Tuesday, December 21, 2004

When the Rich Cheat On Their Taxes, It's Not Their Fault

We saw in our last post that the pure economic model of tax compliance is a failure. Although game theory explains why cheating may at times be a rational strategy, it does not illuminate the actual world that surrounds us. Face-to-face tax audits are, at present, somewhat less likely than winning roulette bets, and infrequent penalties make the cost of being detected only slightly greater than the tax payment that would have to be made anyway, so economic logic implies success for crooks. Lax enforcement means that the burden of taxation shifts away from income or wealth, and becomes instead a charge on honesty.

monopolyIn such an environment, bad behavior should drive out the good. What use is it to be upright, if all our neighbors are scoundrels? Yet in disregard of the canons of self-interest, voluntary compliance with the income tax system still hovers somewhere around 85%, and the (admittedly dated) last Taxpayer Compliance survey in 1988 found that those with incomes above $500,000 remembered to report fully 97% of the income they received. People, it seems, act better than they ought to: they file and pay even when they shouldn't.

Classic economic analysis takes its strength from the use of Newtonian physics as a model. Both disciplines use the calculus of integrals and differentials to compute changes in the magnitude and position of observeables acted on by quantifiable forces in a frictionless vacuum. However, we do not live, work or consume in outer space, but in a world full of friction -- were it not for friction, we could not even stand upright. Economic abstractions often have the same names as things that are real, so it is easy to confuse them, and attempts to apply classical economic analysis to public policy all too often consist of implausible assumptions followed by brilliant mathematics leading to unpalatable conclusions.

Let's see what we can learn from those who study the friction-filled world of social structures.

The key insight is that tax evasion, at least at the high end, is not a solitary endeavor. We are not born with an intuitive sense of the Natural Law way to fill out Form 1040. Rather, compliance and noncompliance are both learned behaviors. The American public environment includes a very visible set of institutions of tax compliance, ranging from the Internal Revenue Service through H&R Block. There is a minor industry of compliance technicians, such as attorneys, accountants and tax preparers, with their own professional associations and specialized publications filled with technical jargon. Almost 75 million individual tax returns filed in 2003 (57.2% of the 130.8 million total) were signed by paid preparers, who were responsible for 77.1% of the returns reporting income over $200,000. And even tho the deduction for tax preparation fees is hedged with restrictions (it’s only for those who itemize deductions, and only if such fees plus other miscellaneous expenses exceed 2% of income, and only after a further phase-out if income is over @ $130K, and only if there’s no Alternative Minimum Tax to worry about), there were 21.5 million returns that year claiming a total of $4.8 billion for such expenses.

The elements that induce deviant behavior – the patterns of tax avoidance and gorillaevasion – are less visible. (This may be because our attention is directed elsewhere, as in the classic experiment where subjects told to closely count basketball practice passes did not notice the gorilla-suited woman wandering across the court.) Much of our knowledge comes from the old (the last was in 1988 with a 1992 update) Taxpayer Compliance studies, which found that only about 82% of true tax liability was voluntarily and timely paid. (IRS Publication 1415, Rev. 4/96, p. 11.) The largest part of this tax gap – over half -- was due to under-reported income, and most of the money that “hit the floor” rather than the returns was business income: in year 1992 an estimated $188 billion of such income was "misreported", resulting in a $39 billion understatement of tax.

The largest single part of the tax gap was the $73 billion omitted from "nonfarm proprietor income". In other words, small independent business entrepreneurs, the keepers of the American dream, paid about $17 billion less than the amount the law required. In the aggregate, according to the TCMP, only 68% of their income found its way onto tax returns. (Id., p. 18, table 3). farmThey shared a common attitude towards tax reporting fidelity with that other icon of traditional values, the American farmer. Farmers, on average also reported only 68% of income, understating the total by $18 billion so as to retain roughly $3.3 billion of public money for their own benefit. E-I-E-I-O. The least compliant of all were the "informal suppliers”, who:

"Provide products or services through informal arrangements which frequently involve cash-related transactions or "off the books" accounting practice. Child-care providers, street-side vendors, and moonlighting professionals are among this type of nonfarm sole proprietor." (id, p. 53)

According to the TCMP, the informals reported only 19% of their income, an understatement of $60 billion which avoided $12 billion in tax payments. By contrast, landlords 'fess up to 83% of what they actually collect, sellers of real property and other business assets report 72% and alimony recipients report 87%. The sucker prize for compliance goes to the employed middle class. The IRS found that over 99% of wage and salary income is correctly reported on tax returns. (Employee pay constitutes 82% of all income for those making between $20,000 and $100,000 -- by contrast, those with $200K plus have a large share of income generated by capital, so only about half their income represents pay for work.)

