In 2010, as Greece signed a bailout deal with the International Monetary Fund, forecasts by the IMF and the European Commission suggested the country’s debt-to-GDP ratio would peak below 150 percent of gross domestic product in 2012. The forecasts also projected that Greek GDP in 2015 would be 8 percent larger than in 2011. This optimistic vision of the future was based on underlying assumptions that Greece would go from having the lowest productivity growth in the euro zone to among the highest, alongside the highest labor force participation rates and employment rates equal to Germany’s.
No one—including the IMF—believes those assumptions anymore. The latest estimates suggest Greek GDP will be 10 percent smaller than in 2011. Why were those early prognostications so rosy? It turns out that the IMF, the EU, and other institutions have a tradition of consistently overoptimistic growth forecasts in times of crisis. Dealing with that tendency would significantly reduce the harm done by future financial crises.
Debt climbs when economies falter. Using a sample of developing countries, New York University economist Bill Easterly showed in a recent paper that the pace of growth has a strong association with debt levels. He suggested that developing economies that experienced accelerated growth between 1975 and 1994 saw debt-to-GDP ratios increase by 1 percent per annum. Countries that saw growth rates fall, on the other hand, saw debt-to-GDP climb far faster—between 3 percent per annum for those that saw modest slowdowns to 7 percent for those that saw the sharpest declines. In his analysis, he demonstrated how the same pattern applied to countries in the euro zone after the financial crisis.
At the same time, Easterly showed that countries in debt crises are prone to especially optimistic growth forecasts. Just as the IMF and the European Commission have for Greece, the World Bank and IMF drew up growth forecasts for these countries to predict debt sustainability. Easterly found that forecasts for highly indebted poor countries in the 1990s and early 2000s were consistently too positive in their estimates–overpredicting growth by an average of 1 percent a year. IMF researchers agree with him: Hans Gemberg and Andrew Martinez of the fund’s Independent Evaluation Office suggest the organization’s forecasts “tended to be consistently over-optimistic in times of country-specific, regional, and global recessions.”
Looking at European Commission forecasts of government expenditure as a percentage of GDP, António Afonso and Jorge Silva of the University of Lisbon find (PDF) that the commission similarly and systematically forecast lower government expenditure-to-GDP ratios than occurred. The commission’s forecasts for one year out were off by an average of 0.4 percent of GDP over the period 1999 to 2011—and around 84 percent of the error was due to overconfidence about GDP growth rates. The optimistic inaccuracy about the government’s fiscal situation fed into similarly overoptimistic government debt-to-GDP predictions. As with the IMF, the commission’s forecasts have been particularly distorted for Greece, Ireland, and Portugal—the crisis countries receiving IMF and European support.
It isn’t hard to figure out why it's tempting to forecast faster growth when debt levels are rising: It's a way to avoid painful choices. Easterly showed that the U.S. Office of Management and Budget cranked up growth forecasts after 2001 so that lower taxes combined with increased spending on defense and homeland security could look fiscally responsible.
For highly indebted poor countries and the World Bank, as much as for Greece and its creditors, strong growth forecasts make higher debt levels look sustainable. If the economy grows, so will government revenues. From the point of view of the debtor country, the bigger revenues grow, the less need to cut spending to allow for debt payments. From the point of view of creditors, the bigger revenues grow, the less need to forgive debt to make payments a sustainable burden.
In turn, that's a reason to separate the process of creating growth forecasts from the process of negotiating rescue packages. For the IMF, for example, the same teams that work on forecasting growth and debt levels help to design lending and reform packages, reporting to the same leadership team. That means the fund only needs one set of country experts. But it also makes life for those experts far more straightforward if they err on the side of optimism in their forecasts. The economic monitoring and forecast teams should be in a completely separate institutional structure from those deciding on lending and loan conditions.
If only Greece, Europe, and the IMF had accepted a role for external and independent growth forecasts rather than cooking up their own, perhaps they could have arranged a more sustainable package of relief earlier, rather than shunting the problem–and Greece’s recovery—down the road. The same applies to numerous other debt discussions–and the problem will continue until the people making the predictions don’t have every incentive to use a rose-tinted crystal ball.