Date
26 September 2017
Some signs suggest limited slack in the US labor market. Photo: AFP
Some signs suggest limited slack in the US labor market. Photo: AFP

The riddle of the risk of global deflation

Declining inflation in many major economies is consistent with negative output gaps (which occur when actual output is lower than potential output). It may also indicate that output gaps are being underestimated. Moreover, even as negative output gaps start to close in economies such as the United States over the next two years, those in some emerging economies are likely to expand so that the global negative output gap remains very wide.

Against this background, disinflationary pressures will remain, and as global inflation is already subdued a slide into deflation in some countries is a significant risk. Not all measures of economic slack line up with measures of the output gap. This makes the task of policymakers in the advanced economies even more complex, but the biggest risk is still that they withdraw policy stimulus too quickly.

Inflation is very subdued in the advanced economies. Consumer price index inflation is just 0.7 percent in the eurozone, with several member states seeing declining price levels, and is around 1.5 percent in the US, Britain and Japan. Chinese inflation in April declined sharply, to just 1.8 percent.

Producer price inflation, meanwhile, is even lower at around 1 percent in Japan, with negative rates in the eurozone and China. Export price growth is also very weak, so overall producer pricing power seems very limited.

This weakness in prices comes despite the fact that the global expansion is now in its fifth year. How can it be explained?

A key framework for analysing global inflation is the study of output gaps – the difference between actual output and estimated potential output. If actual output is below potential (a negative output gap) there is slack in the economy, which implies downward pressure on inflation, and vice versa.

Available estimates suggest the output gap is indeed negative in most advanced economies –- substantially so in many cases. We estimate that the negative output gap was around 2 percent of GDP in Japan last year, around 4 percent in the eurozone, about 4.5 percent in the US and roughly 5 percent in Britain.

Present estimates of output gaps are unusually large given the maturity of the global upswing. The OECD thinks the negative output gap for advanced economies as a group was around 3 percent of GDP last year. This is in stark contrast to similar stages in previous recoveries.

It estimated a positive gap of 1 percent of GDP four years after the last trough in 2003, one of around zero four years after the 1993 trough and a negative output of just over 1 percent of GDP four years after the 1982 trough, which was the deepest of the three previous recessions.

The output gap is an unobservable variable. Potential GDP can only be inferred from other indicators, so estimates of the output gap are subject to considerable margins of error and tend to be revised substantially over time. As such, we should be cautious when interpreting current estimates.

What about output gaps globally?

Using our estimates for output gaps for the US, Britain, Japan, the eurozone and the BRIC emergers, we can construct a “global” output gap. We estimate a negative global gap of over 3 percent of GDP for 2013, weighting the gap using GDP at prices and exchange rates of 2005. This estimated gap has shifted little since 2010, underlining the relative weakness of this global upturn. The gap is forecast to narrow in 2014-15 but will still be 2.5 percent of GDP in 2015, historically quite wide and implying continued global disinflationary pressures.

One reason for this is that while negative output gaps are set to shrink in the US, Japan and Britain, the gap is forecast to barely shift in the eurozone while in the BRICs, negative output gaps are expected to widen thanks to below-trend economic growth.

Central banks, especially the Fed and the Bank of England, have begun to emphasize the importance of measures of slack or spare capacity in setting policy. But measuring output gaps is always hard and there is some evidence that estimates have become more prone to revision over time. Moreover, different measures of economic ‘slack’ do not tell a coherent story, complicating the issue considerably.

This analysis suggests that relatively wide negative output gaps exist in the major economies, and that there is a risk that these might even be understated. The recent fall in inflation seems to support this interpretation, especially for the eurozone where if anything recent output growth has modestly outperformed expectations (consistent with the idea that it is the level of the output gap that matters for inflation).

But it may reasonably be asked why inflation has not fallen even further given the estimated size of the output gaps. Part of the answer may be rigidities in setting nominal prices and wages. Looking back to the Japanese example of the 1990s, sustained deflation in consumer prices did not set in until 1999 –- long after the initial crash (though the GDP deflator began to decline from 1995). It may just be a question of time lags.

Meanwhile, in the US and Britain some indicators suggest that the degree of slack in labour markets is quite limited, which seems hard to square with large overall output gaps. One reason for the apparent inconsistency may be the particular impact of the global financial crisis. The academic literature suggests that output losses from financial crises mostly result from negative productivity shocks, rather than through the labour market or capital stock. Moreover, the shock may be centred narrowly on the financial sector itself, not a very labor-intensive sector.

So an output gap caused by a very sector-specific productivity shock (eg. in the financial industry) might be less disinflationary because the broader labor market could remain more robust. This would fit with recent US and British experience, and also with the experience in the eurozone. There, where the labor market picture is very weak and much more in line with large output gap estimates, inflation is also lower.

The different messages emanating from different measures of spare capacity explain why central banks have retreated from using specific measures of it over recent months in favor of a broader (and vaguer) approach. Vaguer communications by central banks seem likely to continue, especially as they grant central banks considerable leeway to adapt to new information and try to make sense of it. There are plenty of historical examples of policy mistakes resulting from the mismeasurement of output gaps (such as in the US and Britain in the 1970s) and central banks will be keen to avoid adding to these.

Arguably, the bigger risk is still removing monetary accommodation too quickly. In this respect, it is not only domestic indicators that matter, but also the broader international context. The forecast widening of output gaps in the emerging economies could add a new layer of disinflationary pressures to the global economy.

If negative output gaps are large, the world economy may also be particularly vulnerable to any additional economic or financial shocks. Businesses already operating below capacity might react more negatively to demand shocks than if recent capacity use had been high and orders strong. Financial market shocks might also have larger effects on output than would be the case if the world economy were operating closer to capacity.

These factors support our view that rate increases in the major economies will not occur until the second half of 2015 at the earliest, and also that a slide into deflation is a significant risk in some economies – most notably the eurozone.

The writer is a senior economist at Oxford Economics.

– Contact us at [email protected]

SK

 

Inflation is very subdued in the advanced economies, including Japan where CPI inflation is around 1.5 percent. Photo: Bloomberg


Senior economist at Oxford Economics

EJI Weekly Newsletter

Please click here to unsubscribe