The tax bill signed into law by President Trump in December 2017 brought to light a number of long-running debates in fiscal policy. Most of these debates are familiar to those who follow politics: Republicans expect the tax reforms to incentivize productive behavior, such as investment, that boosts economic growth, while Democrats see it as increasing the fortunes of the well-off when money could be directed to the needy instead. Both views are true, more or less, which explains why they are so strongly believed. Republicans are also generally happy that the act reduced government revenues, while Democrats are unhappy for the same reason. These debates, while worthwhile, are predictable and often tiresome.
Quietly, though, beneath the surface, the tax legislation included some interesting cross-partisan ideas with implications that do not fall along predictable ideological lines. Chief among these was a series of prudent limitations on interest deductibility.
Limiting interest deductibility is a serious bipartisan idea that will increase U.S. macroeconomic stability while making crises like the 2008 meltdown less likely. These limitations will reduce the substantial tax incentives for leverage in the economy and move more enterprises towards equity financing, which is far more flexible in times of turmoil. Ultimately, the effort to curb interest deductibility will be one of the most valuable aspects of the reform.
Furthermore, it will be politically durable. It will endure precisely because it does not fall along the partisan lines of traditional tax debates. It is important to note that these prudent moves are not changes in tax rates but, rather, changes in tax bases. In other words, they change the composition of what counts as “income” for tax purposes, rather than the rates at which the income is taxed. While rates are almost always a partisan issue, tax bases are not necessarily so. These changes to tax bases could easily be adopted and continued by a Democratic administration, even if Democrats ultimately reverse the tax rate cuts that President Trump signed into law.
I worked in tax policy for several years. (In the interest of disclosure, I gave advice and revenue estimates to both House leadership and President Trump’s 2016 campaign on early versions of their tax proposals, which evolved into the bill signed in December 2017.) The longer one works in tax policy, the more interest one takes in tax bases rather than tax rates, and the more one notices bipartisan agreements among experts. Interest deductibility is one of these issues. Career policy analysts on both sides have long considered limiting it, and it is largely a product of chance that a Republican president ended up signing this change into law. Now that these provisions are law, I expect that they will not be reversed.
President Trump will never be popular among the liberal expert class. Even now, however, some members of that class have publicly praised the limitations on interest deductibility in the bill. Even more of them do so sotto voce. Over the long run, as distance from the partisan rancor over the bill fades, Democratic policy experts are likely to defend and entrench this aspect of the tax bill, even as they hope to unwind many others. Ultimately, limiting interest deductibility may become the best and longest-lasting policy legacy of the Trump administration.
New Limitations on Interest Deductibility
The recently passed tax legislation limited interest deductibility for both business debt and personal mortgages in four main ways. For business debt, the change was relatively simple: businesses taking a deduction for interest paid may not use it to reduce their earnings by more than 30 percent.
For personal mortgages, the legislation takes a three-pronged approach. First, mortgage interest is only deductible up to $750,000 of debt principal (as opposed to $1 million prior to the law’s enactment). Second, the standard deduction was nearly doubled. Since taxpayers choose between taking the standard deduction or taking a set of itemized deductions, any increase in the value of the standard deduction reduces the implicit value of the itemized deductions. Third, the itemized deduction for state and local taxes paid was substantially reduced. Since this provision works in tandem with the mortgage interest deduction, weakening it also weakens its partner by making itemization relatively less attractive and the standard deduction relatively more so.
In sum, all of these changes point in the same direction—towards reducing the deductibility of interest payments.
Equity, Debt, and Taxes
It is important to understand business debt finance in the context of its alternative: equity finance. Both debt and equity finance are means of connecting savers to businesses in order to help both parties. The savers provide the business with cash to fund its immediate priorities, and in return they get a claim on the business’s future profits. Lenders typically get a fixed claim on just the principal plus interest, while shareholders get a broader residual claim on whatever profit is left after lenders are paid. Debt is lower risk than equity in part because it is a claim on a fixed amount of resources, and in part because it is senior to equity. It is not risk free, however.
Businesses face a choice between debt and equity finance. The optimal mix between these two means of financing is a complex area of study for scholars of corporate finance, with no clear answer. A basic, simplified model of corporate finance pioneered by Franco Modigliani and Merton Miller in 1958 would suggest that the financing mix does not ultimately change the value of the enterprise. In a simple model, this is correct: shifting between debt and equity finance does not change the underlying business, nor the risks involved in it. It simply reallocates the claims and the risk.
