With Employment Data, the Devil’s in the Details

When I speak to groups in various parts of the country, I often hear that things there don’t reflect the national employment averages and trends I talk about. This usually leads to some interesting discussions about why that might be the case.

The problem with looking at region-specific trends is the lack of research in that area. So I was glad to discover a recent report by Deloitte University Press that breaks out some numbers on a regional basis, with interesting results.

The geography of jobs

The report makes several key points worth noting:

• The recovery is not only slow but much more concentrated than generally thought. Only 16 states and the District of Columbia have employment of more than 1 percent above their pre-recession levels.
• The drivers of employment growth are quite limited, geographically. Many of the states that have gained are in the northeast, such as Massachusetts and New York, which have benefited from growth in business services, health, and education. Other states in the middle and west of the country, such as North Dakota and Texas, have benefited from the energy renaissance.
• The weakness in the rest of the country is due largely to declines in two sectors—construction and manufacturing. Only three states and DC have shown growth in the construction sector, while many states have posted double-digit declines. For manufacturing, only two states have grown, and many have seen significant declines.

Keeping things in perspective

I don’t question the numbers here, but we should take some other factors into account when interpreting this data.

First, using the peak numbers from December 2007 to measure a decline is somewhat misleading. Those figures, particularly in construction, were based on an unsustainable boom; we simply didn’t need all of the houses that were being built.

Second, the state-level perspective neglects relocation effects and relative state sizes. Percentages can mislead. A 9-percent job gain in the large state of Texas, for example, can offset much greater percentage declines in smaller-population states, simply because of the numbers involved. The state numbers don’t negate the national ones; they simply provide a different perspective.

Third, absolute changes in employment don’t necessarily take into account sectoral changes. As housing construction declined, for example, growth from other, more sustainable sectors might have filled the gap, resulting in a healthier economy, even if overall employment did not grow.

Is the recovery as sustainable as we thought?

Those caveats notwithstanding, the report prompts a number of questions. Although national trend growth has returned to normal, the distribution of the job gains raises serious concerns about the sustainability of consumer spending growth. With more than two-thirds of states showing no real employment growth, the recovery certainly looks less durable than it does from a national perspective. 

Kudos to the authors for putting out an accessible report on data sets that should get more attention.