Support Ratio More Accurate Than Unemployment Numbers

US unemployment data is actually a pretty sad indicator of economic health. It ignores too many things.

Unemployment is simply the number of folks unemployed divided by the civilian population aged 16 and over. There are many problems with the measure.

1. It ignores institutional employees. Civil servants make up a very significant and stable part of the consumer population who contribute to GDP. Why ignore them?
2. The unemployed exclude new entrants into the labor force who have been unable to find jobs as they are not considered part of the labor force. It also does not include people who have simply removed themselves from the labor force because they have been unable to find a job. It also excludes people who have headed for re-training or higher education. And of course, it excludes retirees and kids below 16. The folks who are "not part of the labor force" are important; those in employment (both civilian and institutional) must support them.
3. It also ignores demographic decay. A country can have 100% employment and remain in a very high risk position where, for example, the work force supporting the population is very small (i.e. a high retired population and low working population).

For this reason, I tend to follow the "Support Ratio" as a better indicator of economic health. The Support Ratio is simply the number of people employed (institutional & civilian) divided by the total population. This tells you the % of the population which supports the consumer.

As of now, the support ratio is at 70.2%. Since 1947, the Support Ratio has varied in a range between 68.60% and 73.11% with a mean of 71%. Presently we are at a level well below average and in the first quartile range; we are also 0.43% below trend. Clearly the situation is not pretty. So what does the future hold?

In my view, the base case should see the Support Ratio falling further to 69.9%; at this level, close to 4 million further job losses can be expected to arise which will lead to a conventional unemployment rate of 9.1%. As an unlikely worst case scenario, we could see the Support Ratio slip to 67.75%. This number is the mean Support Ratio less 3 standard deviations, so the probability of anything worse occurring is near zero. (However, one should bear in mind that statistic modeling does not always work; all it can do is signal a near zero probability. An event can make 100% of that remote likelihood become reality, no matter how improbable.)

With a Support Ratio of 67.75%, we will be more than 4% below trend, which is about the worst since 1947. At this level of Support Ratio, conventional unemployment will measure at 9.8% and lead to an additional 5 million job cuts over the coming months. Some of the pain will be alleviated through productivity gains and lower inflation in the short term; rising inflation will reappear once employment starts ticking upwards once more.

In my view, the market has priced these risks.

Data junkies can have a look at the numbers at http://www.maxkapital.com/USEmploymentData.pdf.