Harder than forecasting the economy

Forecasting the economy is hard enough already, but when it comes to economic developments, there is one variable that is even harder to forecast than all the others: Debt/GDP. The problem with debt/GDP forecasts is that you not only have to forecast the economy (in the form of GDP) but also the national debt level and that depends heavily on politicians and their spending habits.

One may say that this shouldn’t be too hard. Just assume that politicians are always more profligate and willing to spend more money than they have, and one should have a pretty good handle on the deficit. But it turns out that the spendthrift of politicians is constantly surprising to the upside even if we make that assumption. I know of no economist who is under the impression that official deficit forecasts by governments are anything but a work of fiction. Sometimes it may be science fiction in the sense that it is based on sensible economic assumptions, but more often the forecasts should be categorised as a fantasy novel.

Yet, even cynical economists are constantly surprised by the amount of excess spending by governments as shown in a recent CEPR study that looked at the forecast error of the IMF and the Economist Intelligence Unit for debt/GDP-ratios of 127 countries. The further in the future the forecast, the bigger the forecast error gets, but it is about the same in developing and industrialised countries. And no, the forecast error is not large because economists are too optimistic about economic growth in the future. The study showed that the problem is that economists are constantly surprised by the amount of debt piled on over time. So, my practical advice is if you are forced to forecast debt/GDP, just make your worst assumption, and then double it. Maybe that will get you somewhere near reality.

Forecast error for debt/GDP-ratios

Source: Flores et al. (2021)