The median wage in 2012 (about $16.28/hr) was only 5.7 percent higher (in real, inflation adjusted dollars) than it had been in 1973. Part of this is explained by a 3.5 percent drop in the downturn, but the bigger story was the three prior decades of paltry wage growth.

The culprits in this story are legion and include--over the long haul--a collapse in private sector union density, a meager (in value and scope) minimum wage, workers-be-damned trade policy, and job growth crowded into low-wage service occupations.  But a big part of the story is simply slack in the labor market.  Over the past generation, the only respite from unrelenting downward pressure on wages came during a brief spell of full employment in the late 1990s.  Those years saw wage gains across the board, closely resembling the shared prosperity of the 1947-1973 era.  But on either side of that boom, when high rates of unemployment were the norm, wages (especially for those at the median and below) fell steadily.

We can see the importance of full employment at work in the relationship between wage growth and unemployment in the states across the boom of the late 1990s and across the last business cycle.  The graph below plots the change in real wages for two seven year periods: from 1996-2002 and from 2006-2012 (the latter running from the year before the recession to the most recent available annual data) against each state’s unemployment rate at the midpoint of the period (1999 and 2009).

In the late 1990s, state unemployment rates clustered around that national rate of just over 4 percent, and only three states suffered unemployment rates over 6 percent.  Workers at most deciles, accordingly, enjoyed robust wage growth.  Full employment was especially important for workers at the bottom of the wage distribution.  Wage growth—and the relationship between full employment and wage growth—was strongest at the 10th decile and weakest at the 90th.

Over the last seven years, wage losses are both steeper at the lower wage deciles, and more clearly shaped by background patterns of unemployment (which vary more widely, by state, around the national rate).  We can see this in the general pattern (marked by the red trendline), in which states with higher rates of unemployment show steeper wage losses.  And we can see it in the outliers: North Dakota, whose energy boom largely insulated it from the national recession, sustained impressive wage gains at an unemployment rate of 4.2 percent.  Michigan, whose 2009 unemployment rate of 13.3 percent was the highest in the nation, also saw the steepest wage losses at the 30th, 40th, and median wage deciles.

Colin Gordon is a professor and director of Undergraduate Studies, 20th Century U.S. History, at the University of Iowa.