Urban Wire Stimulus vs. austerity across the top 100 metros
Josh Mitchell
Display Date

The drop in public-sector employment and the possible negative effects on the recovery have gotten a lot of attention recently.  It is important to realize that while nationally public sector employment has declined, employment trends differ both across metros and by level of government. Public-sector employment has declined in 57 out of the 100 largest metro areas since the onset of the Great Recession (all calculations are from BLS Current Employment Statistics 4th quarter of 2007 and 2011). Across the top 100 metros, federal employment has increased by 50,000 jobs (3 percent) while state employment has declined by 33,300 jobs (1 percent) and local employment has dropped by 161,000 jobs (2 percent). While these changes have garnered attention, it’s important to note that during this time, private-sector employment has fallen by 4.5 million jobs (6 percent).

Employment trends also differ across metros. The table below lists metros with the biggest changes in public-sector employment and tracks their private-sector employment performance. Two metros in Texas  (McAllen and Austin) had the highest percent growth in public sector jobs and also saw growth in private-sector jobs. Three metros in California (Santa Rosa, Modesto, and Stockton) had some of the largest declines in public-sector jobs and also saw very large declines in the private-sector.

Metros with Highest/Lowest Percent Growth in Government Employment (2007-2011)

Source: Author’s analysis of BLS Current Employment Statistics (2007 and 2011)


Meanwhile, the national debate over temporary government expansion (stimulus) versus contraction (austerity) continues. From the American Recovery and Reinvestment Act (ARRA) of 2009 to extensions of unemployment insurance and the payroll tax cut, proponents of stimulus have sought to boost public and private spending to accelerate the recovery. Opponents of stimulus have advocated just the opposite: cutting government spending, they argue, will increase confidence, spur investment, and lead to higher growth.

Can we learn anything from what metros have done? Before considering the possible stimulative or contractionary effects of changes in government employment, it’s important to note the relative size of each sector. As of the end of 2011, the private sector represents 85 percent of total metro employment, the federal government represents 2 percent, state government represents 3 percent, and local government represents 10 percent.

How have changes in public employment affected overall (and mainly private) employment? Below is a simple scatterplot relating the change in each metro’s public-sector employment (x-axis) to the change in its overall employment (y-axis) expressed as a percent of pre-recession employment.

Change in Government and Total Employment Across Metros

Source: Author’s analysis of BLS Current Employment Statistics (2007 and 2011)


The strong positive relationship indicates that every additional government job is associated with about three additional total jobs (meaning two additional private-sector jobs). This relationship is similar to what Paul Krugman finds for public-sector spending across European countries. (In fact, the slopes are the same!) However, one needs to be cautious in interpreting this relationship. Causality could run in either direction: government employment gains from ARRA and elsewhere could be responsible for boosting a depressed economy. Alternatively, the scatterplot could simply illustrate that those metro areas hardest hit by recession were forced to meet balanced budget requirements by slashing government payrolls.

We cannot sort out cause from effect from the graph alone. However, recent research provides strong evidence that government funds allocated across U.S. states and localities did indeed boost the local economy. Using situations where state and local governments ended up with more or less funding for reasons unrelated to current economic conditions and beyond their control, Gabriel Chodorow-Reich of the University of California, Berkeley, and colleagues found that states that received higher Medicaid matching funds experienced employment boosts of 3.8 job years per $100,000, with the vast majority of jobs created in the private sector. Studies by Feyrer and Sacerdote at the National Bureau of Economic Research and by Wilson at the Federal Reserve Bank of San Francisco find more modest but still substantial job-saving effects.

So both the metro experience and recent research suggests protecting public-sector jobs might ensure that the current recovery continues.

Research Areas Neighborhoods, cities, and metros
Tags Employment and income data Job opportunities Labor force Unemployment and unemployment insurance Community and economic development