How Shifting Occupational Composition Has Affected the Real Average Wage: OES Data from 2002-2007 Reveal That an Overall Shift in Employment towards Occupations with Lower Mean Wages Hindered Growth in the U.S. Real Average Wage and That Wage Growth Was Concentrated in Higher Paying Occupations; the Data Also Show a Shift in Employment from the Middle-Paying Occupations to the Highest and Lowest Paying Occupations

Article excerpt

Between November 2002 and May 2007, the cross-occupational average hourly wage in the United States increased by $2.46, from $17.10 to $19.56, or by about 14 percent, according to the Occupational Employment Statistics (OES) program. Adjusting the 2002 figure to May 2007 dollars (1) shows the real average hourly wage increased from $19.48 to $19.56, approximately a .41-percent increase.

There have been numerous studies and programs devoted to understanding this recent slow growth in the Real Average Wage (RAW). Many studies attribute slow wage growth to the increasing cost of employee benefits and health insurance--a phenomenon that results in employees' wages becoming a smaller part of their total compensation. (2) Other studies have analyzed how wage growth relates to income or wage inequality. (3) This article seeks to contribute towards an understanding of RAW growth by quantifying how changes in the occupational composition of U.S. employment have affected the average wage.

This article analyzes occupational wage and employment data from the OES program to understand how changes in occupations' wages and changes in occupations' levels of employment each have contributed to growth in the U.S. RAW. Overall wage growth could stem from increases in the mean wages of particular occupations, from a shift in employment towards occupations with higher wages, or from a combination of the two factors. This article's analysis of OES data from November 2002 to May 2007 finds that a shift in employment towards lower paying occupations hindered U.S. RAW growth, that increases in the real mean wages of individual occupations was the only factor that caused growth in the U.S. RAW, and that most of the average wage growth was due to increases in the wages of the highest paying occupations. This analysis also finds a shift in employment towards the highest paying and lowest paying occupations and away from middle-paying occupations. This article will show which occupations experienced growth and which experienced decline in real mean wages or in share of employment, and how these changes influenced the U.S. RAW. It will also reveal patterns of lower and higher paying occupations and of education and training categories, and give a brief analysis of changes in the average wages of U.S. States.


The OES program estimates national employment and wages by occupation and provides a data set for understanding changes in the average wage over the medium term. The OES program surveys 1.2 million business establishments, using 3 years of data collected in six semiannual panels to produce estimates for over 800 occupations. (4) Because of the survey methods employed, it can be difficult to use the data for time-series analysis, but this study mostly overcomes the issue because it compares wage and employment data 4 1/2 years apart and analyzes cross-industry wage and employment estimates that have been retabulated on the basis of a common coding system. (5) However, between November 2002 and May 2007, OES implemented refinements in occupational coding procedures that have caused some management workers to be moved from one occupation to another. Therefore, some results of this analysis may have been affected by this worker classification change and must be interpreted cautiously.

Change in the U.S. average wage may be due to changes in the mean wages of individual occupations or to shifts in employment among higher and lower paying occupations. An occupation's share of national employment is the percent of total jobs in the Nation for which the occupation accounts. This article uses a "shift-share analysis" of OES data to quantify the effect of changes in mean wages and the effect of changes in employment share on the U.S. RAW from November 2002 to May 2007. (6) OES data previously have been employed to examine the role of occupational composition, or the assortment of shares of national employment held by occupations, in the average wage differentials of U. …