Employees have taxes withheld before the paycheck touches their fingers. Those taxes are remitted to the government by employers or employers' payroll agents, along with periodic reports in a format for computerized matching. Most employers have an incentive to fully report employee pay, because this amount is deducted when the employers calculate their own tax liability. (Incentives are weaker is the employer is a tax-exempt non-profit or when employees are also the owners.) Third-party self-interested reporting combined with withholding at the source (using tables designed to over-withhold, with annual refunds as the reward for filing returns) is the key to successful income tax administration in the US. Two-thirds of all taxpayers receive refunds, and for them April 15th is more like Christmas than Doomsday. Third-party reporting without withholding is not quite as effective: the TCMP tables show compliance at 97.8% for interest income and 92.3% for dividend income. The numbers for pensions, unemployment compensation, and Social Security benefits fall in the same range.

What accounts for these differences in tax behavior between street vendors, wage earners, pensioners and small-business entrepreneurs? One could search for inbred patterns of criminality or psychological roots of deviance, but tax cheating seems instead to be the result of opportunity and example. It is a classic form of "white collar" deviance, akin to criminal behavior by those who, due to their status, do not conceive of themselves as criminals. The landmark description of the sociology of criminal behavior of business corporations and high-status executives is Edwin Sutherland's work "White Collar Crime" (Yale Press, 1983 ed. orig. pub. 1949), a work worth quoting at length. Sutherland's "differential association" hypothesis was:

"That criminal behavior is learned in association with those who define such criminal behavior favorably and an isolation from those who define it unfavorably and that a person in an appropriate situation engages in such criminal behavior if and only if the weight of the favorable definitions exceeds the weight of the unfavorable definitions...."(p. 240)

"In the process of learning practical business a young man with idealism and thoughtfulness for others is inducted into white collar crime. In many cases he is ordered by managers to do things which he regards as unethical or illegal while in other cases he learns from those who have the same rank as his own how they make a success. He learns specific techniques of violating the law together with definitions of situations in which those techniques may be used.... also, he develops a general ideology to assist the neophyte in business to accept the illegal practices and provide rationalizations for them." (p. 245)

Sutherland offers examples of the diffusion of illegal practices from one firm to the next, such as fraudulent advertising of interest rates by automobile manufacturers and the role of the American Gas Association in helping its members misrepresent the heating value of their product. His key point -- one often missed -- is that white collar crime is not the result of individual "bad apples", but is a normal consequence of the “rationalistic, amoral and non-sentimental behavior” of the corporate form of business, organized as a pure economic entity. (p. 236) Position, not personality, is what matters:

"Variation among corporations in the violations of law are due principally to position in the economic structure…. many corporations violate the antitrust law in certain industrial areas and not in others …. Many corporations which violated the antitrust law 40 years ago are still violating that law although the personnel of the corporation has changed completely" (p. 262)

"The rationality of the corporation in relation to illegal behavior leaves it to select crimes which involve the smallest danger of detection and identification and against which victims are least likely to fight.... A second aspect of corporate rationality is the selection of crimes on which proof is difficult.... The rational corporation adopts a policy of "fixing" cases.... The fixing of white collar crimes is much more inclusive than the fixing of professional thefts. The corporations attempt to prevent the implementation of the law and to create general goodwill as well as deal with particular charges.”

"Cui Bono?" -- "So Who Benefits?" We'll have more to say in a later post about corporate tax shelter cheating and how wealthholders, investors, pensioners and endowments may innocently benefit from corruption by proxy. Let's focus for the present on individual business owners and other less well organized wearers of white collar status.

