Who’s most likely to dodge taxes?

The Panama Papers leaked the names of tax avoiders, but they’re only part of the problem. Researchers are on the case.

Credit: Kyle Platts

Alina Dizik | Nov 21, 2016

Sections Economics Public Policy

Collections Tax Policy

Take a peek at the tax returns of Greece’s self-employed, and many of the people who filed those returns will seem like poor contenders for a loan. On paper, Greece’s lawyers, doctors, and accountants are struggling, yet banks continue to take a chance on lending to them at interest rates that don’t appear to reflect the risk of default implied by the tax returns. 

The reason: Greek citizens understate their earnings to tax authorities, but tell banks a more realistic income. Armed with that knowledge, UMass Amherst’s Nikolaos Artavanis, University of California at Berkeley’s Adair Morse, and Chicago Booth’s Margarita Tsoutsoura used bank figures to back out how much money is going unreported to Greece’s tax authorities. They estimate that the country had €28 billion in unreported income in 2009, which added up to 31 percent of the annual deficit, and an additional €11.2 billion that could have been collected to offset the deficit. 

Welcome to the murky world of tax evasion and avoidance, where the amount people avoid in taxes is hard to track because, after all, “it’s an activity that’s meant to be hidden,” says Tsoutsoura. In the past seven years, she and other researchers in academia and in the government and nonprofit sectors have been on a global hunt to document what individuals and companies are evading—or legally avoiding paying. 

Eschewing official tax returns, researchers are calculating missing taxes by creatively slicing data and working with third-party sources to cross-reference information. The end goal is to push governments to change complicated and often-ineffective tax systems. 

The Panama Papers 

On a global scale, the amount of capital floating out of regulators’ sights is staggeringly large, says Copenhagen Business School’s Brooke Harrington, author of Capital without Borders: Wealth Managers and the One Percent. 

Some of that is due to tax avoidance practiced by the very rich, which the “Panama Papers” have helped corroborate. In the Panama Papers, the names of 140 officials from 50 countries were leaked from a single wealth-management firm, Mossack Fonseca. “We don’t appreciate how much they have in common with each other—whether it’s a Russian oligarch or some Asian dictator—it’s like they are all going to the same hairdresser,” says Harrington, who has studied the professionals who create tax-avoidance strategies.

Many wealth managers, including those at Mossack Fonseca, work with the world’s 168,000 ultra-high-net-worth individuals, who have at least $30 million. Most practices of wealth-management firms are legal and make use of offshore legislation to pursue tax avoidance—sometimes in countries that are not known as tax havens. “Wealth managers are like hackers,” Harrington says. “They are hacking the legal systems of other countries.” 

This year, several EU member states have put more pressure on tax havens to encourage transparency by creating a blacklist of territories with weak banking laws that don’t comply with the Organisation for Economic Co-operation and Development’s (OECD) Automatic Exchange of Information standard, says Vanessa Mock, European Commission spokesperson for the Taxation and Customs Union. Because being listed as a tax haven can affect long-term investment in a country, Panama, Lebanon, and the South Pacific island nation of Vanuatu have all recently signed up to implement international tax standards.

While avoidance is legal, evasion is not. And the price of evasion is high. The United States alone had an average $458 billion gross tax gap annually for the 2008–10 tax years, with $319 billion of that due to individuals, according to the Internal Revenue Service. Evasion can be particularly problematic for developing countries. About $1 trillion illicitly leaves developing countries every year, according to Global Financial Integrity (GFI), a Washington-based nonprofit. “A good chunk of [illicit financial flows] is likely related to tax evasion, but we don’t have the resources here to go further to determine how much,” says Matthew Salomon, a senior economist at GFI.

Besides draining government coffers and inhibiting development, “evasion leads to distortions in the economy between activities that are taxed, and others that are not,” says Harvard’s Dina Pomeranz. “It can push people into activities that are easier because of evasion but economically inefficient.” For example, smaller companies—not obligated to track their revenues or spending in much detail—may stay purposely small to evade tax collectors.

But despite the problems caused by tax nonpayment, “there’s hesitation on the part of some political elites to look at this,” says Gail Hurley, a New York–based policy specialist for development finance at the United Nations Development Programme (UNDP). “Governments are wary if you start digging into these activities.”

