By Most Measures, Changes in District Employment Are Closely Following the U.S. Average
At the national level, the Labor Department tracks employment using two different surveys. One survey asks business establishments how many people they employ, while the other asks households how many individuals in the home have jobs. Differences in the sample size of each survey and the way they define employment can lead to different estimates for employment. For instance, someone who holds two jobs will show up once in the household survey, but twice in the establishment survey. The establishment survey also can’t capture self-employed individuals.
Subtle differences such as these can lead the two series to diverge, especially at transitions in the business cycle. This is evident in the recovery that followed the 2001 recession.
In the most recent recovery, the two series also began to diverge somewhat around the beginning of 2009. A roughly 1 percentage-point gap has persisted since.
For regions within the country, employment measures can be constructed that are conceptually similar to each of the national series. Establishment- and household-based measures for the Fourth District have followed the U.S. measures closely. Toward the end of last year, the establishment-based measures for the U.S. and Fourth District were about 3 percent below their respective December 2007 levels, when the recession began. For the household-based measures, employment in both the U.S. and the District was about 2 percent below December 2007 levels. (Technical note: The smallest geographic area for which establishment-concept employment measures are available is the metropolitan area. Accordingly, the District measure aggregates employment from metropolitan areas that are fully or partially contained in the District, but excludes employment from nonmetropolitan areas.)
It is a little surprising that changes in District employment have so closely followed the national pattern, especially in light of the 2001 recession and recovery episode. Over the roughly five-year span following the business-cycle peak in March 2001—about the same amount of time that has elapsed since the start of the last recession in December 2007—national and District employment measures exhibited much different growth trajectories. Either type of employment measure suggested that the District had seen employment growth that was about 4 percentage points lower than the nation’s over this period.
The weaker employment recovery that the District experienced in the 2000s—adding almost no net new jobs during the expansion—was broad based. Essentially every major industry group grew its payrolls faster (or reduced them less aggressively) outside of the District. Employment in industries like education and health care, professional and business services, and leisure grew in the District, but more slowly than outside of the District, while manufacturing and information shed proportionately more workers here. Perhaps most notable is the collection of industries in which employment shrank here but grew in the rest of the country—among which were wholesale and retail trade, extraction and construction, and financial services.
In the current recovery, this pattern has so far not arisen. Instead, there are minor differences in employment growth across industries, with the District sometimes faring better and adding proportionately more workers than the rest of the country, and sometimes not. On balance, the overall change in the establishment-based measure is nearly identical in the District and the nation, rounding to 2.7 percent in both cases.
The very even recent performance suggested by the foregoing comparison of household- and establishment-type employment measures, as well as by the comparison of employment changes across industries during the recovery, is contradicted by the District’s unemployment rate. As of October, the latest month for which these data are available, the District’s unemployment rate was almost a full percentage point lower than the national average—7.0 percent versus 7.9 percent. (The most recent estimate for the U.S. rate is 7.8 percent for December.)
The rates began to diverge in the summer of 2010, and since the summer of 2011, the District’s rate has been at least half a percentage-point lower than the national average. How could this be the case, when the household-based employment measures, which are used to calculate the respective unemployment rates, have behaved so similarly? The answer is that the labor force in the Fourth District has followed a different path during the recovery than the nation’s labor force. Just as that gap began to widen in the middle of 2010, so too did the unemployment rates. If the District’s labor force had followed the same path as the national labor force since December 2007—that is, changed in proportionately the same way since—the two unemployment rates would be almost equal—8.0 percent in October for the District and 7.9 percent for the U.S.
That the divergence between the District’s unemployment rate and the national average is being driven largely by labor force declines should give us pause. These declines aren’t indicative of a strong labor market. Accordingly, it would be inappropriate to interpret the District’s below-average unemployment rate as suggesting as much. Alternatively, perhaps the District’s labor force is simply being mismeasured and is tracing a path more like what we’re seeing for the nation. In that case, the District looks like an average performer, rather than an above-average performer. Either way, the District’s unemployment rate should be interpreted with caution.