Back to

related links:

Pay Comparability (NALC & Professor James L. Medoff, Harvard University)

USPS Chief Financial Officer  Michael Riley addresses Postal Wages-

Statement submitted to Postal Commission by Michael L. Wachter

APWU Responds to Papers by Michael Wachter and Michael LeRoy (pdf)

The Relative Size of Labor Costs AT UPS, FEDEX & USPS -Workplace Economics

U.S. Postal Service productivity and the postal business (NALC)

Testimony of consultant from Watson Wyatt Worldwide-the firm providing report on compensation and incentive system design 

Postal commission considers new agency pay system






(February 12, 2003)

Dr. Joel Popkin


Joel Popkin and Company

1155 15th Street, N.W., Suite 614

Washington, D.C. 20005

(202) 872-0990


Labor Costs as a Percentage of Total Costs Postal Workers’ Pay Comparability
Postal Pay Has Tracked CPI and Lagged ECI  

The purpose of this submission is to address the numerous claims that have been made that the pay of postal workers is too high and that labor costs are about 80 percent of total USPS costs. I am an economist who has analyzed postal pay and other issues for almost 25 years. On occasion I have been invited to testify before Congress, and have otherwise made my views known through Op-Ed articles.

I am president of Joel Popkin and Company, an economic consulting company established in 1978. Before that, I was director of the Washington Office of the National Bureau of Economic Research headed by Martin Feldstein, and served at the President’s Council of Economic Advisers (CEA) when Herbert Stein was its chairman. At the CEA I focused in particular on the wage and price controls of 1971-74. That responsibility resulted from experience gained in my prior post as BLS Assistant Commissioner for Prices and Living Conditions in which I was the senior civil servant charged with the compilation and publication as the Consumer (CPI) and Producer (PPI) Price Indexes. I hold a BS in Economics from the Wharton School of Finance of the University of Pennsylvania, and a Ph.D. in economics from the same university.

Labor Costs as a Percentage of Total Costs

My comments address three issues. The first is the widespread popular characterization –spin, if you wish-- of labor costs as representing 80 percent of the total costs of USPS. That is blatantly incorrect. Publicly available USPS records (summarized since 1984 in Chart A) show that wages and benefits allocated to workers covered by collective bargaining represent 56.8 percent of total costs. Costs associated with workers covered under APWU bargaining agreements, largely mail processing and retail station clerks, accounted for only 26.0 percent of USPS costs in the 2002 fiscal year. In addition, the data in Chart A show that over the past couple of decades the labor share of postal costs has declined, the share of the APWU most dramatically. I would hope your report will reflect the correct data on the labor cost share of USPS costs and recognize that today’s share reflects a substantial decline from 1984.

Chart A

Compensation of All Bargaining Employees and APWU Bargaining Employees as a Share of Total USPS Expenses

Some might argue that total compensation costs for all postal employees, including managers, at 76.8 percent, are high as compared with competitors. But if the comparison is properly made there is not much difference in compensation cost share between the USPS and two of its important competitors, UPS and Federal Express. The makeup of operating expenses is heavily dependent on how the business is structured. The capital intensity among USPS, UPS and FedEx differs considerably because of the nature of each business. For example, FedEx owns and operates a significant fleet of airplanes. Overnight speed is the major factor consumers are paying for when they purchase FedEx's services and that speed is produced by its airfleet. However, the fact that it owns such a fleet means that its total operating expenses are much more heavily weighted toward capital than toward labor and therefore a cursory inspection will show a relatively large difference in the labor share of operating expenses for FedEx and for USPS. FedEx's labor share is the lowest of the three. But the 2002 gap between FedEx and USPS narrows to about 13 percentage points if operating data are treated consistently to take account of differences in capital intensity. Such an adjustment is made in a paper by Workplace Economics which I understand will be posted on the APWU website for anyone interested in analyzing the full set of numbers.1 On this same consistent basis, the UPS labor cost share was only 7 percentage points lower than the USPS share in 2001. However, this analysis still has made no adjustment for the fact that the USPS must send a postal worker past every residential address, and many business addresses, six days a week, which is a very labor intensive requirement.

