Accelerated Gains in the Job Market

For the first time in several months, the BLS has released an employment report in which the employer survey and household survey tell a consistent story: The U.S. job market is improving. The employer survey showed job gains of slightly more than 190,000 in February, while the household survey showed an increase of 250,000 in the number of adults who say they are employed. This is the best combined picture of employment gains since early last spring.

February’s employment gains produced another small dip in the unemployment rate. Since November the unemployment rate has dropped 0.9 percentage points, falling below 9 percent for the first time since April 2009. All the decline was due to employment gains. The labor force participation rate has actually edged up 0.2 percentage points since November.

The recent improvement in the job market should not obscure the painfully slow pace of improvement since employment started to grow in early 2010. Both the employer survey and household survey show only modest monthly gains coming out of the nation’s worst post-war recession. Since December 2009 payroll employment in the employer survey has increased slightly less than 1.2 million, or about 85,000 jobs a month. The number of adults who say they are employed in the household survey has increased about 1.6 million, or 115,000 a month.

All of the job gains have been in the private sector. Government employment has dropped 268,000 (or 1.2%) since December 2009. Most of the falloff was in local government employment, with the decline split equally between jobs in education and in all other functions of local government. Government job losses continued in February, and I expect them to continue for a number of months. Federal fiscal assistance to state and local governments, generously provided in the 2009 stimulus package, is nearly at an end. States and localities must plan to finance more of their operations out of their own taxes and fees. Because many state governments are reluctant to boost local taxes, it is likely they will continue to trim their payrolls.

The improving pace of job gains in the private sector is good news for the unemployed. The number of unemployed fell 190,000 in February. It has fallen by a monthly average of 456,000 since November 2010. Over the same time span, the number of unemployed workers who have been jobless longer than 6 months has dropped 112,000 per month. The median duration of unemployment spells in progress is still extremely high by historical standards, but it has improved in recent months. People suffering very long spells of joblessness have borne an extraordinary share of the labor market pain inflicted by the recession and slow recovery. The average duration of an unemployment spell in progress hit a new all-time high in January and February, a spell length of more than 8 months. Measured on a comparable basis, the average duration of an unemployment spell at the beginning of 2011 was 68% longer than the average unemployment duration in the 1981-1983 recession, when the average unemployment spell hit a peak of a little less than 5 months.*

The recent good news about the job market is heartening, but we should not lose sight of the slow overall progress made since the economy began to grow in the summer of 2009. The nation has not come close to replacing the jobs lost in the recession. Payroll employment remains almost 7.5 million below its level when the recession began, and the number of adults who report being employed is still 6.7 million below the pre-recession peak. The rate of job recovery is slow because the pace of economic expansion has been anemic by the standard of post-war recoveries that follow deep recessions. The modest improvement in aggregate demand combined with continued robust growth in labor productivity means that many employers have seen little reason to expand their payrolls.

* Correction:  In an earlier version of this comment, Burtless stated that the average duration of an unemployment spell reached 37.1 weeks in February 2011, an average duration that was 75% longer than the peak average duration in the 1981-1983 downturn.  This statement was based on the estimates published by the BLS.  However, the statement is a bit misleading.  In January 2011 the BLS began to use a new method to calculate the average unemployment spell. The new method tended to increase the measured duration in comparison with the method used before January 2011.  Using the same method used by BLS to calculate average unemployment durations in December 2010 and earlier months, the average duration of unemployment in February 2011 was just 35.6 weeks.  Even using the old calculation method, however, the average duration of unemployment spells reached a new post-war peak in January and February 2011, attaining a level that was 68% higher than the peak level attained in the 1981-1983 downturn.