Unobservables, uncertainty, and exit

Helge Berger, Emil Stavrev 10 October 2009

a

A

Most conventional macroeconomic models predict that an increase in actual output relative to potential output (that is, an increase in the output gap) is positively associated with higher inflation. Thus, central banks spend considerable time trying to produce reliable estimates and forecasts of the output gap to inform their monetary policy decisions (Svensson et al, 2008). However, by accounts of monetary policy practitioners and academics alike, this is a difficult task, as estimates of output gaps are bedeviled by limits to data availability in real time, large data revisions, and conceptual problems, including the fact that the output gap will change as monetary policy changes. And none of these problems have been made any easier by the crisis and its aftermath.

Dealing with uncertainty

Take the problem of measurement, for instance. Estimates of the real-time output gap tend to include sizable and persistent errors. A comparison of simple filter-based quasi “real-time” estimates of the output gap in the euro area with the actual or “true” gap computed with the advantage of hindsight over a longer period suffices to illustrate this point (Figure 1). The unconditional standard deviation of the error is almost one percentage point, and its autoregressive coefficient is very high – about 0.9. Moreover, because at times the real-time output gap takes the opposite sign of the true gap, it could potentially lead policymakers to mistake the direction of price pressures.

Figure 1. Real-time and “true” output gaps

Monetary policymakers and others interested in the output gap thus face a difficult task – separating the information contained in measures of potential output from the noise surrounding it. In practice, a central bank like the ECB looks at a wide array of indicators, cross-checking the information where possible (ECB 2004, Fischer et al. 2008). However, even with the most broad-based assessment, some uncertainty about the true level of potential output relative to actual output will remain. One consequence of this is that there can be diverging views on the state of the economy, driving an informational wedge between price setters and policymakers (Gorodnichenko and Shapiro, 2007). In the present context, for example, the central bank could be more pessimistic about the crisis’ impact on potential output than the general public or vice versa.

So should central banks ignore the output gap altogether? The short answer is no. Although this discussion is ongoing, many observers agree that the output gap should inform monetary policymaking even in the presence of exceptional uncertainty. For example, Orphanides et al. (2000) using data on historical revisions to real-time estimates of the output gap in the US, show that it is usually optimal to place some weight on the output gap when decisions on the course of monetary policy are being made, even in the presence of measurement error. Ehrmann and Smets (2003) come to a similar conclusion based on a calibrated model of the euro area with incomplete information about potential output.

We also find that, under plausible assumptions, even a noisy output gap measure can help steer monetary policy. Simulations with a macro model estimated on euro area data (see Table 1) show that the volatility of inflation and the output gap following key macroeconomic shocks (i.e., demand, supply, and monetary policy) is higher if the output gap is not taken into account.

Table 1. Macroeconomic performance under output gap uncertainty

Standard deviation, percentage rates
  50 basis points shock to:
  Interest rates Inflation Output
  React to output gap Ignore output gap React to output gap Ignore output gap React to output gap Ignore output gap
Inflation 0.09 0.12 0.10 0.11 0.29 0.33
Output 0.14 0.17 0.05 0.06 0.41 0.43

Source: IMF output estimates.

That said, there is certainly room for costly policy mistakes. Take the example of a central bank observing a recovery of demand after a recession and planning to exit from a period of low interest rates. If monetary policymakers were taking a more pessimistic view of potential output than the public, their decision to raise interest rates could dampen prices more than it is needed to achieve the inflation objective. Indeed, using the simulation model for the euro area, it can be shown that if monetary policymakers underestimated the output gap by 0.5 percentage point relative to the public’s view, monetary policy would be tighter, and interest rates higher by about 150 basis points over a year, while inflation would be lower by 0.7 percentage point compared to a situation where all shared the same view of the output gap.

Implications for monetary policy

Given that policy mistakes cannot be avoided, we should ask how to minimise their costs. One way is for central banks to communicate forcefully their commitment to maintaining price stability in the context of the current uncertainty. Under the present circumstances, reinforcing the central bank’s commitment to price stability with a particular focus on the repercussions of uncertainty about the output gap on inflation would help anchor inflation expectations more firmly as well as more closely align the output gap beliefs of the central bank and the public. In practice, this approach would require a specific communication effort, implying, in particular, that the central bank make transparent its views on the development of potential output upon which it is basing current monetary policy decisions and announce clearly its willingness to adjust these views in light of new information and to act upon it to ensure price stability. An effort along these lines would seem especially helpful in transitioning away from the current policy stance of very low interest rates while avoiding unwittingly creating deflationary expectations.

