Job Growth in Perspective, Part I

Editorial Note: Because this post turned out so long, I’m breaking it into two parts. The current post gives some motivation and then describes and presents the data, but with little interpretation. I’ll give my interpretation and comments in Part II.

In a fantastic post on Friday, Kash showed how truly poorly the Bush II Economy rates on a variety of metrics: employment, GDP, compensation, after-tax income, and government spending. Inspired by Kash’s post, as well as the frequent surfacing of jobs-creation as a debate issue, I decided to examine the following question:

We all know that Bush is the first president since Herbert Hoover to lose jobs over the entire course of his administration. That’s clearly bad, but how bad is it?

First the mandatory disclaimer. Kash, PGL, and I have written at various times and in various ways that the there’s truth to Bush’s claim that he inherited a recession. Where we place the blame is on his failure to effectively do anything about it. Certainly, cutting taxes likely had some stimulative effect, but it was evidently minimal — seemingly by design. By skewing the benefits so dramatically upwards, directed towards the few rather than the many, he virtually ensured that the resulting demand stimulation would likewise be narrow rather than widespread. Since the demand stimulus was minimal, the resulting need for additional people to make stuff (i.e., job creation) was insufficient to offset the recession-induced fall in jobs. Today, we see the results of this ‘policy’: an astoundingly bad, historically bad, record of job creation.

How bad is it? Very bad. I have color-coded the data below it so even the dimwitted will understand this.

The data are from the Greenspan-approved establishment, or “payroll”, survey (actually, nearly every economist not working for or hacking for the administration says the establishment survey is better) which is far more comprehensive than the household survey. The data are aggregated from monthly changes, meaning the annual change figures are the average of the monthly changes within a year (if you instead computed year-over-year changes, the resulting numbers would be close but slightly different.)

Also, since this has been an issue recently, I’ve broken it down into private-sector and total employment; the difference between the two is government employment. A few more notes, and then the data. From the second half of Eisenhower’s administration to the present, the overall average month-over-month rate of employment growth is 0.17% (in today’s numbers, that’s about 180,000 new jobs monthly). The standard deviation of that statistic is .0015, and the distribution is somewhat skewed downward, meaning that employment growth one standard deviation below the mean or more is more common than the reverse. The best year in the data, 1978 (Carter), averaged .41% monthly job creation; the worst year, 1982 (Reagan), averaged -.20% monthly job dis-creation.

Here’s the key to the color-coded data:

And here are the jobs data:

AB