Not all farmers or all entrepreneurs cheat, but a 32% income misstatement rate means that noncompliance is fairly widespread. The distribution of tax noncompliance by occupation is highly uneven, indicating that differences in position and opportunities to learn by example play a role here, as they did for Sutherland's executive crooks. A recent (2003) study shows that folks involved in vehicle sales have the highest level of noncompliance, paying less than half the taxes they should. (This finding is unlikely to hurt the reputation of used car salesmen, since polls show that their public esteem can hardly decline any further.) Some of the other rankings are predictable (accountants and tax preparers themselves are the most compliant, at 95%, with college professors closely behind at 94.1%), while some are suprising (social and religious workers have a 79.2% compliance rate, just above construction workers and below the 80.6% for folks in "real estate, finance and insurance.") Preachers, it seems, are no better than plumbers. Those who wish to see how other occupations stack up should study the charts here and here.

These findings contrast with the high (95% - 97%) compliance rates for high income (>$200K) taxpayers that emerged from the last TCMP round back in 1988. The reasons bear investigation. Perhaps the economy has changed since then. Perhaps entrepreneurs cheat while clawing their way up to the top, but turn honest once they reach the plateau of economic security. Perhaps high income entrepreneurs are not in the sole proprietor tax return classification, but conduct their business activity through closely held corporations, partnerships and limited liability companies. The owner of a business entity may be compliant, in the sense of accurately transcribing the numbers reported by the entity, even if the numbers themselves, as generated by the business entity, are laced with fiction.

So it seems that folks are inclined towards honesty or bent towards crookedness not by rational calculation, but by the osmotic pressure of position and the opportunity to learn from one's associates. Those who work for pay that is withheld and reported are denied the opportunity to evade, and have no use for learning the finer points of fraud. By contrast, when a small business owner receives funds personally, in cash or anonymously negotiable instruments, under circumstances where the person paying has no obligation to tell the authorities anything, the opportunities are obvious. Many no doubt resist temptation because of the warnings of conscience -- “the inner voice that tells us that someone may be looking” -- but some are receptive to learning from friends or family members who are themselves in the know.

One gets a feel for the world of small business evasion from a funny, gritty paper presented by Professors Bankman and Karlinsky at a recent Conference on International Tax Administration(1). Their interviews illuminate the shadowy world of tax deviance.

"We find that cash business owners who misreport net income rely primarily on their own wiles, and on advice of family, friends and associates. However, such business owners believe their accountants are aware of the discrepancy between actual and reporting income; some of these business owners require the active assistance of their preparer in their underreporting.”

They confirm that misreporting is learned behavior:

"When prodded, however, most business owners reveal the practice of misreporting, or at least the mechanics of misreporting, was learned from family and friends. The Storekeeper, for example, credits his father as a role model, and takes pride in the way he has counseled friends and associates. “I tell people everything,” he says. “Like never, ever deposit the cash.” Other business owners showed a similar readiness to advise others, particularly those with whom they are close. “When I saw how stupidly [my cousin] was taking the cash, in front of employees, I almost died,” says one interviewee. “Lucky I got to him in time.”

So why don't all small business owners cheat? It turns out that making money without working for it can be hard work. One must be circumspect: when the boss is seen to have sticky fingers the rest of the workers may follow the example, and it is not easy for a thief to report being robbed. It is hard to plan and operate a successful business on the basis of dishonest records or no records at all, so the successful crook may have to take on the difficult and perilous task of maintaining two sets of books. The Professors point out more problems for evaders, in that business growth may be limited by lack of access to commercial financing (banks want to see tax returns). The successful cheat must also deal with the Midas Curse. How does one enjoy holding hot money? Gold and gold alone is of little use; it's too heavy to carry and too tough to chew.

