with a specific focus on state and local finances and intergovernmental relations. Dr. Bell was Principal Research Scientist at the Johns Hopkins University Institute for Policy Studies and taught in the Institute’s Masters in Policy Studies Program (MAPS). Dr. Bell has edited seven books and published articles in several journals including National Tax Journal, Public Finance, Urban Studies, Journal of Urban Economics, Environment and Plan- ning C: Government and Policy,andPublic Budgeting and Finance.

Charlotte Kirschner, Research Associate, George Washington Institute of Public Policy, George Wash- ington University, 805 21st Street NW, Washington, DC 20052. Charlotte Kirschner is a Ph.D. candidate in Public Policy concentrating in national security with a specific interest in homeland security and terrorism.

Ms. Kirschner’s dissertation research focuses on the impacts of the homeland security grants on state and local public safety spending and preparedness. She works as a Research Associate for the George Wash- ington Institute of Public Policy on the ‘‘Significant Features of the Property Tax’’ project. Previously, Ms.

Kirschner worked for the Administrative Offices of the Pennsylvania Supreme Court conducting macro- level policy research on domestic violence, juvenile court justice, and court house security, and managing the state’s caseload statistics system.

Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 111 creased reliance on residential property to generate property tax revenues coupled with soaring property values between 2000 and 2007, intensified pressure for local officials to respond to concerns about higher property taxes. In 2007, three statesFIndiana, Geor- gia and FloridaFconsidered proposals to eliminate the property tax, especially the portion that falls on owner-occupied residential properties. Many more states have im- plemented various types of limitations intended to curb residential property tax bills. 1 The result has been erosion of the property tax as a source of local revenue through a variety of devices designed to reduce residential property tax burdens (e.g., assessment limits, homestead exemptions, limitations on tax rates or total levies, etc.). In many cases, state and local governments have contributed to the erosion by reducing the property tax base through exemptions and establishing mechanisms that reduce the tax on specific classes of property or for specific classes of owners. 2 In order for decision makers and the general public to have an informed debate about property taxes there is a need for accurate and reliable information. Two types of in- formation would be particularly useful. First, it is important to know what sort of burden property taxes impose on taxpayers. Such burden measures are generated by dividing property taxes paid by some measure of income. This measure provides infor- mation on the share of income devoted to paying property taxes, how that share varies across communities or states and how it varies within communities across households.

Second, it is important to know how heavily the property tax is actually being used.

Such a measure would generally present taxes paid relative to the base of the tax. In the case of the property tax this would involve multiplying the statutory, or nominal, tax rate by the assessed value which is the base of the property tax. Typically, market value is the intended base of the property tax and assessed values are intended to reflect market values. In practice, the ability of local jurisdictions to reflect market values by assessed values varies across states, across jurisdictions in the same state and across individual parcels within land use types within a single jurisdiction. As a result, simply multiplying the statutory tax rate by assessed values will not always generate reliable estimates of the actual effective property tax rate for individual properties.

An alternative approach would be to calculate directly the effective property tax rate in a community which considers the level of taxes paid relative to the actual base of the tax. Because fractional assessment, homestead exemptions, assessment-increase caps, and other measures cause the statutory base to differ from the true base, effective prop- 1. For more information on such limitations, see David Brunori et al., ‘‘Tax Expenditure Limitations and Their Effects on Local Finances and Urban Areas,’’ inUrban and Regional Policy and Its Effects, eds.

Margery Austin Turner, Howard Wial, and Harold Wolman (Washington, DC: Brookings Institution Press, 2008), 109–154 and Bing Yuan et al., ‘‘Tax and Expenditure Limitations and Local Public Finances’’ inErosion of the Property Tax Base: Trends, Causes and Consequences, eds. Nancy Y. Augustine et al.

(Cambridge, MA: Lincoln Institute of Land Policy, 2009), 149–191.

2. For a fuller discussion of efforts to reduce residential property tax burdens, see Nancy Y. Augustine et al., eds.Erosion of the Property Tax Base: Trends, Causes and Consequences(Cambridge, MA: Lincoln Institute of Land Policy, 2009). Public Budgeting & Finance / Summer 2009 112 erty tax rates typically are calculated by dividing property taxes paid by the market value of a property. Such a measure provides information on whether or not property tax rates are actually high in a jurisdiction, how they compare with property tax rates in neigh- boring jurisdictions or states, how they vary across different land uses within a juris- diction and whether or not property tax rates on residential property are out of line with other land uses or jurisdictions and should be reduced through one of the plethora of property tax relief tools available to decision makers.

Finally, to the extent the property tax serves as a benefits tax, the effective property tax rate reflects the ‘‘price’’ of locally provided goods and services to property owners.

This information is useful when a family or firm is choosing a jurisdiction in which to locate. Also, this information is useful in determining the level and quality of services received for a given price which allows the taxpayer to provide feedback to local decision makers.

The purpose of this paper is to provide a reconnaissance of available sources of information on effective property tax rates. This paper proceeds through three distinct sections. The opening section reviews two popular measures of property tax burdens often reported in the press which are typically part of any legislative debate on property tax issues. The subsequent section looks at how effective property tax rates are computed by organizations that publish them regularly. The final section reviews information on effective property tax rates available on state Web sites.

POPULAR MEASURES OF PROPERTY TAX BURDENS Two popular, and often reported, measures of property tax burdens that facilitate com- parison across states and jurisdictions within states are property taxes per capita and property taxes per US$1,000 of personal income.

Typically, these measures of property tax burden start with property tax collections as reported by the U.S. Census Bureau, which are defined to include taxes on all prop- ertyFboth real and personal. Specifically, the U.S. Census Bureau defines property to include real property (e.g., land and structures) as well as personal property (e.g., tan- gible property such as automobiles and intangible property such as bank accounts).

