Chart Detail

Impact of rising and falling unemployment on wage levels and gaps, 1979–2011

1979–1985
(High unemployment)
1995–2000
(Low unemployment)
2006–2011
(High unemployment)
Men Women Men Women Men Women
A. Basic trends (actual change)
Unemployment rate change 1.4 1.4 -1.6 -1.6 4.3 4.3
Cumulative higher/lower unemployment rate* 13.9 13.9 -5.0 -5.0 15.2 15.2
50/10 wage gap (log) 9.6 17.2 -4.1 -1.8 2.7 -0.4
90/50 wage gap (log) 8.7 8.0 3.0 1.0 3.1 2.7
B. Estimated cumulative impact of
unemployment on:
1) Hourly wages**
Low wage -25.2% -18.4% 11.1% 8.2% -29.4% -21.5%
Median wage -11.1 -11.6 4.9 5.2 -13.0 -13.6
High wage -6.6 -10.2 2.9 4.6 -7.7 -11.9
2) Wage ratios (log)
50/10 14.1 6.8 -6.2 -2.9 16.4 7.9
90/50 4.5 1.5 -2.0 -0.6 5.3 1.7
C. Unemployment contribution to change in:
50/10 wage gap (log) 146% 39% 150% 168% 617% -2123%
90/50 wage gap (log) 52 18 -66 -61 169 64

* How much the unemployment rate exceeded or fell below the starting level across the span of the period, measured in percentage points

** Wages at the 10th, 50th, and 90th percentiles of the wage distribution

Source: Authors' analysis of Current Population Survey Outgoing Rotation Group microdata and Current Population Survey public data series using model from Katz and Krueger (1999)

Updated May 18, 2012

Documentation and methodology

Table is based on analyses of yearly wage decile data from Tables 4.5 and 4.6 (see Appendix B), and of unemployment data using model from Katz and Krueger (1999). The unemployment rate is from the Current Population Survey. The simulated effect of change of unemployment presented in the table was calculated by regressing the log-change of nominal wages on the lagged log-change of the CPI-U-RS (but, following Katz and Krueger [1999], the coefficient is constrained to equal 1), the unemployment rate, lagged productivity growth, and dummies for various periods (1989–1995, 1996–2000, 2001–2007). Using these models, wages were predicted for the periods in the table given a simulated unemployment rate series in which unemployment remains fixed at its starting-year level. So in the 1979 to 1985 period, unemployment was fixed at its 1979 level and not allowed to rise (as actually happened) throughout the period. The “estimated cumulative impact of unemployment” shows the difference between actual wages and the wages when unemployment was held fixed in the starting year.

Get the data

Get the data behind the charts.

Explore Our Charts

Subject

Demographic

Select a subject or demographic

Interactive feature

Interactive feature promo image When income grows, who gains?

For the media

Economic Policy Institute Media Relations Department (202) 775–8810 |