“ Perhaps the most important learned piece of information for misreporting business owners is that it is imprudent to deposit unreported cash, or use the cash to make large investments or consumption purchases (e.g., stocks, rental property, personal residences)….. One preparer noted that cash business taxpayers may have homes with ordinary exteriors, but inside will be clothing, jewelry, artwork, rugs and furniture that may be two or three times more valuable than the structure or land. Even still, cash will accumulate, with some business owners accumulating hundreds of thousands of dollars in bank safe deposit boxes, or hidden in the home. “

Bankman & Karlinsky also explain why, contrary to intutition, market forces tend to rob cheaters of their gains from cheating. They reason that widespread noncompliance in the cash business sector of the economy draws labor and capital away from the rest of the economy. Increased competition in the noncompliant cash business sector then tends to drive rates of return down to the point where the tax savings from evasion is offset by lower underlying profitability. {It's the economy as a whole that loses, due to the misallocation of resources}. This theory finds support in “Business Profitability and Income Tax Compliance” by Kim Bloomquist of the IRS (IRS Research Bulletin Pub. 1500 at p. 101 (1999)), which noted a very high statistical correlation between net profit margin and the level of unreported income ( -0.90 for nonfarm proprietors and -0.774 for farm proprietors.) Bloomquist thinks "this may indicate some businesses initiate or increase their level of tax evasion activity in response to lean economic times", but the correlation may also run the other way: businesses in noncompliant sectors have lower rates of return.

Coming next time: When the Rich Cheat on Taxes, They Cheat Honestly.
(1) Bankman, J. & Karlinksy, S. (2002) ‘Developing a theory of cash business tax evasion behavior andthe role of their tax preparers’. Paper presented to the 5th International Conference on TaxAdministration, ATAX, University of New South Wales, Sydney, April.

Friday, December 03, 2004

The Rich Don't Cheat Like They Should

The last posting condensed our review of the literature on the tax compliance behavior of high income people into two paradoxical statements, to which we now add three more:

goldenThe Rich don't cheat as much as they should.
[According to the economists, that is.]

When the Rich cheat, it's not their fault.
[Say the white-collar crime sociologists.]

When the Rich cheat, they cheat honestly.
[Says the Law].

The Rich don't cheat -- they have people to do that for them.
[Say the accountants.]

The Rich don't need to cheat.
[According to the Senate and House of Representatives.]

Let's see what sense, if any, underlies these surprising statements.

1) The classic economic model of tax behavior posits a crowd of rational sociopaths ("taxpayers") who weigh the savings from successful evasion against the expected value of the penalties charged on those occasions when their evasion is detected ("audited"). For example, if some contemplated crooked dodge would reduce tax by $1,000 but have a 20% chance of detection by audit and generate a penalty for evasion equal to 200% of the tax, then the expected benefit from cheating is $800 (80% success x $1,000) and the expected cost is $400 (20% times the additonal $2,000 penalty), so a rational, risk-indifferent taxpayer would go ahead and cheat. More sophisticated models use "utility" (i.e. whatever it is that rational sociopaths try to maximize) instead of dollars, discount to present value, adjust for risk-aversion, introduce uncertainty as to detection probabilities, reflect changes in risk-tolerance and tax rates as a function of increasing income, and so forth.

taxdragonFortunately for the US Treasury, in the real world the amount of evasion tends to be much, much smaller than such models would predict, given actual audit probabilities and penalty amounts. For example, of the 130,341,159 individual income tax returns filed in year 2002, only 849,296 ( 0.65%) were examined by the IRS and only 206,457 of these involved face-to-face meetings, as opposed to correspondence generated by computerized data matching programs. Concern over abuse of the Earned Income Credit (a subsidy for the working poor that is implemented thru the income tax system) means that poor people (returns showing income under $25,000) have the *highest* chance of audit. The audit rate for those who are better off -- such as the 2,084,855 returns showing self-employment receipts over $100,000 -- was 1.47%, and these audits resulted in total proposed additional tax of $581 million, not quite enough to pay one day's worth of interest on the National Debt.

While the Tax Code provides a penalty for negligence equal to 20% of the tax, and a 75% penalty for civil tax fraud, these penalties are rarely charged against individual income taxpayers (2,401 penalties in year 2003 equals once per 50,000 tax returns) and the total amount, after abatement for reasonable cause etc., was only $63.4 million -- which someone still has to collect, and which would only cover the cost of running the IRS itself for about two days. In addition, while the prospect of doing hard time may deter the timid, only 384 people -- one taxpayer in 300,000 -- went to jail in year 2003 for tax crimes involving legal source income, and such tax prosecutions cost the government an average of over $100,000 each. To sum up, with an audit probability of 1.5% and the likely punishment simply being made to pay the tax that was due anyway, rationality demands dishonesty. It's a good thing that most people, including most rich people, aren't rational.