There is substantial variation across the 50 states in terms of how governments tax personal property. Table 1 contains information on which states tax three of the most commonly taxed types of personal property: motor vehicles, inventory, and machinery and equipment. As of 2006, eight states do not tax personal property at all. 3Nineteen states have provisions to tax motor vehicles as personal property; 18 states have pro- visions to tax inventories as personal property; and 40 states have provisions to tax machinery and equipment as personal property. 3. States that do not tax personal property are Delaware, Hawaii, Iowa, Illinois, North Dakota, New York, Pennsylvania, and South Dakota. Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 113 TABLE 1 States Allowing the Taxation of Personal Property in 2006 by Type of Personal Property StatePersonal property tax allowedMotor vehicles InventoryMachinery and equipment Alabama Yes Yes Yes Yes Alaska Yes No Yes Yes Arizona Yes No No Yes Arkansas Yes Yes Yes Yes California Yes No No Yes Colorado Yes No No Yes Connecticut Yes Yes No Yes Delaware No No No No Florida Yes No No Yes Georgia Yes Yes Yes Yes Hawaii No No No No Idaho Yes No No Yes Illinois No No No No Indiana Yes Yes No Yes Iowa No No No No Kansas Yes Yes No Yes Kentucky Yes Yes Yes Yes Louisiana Yes No Yes Yes Maine Yes Yes Yes Yes Maryland Yes No No No Massachusetts Yes Yes Yes Yes Michigan Yes Yes No Yes Minnesota Yes No No Yes Mississippi Yes Yes Yes Yes Missouri Yes Yes No Yes Montana Yes No No Yes Nebraska Yes No No Yes Nevada Yes No No Yes New Hampshire Yes No No No New Jersey Yes No No Yes New Mexico Yes Yes Yes Yes New York No No No No North Carolina Yes Yes No Yes North Dakota No No No No Ohio Yes No Yes Yes Oklahoma Yes No Yes Yes Oregon Yes No No Yes Pennsylvania No No No No Rhode Island Yes No Yes Yes Public Budgeting & Finance / Summer 2009 114 The differential treatment of personal property for tax purposes across states results in significant variation among states in the composition of property tax revenues reported by the U.S. Census Bureau. For example, Table 2 reports personal property tax revenues for local governments within six metropolitan areasFBaltimore, Las Vegas, Miami, Milwaukee, Richmond, and San Francisco. The proportion of total property taxes gen- erated from taxes on personal property varies significantly across these metropolitan areas. Personal property taxes account for 27 percent of all property tax revenues in the Richmond metropolitan area, 11.6 percent of the property tax revenues in the Las Vegas area, and 15.4 percent in the Baltimore metropolitan area. In San Francisco, Miami, and Milwaukee, personal property taxes make up a much smaller proportion of overall property tax revenues, as personal property taxes account for 5.7, 4.3, and 1.6 percent of property tax revenues, respectively. 4 Additional problems interpreting these common measures arise when corporations do business with out-of-state customers. This is because, as is often argued, corporations do not pay property taxes, people do. So while the business entity in a state may actually pay the property tax bill, a share of those payments do not come out of resident income and are not paid by state residents. Corporations will first try to pass the cost of doing business, including property taxes paid, forward to customers in the form of higher prices. If a company is successful in that strategy,and it only sells to in-state customers, then such property taxes would be paid by residents from state personal income. But to TABLE 1 (Continued) StatePersonal property tax allowedMotor vehicles InventoryMachinery and equipment South Carolina Yes Yes No Yes South Dakota No No No No Tennessee Yes Yes No Yes Texas Yes Yes Yes Yes Utah Yes No Yes Yes Vermont Yes No Yes Yes Virginia Yes Yes Yes Yes Washington Yes No No Yes West Virginia Yes Yes Yes Yes Wisconsin Yes No No Yes Wyoming Yes No No Yes Source: Significant Features of the Property Tax, Lincoln Institute of Land Policy and George Washington Institute of Public Policy. 4. Patricia S. Atkins, Leah B. Curran, and Michael E. Bell,Intra-Metropolitan Area Revenue Raising Disparities and Equities(Washington, DC: U.S. Department of Housing and Urban Development, mim- eograph, 2005), 22. Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 115 TABLE 2 Personal Property Tax Revenue as a Percent of Total Property Tax Revenue by Metropolitan Area Baltimore Las Vegas Miami Milwaukee Richmond San Francisco Personal property tax revenueUS$352,972,944 US$127,892,864 US$94,801,204 US$31,202,901 US$260,924,933 US$127,633,301 Total property tax revenueUS$2,294,262,000 US$1,105,352,000 US$2,215,885,000 US$1,972,256,000 US$965,321,000 US$2,223,996,000 Percent of total tax revenue from personal property tax15.40% 11.60% 4.30% 1.60% 27.00% 5.70% Source: Patricia S. Atkins, Leah B. Curran, and Michael E. Bell,Intra-Metropolitan Area Revenue Raising Disparities and Equities(Washington, DC: U.S.

Department of Housing and Urban Development, mimeograph, 2005). Public Budgeting & Finance / Summer 2009 116 the extent such taxes are exported to out-of-state customers; taxes will not be entirely paid by residents from state personal income. Thus, the larger the share of the property tax base attributable to commercial and industrial property, the more likely it will be that some portion of property taxes will be paid by nonresidents thereby undermining the comparability across states of traditional measures of property tax burdens.

Table 3 contains information on the relative importance of various land use types across states. These data reflect the most recent data on the composition of a state’s property tax base available from each state’s Web site from 20 states. Not all states, however, provide data on the composition of their property tax base by land use type on their Web site.

The share of a state’s property tax base attributable to residential property ranges from 85 percent in Massachusetts to just 39 percent in South Dakota. In three of the 20 states listed in Table 3, residential property accounts for 75 percent or more of the total TABLE 3 Composition of State Property Tax Base by Land Use Type a Residential (%) Agricultural (%)Commercial and Industrial (%) Other (%) Alaska 58.35 0.04 25.63 15.98 Colorado 46.08 1.10 30.78 22.05 Florida 50.53 3.52 11.85 34.10 Idaho 66.32 4.43 23.95 5.30 Indianapolis 55.80 5.37 23.03 15.81 Iowa 44.17 18.61 30.17 7.06 Maryland 70.18 1.85 16.15 11.82 Massachusetts 84.58 0.00 13.13 2.29 Michigan 68.20 2.60 20.12 9.07 Missouri 52.02 1.86 20.87 25.25 Montana 43.72 7.59 15.91 32.78 New Jersey 74.78 0.83 19.34 5.05 North Carolina 77.92 0.00 19.30 2.78 North Dakota 41.00 31.18 22.28 5.53 Ohio 73.77 4.03 22.15 0.06 Oklahoma 67.08 23.32 9.60 0.00 Oregon 48.04 0.00 6.23 45.74 Pennsylvania 70.51 3.92 23.51 2.06 South Dakota 39.08 34.80 23.60 2.52 Utah 45.92 0.00 19.79 34.29 aData across states are not perfectly comparable as the type of properties included in each category varies based on the state definition.

Source: Significant Features of the Property Tax, Lincoln Institute of Land Policy and George Washington Institute of Public Policy.

Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 117 property tax baseFNew Jersey, North Carolina, and Massachusetts. In seven states residential property accounts foro50 percent of the property tax baseFColorado, Iowa, Montana, North Dakota, Oregon, South Dakota, and Utah.

The variation across states in the relative importance of commercial and industrial property is not as great as for residential property. The state share of the property tax base attributable to commercial and industrial property ranges from 31 percent in Col- orado too7 percent in Oregon; a 25 percentage-point difference. Approximately 15 percent of the states listed in Table 3 have commercial and industrial property account- ing for425 percent of the state’s total property tax baseFAlaska, Colorado, and Iowa.

In four states commercial and industrial property accounts foro15 percent of the state’s property tax baseFFlorida, Massachusetts, Oklahoma, and Oregon.

Given the complicated property tax landscape across the 50 states, one should proceed with caution when interpreting these two common measures of property tax burden.

