November 12, 2014
The economic collapse in the wake of the global financial crises (GFC) and the weaker-than-expected recovery in many countries have led to questions about the impact of severe downturns on economic potential. Indeed, for several major economies, the level of output is nowhere near returning to pre-crisis trend (figure 1). Such developments have resulted in repeated downward revisions to estimates of potential output by private- and public-sector forecasters. In addition, this disappointment in post-recession growth has contributed to concerns that the U.S. economy, among others, is entering an era of secular stagnation. However, the historical experience of advanced economies around recessions indicates that the current experience is less unusual than one might think. First, output typically does not return to pre-crisis trend following recessions, especially deep ones. Second, in response, forecasters repeatedly revise down measures of trend.
Economic models usually assume that recession-induced gaps will close over time, typically via a period of above trend growth. In our results, growth is not faster after the recession than before, implying that the recession-induced gap is closed primarily by revising estimates of trend output growth lower. Interestingly, much of the downward revision to estimates of trend output happens well into the recovery. In particular, as economies recover and the lower level of actual output persists, potential output is gradually revised down toward actual GDP.
This pattern of revision also holds true if potential is calculated using a growth accounting framework, the method used by policymaking institutions such as the OECD. To see how estimates of potential using this methodology are adjusted around turning points, we use projections from the OECD's bi-annual economic outlook for 62 recessions from 1989 to 2009 in 23 advanced economies and construct a database of various vintages of the OECD's estimates of potential growth--i.e. forecasts made a year prior to the recession trough, at the trough, and three years after the trough. Figure 4 shows these vintages averaged around recession troughs. These data reveal a pattern of downward revisions to the level of potential around turning points. Even three years post-trough, potential growth is still being revised down. This same pattern of systematic underestimation of the impact of recessions on potential and the subsequent downward revision of potential output holds true for other policymaking institutions in the wake of the Great Recession. While it is tempting to attribute this to the impact of the financial crisis on growth, the discussion above suggests that this pattern is long standing. Ironically, despite being known as the dismal science, economists may be too optimistic about the recovery path of output following recessions.
