Keen and Perry: “Our study looks at a number of crisis-related issues and then tries to figure out what it implies for tax policy going ahead” (IMF photo)
Tax system’s debt bias
IMF Survey online spoke to Michael Keen and Victoria Perry of the IMF’s Fiscal Affairs Department about the study’s main findings and what role tax policy had in the global economic crisis.
IMF Survey online: Did tax policy play a role in bringing on the global economic crisis?
Keen: The bottom line of our current research is that, while taxation policies did not cause the crisis, they probably contributed to key vulnerabilities that became evident during the financial crisis. In particular, we see tax fingerprints on high leverage, on complexity, on the lack of transparency of financial arrangements, and on risk taking. In our paper, we study these issues and then try to figure out what that implies for tax design going ahead.
IMF Survey online: Did taxation policy play a role in the collapse of the housing market?
Perry: As far as we can see from evidence of others and from looking at these things ourselves, tax generally did not underlie the housing price bubbles. But it is really quite clear that the tax system has in some cases had a large effect on the overall level of household indebtedness, and also on the level of debt relative to home values. So if, for example, somebody with a $200,000 house would have a $100,000 mortgage in the absence of tax, the econometric evidence is that the same borrower would, with the tax system as it now stands in the United States, have a fairly significantly higher mortgage for that same house. This reflects the incentive, in the United States and in many other countries, for borrowers to reduce their income taxes through a deduction of mortgage interest. We don't draw from this the conclusion that we should instantly remove all these housing tax privileges. That’s clearly not something one would want to do until the housing market has come back to some kind of normality and health. But on the other hand, we certainly would be cautious in adding new privileges of this kind, which may lead to future problems. It could be useful now, in fact, to reduce housing transaction taxes, taxes that are paid when housing changes hands. These taxes are distortionary, in the direction of reducing the liquidity of the housing market.
IMF Survey online: What stands out as your most important findings?
Keen: One that we think is quite important is the incentive in almost all countries to use debt finance rather than equity finance at the corporate level. That’s because companies can deduct their interest payments against taxable income, but they can’t deduct the return paid to equity holders. So, at a corporate level, that causes a large bias to borrow rather than finance yourself through equity. Clearly there’s a risk here of excessive leverage. All the evidence suggests that leverage and, therefore, debt equity ratios are likely to have been higher for tax reasons than they otherwise would have been, with fairly obvious implications for the vulnerability and the depth of recession and crisis.
IMF Survey online: How could taxation policy respond to this situation?
Keen: Broadly speaking, legislators need to level the playing field between debt and equity, which could be done in one of two ways. You could either eliminate the deductibility for interest—and some countries have in recent years tightened interest deductibility conditions—or you could take the opposite approach: keep interest deductibility, but also allow a deduction for a “notional return” on book equity, just as the company deducts its interest. This is known as an Allowance for Corporate Equity (ACE) system. Some countries have experimented with an ACE system, and we think those experiences have been fairly encouraging. At the same time, we wouldn’t want to rule out other ways of achieving neutrality in corporate taxation, whether by restricting interest or by moving to what is sometimes called a cash flow form of corporate tax.
IMF Survey online: Are there drawbacks to moving to a tax system that allows deduction of the “notional rate of return” on book equity?
Keen: One of the main concerns that countries might have is the narrowing of the tax base. Clearly, corporate tax revenues are taking something of a hit anyway, and clearly, also, the cost of an ACE does depend on what the imputed rate of return would be. That would be calibrated relative to government borrowing rates. When those are low, the narrowing of the base would be correspondingly low.
Simply looking at the impact of reform on corporate tax revenues though may be too narrow a perspective, though the point here is a bit tricky. The question is whether the corporate tax is ultimately borne by labor—when capital is mobile across countries, it’s not going to bear the tax levied by one country. So the real burden of the tax is going to be passed on to whatever can’t move to escape it, which typically means labor. The reform would in principle generate an efficiency gain that would take the form of increased return to these immobile factors of production. So it might be more sensible and efficient to recover the lost revenue by taxing labor. But while that is an argument that many economists would buy, it’s one that politicians might find hard to sell.
There is also the question of transition arrangements. It would make a big difference to revenue whether you gave this allowance immediately with respect to all existing equity, or whether you gave it looking forward to expansions of the equity base occurring from now on. There are also some technical questions as to how you would accomplish the transition.
IMF Survey online: Why would you use a “notional rate of return” as opposed to trying to come up with an actual one?
Keen: There are two main arguments. One, the argument from principle, is that what you’re really doing is giving companies a tax break, and so long as they’re perfectly sure they’re going to get that tax break at some point in the future—they may not get it today because they may have losses, and so on—for them there’s no uncertainty about the income stream associated with it. And if that’s the case, then this is a risk-free asset, and you would value it at a risk-free rate.
The second, more practical, argument is that if you try to have a notional rate that varies across firms, the system could become an administrative horror. Largely for this reason, countries that have adopted this approach used a single benchmark rate.
IMF Survey online: Is it consistent to codify an actual deduction on the debt side and a notional deduction on the equity side?
Perry: In most instances, when the ACE has been implemented, there hasn’t been a restriction on the deduction of interest paid. However, logical consistency and neutrality between debt and equity would argue that you would only allow the deduction for interest at the risk-free rate as well. This would be viewed as a more radical change, but it’s the logical implication of this system. And, it has other attractions. It would not only reduce the revenue loss, but also eliminate, to a much greater extent than the ACE alone, the huge amount of effort, tax planning and transaction costs that arise in attempting to define and distinguish whether an instrument is debt or equity. Then the two things would essentially be functionally equivalent. But now we are talking very radical reform indeed.....
IMF Survey online: Do you see any connection between tax policy and the use of complex financial instruments or securitization?
Perry: In the paper, we talk about complex financial instruments and the role that tax distortions play in encouraging their use. We also touch on some of the emerging arguments that tax may have had a role to play in securitization of the subprime market—people are still trying to figure out precisely what’s been going on on the tax side.
One of the things that securitization can do by unbundling some asset, for instance, is to allow a deduction to be taken or a receipt to be taxed at a different rate from the one that would apply if the original asset were held. So it may be possible, for instance, to take a capital loss against interest income, which would be at a higher rate—making the loss worth more to the taxpayer—than otherwise would be the case.
Comments on this article should be sent to imfsurvey@imf.org
