Sorry, Wrong Number: Job Growth Nonsense

Sorry, Wrong Number: Job Growth Nonsense

My wife says, correctly, that I don’t suffer fools well. More accurately, and more painfully, I don’t suffer people who I think are fools very well . . . though often it turns out that they are not the fools in the relationship.

I have even more trouble with foolish statistics than I have with foolish people. At least I know that someone I consider foolish is still, at the end of the day, a person, entitled to my respect and fair treatment. Not that I achieve that all the time; I just recognize its truth, and when I have failed to honor it.

Numbers are entitled to no such special regard.

So let me just say, straight out, that using the number of jobs created to compare states’ economic growth or vitality is . . . well, incredibly foolish.

It was just a short note in the Business section of the Tampa Bay Times today:  Net jobs in Florida rise.  The story included a nice little table listing the five states that produced the most jobs in the last year, and the five that produced the least.

It should come as absolutely no surprise that the states that produced the most jobs also are among the states with the largest populations. All five are in the top ten in population, and only one (North Carolina) is not in the top five.

Similarly, the states that produced the least jobs also are among the states with the least population. The relationship is a little less strong here. Three of the five states producing the least jobs are among the 10 states with the least population (one is among the five smallest states). The other two are somewhat smaller states (ranked 32nd and 30th), but not the smallest states.

I trust it is obvious why one would expect a strong statistical relationship between the size of a population and the size of job growth (or, for that matter, job loss in bad times). But just to illustrate the point:

The state with the smallest job growth (Montana, with 8,800 new jobs) has an estimated 2013 population of just over one million. The state with the largest job growth (Texas) saw 396,200 new jobs between July 2013 and July 2014. In order for Montana to have rivaled Texas’s raw job growth numbers, nearly 40 percent of the state’s population would have had to get new jobs.

But for Texas to generate 396,200 new jobs means roughly 1.5 percent of the population ended up in a newly created job last year.

40 percent growth, or 1.5 percent growth.

Makes a big difference what the base is.

A better base than population is number of jobs. According to the Bureau of Labor Statistics, 396,200 new jobs in Texas in the last year amounts to a 3.5 percent growth in total nonfarm employment. For Montana, their 8,800 new jobs looks like almost two percent growth overall.

So yes, Montana did see less job growth than Texas in relative, as well as absolute, terms. But a 1.5 percent gap constitutes a much closer race than a gap of 387,400.

If we were to re-sort this selection of 10 states (the five with the largest number of new jobs and the five with the smallest number of new jobs, July 2013 to July 2014) by percentage growth in total nonfarm employment, two of the “least growth” states would be among the five with the most growth. Indeed, North Dakota would top the list, posting a 4.4 percent growth in employment year-over-year (almost a full percentage point better than Texas’s 3.5 percent). California and New York would drop to sixth and a dismal ninth, respectively. (Florida, by the way, would remain in third place in this analysis, with 2.7 percent growth).

Some other states also would be clamoring for recognition, like Nevada (3.8 percent growth), Utah, (3.6 percent growth), which both beat Texas, and Colorado (2.7487 percent growth, marginally better than No. 3 Florida).

The simple point: total number of jobs generated is interesting, but has no value if we are trying to compare the performance of different states (or, for that matter, different regions within a state). Success in economic development only can be measured meaningfully if a relative base of comparison (like total number of jobs) is used.

Presenting absolute numbers of jobs generated is foolish. It also can make successful states seem much less so, and states that are struggling seem like winners.

2 Responses to Sorry, Wrong Number: Job Growth Nonsense

  • Richard L. Block

    Dear Scott. Why bother to waste your time on stats? They are only useful for the ones who are employed to produce them. Water seeks is own level. States which are employing will be known by word of mouth. Demographics have changed without reading employment statistics. No need to spend time burning up you precious God given time to read and analyze cooked statistics.

    • Dr. Scott Paine

      I understand the perspective, but I think dismissing statistics of this sort as ‘cooked’ is unfair and misleading. No statistical data is perfect, and certainly one can ‘cook’ statistics. But statistical data of the sort I’ve discussed here is useful as one way of examining developments.

      Statistics also can be politically powerful. That’s a very good reason to examine them closely and critically, so that such power is not given to them inappropriately.

      Thanks for reading and responding!