Follow the paper trail

But researchers are digging anyway as they attempt to quantify the problem. 

Traditionally, authorities have attempted to track tax evasion through random auditing, but private-sector data are allowing a new approach. “Because the private sector adapts to a culture of tax evasion, private sector data offer a window into the magnitude of, distribution of, and motivation for tax evasion,” Artavanis, Morse, and Tsoutsoura write.

One of the first studies to use private data was conducted in 2010 by Carnegie Mellon University’s Serguey Braguinsky (currently at the University of Maryland), Clemson University’s Sergey Mityakov, and then–Harvard PhD candidate Andrey Liscovich, and compared car registries in Moscow with car owners’ salary records. By looking at the price of the car, the researchers could detect unreported income. This kind of data provides the ability to cross-check official sources of income with private, and perhaps unexpected, data sources.

Not all the activity that turns up in cross-checking is nefarious. In a 2012 study, Chicago Booth’s Erik Hurst sought to determine whether researchers could trust household data. In comparing spending data to reported income data for self-employed Americans, he finds that the self-employed generally underreported income by 25 percent; but he doesn’t assume that’s due to evasion, which implies intentionally either understating your income or overstating your deductions.

Computing the incomes of self-employed workers gets complicated, as they or their accountants have to identify business and personal expenses. Income-computing mistakes in this group often result in lower taxes. “Given the complexity, errors tend to accrue in a way that leads income to be underreported,” says Hurst. “Some of it may be nefarious, while some of it may be just systematic errors when income is being reported. My data does not allow me to distinguish between the two. Given this, I cannot conclude that explicit evasion is taking place.”

Hurst finds that privately collected data about how the self-employed spend their money are more accurate than government data. “People don’t skew their expenditure data,” he says. “They eat a certain amount of food, drive a certain car, and [from that] I can infer their income.”

Artavanis, Morse, and Tsoutsoura collected private banking data about people who applied for credit. The anonymized data included incomes, debts, occupations, and more. 

They saw that professions that generated the least amount of data—leaving little in the way of paper trails—had the highest incidence of evasion. That was particularly true for the self-employed. Self-employed doctors, engineers, and educators generated relatively few receipts and documents, and were more likely to evade taxes. People working in the pharmacy and transport industries, meanwhile, had naturally higher paper trails and were less likely to evade taxes.

Tax evasion across industries
In a study of the magnitude of tax evasion in Greece, researchers developed a measure of how well income can be traced from industry to industry. They find a weaker paper trail in many industries that see higher amounts of tax-evaded income.

Tax-evaded income

Paper-trail intensity

Scale: 35,000 euros

2.5 (high)

2.16

Medicine

Law

Engineering

Education

1.28

1.12

1.04

0.6

Media

0.28

Lodging/Tourism

-0.03

Accounting/Finance

-0.11

Business services

Construction/

Transport

-0.84

Retail

-1.12

-1.22

Manufacturing

-1.26

Personal services

Farming

-2

0 euros

-2.5 (low)

Source: Artavanis et al., 2015

Tax evasion across industries
In a study of the magnitude of tax evasion in Greece, researchers developed a measure of how well income can be traced from industry to industry. They find a weaker paper trail in many industries that see higher amounts of tax-evaded income.

Paper-trail intensity

2.5 (high) to -2.5 (low)

Tax-evaded income

Euros

Medicine 32,548

Law 30,979

2.16

Engineering 29,565

Education 27,116

1.28

1.12

1.04

0.6

Media 20,349

0.28

Lodging/Tourism 17,237

-0.o3

Accounting/Finance 16,727

-0.11

Business Services 16,665

Construction/

Transport 12,862

-0.84

Retail 11,855

-1.12

-1.22

Manufacturing 7,271

-1.26

Personal Services 6,374

Farming 6,121

-2

Source: Artavanis et al., 2015

The likelihood of evasion increased alongside wealth. Of the professions studied, doctors, private tutors, engineers, lawyers, accountants, and financial-service agents had the highest likelihood of evasion, with an average €24,000–€30,000 evaded per person. 