The next section turns to a discussion of alternative measures of effective property tax rates.

EFFECTIVE PROPERTY TAX RATES: ALTERNATIVE MEASURES The previous section reviewed two popular measures of property tax burden often pub- lished in the press and utilized in policy debates. Neither measure relates property tax liabilities, or payments, to the actual base of the taxFthe market value of property.

Thus, a preferred, and more typical, means of exploring the differential impact of prop- erty taxes across properties is to look at effective property tax rates across properties or jurisdictions. Effective tax rates relate tax liabilities to the actual base of the tax. For example, effective income tax rates look at total income tax liabilities relative to the base of the tax, personal income. Similarly, effective property tax rates compare tax liabilities to the base of the tax, assessed value. This is in contrast to measures that compare property tax liabilities, or payments, to income, which is a measure of burden. Property tax burdens, relative to income, are useful for some purposes, but they do not tell you anything about how heavily the tax is being used. For that purpose we need to compute effective property tax rates that compare the tax liability with the tax base.

AARP: State and Local Property Tax Burdens in 2005 5 In May 2007, the AARP Public Policy Institute published a report on state and local property tax burdens in 2005. The purpose of the report is to provide policymakers with information on residential property tax burdens across states and demographic groups, so as to assist them with their policy deliberations. 5. The methodology summarized in this section is drawn from David Baer,State and Local Property Tax Burdens in 2005,No. 2007–09 (Washington, DC: AARP Public Policy Institute, 2007). Public Budgeting & Finance / Summer 2009 118 For the purposes of the AARP study, property tax burdens were defined as the ratio of property taxes divided by a measure of family income. Data were obtained from the 2005 American Community Survey (ACS) conducted by the U.S. Census Bureau. The study calculates the overall median state and local residential property tax burden for three groups: all homeowners, homeowners under age 65, and homeowners age 65 and older. AARP used property taxes and income for each household to calculate the median property tax burden in each state. They calculate property tax burdens by state by the following steps:

(1) Income for each household was calculated by including income from family members and any unmarried partners in the household.

(2) Household property taxes were reported in ranges rather than actual amounts and the midpoint of each range was assumed to be the amount each household paid in property taxes. 6 (3) After calculating household income and property taxes, the property tax burden was calculated for each of the three groups. The median property tax burden for each group was used as the best estimate for the overall state property tax burden.

Table 4 reports the results from the AARP study of the ten states with the highest and lowest property tax burdens. These property tax burden measures relate property tax liabilities to household income.

Data from the ACS is self-reported data. Question 20 on the Housing portion of the survey simply asks ‘‘What are the annual real estate taxes on this property?’’ The ac- curacy of such self-reported data must always be interpreted with caution, and any response bias may not be consistent across respondents. This is a particular concern with respondents over 65. For example, seniors may be eligible for property tax relief pro- grams that do not directly reduce the property tax bill. It is important, therefore, to know if the numbers being reported are gross or net property taxes; that is, have any property tax relief measures, especially those targeted at seniors, been netted out of the property tax liability? The answer to this question will vary across responding households de- pending, in part, on how much time they take to determine the answer to the survey question. This is an important issue relative to comparability of data and trustworthiness of conclusions, because the question asks what the real estate taxes are on a property, not how much an individual household paid in property taxes. Households may, however, incorrectly respond with the latter rather than former interpretation, skewing the results.

Similarly, sometimes property tax bills include charges for specific services like solid waste collection and disposal. Those costs should not be included in the numbers re- ported to the American Community Survey.

6. Ranges are in US$50 increments from US$1 to US$999; US$100 increments from US$1,000 to US$4,999; US$500 increments from US$5,000 to US$5,999; and US$1,000 increments from US$6,000 to US$9,999. The highest range is US$10,000 or more in property taxes. Taxes falling in this highest range were estimated to be US$15,000. Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 119 National Association of Home Builders The National Association of Home Builders (NAHB) is another organization which estimates effective property tax rates on a regular basis. Like the AARP, the NAHB utilizes data from the 2005 ACS and calculates effective property tax rates for each state and the District of Columbia. TABLE 4 Ten Highest and Lowest Property Tax Burden States in 2005 a All homeownersHomeowners under age 65Homeowners age 651 StateMedian property tax burden (%) StateMedian property tax burden (%) StateMedian property tax burden (%) Ten Highest Property Tax Burden States New Jersey 6.5 New Jersey 5.8 New Jersey 10.5 New Hampshire 5.6 New Hampshire 5.2 New Hampshire 8.4 Vermont 4.9 Vermont 4.4 Connecticut 8.3 Connecticut 4.9 Connecticut 4.4 Vermont 7.6 Wisconsin 4.7 New York 4.3 Rhode Island 7.3 New York 4.7 Wisconsin 4.3 Wisconsin 7.3 Rhode Island 4.5 Illinois 4.2 Massachusetts 7 Illinois 4.4 Rhode Island 4.1 New York 6.3 Massachusetts 4 Texas 3.6 Illinois 5.4 Pennsylvania 3.5 Massachusetts 3.5 Pennsylvania 5.2 Ten Lowest Property Tax Burden States Louisiana 0.3 Louisiana 0.3 Louisiana 0.3 Alabama 0.6 Alabama 0.6 Mississippi 0.4 Mississippi 0.9 West Virginia 0.9 Alabama 0.6 West Virginia 0.9 Arkansas 1 West Virginia 0.9 Arkansas 1.1 Mississippi 1 Delaware 1.4 Hawaii 1.2 Hawaii 1.2 Arkansas 1.5 Delaware 1.3 Wyoming 1.2 South Carolina 1.5 South Carolina 1.3 Delaware 1.2 Hawaii 1.6 Wyoming 1.3 South Carolina 1.3 Oklahoma 1.6 Oklahoma 1.4 Oklahoma 1.3 Kentucky 1.7 aThe property tax burden for each household equals the ratio of property taxes divided by the combination of family income plus income of any unmarried partner.

Source: David Baer,State and Local Property Tax Burdens in 2005,No. 2007–09 (Washington, DC: AARP Public Policy Institute, 2007). Public Budgeting & Finance / Summer 2009 120 Unlike the AARP, however, the NAHB expresses property taxes in relation to the value of property, not household income. The NAHB determines the median home value for each state from the ACS. It then determines the median real estate taxes paid per home and presents that as tax rates per US$1,000 of home value. 7 Table 5 presents data for each state on property tax burdens and effective property tax rates for the AARP and the NAHB. Effective property tax rates relative to personal income as calculated by the AARP range from 6.5 percent in New Jersey to 0.3 percent in Louisiana. Effective property tax rates relative to median home value as calculated by NAHB range from a high of 1.8 percent in Wisconsin and Texas to 0.2 percent in Hawaii and Louisiana. While the state rankings are not identical under the two approaches, there is a correlation coefficient of 0.802 between the two rankings.

