A popular belief these days is that the recession has opened “output gaps” in many countries. Output gaps measure the difference between actual output and potential output and have an attractive simplicity for policymakers. If capacity persists across an economy, interest rates can be kept low with little fear of rising production and labour costs.

Trouble is, it is almost impossible to measure an output gap with any accuracy. How on earth do you measure an economy's potential? In a 1999 paper, Athanasios Orphanides and Simon Van Norden cite three big problems. First, the construction of output data is continually being updated – real time apples end up being compared with future pears. Second, economic models often revise historical estimates of potential output with the benefit of hindsight. Third, countries undergo structural changes, making comparisons between actual and potential estimates of output inaccurate.

Indeed, the research reveals a 4 percentage point range between the various ways economists traditionally model the output gap in the US. That difference is of the same order of magnitude of the business cycle itself. Worse, the margin of error is even wider when output gaps are estimated during turning points in a cycle – just when policymakers tend to look at using these measures the most.