A substantial part of the American public seems to believes that the rich don't pay much tax. Professor Slemrod reports one survey where "respondents believed that 45% of millionaires paid no income tax at all, when IRS statistics showed that the actual figure was less than 2%." (1) The public misconception was reinforced by President Bush himself during the recent election campaign when he explained to the assembled rubes:

"The really rich people figure out how to dodge taxes anyway."
"Most rich people are able to avoid taxes, and if you can't raise enough money from taxing the rich, guess who pays the taxes? Yes, you do. "
"The rich hire lawyers and accountants for a reason, to stick you with the tab."

This kind of Presidential nonsense only feeds the cynics. Of course it's not true.

Let's look at the numbers compiled by the IRS Statistics of Income Division. In year 2002 a total of 2,414,127 returns showed adjusted gross income over $200,000. Only 1.86% of all returns were at this income level, but with total income of $1.251 trillion they represented 20.75% of all income. After taking all deductions, 26.4% of all taxable income was on these high-income returns, and their income tax came to $323,977,221,000 -- 40.6% of all income tax that year was on these 2.4 million high-income returns.

"But Professor Tax", you say, "those numbers are for ordinary high income people. What about the real millionaires? What about the 168,977 tax returns that showed income of more than $1,000,000 in year 2002?" Indeed, 0.13% of all returns -- one out of every 770 -- were filed by such folks, who got 7.89% of all the income, 10.34% of all income after deductions, and paid 17% of all the tax, amounting that year to $135,841,970,000. Ouch! No wonder they hate the income tax so much. At the top of the top we've the 5,309 returns showing income over ten million dollars -- 0.0041% or one in every 25,000 -- who had 2.41% of all the income, 2.78% of the taxable income, and paid 4.23% of all the tax. The amount of tax was $33,737,749,000, which works out to an average $6.5 million each. The hand bleeds at the very thought of signing such a check.

chartThis, however, is only what the rich admit to in their tax filings. Maybe there are huge pools of underground cash still hidden from the eyes of the revenuers? Isn't there some substance to all anecdotal evidence about abusive tax shelters and billionaire criminal kingpins?

The IRS' "Audits from Hell", the Taxpayer Compliance Measurement Program in operation between 1969 and 1988, estimated overall income tax compliance at 85% to 90%, which meant a "tax gap" of 10% - 15% between what should have been paid and what actually was paid. An estimate of the components of the tax gap for year 1992 (2) found it was due to:

24 percent - self-employed individuals who do not report all income subject to tax

24 percent - other individuals who do not report all taxable income

19 percent - corporations with assets of $10 million or more that understate their tax liability

9 percent - individuals who do not remit all taxes reported due on their returns

8 percent - individuals who do not file a return

6 percent - individuals who take excessive deductions

6 percent - corporations with assets of less than $10 million that understate their tax liability

4 percent - attributable to other reasons

One study that analyzed data from the last TCMP in 1988 concluded that high income (over $100,000 AGI) returns had a higher rate of voluntary compliance than other returns, and caponereported over 95% of true income. Returns with income over $500,000 had an even higher compliance rate, reporting over 97% of true income. (3) (Note tho, that this was at a time when the 1986 Tax Reform Act had shut down one generation of tax shelters, and the modern abusive shelters were still taking shape.) And the aging Godfathers and rising capos of the criminal underworld are well advised to reach a truce with the Treasury, since one of the proudest claims in the history of the tax police is still: "We got Capone."

Next time: Part 2 - The Sociology of Tax Noncompliance

Unlinked notes:
(2) Government Accounting Office Report GAWGGD-95-157 on Taxpayer Compliance, June 1995

(3) Slemrod, n. 1 at 10, citing Charles Christian article in IRS Publication 1500 (1993-94)

This page is powered by Blogger. Isn't yours?