Minnesota Taxpayers Association 50-State Property Tax Comparison 8 The Minnesota Taxpayers Association (MTA), in cooperation with other member states of the National Taxpayers Conference, prepares periodically a study comparing effective property tax rates across states. In contrast to statutory rates that are typically applied to taxable values to determine tax liabilities for individual properties, the MTA uses effec- tive tax rates to express the relationship between net property taxes and true market value of individual properties. These measures of effective property tax rates allow for more meaningful comparisons across jurisdictions because they include the effects of all statutory tax provisions as well as the effects of local assessment practices.

Reports have been prepared for tax years payable 1995, 1998, 2000, 2002, 2004, and 2005. The study compares effective property tax rates for the largest city and a typical rural city in each state. The study computes effective property tax rates for four classes of propertyFhomestead, commercial, industrial, and apartments. For owner-occupied housing, the most recent years look at effective property tax rates for homesteads valued at US$70,000, US$150,000, and US$300,000.

The MTA approach to calculating effective property tax rates assumes that the property tax calculation has five distinct components:

a ‘‘true’’ market value (TMV) a local assessment/sales ratio (SR) a statutory classification rate to determine the proportion of the assessor’s esti- mated market value that is taxable (CR) the total local property tax rate (TR) applicable property tax credits (C) 7. Natalia Siniavskaia,Real Estate Tax Rates in the American Community Survey, Special Studies (Washington, DC: National Association of Home Builders, 2007).

8. The discussion of the methodology used by the Minnesota Taxpayers Association to calculate effec- tive property tax rates is summarized from Minnesota Taxpayers Association,50-State Property Tax Comparison Study(St. Paul, MN: Minnesota Taxpayers Association, 2005). Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 121 TABLE 5 Effective Property Tax Rates from the ACS, 2005 StateAARP effective property tax rates relative to personal income (%)NAHB estimate of effective property tax rates relative to median home value (%) Alabama 0.6 0.31 Alaska 2.8 1.137 Arizona 1.9 0.611 Arkansas 1.1 0.525 California 3.1 0.477 Colorado 2 0.581 Connecticut 4.9 1.424 Delaware 1.3 0.395 District of Columbia 1.8 0.376 Florida 2.8 0.789 Georgia 1.8 0.712 Hawaii 1.2 0.204 Idaho 2.4 0.909 Illinois 4.4 1.579 Indiana 2 0.943 Iowa 2.6 1.271 Kansas 2.4 1.24 Kentucky 1.5 0.667 Louisiana 0.3 0.172 Maine 3.3 1.122 Maryland 2.8 0.771 Massachusetts 4 0.823 Michigan 3.2 1.236 Minnesota 2.6 0.814 Mississippi 0.9 0.503 Missouri 1.9 0.822 Montana 2.8 0.995 Nebraska 3.3 1.669 Nevada 2.3 0.51 New Hampshire 5.6 1.633 New Jersey 6.5 1.603 New Mexico 1.5 0.563 New York 4.7 1.188 North Carolina 1.9 0.757 North Dakota 2.6 1.497 Ohio 3 1.233 Oklahoma 1.4 0.713 Oregon 3.3 0.949 Public Budgeting & Finance / Summer 2009 122 Thus the net local property tax for each parcel of property can be written:

Net Property Tax¼TMV SR CR TR Cð1Þ These calculations start with the true market value of a parcel of property which is determined in the local real estate market consisting of arm’s length transactions between willing buyers and sellers. Three homestead values are used as constants across states because the goal of the study is to compare the effects of property tax structures across states.

Starting with the assumed true market values, the study then adjusts those values with the use of assessment/sales ratios applicable to the location and type of property being studied. These ratios are typically county-level ratios for specific classes of property. 9By applying assessment/sales ratios, the MTA is recognizing that a homestead with a true market value of US$70,000 may be carried on the tax roll of individual jurisdictions at US$65,000 or US$50,000 depending on assessment practices unique to each jurisdiction.

The next step in the process of calculating effective property tax rates is to apply statutory classification or differential assessment schemes to the assessor’s estimate of market value. Not all states have differential assessment schemes or classification by land use type. In those cases, a homestead assessed at US$100,000 and a business with the same assessment would pay identical property taxes and their effective property tax rates would be the same. TABLE 5 (Continued) StateAARP effective property tax rates relative to personal income (%)NAHB estimate of effective property tax rates relative to median home value (%) Pennsylvania 3.5 1.469 Rhode Island 4.5 1.092 South Carolina 1.3 0.568 South Dakota 2.7 1.381 Tennessee 1.7 0.696 Texas 3.4 1.817 Utah 1.9 0.676 Vermont 4.9 1.635 Virginia 2.3 0.668 Washington 3.3 0.988 West Virginia 0.9 0.461 Wisconsin 4.7 1.82 Wyoming 1.3 0.546 9. Some states, however, do not calculate ratios for each land use type. For example, New Hampshire calculates assessment-sales ratios using all sales from all types of properties. Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 123 Once taxable values for each parcel are determined, as discussed above, they are multiplied by the statutory tax rate applicable in each jurisdiction. The tax rate used is the total aggregate tax rate which includes the tax rate for all taxing jurisdictions in the state which typically levy property taxesFe.g., cities, counties, and school districts.

Special assessments were excluded from these calculations.

Multiplying the tax base by the applicable rate determines a gross tax liability for each property. The final step in the tax calculation is to apply any general deductions from the gross property tax calculations. This might include such things as tax credits and circuit- breaker refunds. These net tax liabilities are then compared with true market value for each individual property to determine effective property tax rates. Table 6 presents effective property tax rates for the median priced residential property in the largest city in each state for taxes payable in 2005.

According to this measure of effective property tax rates, Honolulu has the lowest effective property tax rate at just 0.36 percent. The highest effective property tax rate is in Detroit, Michigan at 3.23 percent. Nationally, the average effective property tax rate is 1.35 percent.

Table 7 presents data on effective property tax rates for eight cities in New Hampshire for taxable 2006. The MTA study includes the largest city in New Hampshire, Man- chester. Using the MTA methodology, the effective tax rate in Manchester was 1.492 percent in 2006. Other than Salem, this is the lowest effective property tax rate reported in Table 7, in part because large cities tend to have more diverse revenue structures than smaller cities and therefore rely less on property taxes. Effective property tax rates for these cities range from a high of 1.919 percent in Derry to a low of just 1.147 percent in Salem. In other words, the effective property tax rate in Derry is nearly two-thirds greater than the effective property tax rate in Salem and nearly 30 percent greater than the effective property tax rate in Manchester. Within-state variation in effective property tax rates across cities is not captured by the MTA study and these differences can be substantial as illustrated in Table 7 for eight cities in New Hampshire.

