Yesterday was the first session of the 17th (!) NYU Tax Policy Colloquium. I am doing it with Alan Auerbach again, for the 3-½th time, and we have 27 students (the max allowed). Our Week 1 guest was Michelle Hanlon, with respect to her paper (co-authored with Maydew and Thornock), Taking the Long Way Home: Offshore Investments in U.S. Equity and Debt Markets and U.S. Tax Evasion.
The paper seeks to get a handle empirically on the existence and magnitude of illegal tax evasion by U.S. individuals that takes the following form. You want to invest in U.S. securities without paying U.S. income tax, so you establish a corporation in a tax haven (say, the Cayman Islands) and have it invest in U.S. securities. But you ignore the fact that such "round-tripping" gives you current year tax liability (or its present value equivalent) under the passive foreign investment company (PFIC) rules, obviously counting on the prospect that the U.S. tax authorities will not learn of your ownership interest in the offshore company.
Since researchers, no less than the IRS, can 't directly observe this illegal activity, one needs an empirical strategy to try to estimate it. The Hanlon paper finds two main things. First, when relevant U.S. tax rates go up (such as for individuals' ordinary income or long-term capital gains), inbound investment to the U.S. from tax havens tends to go up. The suggested explanation is that the higher tax rate increased U.S. investors' incentive to engage in fraud, but did not directly affect investors from other countries. Hence the surmise that perhaps the increased inbound capital flow may actually reflect round-tripping, and associated tax fraud, by U.S. investors.
Second, when certain agreements are reached between the U.S. and a tax haven country that may indicate an increased probability of detection for fraudulent evasion of the PFIC rules, inbound capital flows to the U.S. from the haven tend to decline. Once again, the surmise would be that, if only U.S. investors are being affected, it is a decent surmise that they are responsible for the change.
The paper makes a nice contribution by shedding some light on this issue, but it remains hard to tell just how much fraudulent round-tripping is going on. For example, one could concoct a scenario in which the U.S. tax rate increase induces U.S. investors shift to municipal bonds and tax-favored housing. So they sell securities to non-U.S. investors, who may invest through tax haven entities if they have other reasons for doing it this way.
Also, a higher U.S. tax rate could lead to round-tripping that is not associated with tax fraud. A paper that David Miller presented at the colloquium last year explored the various reasons why U.S. investors may choose to use tax haven entities, even when fully complying with the law (including the PFIC rules). For example, this structure may enable them to avoid itemized deduction disallowance or limitation rules that only apply to individuals. So we could potentially have the round-tripping story that the paper shows, minus the implication that it is mainly about tax fraud.
Likewise, suppose that there are U.S. taxpayers who have figured ways to engage in legal tax avoidance with respect to the PFIC rules, rendering them inapplicable in circumstances where one might argue that as a matter of policy they ought to apply. Once again, we would have the round-tripping story without the fraud.
Insofar as the data do indeed show non-trivial levels of tax fraud, the issue of what to do about it remains. Open questions include what sort of revenue estimate one might get for particular crackdown measures, and how things will change when the Foreign Account Tax Compliance Act (FATCA) takes effect next year.
Still, the paper presents a suggestive initial look at an area where there have been large gaps in our empirical knowledge.
I was at the evening session and it was a very interesting discussion. The one aspect of the paper that I do not fully understand is how the tax rate effect in this context can be isolated from the effect of the other variables. In domestic context it is often hard to separate the tax rate effect from the income effect. But here it seems impossible to look at the estimated evasion within different tax rate margins separately. Potentially, there could be some additional variables that are hard to measure specifically in this context. In particular, as was suggested during the discussion, it could be that the tax evader is choosing between different options of evasion, and thus particular tax rate could be influencing his choice of an evasion mechanism and not necessarily her decision to evade or comply. Also, although it is contrary to an intuitive notion, Allingham-Sandmo theoretical model of tax evasion that is used in the study (especially if supplemented by Yitzhaki’s changes) appears to be not conclusive on the issue whether tax rate increases have positive effect on evasion.
ReplyDelete