This refers to the article entitled "The public-private pay debate: what do the data show?" by Michael A. Miller., appearing in the May 1996 issue of the Monthly labor Review.
Mr. Miller concluded that government employees, at least in the white-collar area, are paid less per week than employees in the private sector. However, he does not answer the real question of whether compensation for public employees per hour worked is higher than the compensation paid per hour worked for comparable employees in the private sector.
By compensation, I mean the traditional measure of all elements of pay including fringes, health and welfare, and payroll taxes. By per hour worked, I mean the number of hours that an employee actually works and this excludes holidays, vacations, sick leave, and any other type of paid leave.
It should not be difficult to get credible estimates of total compensation and hours worked, by occupation, for both the private and public sectors. This should be done, and if is, it will settle the controversy once and for all. I would be very much interested in whether Mr. Miller's conclusions remain the same after he takes all of the relevant factors into account.
[signed] John Roberts Support Services for Collective Bargaining/Litigation
Dear Mr. Roberts:
The two issues you raise--using wages or total compensation, and comparing costs per hour paid or costs per hour worked--are important when making comparisons between sectors of the economy. The studies we have made to date, however, have been limited by the underlying survey data.
Our studies have used data from two survey programs that differ in what elements of compensation are included, how costs are calculated, and how occupations are defined. Your letter asks about the first two of these differences. The primary data used in Mr. Miller's article were from the Occupational Compensation Survey Program (OCSP), and are limited to straight-time wages and salaries. Pay rates are expressed in terms of the standard hours employees regularly work, but do not include adjustments for paid leave.
On pages 20 and 21 of his article, Mr. Miller summarized the results of an earlier study that did include a comparison of total compensation (wages and benefits). This earlier study, "Cost of employee compensation in public and private sectors," was written by Bradley R. Braden and Stephanie L. Hyland and appeared in the May 1993 issue of the Review. The Braden/Hyland article primarily used cost per hour worked data from the Employment Cost Index (ECI) to compare compensation differences in the public and private sectors. In calculating the costs per hour worked, costs for wages and salaries (including overtime pay) and employee benefits were divided by hours worked (scheduled hours plus overtime hours, less leave hours).
Direct analysis of the two survey datasets is difficult, largely because of the differences noted above; but some rough comparisons can be made. One major difference is that the ECI data contains employee benefits costs, whereas the ocsP is limited to wages and salaries. According to the ECI data, employee benefits (including overtime pay) comprise roughly the same proportions of total compensation in private industry as in State and local governments. In 1992, benefits represented 28.2 percent of compensation costs in private industry and 30.2 percent in State and local governments.
The second major difference you note involves the ECT calculating costs per hour worked, and the OCSP publishing wages and salaries for the regular straight-time work schedule. Because overtime pay (about 1 percent of total compensation costs in 1992) is small relative to paid leave (about 7 percent), we believe the main difference between the two series would be in paid leave costs. The proportionate share for paid leave--6.8 percent of total compensation costs in private industry and 7.7 percent in State and local governments--was comparable between the two sectors in 1992.
For white-collar occupations, average standard weekly hours are reported in the ocsP National Summary, 1993 (Bureau of Labor Statistics Bulletin 2458), in the second column of tables A-1 to A-3. These weekly hours are similar for private industry and State and local government workers in most of the occupations that could be compared. Data for blue-collar workers are published as hourly rates in ocsP; so information on work schedules is not available for comparison.
The third major difference is in how occupations are defined. The ECI covers all workers, but the size of the sample limits occupational comparisons to nine major occupational groups. While this allows some significant comparisons, the data from the OCSP reported by Mr. Miller is unique in its occupational detail. On the other hand, as Mr. Miller noted on page 21 of his article, the ocsP occupations covered only about 8 percent of the surveyed work force.
We are currently in the process of combining our three major surveys--OCSP, ECI, and the Employee Benefits Survey--into one major program, called COMP 2000, which will allow for more detailed comparisons of total compensation for individual occupations or occupational groups than are currently possible. As part of the COMP 2000 program, we are also considering publishing both measures of cost per hour worked and cost per hour paid.
[signed] Kathleen M. MacDonald Assistant Commissioner for Compensation Levels and Trends Bureau of Labor Statistics
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||methodologies of computing differences in public sector and private sector compensation|
|Author:||Roberts, John; MacDonald, Kathleen M.|
|Publication:||Monthly Labor Review|
|Article Type:||Letter to the Editor|
|Date:||Nov 1, 1996|
|Previous Article:||Bangalore: India's silicon city.|
|Next Article:||Reflections on Human Development.|