District of Columbia Nationwide Comparison of Tax Rates and Tax Burdens 10 The government of the District of Columbia publishes annually a nationwide compar- ison of tax rates and tax burdens. The report starts with the recognition that a juris- diction’s revenue system reflects its revenue needs, tax base, the intergovernmental system within which it operates, constitutional and legal limitations it faces, taxpayer demand for services, and other factors. To capture these differences, the study compares tax burdens in 51 different locations for a hypothetical family of three. Specifically, the 10. The discussion of the methodology used by the government of the District of Columbia in their study of effective property tax rates across the county is drawn from District of Columbia: Office of the Chief Financial Officer,Tax Rates and Tax Burdens in the District of ColumbiaFA Nationwide Comparison, 2005(Washington, DC, 2006). Public Budgeting & Finance / Summer 2009 124 TABLE 6 Effective Property Tax Rates for the Median Priced Residential Property in the Largest City in Each State, 2005 State Largest city Effective tax rate (%) Ranking Alabama Birmingham 0.66 47 Alaska Anchorage 1.42 20 Arizona Phoenix 1.04 37 Arkansas Little Rock 1.18 27 California Los Angeles 1.14 30 Colorado Denver 0.51 50 Connecticut Bridgeport 1.81 11 Delaware Wilmington 1.03 38 District of Columbia Washington 0.83 43 Florida Jacksonville 1.52 15 Georgia Atlanta 1.16 29 Hawaii Honolulu 0.36 51 Idaho Boise 1.17 28 Illinois Chicago 1.50 16 Indiana Indianapolis 1.39 21 Iowa Des Moines 1.73 12 Kansas Wichita 1.19 26 Kentucky Louisville 1.13 32 Louisiana New Orleans 0.92 41 Maine Portland 1.87 9 Maryland Baltimore 2.29 4 Massachusetts Boston 0.68 46 Michigan Detroit 3.23 1 Minnesota Minneapolis 1.29 23 Mississippi Jackson 1.32 22 Missouri Kansas City 1.45 18 Montana Billings 1.05 36 Nebraska Omaha 2.06 7 Nevada Las Vegas 1.13 31 New Hampshire Manchester 1.59 13 New Jersey Newark 2.17 6 New Mexico Albuquerque 0.97 39 New York New York City 0.64 49 North Carolina Charlotte 1.13 33 North Dakota Fargo 1.99 8 Ohio Columbus 1.49 17 Oklahoma Oklahoma City 1.10 34 Oregon Portland 1.24 25 Pennsylvania Philadelphia 2.23 5 Rhode Island Providence 1.43 19 Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 125 state and local tax burdens for the family in the District of Columbia are compared with comparable burdens in the largest city in each state.

The four major taxes included in the study are the individual income tax, real property tax on residential property, general sales and use tax, and automobile taxes, including gasoline tax, registration fees, excise tax, and personal property tax. All tax burdens reflect state and local tax rates. Tax burdens are compared for a hypothetical family that consists of two wage-earning spouses and one school-age child. The gross family income levels are assumed to be US$25,000, US$50,000, US$75,000, US$100,000, and US$150,000. Housing values across income levels are based on data from the Amer- ican Community Survey conducted annually by the U.S. Census Bureau and are adjusted by linear regression for the different income levels. TABLE 6 (Continued) State Largest city Effective tax rate (%) Ranking South Carolina Columbia 1.05 35 South Dakota Sioux Falls 1.25 24 Tennessee Memphis 1.87 10 Texas Houston 2.33 3 Utah Salt Lake City 0.80 45 Vermont Burlington 1.56 14 Virginia Virginia Beach 0.84 42 Washington Seattle 0.97 40 West Virginia Charleston 0.82 44 Wisconsin Milwaukee 2.47 2 Wyoming Cheyenne 0.65 48 Source:Minnesota Taxpayers Association,50-State Property Tax Comparison Study(St. Paul, MN, 2005). TABLE 7 Effective Property Tax Rates for Eight Cities in New Hampshire, 2006 for Residential Homestead Property with a True Market Value of US$150,000 True market valueAssessment/ sales ratioStatutory tax rate Total taxEffective tax rate (%) Manchester US$150,000 0.526 0.02836 2,237.60 1.49 Nashua US$150,000 0.992 0.01632 2,428.42 1.62 Derry US$150,000 0.71 0.02703 2,878.70 1.92 Merrimack US$150,000 0.812 0.02307 2,809.93 1.87 Concord US$150,000 0.947 0.01977 2,808.33 1.87 Dover US$150,000 0.879 0.01942 2,560.53 1.71 Salem US$150,000 0.52 0.02206 1,720.68 1.15 Rochester US$150,000 0.955 0.0173 2,478.23 1.65 Source:Calculations by Bethany Paquine. Public Budgeting & Finance / Summer 2009 126 Real property tax burdens in the 51 cities included in the study are a function of each city’s residential real estate values, the ratio of assessed value to market value, the tax rate, and various homeowner exemptions and credits. ACS data were used to determine the median house value at specific income levels.

Table 8 reports the effective property tax rates for the largest city in each state and the District of Columbia using the methodology just described. The effective property tax rates range from 3.21 percent in Indianapolis, Indiana to just 0.38 percent in Honolulu, Hawaii. Nationally, the average effective property tax is 1.64 percent.

The Washington, DC and MTA studies both calculate effective property tax rates for the largest city in each state and the District of Columbia. Different methodologies and different data are used, albeit these estimates are both for 2005. It is not surprising then that the two studies present somewhat different estimates of effective property tax rates.

In a sense, the DC estimates reported seem to be gross effective property tax rates because they are derived before various property tax relief measures are applied. Al- ternatively, the MTA estimates are net effective property tax rates because property tax relief measures have been explicitly accounted for. Also, the results reported for the MTA are for the median priced home in each city and the numbers reported by the District of Columbia are essentially average numbers for each jurisdiction.

As a result, the absolute estimates of effective property tax rates vary between the two studies. The correlation coefficient between the two estimates of effective property tax rates is 0.694. However, the correlation between the rankings of the states by these two measures of effective property tax rates is 0.815Fnot a perfect correlation, but the relative rankings are closer than the absolute estimates of the effective property tax rates.

District of Columbia Fiscal Policy Institute 11 The District of Columbia Fiscal Policy Institute (FPI) is another organization, in ad- dition to the government of the District of Columbia, which annually calculates tax burdens for hypothetical families. The FPI, however, only calculates tax burdens for families in the Washington, DC metropolitan area.

The analysis calculates the total tax burden from property, income and car taxes for hypothetical families in the District and four suburban jurisdictionsFMontgomery and Prince George’s counties in Maryland and Fairfax and Arlington counties in Virginia.

