Beware The Government Debt-To-GDP Ratio

Quoting government debt levels as a percentage of GDP is misleading, writes Paul Sheard.

A manager weighs banknotes on a scale at a bakery in Venezuela. (Photographer: Manaure Quintero/Bloomberg)
A manager weighs banknotes on a scale at a bakery in Venezuela. (Photographer: Manaure Quintero/Bloomberg)

Government debt levels around the world are rising fast as governments run up large budget deficits because of the pandemic-induced recession. In judging the level of government debt, and how much of a potential problem it might be storing up for the future, a universally cited metric is the nominal government debt-to-GDP ratio, the amount of government debt expressed as a percentage of gross domestic product.

Quoting government debt levels as a percentage of GDP is misleading. An elementary rule of data analysis is to use the correct units. Expressing government debt levels as a percentage of GDP violates this basic rule and risks leading policymakers up the fiscal and monetary garden path. The correct unit of a debt-to-GDP ratio is the time unit of the denominator: number of years. Using the United States as an example, the unit of the numerator is dollars and the unit of the denominator is dollars per year, meaning that the unit of the ratio is “number of years”, of current GDP.

The numerator of the debt-to-GDP ratio is a stock at a point in time while the denominator is a flow per unit of time. This apples and oranges aspect of the metric alone should caution against relying on it too unquestioningly.

Analysts and policymakers commonly use debt-to-GDP ratios as an indicator of “fiscal sustainability,” meaning that high and rising ratios are prone to set off the alarm bells, and often presage fiscal austerity ahead. Not so fast.

A debt-to-GDP ratio of 100%, let alone 200%, carries an unsustainable-sounding connotation: the level of debt equals, or is double, that of GDP. A more accurate interpretation is that, if the whole of the government debt had to be paid off all at once, it would require one year’s worth of GDP to do so.

This is much less scary. For one thing, unlike an individual citizen or even business, the government is an entity that has what economists call an “infinite horizon.” An amount of debt equivalent to just one or two years of the life of an entity that, in principle at least, lives forever does not sound so daunting.

For another, unlike the debts of an individual or company, surprisingly enough the government does not really ever have to repay its debt. The appearance that it does is just an artifact of the way governments have structured their monetary and fiscal affairs, reflecting lessons learned in the twentieth century: making the central bank ‘independent’ can help the government commit not to abuse its access to its printing press. This previously obscured fact has been revealed by the now commonplace quantitative easing of central banks. Every time the Federal Reserve, say, buys a dollar of Treasuries it is converting a federal government liability that has a repay-by date into one that does not. What the government, via its central bank, can do for one dollar of its debt in principle it could do for any or all of it.

Another interpretation of a debt-to-GDP ratio measured in this way is that this is the number of years of GDP that would be requisitioned if the holders of the debt unleashed the associated purchasing power on the economy all at once.

Government debt is an asset for those who hold it. For them, government debt represents purchasing power.

Again, this makes things sounds a good deal less scary. The stock of government debt, from the viewpoint of those who hold it, represents that portion of their cumulative prior savings that they hold in that liquid safe form. Because individuals both build up and run down their savings gradually, over a lifetime, it is never going to be the case that the holders of government debt securities all decide to unleash that purchasing power at the same time.

But suppose for argument’s sake that they did. What would happen? Suddenly there would be too much money chasing not enough goods, the classic recipe for inflation to get out of control. If it is to keep its contract with its citizens, the government, including importantly the central bank, would need to rein in the excess spending. It could do so by implementing a judicious combination of monetary and fiscal tightening, the central bank raising its policy interest rate (and unwinding QE) and the government raising taxes and cutting spending.

In the normal way of macroeconomic thinking, the government uses taxes in order to help finance itself. In reality, governments use taxes for three purposes: to redistribute income; to address negative or positive externalities; and to modulate purchasing power. Pedantic as it may seem, citing the popular government debt-to-GDP metric in its correct units helps government finances to be seen in their proper light: as part of a means to achieve economic prosperity, not as an end in themselves.

Paul Sheard is a research fellow at Harvard Kennedy School and former chief economist at Lehman Brothers, Nomura Securities, and S&P Global.