How The Chronicle’s Rankings of Generosity Were Compiled

One basic premise underlies The Chronicle’s analyses of charitable giving: Most previous analyses of

individual philanthropy have ignored differences in the cost of living in various parts of the country. Put simply, a dollar in New York City does not have the same value as a dollar in Tuscaloosa, Alabama.

The Chronicle conducted two studies designed to look at charitable giving.

The goal of the first study was to determine, on average, how much people in each of the nation’s more than 3,000 counties give to charity as a percentage of their discretionary income — defined as the amount of money left over after basic living expenses for such items as housing, taxes, food, transportation, and clothing are paid.

It also examines demographic differences among the counties to determine what effects those differences have on giving trends.

The analysis is based on an Internal Revenue Service compilation of tax returns for 1997, along with demographic data from the U.S. Census Bureau, cost-of-living data from the U.S. Bureau of Labor Statistics, and analyses of what people of different incomes and in different regions of the country pay in federal, state, and local taxes.

The second study, which also examines the percentage of discretionary income that individuals give to charity, is based on interviews the U.S. Bureau of Labor Statistics conducts quarterly with approximately 7,500 Americans to determine the typical expenses — including donations to nonprofit organizations — of people at all income levels and from all regions of the United States.

Itemized Tax Returns

The first study covers taxpayers earning $50,000 or more annually who itemize deductions on their federal tax returns, the only people for whom reasonably accurate data can be collected from the Internal Revenue Service.

Charitable deductions are reported to the federal government only by taxpayers who itemize their deductions.

While some people at all income levels itemize, the percentages of people who do drops sharply with income. More than 75 percent of taxpayers with adjusted gross incomes of $50,000 or more itemize their deductions, for example, but less than 20 percent of those earning below that amount itemize.

The IRS is barred by federal law from releasing information about individual tax returns, but it did conduct a broad analysis of individual income-tax returns filed during the 1997 tax year, breaking the data into two categories: those for people with incomes below $50,000 and those with incomes of $50,000 and above.

For every zip code, the agency calculated the number of returns filed, the number of returns on which deductions were itemized, the total adjusted gross income for each group, and the total amount of deductions itemizers claimed for charitable donations.

The Chronicle used only the data for people with incomes of $50,000 or more.

Living Costs

To determine living costs, the study started with Census Bureau data for each zip code on the average housing costs paid by people who have mortgages. The census data include annual mortgage payments, property taxes, and other basic housing expenses such as utility payments. The Chronicle used data for homeowners with mortgages (and not data from renters) because 87 percent of taxpayers who itemize their deductions have mortgages.

After converting data by zip code into data by county, the study then calculated the average adjusted gross income, average housing costs, and average charitable donations for each county.

All other living costs — including food, transportation, clothing and other basic needs — were estimated using Bureau of Labor Statistics data. The bureau releases cost-of-living data for 23 major metropolitan areas, and those costs were applied to all the counties falling within those areas.

Other living costs were estimated based on the bureau’s quarterly interviews with a sample of 7,500 Americans. Because of the small size of that sample, the bureau’s data cannot be used to calculate the actual living costs for every county. Instead, the survey numbers can be grouped only into regions (East, Midwest, South, and West), by size of the metropolitan area, and by urban and rural settings.

While this is only a rough way of estimating living expenses, The Chronicle and its statistical consultant, Arthur Kirsch, professor emeritus of statistics at George Washington University and senior consultant to Independent Sector, feel that the estimate is fair.

Tax Payments

To figure the average federal taxes paid, the study used IRS data on people of different income levels. The average tax was determined using the average gross income for each county. Average state and local taxes (not counting property taxes, which are included in the census data) were estimated using a study compiled by the Institute on Taxation and Economic Policy, in Washington.

The average housing costs, living expenses, and taxes were then subtracted from the average adjusted gross income for each county, leaving an average discretionary income. The average charitable donation was then divided by the average discretionary income to figure the average percentage donated to charity.

Census Bureau demographic data were also added for each county. The data included (only for households with incomes of $50,000 or more) the race and ethnicity, ages, gender, and marital status of the heads of households. For each zip code, The Chronicle calculated the percentage of households headed by whites and blacks, the average age of the heads of households, the percentage of households headed by married men or single women, and so on.

Other potentially relevant demographic data — including education level and main language spoken in the household — were not available, because the Census Bureau does not break them out by income level.

To assess the role religion plays in charitable giving, The Chronicle included data from the report “Religious Congregations and Membership in the United States,” published in 2000 by the Glenmary Research Center, in Nashville. The center is affiliated with Glenmary Home Missioners, a Catholic missionary and service group in Cincinnati. The report is based on surveys of religious participation in every county.

Labor Statistics Survey

For the second study, The Chronicle obtained the U.S. Bureau of Labor Statistics quarterly survey data for the second, third, and fourth quarters of 2001 and the first quarter of 2002. (The bureau only began including questions about charitable giving in the second quarter of 2001.) In addition to information on income and spending, the BLS data also include various demographic data about the heads of households in the survey, including race, ethnicity, age, marital status, education, type of employer, geographic region, and community size.

As in the first study, living expenses were subtracted from adjusted gross income to determine discretionary income, and the average percentage donated to charity was calculated by dividing the average charitable donation by the average discretionary income.

The bureau’s survey asks three questions about charitable giving: one each on donations to religious organizations, donations to secular charities, and donations to educational institutions. The Chronicle combined donations to secular charities and educational institutions into a single category: nonreligious charities.