Yet this theory relies on some serious simplifications: it assumes no bankruptcy costs, no taxes, and efficient markets with no agency problems or asymmetric information. All of these simplifications are, of course, false, but the theorem remains a useful framework for discussing the tradeoff between debt and equity finance.
Bankruptcies, for instance, impose costs on firms, over and beyond the costs of the poor investments that lead to it. For example, a firm with some profitable operations and other unprofitable components might be liquidated completely in bankruptcy to pay lenders. Even if a deal is struck to save the profitable components, that deal will require legal fees and due diligence. The more debt finance is used, the more likely bankruptcy is. This is one of the main reasons a firm might not maximize its use of debt finance.
On the other hand, equity finance has its own problems. Simply turning over money to management without the specific and binding demands of debt is dangerous, since management can use that money to enrich itself while doing little for shareholders. For this reason, some investors, especially those in private equity, prefer to have a debt burden to enforce discipline in management.
Finally, taxes are a relevant concern, and this is the point where the 2017 tax reform bill comes in. Interest payments have traditionally been deductible for businesses, while payments to equity, in the form of share buybacks or dividends, have not been. This asymmetry in the taxation of debt and equity instruments encourages the use of debt instruments. The 2017 tax reform modestly reduced that asymmetry.
We do not live in a Modigliani-Miller world, and we never will, but the exceptions to Modigliani-Miller are instructive because they have different implications for social welfare. For example, avoiding bankruptcy through temperate use of debt finance is not just a private good for the business, but a genuine social good as well. As a society, we should be glad that businesses consider bankruptcy costs when determining their capital structure.
Imposing discipline on management through capital structure is a less clear-cut example, but it may have genuine social value as well. For example, management with a large debt burden should be less likely to use scarce capital on perks for top-level executives. As a society, we should be glad that owners impose some discipline on management through capital structure.
In contrast, though, the tax-related exception to Modigliani-Miller has no obvious first-order social benefit. A firm that chooses a more leveraged capital structure primarily to reduce its taxes is simply creating private returns at public expense. Each dollar saved through interest deductibility is a dollar lost for the government.
This is the basic case against interest deductibility: in allowing a deduction for interest payments to lenders, but no similar deduction for payments to shareholders, the government has effectively weighed in on a complicated question of business financing without an obvious need to do so, and in creating this preference for debt it foregoes either revenue or the potential to reduce rates.
The 2017 tax reform changed this: it scaled back the deductibility of debt in order to help offset the costs of lower tax rates. In doing so, it has reduced the tax code’s preference for a leveraged capital structure.
Debt’s Deceptive Risks
The central problem with debt instruments is that they are almost like money, but very much not like money in extreme circumstances where default is possible. They deliver precise streams of cash in all but the extreme circumstances at the left tail of the probability distribution. Debt markets, which seem safe, orderly, simple, and predictable under most circumstances, were at the heart of the global financial crisis.
Human beings are extraordinarily bad at comprehending extreme tails of probability distributions. This is a common finding in psychology; for example, it may be one explanation for why some people use lotteries as an investment strategy, despite the negative expected return. It is worrying, then, that debt is like a reverse lottery. The vast majority of the time it gives a predictable, small, and positive return, and only under extremely unlikely circumstances does it give a sharply negative return. It is as if debt instruments are an idea bred in a laboratory to wreak havoc on human behavioral biases.
People’s weak grasp of unlikely outcomes is a point hammered home repeatedly by statistician Nassim Nicholas Taleb. And Taleb argues a further point, even beyond the simple example of lotteries: it is not just that people are bad at weighting unlikely outcomes when the probability is known; they are terrible at estimating those unlikely probabilities to begin with. A few theoretical statistical distributions, such as the normal distribution, or the uniform distribution, are predictable at their extremes. In practice, however, many events, particularly in finance and economics, do not follow the neat distributions shown in textbooks because they can be dramatically influenced by unlikely but consequential occurrences.
One of Taleb’s books on the subject, The Black Swan, was well timed, released in 2007 just weeks before Bear Stearns attempted to bail out failing funds at the start of the crisis. Taleb can be forgiven for repeating this point so frequently, because it so frequently needs to be made. Mistakes about highly improbable events are not just made by ordinary, statistically untrained individuals, such as those who purchase lottery tickets. Mistakes are also made by people whose profession should lead them to know better, such as those who invested in bonds for a living and did not sufficiently account for default risk prior to the crisis.
Even after the financial crisis, modeling of tail risk remains a difficult problem in finance. Consider this excerpt from CEO Jamie Dimon’s 2015 letter to J. P. Morgan shareholders: “then on one day, October 15, 2014, Treasury securities moved 40 basis points, statistically 7 to 8 standard deviations—an unprecedented move—an event that is supposed to happen only once in every 3 billion years or so (the Treasury market has only been around for 200 years or so—of course, this should make you question statistics to begin with).”