Tax burdens are calculated for hypothetical families with annual incomes of US$50,000, US$100,000, and US$150,000. For homeowners, the study assumes that the family with US$50,000 has a home worth US$250,000; the family with US$100,000 has a home worth US$400,000; and the family making US$150,000 has a home worth US$600,000. 11. Discussion of the methodology used by the District of Columbia Fiscal Policy Institute in calcu- lating its effective property tax rates draws on Ed Lazere and Aleksandra Gajdeczka,Taxes on DC Families Are Now the Lowest in the Washington Region(Washington, DC: DC Fiscal Policy Institute, 2006). Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 127 TABLE 8 Effective Property Tax Rates by State, 2005 State CityNominal rate (%)Assessment level (%)Effective tax rate (%) Ranking Alabama Birmingham 6.95 10.00 0.70 48 Alaska Anchorage 1.63 100.00 1.63 24 Arizona Phoenix 14.09 10.00 1.41 31 Arkansas Little Rock 6.90 20.00 1.38 32 California Los Angeles 1.10 100.00 1.10 41 Colorado Denver 8.36 8.00 0.67 50 Connecticut Bridgeport 4.23 70.00 2.96 4 Delaware Wilmington 2.94 49.20 1.45 29 District of Columbia Washington 0.96 100.00 0.96 45 Florida Jacksonville 1.86 100.00 1.86 17 Georgia Atlanta 4.20 40.00 1.68 22 Hawaii Honolulu 0.38 100.00 0.38 51 Idaho Boise 1.76 98.90 1.75 18 Illinois Chicago 7.30 20.40 1.49 27 Indiana Indianapolis 3.21 100.00 3.21 1 Iowa Des Moines 4.66 48.00 2.24 10 Kansas Wichita 11.35 11.50 1.30 33 Kentucky Louisville 1.26 100.00 1.26 34 Louisiana New Orleans 17.17 10.00 1.72 19 Maine Portland 2.01 81.00 1.63 23 Maryland Baltimore 2.31 100.00 2.31 9 Massachusetts Boston 1.23 100.00 1.23 37 Michigan Detroit 6.77 28.60 1.94 15 Minnesota Minneapolis 1.36 90.70 1.24 36 Mississippi Jackson 17.11 10.00 1.71 20 Missouri Kansas City 7.91 19.00 1.50 26 Montana Billings 1.96 80.00 1.57 25 Nebraska Omaha 2.10 96.00 2.01 14 Nevada Las Vegas 3.11 35.00 1.09 42 New Hampshire Manchester 2.79 100.00 2.79 5 New Jersey Newark 2.30 88.50 2.04 13 New Mexico Albuquerque 3.73 33.30 1.24 35 New York New York City 15.01 4.60 0.69 49 North Carolina Charlotte 1.26 95.10 1.19 39 North Dakota Fargo 48.08 4.40 2.09 12 Ohio Columbus 4.94 29.60 1.46 28 Oklahoma Oklahoma City 10.70 11.00 1.18 40 Oregon Portland 1.91 64.20 1.23 38 Pennsylvania Philadelphia 8.26 32.00 2.64 7 Public Budgeting & Finance / Summer 2009 128 Methodologically, the FPI diverges from the DC government study in calculating effective property tax rates. The DC government study calculates effective property tax rates by taking the statutory rate for each jurisdiction and multiplying it by the assess- ment/sales ratio for the jurisdiction. This approximates an average effective property tax rate for the entire jurisdiction.

The FPI uses actual property tax bills for a sample of houses that sold in each jurisdiction. Specifically, for each home value in each jurisdiction, a sample of 50 recently sold homes was used. For example, for homes in the US$400,000 range in Washington, DC, the analysis reflects the average actual tax bill among 50 homes that actually sold for approximately US$400,000 within the appropriate time frame. The FPI also adjusts for various types of residential property tax relief like the annual cap in assessments for DC as well as the DC’s homestead deduction. Thus, unlike all the other studies summarized here, the FPI is the only one that bases its calculations on actual sales price and tax liability data from homes that sold in each jurisdiction. The product is an average effective tax rate from homes that actually sold in each jurisdiction at various prices.

Using 2006 data, the FPI study found that for a family earning US$100,000 per year, living in a house with a market value of US$400,000 the property tax liability averaged US$1,639, or approximately 41 cents per US$100 assessed value. Using the methodology from the DC government report, the average effective property tax rate in 2006 would simply be the statutory rate (92 cents per US$100 of assessed value) divided by the assessment/sales ratio (100 percent), or 92 cents per US$100 assessed value. This effective property tax rate is more than twice as high as the one estimated by the FPI using actual data and making allowance for residential property tax relief measures. TABLE 8 (Continued) State CityNominal rate (%)Assessment level (%)Effective tax rate (%) Ranking Rhode Island Providence 3.00 100.00 3.00 3 South Carolina Columbia 42.76 4.00 1.71 21 South Dakota Sioux Falls 2.53 85.00 2.15 11 Tennessee Memphis 7.47 25.00 1.87 16 Texas Houston 3.01 100.00 3.01 2 Utah Salt Lake City 1.48 97.20 1.44 30 Vermont Burlington 2.72 100.00 2.72 6 Virginia Virginia Beach 1.22 82.40 1.00 44 Washington Seattle 1.08 95.30 1.03 43 West Virginia Charleston 1.45 60.00 0.87 46 Wisconsin Milwaukee 2.46 96.00 2.36 8 Wyoming Cheyenne 7.60 9.50 0.72 47 Source:District of Columbia: Office of the Chief Financial Officer,Tax Rates and Tax Burdens in the District of ColumbiaFA Nationwide Comparison, 2005(Washington, DC, 2006), 17.

Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 129 STATE ESTIMATES OF EFFECTIVE PROPERTY TAX RATES While most states publish nominal tax rates for their local governments, only a few report effective property tax rates. Upon review of the state governments’ Web sites, only 13 states were found to calculate and report effective tax rates. In a couple of additional cases, effective tax rate reports were found on nongovernmental Web sites, such as state tax payer’s associations or think tanks. The following review is limited to those effective tax rate reports found on state governments’ Web sites. Table 9 provides an overview of the states’ reports and how they calculated their effective tax rates.

Although the majority of the effective tax rate reports are prepared by the state’s Department of Revenue or a Division within the Department, others are prepared by commissions or committees. Typically, in the latter cases, the reports were prepared one time rather than on an annual or biennial basis. Reports that are not done annually are, at times, prepared for a special purpose, such as in Arizona, where the effective tax rates were prepared as part of a report analyzing the possibility of reinstating a statewide property tax. 12 As discussed previously, there are multiple ways of calculating effective tax rates.

Most of the states calculate the effective tax rate simply by multiplying the assess- ment/sales ratio by the nominal tax rate. This equalizes discrepancies that occur from different valuation cycles, but does not address property tax relief measures. This also represents the average effective property tax rate and does not provide any information about within-jurisdiction variation across individual properties in effective property tax rates.

