Sunday, April 20, 2014
Exit from comment view mode. Click to hide this space

Debt and Delusion

NEW HAVEN – Economists like to talk about thresholds that, if crossed, spell trouble. Usually there is an element of truth in what they say. But the public often overreacts to such talk.

Consider, for example, the debt-to-GDP ratio, much in the news nowadays in Europe and the United States. It is sometimes said, almost in the same breath, that Greece’s debt equals 153% of its annual GDP, and that Greece is insolvent. Couple these statements with recent television footage of Greeks rioting in the street. Now, what does that look like?

Here in the US, it might seem like an image of our future, as public debt comes perilously close to 100% of annual GDP and continues to rise. But maybe this image is just a bit too vivid in our imaginations. Could it be that people think that a country becomes insolvent when its debt exceeds 100% of GDP?

That would clearly be nonsense. After all, debt (which is measured in currency units) and GDP (which is measured in currency units per unit of time) yields a ratio in units of pure time. There is nothing special about using a year as that unit. A year is the time that it takes for the earth to orbit the sun, which, except for seasonal industries like agriculture, has no particular economic significance.

We should remember this from high school science: always pay attention to units of measurement. Get the units wrong and you are totally befuddled.

If economists did not habitually annualize quarterly GDP data and multiply quarterly GDP by four, Greece’s debt-to-GDP ratio would be four times higher than it is now. And if they habitually decadalized GDP, multiplying the quarterly GDP numbers by 40 instead of four, Greece’s debt burden would be 15%. From the standpoint of Greece’s ability to pay, such units would be more relevant, since it doesn’t have to pay off its debts fully in one year (unless the crisis makes it impossible to refinance current debt).

Some of Greece’s national debt is owed to Greeks, by the way. As such, the debt burden woefully understates the obligations that Greeks have to each other (largely in the form of family obligations). At any time in history, the debt-to-annual-GDP ratio (including informal debts) would vastly exceed 100%.

Most people never think about this when they react to the headline debt-to-GDP figure. Can they really be so stupid as to get mixed up by these ratios? Speaking from personal experience, I have to say that they can, because even I, a professional economist, have occasionally had to stop myself from making exactly the same error.

Economists who adhere to rational-expectations models of the world will never admit it, but a lot of what happens in markets is driven by pure stupidity – or, rather, inattention, misinformation about fundamentals, and an exaggerated focus on currently circulating stories.

What is really happening in Greece is the operation of a social-feedback mechanism. Something started to cause investors to fear that Greek debt had a slightly higher risk of eventual default. Lower demand for Greek debt caused its price to fall, meaning that its yield in terms of market interest rates rose. The higher rates made it more costly for Greece to refinance its debt, creating a fiscal crisis that has forced the government to impose severe austerity measures, leading to public unrest and an economic collapse that has fueled even greater investor skepticism about Greece’s ability to service its debt.

This feedback has nothing to do with the debt-to-annual-GDP ratio crossing some threshold, unless the people who contribute to the feedback believe in the ratio. To be sure, the ratio is a factor that would help us to assess risks of negative feedback, since the government must refinance short-term debt sooner, and, if the crisis pushes up interest rates, the authorities will face intense pressures for fiscal austerity sooner or later. But the ratio is not the cause of the feedback.

A paper written last year by Carmen Reinhart and Kenneth Rogoff, called “Growth in a Time of Debt,” has been widely quoted for its analysis of 44 countries over 200 years, which found that when government debt exceeds 90% of GDP, countries suffer slower growth, losing about one percentage point on the annual rate.

One might be misled into thinking that, because 90% sounds awfully close to 100%, awful things start happening to countries that get into such a mess. But if one reads their paper carefully, it is clear that Reinhart and Rogoff picked the 90% figure almost arbitrarily. They chose, without explanation, to divide debt-to-GDP ratios into the following categories: under 30%, 30-60%, 60-90%, and over 90%. And it turns out that growth rates decline in all of these categories as the debt-to-GDP ratio increases, only somewhat more in the last category.