Fiscal policy and the crisis

Like central banks, governments across Europe have reasons to worry about the impact of a decline in potential growth on fiscal policy. The crisis has dealt a double blow to fiscal policy – it will have to adjust to a weakening growth outlook – at least in the medium run – and simultaneously deal with the fiscal ramifications of the crisis, including large-scale government interventions in the financial sector.

A negative shock to potential growth, even if only temporary, will lead to budget shortfalls because structural revenues will fall and structural expenditures increase. Indeed, the structural fiscal balance for the euro area can be shown to deteriorate significantly because of the shortfall in potential output. Depending on the assumed scenario over the period 2009–14, the accumulated annual differences from a baseline with unchanged potential growth range from 1 percentage point under an “upside” scenario to 5.5 percentage points under a “downside” scenario. The former foresees 1.5% average annual potential growth, just shy of the baseline growth of about 1.9%, while the latter assumes -0.5% average annual potential growth, approximating a sharp initial fall in potential.

However, the crisis has also forced policymakers to use fiscal resources to intervene in the financial system, both directly and indirectly through guarantees, adding to the debt level between 2% and 5% of GDP on top of the discretionary fiscal programs and cyclical effects (IMF, 2009), estimated to be between 16% of GDP for advanced and 10% of GDP for European emerging countries.

Figure 2. Projected changes in public debt

Returning fiscal policy to a sustainable path will be a formidable challenge. Consider a simulation using the average euro-area country as a baseline case. For such a country, even under a benign set of assumptions – if the debt increase related to the crisis were only 10 percentage points of GDP and the drop in potential growth were limited to the “upside” scenario – the average annual primary fiscal surplus required to bring debt back to the Stability and Growth Pact target by 2020 would have to increase by about 1 percentage point of GDP compared to the pre-crisis world. Using less benign growth or debt assumptions considerably darkens the picture. For instance, allowing for the “downside” growth scenario causes the required primary balance to rise by 1.8 percentage points of GDP. If, at the same time, the crisis increased debt by 20% instead of just 10%, the required primary balance would increase by 2.7 percentage points of GDP. These are large numbers – not least because neither of the scenarios takes into account the increasing costs of population aging in Europe or the fact that, at least for some countries, the starting debt levels are much higher than assumed in the baseline case.

Implications for fiscal policy

The necessary fiscal adjustment should come as soon as the recovery has firmed up, supported by improvements in fiscal frameworks and reforms to raise potential output. While the exact amount of the damage to public finances due to the crisis will remain uncertain for some time, fiscal adjustment is urgent. Thus, policymakers should exercise extreme caution and start the required consolidation as soon as the recovery becomes entrenched, focusing on areas that promise swift and lasting results. Improving fiscal frameworks by introducing new national fiscal rules or strengthening existing ones and enhancing the preventive arm of the SGP will also help mitigate the fallout from the crisis and safeguard fiscal sustainability in the medium run. Finally, there is a pressing need to boost potential growth through structural reforms to limit the fiscal damage threatened by the growth effect of the crisis.

References

Ehrmann, Michael, and Frank Smets (2003), “Uncertain Potential Output: Implications for Monetary Policy,” Journal of Economic Dynamics and Control, Vol. 27, pp. 1611–38.
European Central Bank (2004), The Monetary Policy of the ECB(Frankfurt: European Central Bank, 2nd ed.).
Fischer, Björn, Michele Lenza, Huw Pill, and Lucrezia Reichlin (2008), “Money and Monetary Policy: The ECB Experience 1999–2006,” in The Role of Money—Money and Monetary Policy in the Twenty-First Century, ed. by Andreas Beyer and Lucrezia Reichlin (Frankfurt: European Central Bank).
Gorodnichenko, Yuriy, and Matthew D. Shapiro (2007), “Monetary Policy When Potential Output Is Uncertain: Understanding the Growth Gample of the 1990s,” Journal of Monetary Economics, Vol. 54, pp. 1132–62.
International Monetary Fund (2009), Regional Economic Outlook: Europe (Washington, May).
Orphanides, Athanasios, Richard Porter, David Reifschneider, Robert Tetlow, Frederico Finan (2000), "Errors in the measurement of the output gap and the design of monetary policy," Journal of Economics and Business, vol. 52(1-2), 117-141.
Svensson, Lars E.O., Malin Adolfson, Stefan Laséen and Jesper Lindé (2008) Can optimal policy projections in DSGE models be useful for policymakers?, VoxEU.org, 16 September.

a

A

Topics:  Macroeconomic policy

Tags:  monetary policy, fiscal policy, public debt

Assistant Director, IMF

Deputy Division Chief, IMF Research Department

Events

CEPR Policy Research