The consensus view in markets is that the United States is in the late stage of its economy cycle, that it is very close to full employment and that inflation will accelerate this year. On the back of this, many expect that the Federal Reserve – anticipating an overheating economy – will raise interest rates multiple times this year and scale back their asset purchases.
But what if this view is wrong? While we do see signs that indicate inflationary pressures in the US–from lagged PMI trends and capacity utilisation figures – we have repeatedly pointed out that unemployment is a misleading indicator of tightness in the US labour market. This is because the measure is blind to the large volumes of people who left the labour force after the Great Recession of 2007-9 and who are now returning. Official unemployment is a measure for only those people who are in the labour force at any given point in time but do not have a job.
This is the position that Martin Sandbu at the Financial Times (£) takes yesterday, and it can best be illustrated in the chart below.
Sandbu’s basic idea is that it is possible for the US economy to continue hiring people but for wages to remain muted as more people return to the labour market. We can see in the chart above that labour force participation – the proportion in the population of 25-54 year olds which takes part in the jobs market – is still well below its 1990s-peak, while the employment rate – the proportion of the population that has a job – is below its cyclical peaks in 2001 and 2007. The message, Sandbu concludes, is that there is still spare capacity in the US economy even though unemployment is low; and wage (and price) inflation is subdued because both employment and labour force participation are rising.
How likely is this? There is reason for doubt. Unemployment among 25-54 year-olds as a proportion of the population, implied by the gap between labour force participation and the employment rate in the chart, is also near the cyclical lows of 2001 and 2007. This is not the official measure of unemployment, which measures the proportion of the labour force that is unemployed, as opposed to the proportion of the population. However, official unemployment is also at similar cyclical lows. In both 2001 and 2007 for both measures, unemployment turned upwards within twelve months of peaking. This behaviour can be seen in the chart below at the peak of almost every economic cycle in the US since 1948. The only exceptions are in the late 1950s and late 1960s, when unemployment held steady as the economy added new jobs.
The question then is: which is the better cyclical predictor for labour market tightness – employment or unemployment? While Sandbu’s argument for the employment rate is plausible, it implicitly assumes that the damage wrought by the Great Recession can be largely if not fully unwound. On that point, we are sceptical.
There is evidence that during recessions, firms are more likely not just to fire workers but replace them with technology. Technological change has made large numbers of middle-aged Blue Collar men redundant. 90% of jobs in clothing manufacturing and 40% of jobs in electronics in the US have disappeared since 1990. Indeed, the labour force participation rate itself has been trending down since 2000. These jobs are not coming back and for many of the workers who are affected it is not possible technically or mentally to reskill. The scale of the opioid crisis points to the despair that many of these people feel with regards to their situation.
We don’t know what the scale of the damage has been, but there is good reason to believe that the labour force will not reach the size it did in the late 1990s. This doesn’t mean that labour force participation cannot continue rising in the short-term, delaying inflationary pressures; or that we should not pay attention to other measures of labour market tightness such as employment. But Sandbu’s implicit suggestion that, because employment remains low compared to the last 20 years, the US is necessarily still far from full employment is too easy. It is possible that the US economy has exhausted the gains from increased labour force participation, meaning that employment cannot continue to rise without pushing down unemployment, which is at cyclical lows. And this would most likely be inflationary.
We still therefore expect wage and price inflation to pick up this year as the US labour market tightens, but we would not be surprised if the process takes a bit longer than it does under the consensus view, with signs manifesting only during H2 2018. The Trump Administration’s fiscal stimulus program will, however, act only to hasten inflation’s arrival.
Our quarterly report presents our views on the world economic outlook and equity, fixed income and foreign exchange markets. Please click the link to download.