There is also the issue of reverse causality. Debt-to-GDP ratios tend to increase for countries that are in economic trouble. If this is part of the reason that higher debt-to-GDP ratios correspond to lower economic growth, there is less reason to think that countries should avoid a higher ratio, as Keynesian theory implies that fiscal austerity would undermine, rather than boost, economic performance.

The fundamental problem that much of the world faces today is that investors are overreacting to debt-to-GDP ratios, fearful of some magic threshold, and demanding fiscal-austerity programs too soon. They are asking governments to cut expenditure while their economies are still vulnerable. Households are running scared, so they cut expenditures as well, and businesses are being dissuaded from borrowing to finance capital expenditures.

The lesson is simple: We should worry less about debt ratios and thresholds, and more about our inability to see these indicators for the artificial – and often irrelevant – constructs that they are.

Read more from our "Imperfect Indicators?" Focal Point.

Exit from comment view mode. Click to hide this space
Hide Comments Hide Comments Read Comments (8)

Please login or register to post a comment

  1. CommentedKeven Vance

    I don't understand this at all. Yes, debt/GDP is a unit of time, so a debt ratio of 150 percent is in fact 1.5 years. But there's a natural interpretation: it would take 1.5 years for an economy to produce output sufficient to repay the entire debt. And given the limits that exist on a government's ability to appropriate output, it's easy to see why larger numbers would be bad. And the argument about how the debt ratio would be four times larger if we used quarterly GDP is just wrong. If we used quarterly data, the debt ratio example above would be six quarters, and if we used monthly data it would be 18 months. It's exactly the same.

  2. CommentedMatthew Cowan

    There are a lot of delusions about government debt and deficits, including a popularly held delusion that the debt/GDP ratio is not extremely important.

    We need only look at our own history and the experiences of other countries to prove that it is important. In the US, we experienced an excellent economy for the latter half of the 1990's following congressional bills passed to reform spending and balance budgets.

    Conversely, in the US we have experienced an anemic recovery over the past 3 years despite record high levels of spending and deficits. Keynesian theory would dictate that our economy should have taken off given the level of deficit spending we've engaged in with 4 consecutive $1 trillion+ deficits, yet unemployment remains at historically high levels and real GDP growth is well below historical trends.

    In other countries, the same story proves accurate. The Northern European countries fared very well, comparatively, during the most recent global slowdown when compared to countries that engaged in deficit spending and increased the debt/GDP.

    Japan, in particular, has engaged in the largest scale buildup of debt over the past 20+ years. From 1980-1990, its debt/GDP ratio was relatively stable at about 70%. It experienced a recession in 1990 and began increasing its debt/GDP ratio in order to spend its way out of the crisis. Real GDP growth from 1980-1990 ranged from 3%-5% in Japan and since 1990-present GDP growth has ranged from 0%-2.5%.

    Economists who write off a loss of 1% per year of GDP growth in order to justify higher debt/GDP ratios, typically because it favors their political priorities, forget the most powerful concept of finance and economics: compounding. One percent in any given year is not the end of the world (although it is about $160 B of GDP in the US, not a small amount by any standard), but compound that out over several years and you've missed out on quite a bit of economic growth.

    Debt/GDP ratios do matter. The 1 year convention is what it is and if we used the 10 year convention, the issues would still be the same, it would just be the critical 10% threshold that was surpassed. It is sad to see an economist write off a metric of economic peformance as not a big deal. Perhaps he won't give out grades to his students this year because they aren't that important either.

  3. CommentedStefan Schuetzinger

    I like Shillers' unorthodox approach. But he is wrong in a fundamental point. We don't count debts in GDP per year, because our thinking is "in years", but because interest capitalizes after 365 days. So a year has "a particular economic significance".

    I agree, that the 100%-threshold is - to a certain extent -irrational. But is Mr. Shiller willing to lend his money to an economy, which is not competive and which has already accumulated a huge debt-burden ?

