Join our community of smart investors
OPINION

What inflation worries?

What inflation worries?
August 8, 2014
What inflation worries?

There's a simple reason why this might raise inflation. It's that there has for years been a strong trade-off between unemployment and inflation in the following 12 months, just as the old-fashioned Phillips curve predicts. Since 1997 the correlation between the official unemployment rate and consumer price inflation excluding food and energy in the following 12 months has been a whopping minus 0.79. This is why, when the Fed issued its forward guidance on interest rates in 2012-13, it pledged not to raise rates at least until after unemployment had dropped below 6.5 per cent - because lower unemployment is usually inflationary.

This post-1997 relationship implies that if the Fed’s forecasts are correct and that unemployment falls to between 5.1 and 5.5 per cent in 2016, inflation in 2017 will be around 2.6 per cent; it is now 2.1 per cent.

All this, though, poses questions. If it’s so obvious that inflation will rise, why is the Fed so far from raising interest rates? And why have bond yields fallen recently?

One big reason is that there are doubts about this link between unemployment and inflation, and therefore reasons to suspect inflation might stay low.

One problem is that unemployment is higher than the official headline figures suggest. Since the recession, millions of people have dropped out of the labour market and so don't count as officially unemployed. But many of them want to work, and this potential labour supply could hold down wages and prices. The Bureau of Labor Statistics estimates that if we count as unemployed people outside the labour force who want a job and those working part-time who’d like full-time work, then the unemployment rate is 12.2 per cent. That’s six percentage points higher than the headline rate; before the 2008 crisis, the gap between the two rates ranged between three and five percentage points.

A further complication is that the strong negative correlation between inflation and unemployment since 1997 disguises three separate periods. First, there was a strong negative correlation between the late 90s and 2009; low unemployment in the late 90s led to higher inflation, and high unemployment in 2008-09 to low inflation. Then between 2009 and 2011 inflation rose even though unemployment didn't fall much at all. And thirdly, since then we've seen unemployment fall from 9 to 6.2 per cent whilst inflation has merely wobbled around 2 per cent.

There's a reason for this last two changes. It’s that recessions create a mis-match between workers and vacancies and this mismatch tends to fade away in recoveries. In the early stage of recession unemployed builders (say) can't retrain as software engineers. This means they can't compete for the jobs that are available so don't bid down wages, with the result that high unemployment exists alongside rising inflation.

As the recovery takes hold, however, the mismatch declines, because vacancies re-emerge which the unemployed can fill and because unemployed workers retrain or move to where there are jobs. As this happens, unemployment falls without generating inflation. This is exactly what has happened recently.

Which brings us to our current puzzle. We can't be sure whether falling unemployment is a sign that inflation is about to rise or is instead a sign that the mismatch is continuing to decline.

However, even if it is the former, it doesn't follow that inflation is a big danger, simply because it doesn't respond very much to unemployment. The simple post-1997 relationship between headline unemployment and core CPI inflation implies that a percentage point drop in joblessness only adds around 0.27 percentage points to inflation. "The Phillips curve seems relatively flat" says Ethan Harris at Bank of America Merrill Lynch. One reason for this lies in a form of gift exchange between workers and employers. In the downturn, many employers did not cut wages. But the counterpart to this, says Mr Harris, is that wages won’t rise much in the upturn. Jeremy Lawson at Standard Life agrees. He adds that low imported inflation, plus the fact that households don’t expect inflation to rise, also mean that "inflation is not very responsive to changes in the unemployment gap."

Perhaps, therefore, the Fed and bond investors are correct to be relaxed about inflation.