Three states, however, calculate effective tax rates by determining the net tax liability to some extent. Ohio calculates the effective tax rate as taxes-charged divided by taxable value. Although taxes-charged accounts for tax reduction factors, a reduction in the property tax used to eliminate tax revenue growth resulting only from appreciating property values, it does not account for the 10-percent rollback given to all real property, the 2.5-percent rollback for residential real property, or the homestead exemption given to qualified homeowners. 13 Minnesota notes that its effective tax rates are calculated by ‘‘net tax payable divided by the indicated market value.’’ 14 Finally, Wisconsin calculates its effective tax rate by figuring the ‘‘general property tax less state property tax credit 12. Citizens Finance Review Commission,‘‘Reinstate a State Property Tax’’ [report on-line]; available fromhttp://azmemory.lib.az.us/cgi-bin/showfile.exe?CISOROOT=/statepubs&CISOPTR=308&filename= 309.pdf: accessed 15 March, 2009.

13. Ohio Department of Taxation,2006 Annual Report(Columbus, OH: Ohio Department of Taxation 2006), 137–138.

14. Minnesota Department of Revenue, ‘‘Property Taxes Levied in Minnesota: Table 57: Total Tax Capacity Value, Indicated Market Value and Effective Tax Rates of Counties for Property Taxes Levied in 2003 Payable in 2004’’ [report on-line]; available fromhttp://www.taxes.state.mn.us/taxes/property_tax_ad ministrators/other_supporting_content/ptbull_pay04_sect4currentyrtables.shtml: accessed 14 September, 2007. Public Budgeting & Finance / Summer 2009 130 TABLE 9 States Reporting Effective Tax Rates State Year(s)Government department preparing the reportAnnual, biennial, or one time calculationHow are effective tax rates calculated?Are effective tax rates calculated for real property or real and personal property?Are effective tax rates provided by type of jurisdiction or type of property?

Arizona 2000 Arizona Citizen Finance Review CommissionOne time Not available Real property only Type of property Georgia 2006 Georgia Department of Economic DevelopmentNot available Not available Real and personal propertyType of jurisdiction Minnesota Taxes Payable 2001–2004Minnesota Department of RevenueAnnual Net tax liability Real and personal propertyType of jurisdiction Montana Fiscal Years 1998–2006Montana Department of RevenueBiennial Tax rate multiplied by assessment ratioReal and personal propertyType of property New Jersey 1998–2006 New Jersey Division of TaxationAnnual Tax rate multiplied by assessment ratioNot available Type of jurisdiction North Carolina Fiscal Years 2002–2007North Carolina Department of RevenueAnnual Tax rate multiplied by assessment ratioReal and personal propertyType of jurisdiction Ohio 1996–2006 Ohio Department of Taxation Annual Net tax liability Real property only Type of property and type of jurisdiction South Dakota Taxes Payable 1997–2006South Dakota Department of Revenue and RegulationAnnual Tax rate multiplied by assessment ratioReal property only Type of property and type of jurisdiction Tennessee Tax Year 2000Tennessee Advisory Commission on Intergovernmental RelationsOne time Tax rate multiplied by assessment ratioReal property only Type of jurisdiction Vermont Tax Years 2001–2006Vermont Department of Taxes Annual Not available Not available Type of property and type of jurisdiction Virginia 1993–2005 Virginia Department of TaxationAnnual Tax rate multiplied by assessment ratioReal property only Type of jurisdiction Washington Taxes Payable 1991–2006Washington State Department of RevenueAnnual Tax rate multiplied by assessment ratioReal and personal propertyType of jurisdiction Wisconsin Taxes Payable 2000–2007Wisconsin Department of Revenue, Division of State and Local FinancesAnnual Net tax liability Real and personal propertyType of jurisdiction Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 131 (not including lottery credit) divided by the full value.’’ 15 Although these methods of calculating the effective tax rates are not necessarily comparable, they come closer to capturing the true effective tax rate, by compensating for the state’s property tax relief programs.

Another variation in the calculation of effective tax rates among the states is whether they include real property only or real and personal property. As can be seen in Table 1, South Dakota does not assess a personal property tax. Of the remaining twelve states, six states include both real and personal property in their calculations, while the other four do not include personal property even though it is taxed.

16 States also vary in how they report the effective tax rates that they calculated. Some states report these rates by class or type of property, while others provide the data by jurisdiction, while still others do both. States that report their data by type of property tend to provide statewide average effective tax rates for each class of property. States that report their data by jurisdiction do so either for a single jurisdiction, or for multiple jurisdictions. Ohio, South Dakota, and Vermont report their effective property tax rates by property type and jurisdiction. Not only does Ohio report effective property tax rates by county in their annual report, but they also provide a database that can be down- loaded and users can select both the jurisdiction (by county, city, township, school districts, or special districts) they have an in interest in or the property type (residential and commercial real property; or industrial, commercial, mineral, and railroad real property).

The calculation of effective tax rates by these 13 states allows taxpayers within those states to compare their tax rates to other jurisdictions or their property type to others within the state. Table 10 provides the minimum, median, maximum, and range for effective property tax rates in the four states that include only real property in their calculations of effective tax rates, and report their data by jurisdiction. TABLE 10 Within State Variation of Effective Tax Rates Year Minimum Median Maximum Range Ohio 2005 3.254 4.511 7.242 3.988 South Dakota 2005 0.023 1.77 3 2.977 Tennessee 2000 0.003 0.008 0.016 0.013 Virginia 2005 0.24 0.52 1.35 1.11 15. Wisconsin Bureau of Property Tax,Town, Village, and City TaxesF2006, Taxes Levied 2006F Collected 2007[report on-line]; available fromhttp://www.revenue.wi.gov/pubs/slf/tvc06.pdf: accessed 14 September, 2007.

16. It could not be determined if personal property was included in the calculations completed by New Jersey and Vermont. Public Budgeting & Finance / Summer 2009 132 If one divides the range in rates by the median rate we get an idea of the variation in rates across local governments in each state. According to this measure, Virginia has the most variation in rates and Ohio has the least variation in rates across jurisdictions.

Although they do not include estimates from the whole state, the effective tax rates calculated for the eight cities in New Hampshire, in Table 7, suggest a range of effective property tax rates that is similar to those provided by the states shown in Table 10.

States are more likely to know the unique nuances of the property tax system within their states than organizations that are calculating effective tax rates using standard procedures for each state. Our look at the data supports this assumption. Table 11 compares the effective property tax rates calculated by the states and those calculated by the Minnesota Taxpayers Association and the DC Government for the largest city in each state. 17 The correlation coefficient between each state’s estimate and the estimate provided by the Minnesota Taxpayers Association is 0.9162. The correlation between the state’s estimates and the DC government estimates is 0.7676. These differences are not surprising given the different methods used for calculating effective tax rates by each organization and the states.