  4. CommentedPierluigi Molajoni

    The lesson is simple: we should worry less, but for the fact that governments have borrowed from investors driven by pure stupidity – or, rather, inattention, misinformation about fundamentals, and an exaggerated focus on currently circulating stories.

  5. CommentedKapil Khetan

    All debt including zero coupon debt, can be quoted as having some annual interest cost. So don't see why Prof Shiller makes such a big deal about a rough ratio. Naturally, one can get more accurate by discussing interest coverage (interest payment divided by taxes collected). Interesting to hear Prof Shiller's criticism, but would have been even better to hear about the alternative suggested.

  6. Commentedphilip meguire

    Shiller makes a good point, one grounded in the careful attention to units of measurement that he (and I) were taught in high school. GDP is a flow per unit time. GDP is reported annually and quarterly. This is convenient and conventional, but also entirely arbitrary. Hence the denominator of the debt to GDP ratio features a periodic time unit, the choice of which is arbitrary. The longer the time unit, the lower the ratio. Hence the ratio of debt to GDP is arbitrary and meaningless.

    However, Shiller seems to imply that there is no need to worry about the level of govt. borrowing, and that is an impression I wish to forestall.

    I submit that there are basic three categories of public debt. that held by (1) the central bank and public trust funds, (2) by domestic private parties, and (3) foreign entities. Finally, there is (4), debt used to finance long lived infrastructure projects, about which I will say no more than that this is a very valid use of the govt's borrowing power. (1) is scarcely debt at all, because what is a liability to the Treasury is also an asset of another branch of the Federal govt. (3) is the most troubling, especially when the debt is denominated in a foreign currency (a problem for the PIGS but not for the USA).

    So anything divided by GDP should have the same time unit as GDP itself. I propose the interest paid on the debt, calculated for each of (2) and (3) in the preceding paragraph. This is a measure of the carrying cost of the part of the Federal debt that is owed to parties outside the Federal govt.

    The USA NIPA report the interest paid on (3). The Fed discloses the interest on the Federal debt it holds. So far so good. But calculating the interest paid on debt held by Federal govt. trust funds is not easy. (2) is easily calculated as the residual after calculating (3) and (1), but I bet calculating (1) will prove elusive. I invite the NIPA, especially Table 3.2, to adopt the framework I set out here.

    The only way to evaluate the interest to GDP ratio is relative to the historical record. Recent values of the interest to GDP ratios calculated from the available NIPA data are not too worrying, simply because nominal interest rates are so low. But current nominal rates imply negative real rates. When interest rates rise, the interest to GDP ratio will rise, as will the fraction of every Federal tax dollar devoted to interest on the privately held public debt. Taxes will have to rise, or current government expenditure will have to be cut. This is the sense in which the rise in Treasury borrowing this century amounts to a Sword of Damocles hung over the American economy.

    The problem is that two tax-related ratios currently have values that are unusually low by historical norms. One such ratio is that of Federal income and payroll taxes to nonimputed personal income. The other is of corporate income taxes to corporate cash flows, which is lower now than at any time since the Great Depression. Raising those ratios to the their average levels over 1987-2007 would generate approximately 350B of tax revenue. Further tax increases and expenditure cuts may be needed to get our house in order.

    The Flow of Funds accounts reveal that American corporations currently hold about 2.5 trillion of financial assets having a low risk of capital gain or loss. Corporate profits are at an all-time high, but capital expenditures are anemic. I conclude that the American corporate sector broadly understands that current American fiscal policy is a temporary fool's paradise.

  7. CommentedFernando Fuster-Fabra

    The most sensible analysis I have read in decades that makes macro-Economic more understandable. Without being an economist but with wide experience in project management, I have always argued against the average and median stats delivered by economic experts with little or no considerations of "time" as a resource. It has been my point of view that all projects handle 3 basic resources money & materials, human resources and time. If you are short of one, then you need to compensate with any of the two others. Thus, in a systemic crisis as the one today, we need more time to pay up debt and/or more money to produce growth.