This aside is tongue-in-cheek, and one should not discount the entire practice of statistics, but it is an admission that moves in debt markets are not necessarily normally distributed, and, therefore, modeling them is extremely difficult.
Most distributions in finance, instead of following the normal distribution found in an introductory textbook, are “fat-tailed.” That means events that would seem incredibly unlikely under a textbook statistical model end up being far more likely than those models would predict.
Debt instruments are crafted such that all their risk lies in the left tail of the statistical distribution, where people are least likely to consider it or process the risk fully. Investors insulated from most day-to-day shocks can become complacent against systemic threats because those systemic threats have been partitioned into small, bite-sized pieces; pieces that are small enough to ignore and minimize.
There is no cure to systemic economy-wide risk, but capital structures too dependent on debt finance are most likely to trigger the kinds of painful, sudden moments of realization and panic we saw in 2008, followed by the painful unwinding of a web of financial claims. A structure reliant on more flexible equity investment often distributes these risks more efficiently.
Interest Deductibility and Tax Revenue
One common and legitimate argument for interest deductibility is that interest income is ostensibly taxed when it is received. In theory, this is how income taxes are supposed to work: there is a system where payments that cause deductions are then reported as income by another taxpayer until eventually the income reaches its final owner.
Unfortunately, such systems almost never work out as well in practice as they were designed in theory. For example, a similar system existed in the past for alimony. However, the Treasury Department concluded that deductions claimed for alimony paid often vastly exceeded the income reported as alimony received. In other words, taxpayers tend to make errors mostly in their own favor, and as systems become more complex, the more such errors the taxpayer can make and the more revenue the Treasury loses.
The 2017 tax reform ultimately gave up on this system for alimony in order to raise revenue. It scrapped the deduction, but will also no longer require recipients to report the income.
The tax reform did a similar thing for interest. Much like alimony, deductions for interest paid have often exceeded the taxable interest income reported. There are a number of interest recipients who never pay taxes on that interest income, for a variety of reasons. For example, they may be nontaxable retirement plans, or nonprofit endowments, or foreigners not subject to U.S. income tax.
Even if people do pay taxes on interest received, they may not pay it at the same tax rate against which a corporation deducts its interest paid. For example, in the past corporations often deducted their interest against a 35-percent tax rate, while the interest was received by taxpayers whose personal rates were significantly below 35 percent.
Calculations by the Congressional Budget Office in 2014 attempted to quantify this issue, and found that 33 percent of corporate debt was held in tax-free accounts of some kind. While most businesses had (and most financing activities generated) positive effective tax rates, the effective tax rate on debt-financed corporate investment was actually negative 6 percent.
This is, put bluntly, as horrifying a conclusion as one can find in tax policy. It means that a firm has an incentive to do something that would be unprofitable and unwelcome in a tax-free world, simply so that it can reduce its tax burden. Not only does the provision create inefficient outcomes in the real economy, but it also distributes money away from the Treasury and towards businesses—which, given the needs of the government, means that the money must be made up from other taxpayers.
To be clear, not all businesses deducting interest are benefiting from this hole in the income tax. Many of them are using the system in the spirit in which it was designed: they deduct payments to taxable investors, and the investors pay income tax. Increasing the tax burden on these arrangements—as the tax reform bill did—genuinely distorts the incentive to engage in real, productive business activity. To the extent that the GOP bill does this, it is a legitimate drawback. But it is important to consider this drawback as compared to the alternatives: the government is revenue constrained, and virtually every kind of major tax used by governments creates incentives against productive activity. Under these circumstances, it is worth closing a hole in the income tax even if some legitimate activity is caught in the crossfire.
Interest Deductibility in a Globalized Economy
A related cost to interest deductibility is the difficulty it creates for international tax systems in a globalized world. The same basic principles described above, in which it is useful to deduct interest payments against a high rate and then pay interest taxes at a low or zero rate, applies to the international tax system. It is one of many strategies for “profit shifting,” the phenomenon by which businesses use financial engineering to report their income in places with low taxes.
Corporations can benefit by borrowing in the higher-tax jurisdictions and then lending through subsidiaries in lower-tax jurisdictions. After interest-related tax provisions are taken into account, corporations end up with comparatively low taxable incomes in high-tax jurisdictions. Analysts have quantified this effect, showing that it exists in practice as well as in theory.