Researchers can also cross-check information using other government data, even taxation data. Harvard’s Pomeranz studied Chile’s valued-added tax (VAT), paid by more than 400,000 local companies and assessed on products depending on their stage of production and sale. VAT, she finds, is part of a long-established tax system that generates more information than a retail sales tax, making it harder to game. The structure of the VAT is such that it leaves more of a paper trail. Because input costs are tax deductible, companies have an incentive to request receipts from suppliers.

GFI looks at trade between countries, and in particular at aggregate merchandise trade numbers, to draw conclusions about illicit financial flows. In ongoing analysis, it finds that in some cases, companies underreport imports to evade import duties, while at other times they inflate imports and pay more duties in order to depress income. For example, to avoid export duties, a company may underreport the number of pairs of socks it exports. Another company, to lower domestic revenue and associated taxes, might overreport the number of socks it exports. Tracking trade is “a very gross, crude way of coming up with a snapshot” but does not account for the sizable evasion that happens, for example through crypto-currencies including bitcoin, or the simple smuggling of cash, says GFI’s Salomon.

Changing policy in Greece and beyond

While the growing body of data is helping policy makers create deterrents to tax evasion, those hoping to evade taxes are continuously seeking new ways to escape their financial obligations. For researchers and government officials, the result can feel a little like an ongoing game of cat and mouse. “When you start cross-checking, companies in particular move to other, less monitored forms of evasion,” says Tsoutsoura. If one country uses third-party data to find tax evaders, that can encourage individuals and companies to find better places and ways to cheat.

In 2011 and 2012, George Washington University’s Paul Carrillo, Harvard’s Pomeranz, and University of California Davis’s Monica Singhal studied 8,000 companies in Ecuador that were notified by the country’s Servicio de Rentas Internas about tax discrepancies between previously filed returns and third-party sources such as credit-card sales or exports. The government hoped to get companies to pay up, but the plan didn’t work as intended. When companies received notices about discrepancies, some did at first declare higher income in subsequent returns. But they then found another loophole. “They offset much of this higher declared revenue by an increase in declared costs, resulting in only small changes in their reported profits,” the researchers find. For every dollar in tax revenue adjustment, companies in question increased their reported costs by 96 cents, and thus avoided raising their tax bills. 

Who’s most likely to dodge taxes?

But tougher economic times have inspired governments around the world to push harder to track lost revenue, feeding a growing movement to crack down. In the past two years, there has been more interest in tracking how money flows out of the least-developed economies in parts of sub-Saharan Africa, says UNDP’s Hurley. “Civil society has mobilized quite strongly around tax evasion because damages are more severe when the countries are poorest,” she says. 

In 2015, member states at the UN’s Third International Conference on Financing for Development, in Addis Ababa, Ethiopia, publicly agreed that agencies should focus on stopping illicit financial flows, especially from the world’s poorest economies. The UNDP paired with the OECD to start a Tax Inspectors Without Borders program, which works with academics and tax specialists around the world to help developing countries build out their ability to effectively collect taxes and audit multinational companies. 

Greece lost $261 billion in illicit financial flows between 2003 and 2011, according to GFI. Tsoutsoura’s research has gotten what she says is a surprising amount of publicity and resulted in collaboration with government officials. The data popped up in parliamentary tax-policy deliberations and have played a role in tax-related reforms as the Greek government-debt crisis unfolds. Greeks are no longer turning their backs on tax evasion as they did prior to 2009, she adds. “In the past it was much more acceptable if you were underreporting; now people realize that it should not be acceptable.”

In the past year, the government has started using third-party banking data to uncover potential forms of tax evasion, says Tsoutsoura, a native of Greece who travels between there and the US throughout the year. Recent regulatory changes have made it easier for authorities to access this type of financial data, she adds. 

Greece’s self-employed citizens now need to pay taxes on their entire income—unlike wage workers, who still get €9,000 tax free. Still, there’s plenty of room for improvement. The research suggests that professions with the highest rates of tax evasion are aligned with the occupations of parliamentarians, who may not want to pass more stringent laws because of personal conflicts of interest. 

The emerging body of research is encouraging governments, including Greece, to create better policing strategies. Many of the results from tracking Greece’s self-employed may be applicable elsewhere, Tsoutsoura says, especially in countries with pervasive tax evasion—such as Italy, Turkey, and parts of Latin America. “The more we have research that brings numbers to the table, the more people realize that this aggregates to a large amount.”