Most of my economics professors opened class with some variation of the old in-house joke: “Economists have correctly predicted nine out of the last five recessions.” Their scholarly modesty apparently exhausted, however, they proceeded to spend the next four months teaching precisely how to model and forecast the national economy.
In a world where describing the current state of the economy proves difficult — the Bureau of Labor Statistics always has to revise its estimates of current unemployment — it’s understandable that predictions are unreliable. Even so, the manner and degree to which economists are wrong in their forecasts is illuminating. For example, the Obama White House’s predictions often stem from methodological overconfidence.
As James Pethokoukis blogged in response to the most recent GDP report:
In August of 2009, the White House — after having a half year to view the economy and its $800 billion stimulus response — predicted that GDP would rise 4.3% in 2011, followed by 4.3% growth in 2012 and 2013, too. And 2014? Another year of 4.0% growth.
In its 2010 forecast, the White House said it was looking for 3.5% GDP growth in 2012, followed by 4.4% in 2013, 4.3% in 2014.
In its 2011 forecast, the White House predicted 3.1% growth in 2011, 4.0% in 2012 and 4.5% in 2013, 4.2% in 2014.
In its most recent forecast, the White House predicted 3.0% growth this year and next, and then back to 4.0% after that. Good luck with that.
In reality, the economy grew 1.7% in 2011 and even 3% is now looking way out of reach for this year.
In addition, the administration predicted that without the Recovery Act (which was passed in 2009), unemployment would soar all the way to a peak of 9 percent in 2010; the act was needed to keep unemployment to a peak of 8 percent. But even with the stimulus plan in place, unemployment hit 10.1 percent in October of 2009. Currently, unemployment is more than a percentage point higher than the White House predicted it would be if we didn’t pass the stimulus plan, and about two percentage points higher than it predicted it would be if we did.
Part of this error can be attributed to the hazards of economic forecasting. As the administration hastily admitted when forced to downgrade its 2011 forecast last September, the economy was affected by unforeseeable shocks, such as the Arab Spring and the Japanese Tsunami.
Economic foresight is also obscured by the illegibility of the current situation. Hence the refrain that “the recession was even worse than we thought it was.” But this uncertainty also stems from smaller, more technical aspects of forecasting. For example, updated 2010 census data recently showed that there are more women, people under 25, and people over 54 in America than originally thought. These groups average a lower labor-participation rate, and thus their existence contributed to the unusually low participation rate that has been recorded of late.
Finally, as the CBO has demonstrated, forecasting is hardest during irregular booms and busts. The CBO tends to predict that the economy will recover more quickly from recessions than it actually does, and also underpredicts growth during expansions. The only two times during the last 36 years in which the CBO predicted output correctly to the first decimal were in 1986 and 2004, during periods of relatively stable growth.