While the United States is no longer one of the highest-rate jurisdictions after the passage of tax reform, it will likely never be one of the lowest-rate jurisdictions. Even at the 2017 tax reform’s new lower rate, the U.S. still trails traditional tax havens such as Ireland. The IRS will still need to combat profit-shifting strategies.
The tax legislation’s modest limitations on interest deductibility will help the IRS with this somewhat, though unfortunately not as much as earnest tax reformers would have hoped. A previous version of the GOP tax bill, proposed by House leadership in 2016, would have eliminated interest deductibility entirely.
Furthermore, the proposal would have aggressively ended other profit-shifting strategies: its border adjustment provision would have meant the tax system could no longer consider cross-border transactions. Under this regime, companies could not buy overpriced imports and sell underpriced exports to foreign subsidiaries and expect the mispricing to result in a tax windfall. This scheme is technically illegal and it can be stopped in egregious cases, but the IRS cannot investigate or uncover every case of mispricing.
Ultimately these proposals were too radical to implement and the bill was substantially watered down, but some limitation on interest deductibility was preserved. House Republicans, led by Ways and Means chairman Kevin Brady, were courageous enough to look at long-standing provisions of the income tax that seemed to cause trouble, and ask the legitimate question: “what if we have been doing it wrong this whole time?” For that they deserve praise, even if their vision to solve these problems was not totally realized.
A Lasting Change?
Changes to interest deductibility will be a lasting policy legacy of the Trump administration. Many policy priorities of the major parties, like the top marginal income tax rate or net neutrality, tend to ping-pong back and forth as the Republicans and Democrats seek to undo each other’s work.
One should not expect the same for all aspects of tax reform. Democrats are under tremendous pressure to raise tax revenues in order to pay for spending programs. Unfortunately, tax increases are extremely unpopular. Democrats may therefore see some of the tax increases in the Republican legislation as a gift—an increase they did not have to work for.
This should be especially true for interest deductibility. Policy specialists of all types, not just Republicans, tend to harbor a healthy skepticism of the extent of both the mortgage interest deduction and interest deductions for business. For example, as Republicans were preparing to pass the tax bill, William Gale, a liberal-leaning economist at the Tax Policy Center, offered full-throated praise for limiting the mortgage interest deduction. Furthermore, he suggested the only problem with the proposal was that the bill did not go far enough.
This is certainly not a representative sample of how liberal-leaning experts talk about the Republican tax bill as a whole. Liberal praise for the Republican bill has been relatively scarce, and it will likely remain so for quite some time. Liberals see many problems in both its values, which they perceive as too plutocratic, and its implementation, which they see as shoddy and rushed. The overwhelming majority of liberal experts have provided harsh criticism.
While it would be wrong to diminish the sincere beliefs that have led to the harsh criticism, however, views like William Gale’s are sincere and widely held, too. They are just not as widely written, because of the professional incentives in Washington.
Typically, each party’s professional class offers little in the way of cross-partisan praise, especially when single-party legislation is being passed. Republicans chose to shut Democrats out of the tax reform process, and professional Democrats are under no obligation to help Republicans win political fights. Now that the legislation is passed, the Democrats will use its most unpopular provisions as rallying points in order to win upcoming elections. For now, there is little room in Democratic discourse to praise the Republican bill.
However, when the Democrats once again find themselves in the position to shape fiscal policy, one should expect Gale’s view to be repeated both more loudly and more often. The mortgage interest deduction is the tax provision that every economist dislikes—and it was limited in the bill signed by Trump. Furthermore, serious academics like Harvard Business School’s Robert Pozen had long bandied about ideas to limit interest deductibility, for reasons similar to the ones outlined in this piece.
While President Trump will never be beloved by earnest policy wonks, he has delivered a serious item on their tax wish list. In the end, they will elect to keep it.
A Reason for Optimism?
Pessimists in our current political environment have many grievances: they argue serious policy has taken a back seat to media feuds, or that there is no room for bipartisanship anymore, or that expertise has been ignored.
These critiques are often true. But they are not true always and everywhere. The tax reform’s limitations on interest deductibility count as evidence against every one of them. While media feuds raged, a tax idea with little flair but real policy merit managed to navigate its way through the legislative branch. While partisan polarization increased in many respects, a single-party bill quietly adopted a bipartisan idea. And finally, serious scholars of tax policy had their voice heard on the issue.
Limiting interest deductibility will help solve important problems: it will reduce distortions in firms’ capital structure incentives; it will make the economy more stable; it will strengthen the international tax system; and it will raise revenue. Even those most pessimistic about our current political moment can take heart: policymakers can indeed learn from the mistakes of the past and make genuine attempts to fix them.