Sources for the Unskilled Wage

by

Samuel H. Williamson
President of MeasuringWorth
and
Professor of Economics Emeritus
The Miami University
sam@mswth.org

In order to get a complete index of the money wage rates paid for common or unskilled labor from 1774 to the present, three sources are used: a series published by Paul David and Peter Solar in 1977, the work of Robert Margo published in 2000, and various publications of the Bureau of Labor Statistics (BLS).

For 1774 to 1824, the David and Solar (1977) series is used. (Table B.1 p. 59)

For 1825 to 1860, the Margo (2000) series is used. (Table 5B4 p.117).

For 1861 to 1973, the David and Solar (1977) series is used. (Table B.1 p. 59)

For 1974 to 2000, various sources for the BLS as explained below. These data are from the Current Employment Statistics.

For 2000 to the present, the BLS series of the median usual weekly earnings (second quartile) of those wage and salary workers employed full time with less than a high school diploma, and are 25 years and over is used. These data are from the Current Population Survey.

Sources:

David, Paul A. and Peter Solar, (1977).  "A Bicentenary Contribution to the History of the Cost of Living in America," Pages 1-80 in Volume 2 of Research in Economic History, Greenwich: JAI Press, Inc.

Margo, Robert A., (2000) Wages and labor markets in the United States, 1820-1860, Chicago : University of Chicago Press

U. S. Department of Labor, (1976), Area Wage Surveys: Metropolitan Areas, United States and Regional Summaries, 1973-74, BLS Bulletin 1795-29, Washington D. C.: U. S.  Government Printing Office.  Table A-28, p. 88, "Earnings Index-all metropolitan areas," "unskilled plant labor."

_____ . (1982), Labor Force Statistics Derived from the Current Population Survey: A Databook, Vol. 1, Bulletin 2096, Washington D. C.: U. S.  Government Printing Office.  Table C-26, p.736, "Median hourly earnings of wage and salary workers paid at hourly rates by detailed occupations, May 1973-78."

_____ . (1983), Handbook of Labor Statistics, Bulletin 2175, Washington D. C.: U. S.  Government Printing Office.  Table 100, p.258, "Average earnings for selected plant occupations in metropolitan areas by sex, industry division, and region, 1967-81," "Janitors, porters, and cleaners."

_____ . (2000), Employment Cost Indexes, 1975-99, Bulletin 2532, Washington D. C.: U. S.  Government Printing Office.  Table 10, p.85, "Wages and Salaries, private industry workers: Employment Cost Index by occupational and industry group, 1975-99," "Handlers, equipment cleaners, helpers, and laborers."

_____ . (2006), Employment Cost Index.  [On-line Database]  Available: http://www.bls.gov/ncs/ect/home.htm

_____ . (2009) Quartiles and selected deciles of usual weekly earnings of full-time wage and salary workers by selected characteristics, quarterly averages, not seasonally adjusted.[On-line Database]  Available: http://www.bls.gov/cps/

The method of extending the series from 1973 to 2000

In order to continue the David and Solar series after 1973 using current data published by the BLS, two different data series must be merged, using a third series to tie them together.  For the period from 1961 to 1974, David and Solar use the index of hourly wage rates of "unskilled plant workers", from the BLS Bulletin #1775-98.  In 1974 the BLS stopped publishing this series, however, so I chose as the closest comparable series, the BLS Employment Cost Index of "Handlers, equipment cleaners, helpers and laborers," which is available annually from 1976 to the present (U. S. Department of Labor, 2000, p. 85).  The problem is the missing years between when the David and Solar series ends and when the BLS Employment Cost series begins.  Below is an explanation of how these missing numbers were created and the data series were linked.

After the publication of the David and Solar article, the BLS published a Handbook of Labor Statistics in 1983 that included national averages of the series "Janitors, porters, and cleaners" and "Material handling labors", for the year from 1967 to 1981. A complication of using these data are that during the period I am trying to link, the BLS changed the month the data "relates" to from February for the years up to 1975, to July for the years 1976 and after.* This was also a period of rapid wage increases.

Therefore, to give these data observations a "better" annual fit, a fourth order polynomial was fit to the 1973 to 1977 monthly observations with the average earnings as the dependent variable and the month as the independent variable.  The resulting fourth order equation was then used to estimate the other months of the years 1973-1977.  From these estimates, an annual average was calculated for each of these five years.  These are in Column A of the table below.

To use these data to link the years between the David and Solar series and the Employment Cost series, I began with the first full two years of observations in the Employment Cost series (Column B1 below) and took the average of 1976-77 observations divided by the average of the 1976-77 annual observations from column A.  This gave a value of 11.70.  This number is then multiplied times the observations in column A for the years 1973-1975 to estimate of the employment cost index for those three years.  The complete observations are in Column B2.

The final step to merge with the David and Solar series is to divide the value in their series for 1973** by the Employment Cost Index of the same year found in Column B2, resulting in a value of 94.32, and then multiplying that number times the rest of the values in Column B2 for the years 1974-2000, giving the yearly index of money wage found in Column C.  This completes the index for all the years 1774 to 2000.

	Column A	Column B1	Column B2	Column C
1973	3.387		39.6	3734
1974	3.636		42.5	4008
1975	3.949		46.2	4354
1976	4.316	50.6	50.6	4768
1977	4.679	54.6	54.6	5150
1978			59.1	5570
1979			64.1	6042
1980			70.5	6650
1981			76.4	7209
1982			80.5	7588
1983			83.9	7914
1984			87.0	8202
1985			90.0	8485
1986			91.7	8647
1987			93.7	8838
1988			97.3	9175
1989			100.6	9484
1990			104.7	9878
1991			108.7	10255
1992			111.7	10531
1993			114.6	10812
1994			117.7	11100
1995			121.4	11446
1996			125.6	11842
1997			129.9	12255
1998			134.5	12682
1999			138.9	13097
2000			144.2	13597

_____________

*The explanation in footnote on page 263 states, "Because surveys are conducted at various times throughout the year, combined data relate to an average payroll month of reference.  For 1975 and subsequent years, data relate to July (individual area data were combined on a calendar year basis). Before 1975, the data relate to February (individual area data were combined on a fiscal July through June - basis)."

**At the time, David and Solar did not have the data to compute the final index on a calendar year basis in 1973 and 1974 the same way they computed 1952-1972.  They used a regression to estimate these last two observations.  Since the publication of their work, updated data from the BLS "Earnings index" allows for the value of 1973 to be observed from the source.  The 1973 observation is 3734 and not 3753.