Making Monetary Policy Decisions in the Dark

IMF Blog, August 12, 2015

Source details

In the wake of the global financial crisis, monetary and fiscal policies were used aggressively to counteract the effects of the crisis on economic activity.
Policymakers rely on a number of indicators to assess an economy’s level of activity relative to its productive capacity, but assessing the economy’s position in real time is often quite challenging and has consequences for setting policy.
The experience of Brazil in 2011 illustrates real-time mismeasurement: policymakers initially estimated the economy was at a level consistent with its productive capacity, but later revisions showed the economy was overheating and facing inflationary pressures requiring policy tightening.

Potential output is not directly observable and is difficult to measure, but it is used as a benchmark against actual output.
The output gap (actual GDP relative to potential) is commonly used to indicate cyclical position:
- Positive output gap = GDP above potential (overheating risk).
- Negative output gap = excess capacity (slack).
Economists revise output gap estimates as new information arrives; the gap for any given year can change after its initial estimate.
Key empirical observations from revisions to initial output gap estimates:
- Compared to initial estimates, nearly a third of assessments of output relative to potential reach the opposite conclusion based on subsequent information.
- In real time, economists tend to think slowdowns are temporary in nature, but they are frequently longer-lasting, which only becomes apparent after the fact. Thus, an economy’s potential productive capacity tends to be persistently overestimated.
- Estimates of output gaps are less reliable in recession years, precisely when they are needed the most to enact sound countercyclical policies.
- A low share of the variation in output gap revisions can be predicted ahead of time, making it difficult for policymakers to adjust for this uncertainty when policies are set.

Most central banks set monetary policy according to the information available at the time about the economy’s position relative to potential output and the outlook for inflation; many central banks in the region operate with an inflation target.
Empirical findings when using revised output gaps to emulate central banks’ reaction functions:
- Interest rates implied by reaction functions using revised output gap estimates turn out to be substantially different from those that were set in real time.
- These output gap revisions explain almost half of deviations between the actual level of inflation and the central bank’s target.
- Revisions bring to light new information not accounted for in policy decisions, reflecting the difficulty in distinguishing in real time whether fluctuations are cyclical or represent a change in the trend.

Policymakers should be cautious in attributing all movements in economic activity to short-term fluctuations; movements in the trend rate of growth are common and should be adjusted to more quickly by policymakers.
Methods for the real-time assessment of output relative to productive potential need to be improved to reduce uncertainty and better inform policy decisions.
Improving real-time measurement and being more responsive to potential trend shifts would help avoid episodes where policy would likely have been different if policymakers had known then what is known now.

Francesco Grigoli, Alexander Herman, Andrew Swiston, Gabriel Di Bella — August 12, 2015