Friday, 6 December 2013

The good, the bad and the ugly of the US labour market report.

I'm not sure whether the economics profession has ever spent quite so much time arguing about whether the monthly US labour market data are good or bad, as they do now. We used to debate whether the figures meant very much, given that they are subject to big revisions, but we did at least agree whether they were good, boring, or awful.

So, two charts below to provide a very short interpretation of the issues with the data. The top chart shows the unemployment rate (in yellow, inverted, since 1980), and the employment rate. Unemployment is the percentage of people in the labour force that have not got a job. The employment ratio is the number of people with jobs, divided by the total population. From 1983 until 2004, the two  moved together, falling unemployment seeing the employment rate rise, since then, things have got tougher. And since 2010, they tell very different stories. The unemployment rate has fallen steadily. Employment is growing at a rate of 1.7% per annum and payroll growth has been averaging about 170,000 per month for 5 years. The labour force has been growing much more slowly, so the unemployment rate has fallen. And you can draw a straight line through the data and conclude that a 3% fall in 4 years will get the unemployment rate under 6% in 2015. But while growth in the labour force is very weak, the population IS still growing and the employment rate is going nowhere. This is why so many people will tell you today that the fall in unemployment is the 'wrong sort of fall.

The second chart shows the two employment surveys, the household survey which asks people if they have a job and the establishment survey which asks companies how many people they employ. The headline non-farm payroll figure comes form the establishment survey, which is a larger sample. The unemployment rate comes from the household survey. The October data for the household survey were distorted by the Federal shutdown, and that caused the unemployment rate to rise, while employment fell. That is the white line. The Two lines mostly move together, though you can see they have crossed over since 2008. You can also see, I hope, that the household measure of employment has bounced in November, but has not recovered all the ground lost in October. So, if you take the household survey and look at the September-November outcome, jobs were not added. And in both surveys, you can see that employment is merely approaching the levels it was at in late 2007.

Overall, 1.7% employment growth is in line with the average of the last 50 years. That would be fine if we were not trying to recover from the worst recession of the last 50 years. The US economy is growing, but is not catching up lost ground. The unemployment rate is one measure which suggests that some, even much of the output lost in the recession has been lost for ever, and the future will simply see a normal growth path from a low level. Some US economists, including Larry Summers very publicly but also Janet Yellen, simply don't accept this. There is a belief that if the central bank maintains accommodative policy settings for long enough, they can recover some of this 'lost' output. Put it another way - if they ignore the falling unemployment rate maybe some of the people who have become disillusioned and left the labour market will go back to work. If they are wrong, US wage growth and then US inflation will go up if the economic recovery continues.

Personally, I admire the willingness of US policy-makers to throw away textbooks, re-write theories and refuse to accept that a 'new normal' economy is simply not as vibrant as the old one. The contrast with the lack of policy reaction to 0.1% GDP growth and 12% unemployment, is incredible. Refusing to accept the status quo is a bit like Oliver Twist asking for more gruel... it shakes things up.

No comments:

Post a Comment