30

When Is Government Debt Risky?

BERKELEY – A government that does not tax sufficiently to cover its spending will eventually run into all manner of debt-generated trouble. Its nominal interest rates will rise as bondholders fear inflation. Its business leaders will hunker down and try to move their wealth out of the companies they run for fear of high future corporate taxes.

Moreover, real interest rates will rise, owing to policy uncertainty, rendering many investments that are truly socially productive unprofitable. And, when inflation takes hold, the division of labor will shrink. What once was a large web held together by thin monetary ties will fragment into very small networks solidified by thick bonds of personal trust and social obligation. And a small division of labor means low productivity.

All of this is bound to happen – eventually – if a government does not tax sufficiently to cover its spending. But can it happen as long as interest rates remain low, stock prices remain buoyant, and inflation remains subdued? I and other economists – including Larry Summers, Laura Tyson, Paul Krugman, and many more – believe that it cannot.

As long as stock prices are buoyant, business leaders are not scared of future taxes or of policy uncertainty. As long as interest rates remain low, there is no downward pressure on public investment. And as long as inflation remains low, the extra debt that a government issues is highly prized as a store of value, helps savers sleep more easily at night, and provides a boost to the economy, because it assists deleveraging and raises the velocity of spending.

Economists, in short, do not watch only quantities – the amount of debt that a government has issued – but prices as well. And, because people trade bonds for commodities, cash, and stocks, the prices of government debt are the rate of inflation, the nominal interest rate, and the level of the stock market. And all three of these prices are flashing green, signaling that markets would prefer government debt to grow at a faster pace than current forecasts indicate.

When Carmen Reinhart and Ken Rogoff wrote their influential study “Growth in a Time of Debt,” they asked the following question: “Outsized deficits and epic bank bailouts may be useful in fighting a downturn, but what is the long-run macroeconomic impact of higher levels of government debt, especially against the backdrop of graying populations and rising social insurance costs?” Reinhart and Rogoff saw a public-debt “threshold of 90% of [annual] GDP,” beyond which “growth rates fall…. in [both] advanced and emerging economies.”

The principal mistake that Reinhart and Rogoff made in their analysis – indeed, the only significant mistake – was their use of the word “threshold.” That semantic choice, together with the graph that they included, has led many astray. The Washington Post editorial board, for example, recently condemned what it called the “Don’t worry, be happy” approach to the US budget deficit and government debt, on the grounds that there is a “90% mark that economists regard as a threat to sustainable economic growth.”

To be sure, The Washington Post editorial board has shown since the start of the millennium that it requires little empirical support for its claims. But the phrasing in “Growth in a Time of Debt” also misled European Commissioner Olli Rehn and many others to argue that “when [government] debt reaches 80-90% of GDP, it starts to crowd out activity.” Reinhart and Rogoff, it is widely believed, showed that if the debt/GDP ratio is below 90%, an economy is safe, and that only if the debt burden is above 90% is growth placed in jeopardy.

Yet the threshold is not there. It is an artifact of Reinhart and Rogoff’s non-parametric method: throw the data into four bins, with 90% serving as the bottom of the top bin. In fact, there is a gradual and smooth decline in growth rates as debt/GDP ratios increase – 80% looks only trivially better than 100%.

And, as Reinhart and Rogoff say, a correlation between high debt and low growth is a sign that one should investigate whether debt is a risk. Sometimes it is: a good deal of the relationship comes from countries where interest rates are higher and the stock market is lower, and where a higher debt/GDP ratio does indeed mean slower growth.

Still more of the relationship comes from countries where inflation rates are higher when government debt is higher. But some of it comes from countries where growth was already slow, and thus where high debt/GDP ratios, as Larry Summers constantly says, result from the denominator, not the numerator.

So, how much room is left in the relationship between debt and economic performance for a country with low interest rates, low inflation, buoyant stock prices, and healthy prior growth?

Not much, if any. In the United States, at least, we have learned that there is little risk to accumulating more government debt until interest and inflation rates begin to rise above normal levels, or the stock market tanks. And there are large potential benefits to be gained from solving America’s real problems – low employment and slack capacity – right now.

Read an extended version of this argument in J. Bradford DeLong’s blog post, “Risks of Debt: The Real Flaw in Reinhart-Rogoff?”