The taxpayer advocate, Nina Olsen, runs an independent office within the IRS. She got access to the data as part of an effort to learn more about why some taxpayers are more likely to cheat than others.
The study also looked at tax compliance in different industries, and found that people who own construction companies or real estate rental firms may be more likely to fudge their taxes than business owners in other fields.
Many of the communities identified by the study are very wealthy, including Beverly Hills and Newport Beach in California. Others are more middle class, such as New Carrollton, Md., a Washington suburb, and College Park, Ga., home to a section of Atlanta's massive airport.
Steve Rosansky, president and CEO of the Newport Beach Chamber of Commerce, said business owners in his city are probably targeted because many have high incomes. The likelihood of an audit does increase with income, according to IRS data.
"I imagine it's just a matter of them going where they think the money's at," Rosansky said in an interview. "I guess if I was running the IRS I'd probably do the same thing."
The study focused on small-business owners sole proprietorships, to be specific because they have more opportunity than the typical individual to cheat on their taxes. Many small businesses deal in cash while most individuals get paid in wages that are reported to the IRS.
The IRS only audits about 1 percent of tax returns each year, so the agency tries to pick returns that are most likely to yield additional tax money.
The IRS will not say much about how agents choose their targets. But as millions of procrastinators scramble to meet Monday's deadline to file their taxes, the agency is running every tax return through a confidential computer program to determine the chances of collecting more money from an audit.
Each tax return is assigned a score. The higher your score, the more likely you are to get audited because, according to the IRS, the more likely you are cheating on your taxes.
The score is called the Discriminant Inventory Function, or DIF. A high DIF score does not guarantee you are a tax cheat but the IRS claims it's reliable.
"If your return is selected because of a high score under the DIF system, the potential is high that an examination of your return will result in a change to your income tax liability," says an IRS publication that explains the auditing process.
How do you get high score? The IRS won't say, but veteran tax preparers and former IRS workers believe they have a pretty good idea.
"If you're reporting $8,000 of charitable contributions when you're only making $50,000, that's a red flag," said Bob Meighan, vice president of TurboTax, an online tax preparation service. "Likewise if you're reporting business or employee expenses that are out of the ordinary for your income range, that would attract the interest of the IRS as well."
The bottom line, according to the experts: People who take unusually large deductions for their income get a high score. Also, business owners who claim unusually large expenses for the size and type of their business get a high score.
"I had a case here where the person made about $40,000 and they claimed $25,000 of employment-related expenses," said Elizabeth Maresca, a former IRS lawyer who now teaches law at Fordham University. "Most people don't spend $25,000 to earn $40,000. That's an unusual number."
DIF scores can vary across industry, according to the study by the taxpayer advocate. For example, people who owned construction and real estate rental companies were more likely to have high scores. Lawyers, accountants and architects and people who provided other professional services were more likely to have low scores.
Olsen said construction and real estate rental companies probably deduct more expenses that are not independently reported to the IRS. The IRS does not like those kinds of expenses because they are harder to verify without an audit.
"Construction for sole proprietors has been historically a cash business," Olsen said.
The study, which was included in Olsen's annual report to Congress in January, used data from 2009 tax returns to plot the DIF scores for sole proprietorships across the country. The city where you live is not a component of the score, according to the study. Nevertheless, researchers were able to identify clusters of likely tax cheats.
Sole proprietorships make up about two-thirds of all U.S. businesses. Sole proprietors report business income on their individual tax returns and, the IRS says, they account for the biggest share of the tax gap, which is the difference between what taxpayers owe each year under the law and what they actually pay.
The tax gap was $345 billion in 2006, according the latest IRS estimate.
In all, researchers identified clusters of potential tax cheats in more than 350 communities in 24 states, mostly cities and towns but some neighborhoods, too. About one-third of them were in California, with most near Los Angeles and San Francisco.
Most of the others were in communities near Houston and Atlanta, and in the Maryland suburbs of Washington. There were relatively few in the Midwest or the Northeast.
The researchers also looked for areas with high concentrations of small business owners who were very unlikely to cheat on their taxes.
They came up with four: the Aleutian Islands in Alaska; West Somerville, Mass., a neighborhood in Somerville, a suburb of Boston; Portersville, Ind., an unincorporated town in the southern part of the state; and Mott Haven, a neighborhood in the Bronx, one of New York City's boroughs.
Stephen Mackey, president and CEO of the Somerville Chamber of Commerce, said he's glad the business owners in his community excel at civic virtue. But he was at a loss to explain why they stood out from so many others across the country.
"I'd like to think we're not alone in terms of the civic engagement of business people," said Mackey. "But I would say two things. One is they are very close to the community inside and outside their businesses. At the same time, it's not small town America. It's minutes from downtown Boston."
AP Director of Polling Jennifer Agiesta contributed to this report.
Online: National Taxpayer Advocate study: http://tinyurl.com/cjtgpt5
Follow Stephen Ohlemacher on Twitter: http://twitter.com/stephenatap
Who gets audited by the IRS?
By the Associated Press
The Internal Revenue Service only audits about 1 percent of individual tax returns each year. But the more money you make, the more likely you are to get audited. A look at which returns were audited in the 2012 budget year:
Individual returns filed in the previous year: 143 million.
Audited by mail: 1.1 million.
Audited in person: 360,000.
Audit rate: 1 percent.
Individual returns with incomes above $200,000: 4.8 million.
Audited by mail: 109,000.
Audited in person: 70,000.
Audit rate: 3.7 percent.
Individual returns with incomes above $1 million: 337,000.
Audited by mail: 23,000.
Audited in person: 18,000.
Audit rate: 12.1 percent.
Small corporation returns (assets under $10 million): 1.9 million.
Audit rate: 1.1 percent.
Large corporation returns (assets $10 million and higher): 60,500.
Audit rate: 17.8 percent.
Source: IRS enforcement: http://tinyurl.com/agwzcon