SUMMARY AND CONCLUSION This reconnaissance has turned up several measures of property tax burdens such as generic property taxes per capita and generic property taxes per US$1,000 of personal income; and five major measures of effective property tax rates, including studies by AARP, NAHB, MTA, the DC government, and the DC FPI. TABLE 11 Comparison of Effective Tax Rates by Organization State City YearState’s estimate (%)Minnesota Taxpayers Association (%)DC government (%) Georgia Atlanta 2006 1.68 1.16 1.68 New Jersey Newark 2005 2.04 2.17 2.04 North Carolina Charlotte 2005 1.13 1.13 1.19 Ohio Columbus 2005 1.47 1.49 1.46 South Dakota Sioux Falls 2005 1.47 1.25 2.15 Vermont Burlington 2005 1.72 1.56 2.72 Virginia Virginia Beach 2005 0.77 0.84 1.00 Wisconsin Milwaukee 2005 2.35 2.47 2.36 17. Minnesota and Washington only report data by county so their effective tax rates were not included in this table. Likewise, Tennessee was not included as the most recent data are from tax year 2000. Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 133 The AARP report expresses property tax burdens relative to personal income, so it is really more of a property tax burden measure than a measure of effective property tax rates. This measure can be used to identify situations where the share of income des- ignated for paying property taxes is particularly high and might suggest the need for some type of relief. For example, the study documents the relatively high property tax burden on the elderly.

The other major studies reviewed hereFNAHB, MTA, the DC government and the DC FPIFreport property tax payments, or liabilities, relative to some measure of property valueFe.g., the median housing value of a jurisdiction as computed from data in the American Community Survey. These studies represent different approaches to estimating effective property tax rates. The data in Table 12 summarize major differences between the various measures of effective property tax rates. For example, AARP and NAHB rely on data collected by the Census Bureau for their calculations and AARP uses a comprehensive measure of household income to compute property tax burdens, while NAHB uses a measure of property value to compute effective tax rates. The MTA and DC government studies report effective tax rates for the largest city in each state and the District of Columbia, while the AARP and NAHB studies report average effective property tax rates for the entire state. All of the studies except AARP use some measure of property value to calculate effective property tax rates.

To the extent effective property tax rates reflect prices paid for locally provided goods and services, such measures can be used to explore cross state variation in various issues like the level and quality of public services provided, but they provide no information TABLE 12 Summary of Characteristics of Various Measures of Effective Property Tax Rates Source of data Base of comparisonJurisdiction for which effective tax rate is computed Property taxesProperty value/income AARP ACS ACS Income Average for state a NAHB ACS ACS Property value Average for state MTA Calculated for each stateCalculated for each stateProperty value Largest city and one rural jurisdiction in each state DC Government Calculated for each stateACS Property value Largest city in each state Fiscal Policy InstituteCalculated for individual propertiesActual sales dataProperty value Individual jurisdictions in the DC metro area aProperty tax burdens are computed for three groups in each stateFall homeowners, homeowners under 65 years of age, and homeowners over 65 years of age. Public Budgeting & Finance / Summer 2009 134 that can be used for within-state, or within-jurisdiction, variation in effective property tax rates which could be utilized to explore issues of equity.

Because of the different methodologies used in these studies, the results vary across studies, both in terms of absolute estimates of effective property tax rates and the relative ranking of states. Table 13 reports correlation coefficients for these studies and the two traditional measures of property tax burdens discussed aboveFproperty taxes per capita and per US$1,000 personal income.

The AARP study reports median property tax burdens relative to personal income by state. This measure is relatively highly correlated with the traditional measures of prop- erty tax burdensFper capita property taxes (0.796) and property taxes per US$1,000 personal income (0.777). However, this measure is not strongly correlated with the MTA or DC government measures of effective property taxes calculated for the largest city in each state and relative to some measure of property value.

The MTA and DC government studies are somewhat correlated with each other (0.694), but not correlated with traditional measures of property tax burdensFper cap- ita (0.289 and 0.367, respectively) and property taxes relative to US$1,000 personal income (0.380 and 0.480, respectively).

If one accepts the notion that when computing effective tax rates the tax liability, or payment, should be expressed relative to the appropriate tax base, then the MTA and DC government measures of effective property tax rates are preferred estimates of effective property tax rates across states. The down side is that the two approaches only report effective property tax rates for the largest city in each state, which tend to rely less on property taxes than other cities. Also, the DC government study reports what might be characterized as a gross effective property tax rate before any deductions. However, the DC study has the necessary information to calculate a net effective property tax rate similar to the MTA estimates. TABLE 13 Correlation Coefficients across Organizations AARP NAHB MTADC GovernmentProperty Taxes per CapitaProperty Taxes per US$1,000 Personal Income AARP 1 0.802 0.483 0.49 0.796 0.777 NAHB 1 0.685 0.651 0.579 0.69 MTA 1 0.694 0.289 0.38 DC Government 1 0.367 0.48 Property Taxes per Capita 1 0.92 Property Taxes per US$1,000 Personal Income1 Bell and Kirschner / Alternative Measures of Effective Property Tax Rates 135 In an ideal world, one would want to calculate effective property tax rates using actual information for individual properties, not rely on statewide or citywide averages. This is the approach of the District of Columbia FPI which bases its calculation of effective property tax rates in the Washington, DC area on actual sales data collected from individual jurisdictions. A state could replicate this methodology to calculate effective property tax rates across taxing jurisdictions within the state.

While it might be costly to utilize this approach for all jurisdictions with a property tax within a single state, this methodology could be utilized to calculate effective prop- erty tax rates for selected cities in each state. The issue then becomes which cities in each state should be used for comparison purposes. The MTA and DC studies calculate effective property tax rates for the largest city in each state. These two studies compare effective tax rates in Anchorage, Alaska, Portland, Maine, and Manchester, New Hampshire with Los Angeles, California, and New York City. In other words, very large cities with diverse revenue structures are being compared with smaller cities with more limited revenue options.

Rather than focusing on the largest city in each state, which to some extent represents comparing apples to oranges, the effective property tax rate might be calculated for cities that are more comparable in terms of the relative importance of property taxes. One option might be to calculate effective property tax rates for the capitol city in each state, albeit this would still involve some larger cities. Alternatively, one might want to look just at cities with populations of 100,000 or some comparable measure of size.

The studies reviewed here apply a consistent methodology across states to determine effective property tax rates for specific types of properties. At this point, these are the best numbers available for cross state comparisons. Future research can improve on these estimates by employing a methodology which uses actual data for individual properties. Such an approach will generate better estimates of effective property tax rates and will also provide some initial evidence on variation in effective property tax rates with land use types within a jurisdiction. Future research could also improve on the value of such cross state comparisons by looking at cities that are more alike across states than the studies reviewed here. As the debate over the future of the local property tax heats up in the very near future, such information will be critical to ensure that the public dialogue is accurately informed about the actual current use of the property tax.

NOTES The authors would like to thank Pat Atkins, John Bowman, Daphne Kenyon, and Joan Youngman for comments on an earlier draft of this paper. We would also like to thank Robert Ebel and his colleagues at the Office of Revenue Analysis in the DC Office of the Chief Financial Officer for comments made and feedback provided on an earlier version of this paper at a seminar they or- ganized to discuss this topic. Public Budgeting & Finance / Summer 2009 136