Postal Workers’ Pay Comparability

The second issue concerns the allegation that postal workers are overpaid. This allegation was first made in the negotiation context by the USPS in a presentation made at the bargaining table in 1980. I was present as a consultant to the APWU, a position I continue to hold. I inquired whether this model-based analysis took account of the sex and race composition of the USPS vis-à-vis the rest of the private sector workforce and was informed it did not. So I proceeded, using the same model, to introduce race and sex variables. Our first result was published in an academic journal in 1984. I found that the wages of white male, postal employees were at a level not statistically different from those of white males at work in the private sector. That was not the case for women or non-white men. In subsequent analyses using more recent data, this pay pattern, while mitigated over the years by private and public sector initiatives, continues to prevail.

Source: Joel Popkin and Company regression analysis of Current Population Survey data from February 2000,March 2000, April 2000, and May 2000.

Chart B shows the results of our first and latest of the issue. Thus, the evidence shows that wages of white male postal employees are at or below wages of comparable white males in the private sector. Women and non-white male postal workers are, of course, paid equal pay when they perform the same work as that of their white male co-workers. You will note, in the chart, that the result for white males continues to show private sector parity, and perhaps even some premium in favor of the private sector according to the latest analysis.

This result was first presented in 1984 to an arbitration panel chaired by Clark Kerr, an internationally renowned labor economist. The analysis was introduced to rebut results introduced by the USPS, by then using a model different from that presented in 1980 during collective bargaining. After hearing both sides present their models, and the results they yielded, Dr. Kerr concluded “Comparability, like beauty, quite obviously, is in the eyes of the beholder”. He went on to fashion an award that, while providing restraint, was also designed to provide postal workers with real wage gains. While the statistical models were debated in the subsequent three arbitrations, no arbitrator has tied his findings to those models.

Postal Pay Has Tracked CPI and Lagged ECI

Because discussions about statistical models of comparability are moot, I think it may be useful to present actual data on how postal wages have changed in comparison with those of other private sector workers. That is the third purpose of this submission. Chart C-1 first shows how APWU wages have changed since the 1984 arbitration as compared to changes in private sector wages, as measured by the BLS Employment Cost Index (ECI). Two wage series are shown for the APWU bargaining unit. One is the average straight-time hourly wage of all employees in the APWU bargaining unit. The second is for one slot in the APWU’s pay schedule – that for level 5, step O, the pay step with the most workers in it, or the “modal” slot. It is also the top of the pay scale for level 5 workers. The data show that, after a period in which all three measures tracked rather well, a widening gap has opened in favor of the private sector since the early 1990s. Private sector wages have risen faster than either APWU measure. Since June of 1992, when the divergence began, the ECI has risen 41 percent, the APWU straight-time hourly wage has risen 27 percent, and the level 5 step O wages, have risen 23.3 percent.

C- 1

Growth in APWU Wages Compared to Wages in the Private Sector

June 1984 through December 2002


Chart C-2, on the next page, analyzes the same time wage measures but divides each by the CPI in order to measure real, inflation-adjusted wages. The chart shows that by either APWU wage measure postal workers gave up real wages during the 10 year period; one percent for the bargaining unit as a whole, almost four percent in level 5, step O. That is in stark contrast to a 10 percent real wage gain enjoyed by the private sector workers. But the lack of any significant real wage gain for postal employees is not a phenomenon of the period since the 1984 Kerr arbitration. As Chart C-3 shows, real wages have been virtually flat since the PRA became the law in the early 1970s.

I hope this information will be used by the Commission to dispel the pervasive myths about the pay of postal workers. If I can be of further assistance, please let me know.

Source: Testimony Of Joel Popkin (.pdf)

Dr. Popkin received his BS degree from the Wharton School of Finance of the University of Pennsylvania and his Ph.D. in economics from the same university. He is a fellow of the National Association of Business Economists and the American Statistical Association, past Chairman of the Conference of Business Economists and a member of the Committee of Visitors to the economics department of the University of Pennsylvania. He has been Chairman of the Board of the National Economists Club and a member of the boards of the American Statistical Association, the National Bureau of Economic Research, and the Economic Strategy Institute.

back to postalreporter