Monday, February 21, 2011

Wisconsin public workers already under-compensated

A new report from the Economic Policy Institute concludes that Wisconsin public employees are under-compensated.
Comparisons controlling for education, experience, organizational size, gender, race, ethnicity, citizenship, and disability reveal that employees of both state and local governments in Wisconsin earn less than comparable private sector employees. On an annual basis, full-time state and local government employees in Wisconsin are undercompensated by 8.2% compared with otherwise similar private sector workers. This compensation disadvantage is smaller but still significant when hours worked are factored in.
Conservatives love to point out that the average compensation for public workers typically exceeds that of private workers. This is no different in Wisconsin where the report indicates that full-time public workers cost their employers just over $63,000 per year on average compared to just under $62,000 for full-time private workers.

However, public workers tend to have significantly more education (the proportion of public workers with college degrees is double the proportion of private sector workers with degrees). Within education groups, the report finds that public sector employees in Wisconsin receive less compensation than private employees, except for those with less than a high school education.

The report also uses regression methods to adjust for other measurable differences between employees, including education, experience, gender, race, citizenship, and employer size. Once those controls are included, average annual compensation for public workers is 8.2 percent less than for comparable private workers. Average hourly compensation is 4.8 percent less.

Like many other states, Wisconsin faces serious and real budget problems. Difficult and painful choices are required. One problem, however, that Wisconsin does not face is over-compensated public employees. Describing them as such is misleading, and treating them as such will ultimately be self-defeating.


jon said...

C'mon Dave get real.

The EPI is so left wing no economic analysis coming out of it can be given any credence. Why do you continue to post such drivel? Keep digging the hole deeper?

The EPI Board is loaded with union sleazbags like Andy Stern and Richard Trumka.

The Board of Directors:

Barry Bluestone
Northeastern University

R. Thomas Buffenbarger
International Association of Machinists & Allied Workers (IAMAW)

Anna Burger
Service Employees International Union (SEIU) and Change-to-Win

Larry Cohen
Communications Workers of America (CWA)

Ernesto J. Cortes, Jr.
Ernesto Cortes, Jr.Industrial Areas Foundation

Jeff Faux
Senior Fellow, Economic Policy Institute

Leo W. Gerard
United Steelworkers of America (USWA)

Ron Gettelfinger
United Auto Workers (UAW)

Teresa Ghilarducci
The New School for Social Research

Alexis Herman
New Ventures

Robert Kuttner
The American Prospect

Julianne Malveaux
Bennett College

Ray Marshall
University of Texas, Austin

Gerald W. McEntee
American Federation of State, County, and Municipal Employees (AFSCME)

Lawrence Mishel
President, Economic Policy Institute

Debra Ness
National Partnership for Women and Families

Pedro Noguera
New York University

Jules O. Pagano
American Income Life

Manuel Pastor
University of Southern California

Bernard Rapoport
Bernard & Audre Rapoport Foundation

Bruce S. Raynor

Robert B. Reich
University of California, Berkeley

Rep. Linda T. Sánchez
U.S. House of Representatives

Andrew L. Stern
Service Employees International Union (SEIU)

Richard L. Trumka

Randi Weingarten
American Federation of Teachers

Raul Yzaguirre
Arizona State University

Dave Ribar said...

And the substantive critique of the analysis is ...

Preston said...

When I first read this post, I also wondered about the political point-of-view (and biases) of the EPI since this analysis runs counter to much of what is written on the subject elsewhere. As to substantive critique, I'm not a PhD in economics nor a statistician, but I do recall Mark Twain's comment about lies, damned lies, and statistics. Perhaps you, as an educator and economist, can point out some of the areas in which an article like this could be slanted to make the situation fit EPI's point-of-view when the actual facts were different.

Dave Ribar said...


If you can find a non-partisan study that finds huge compensation advantages for public workers, I'd love to see it.

A critical analysis of the EPI study would focus on a) the selection of the comparison group (are full-time workers the right comparison?), b) the assignment of non-wage compensation, and c) the use of other controls in the regression adjustment.

With respect to the first point, full-time and part-time jobs are very different. Comparisons of full-time public jobs to full-time private jobs are appropriate; however, the study could be strengthened by also comparing public and private part-time jobs.

With respect to benefits, the study has done a reasonable job with the data at hand. In any case, Table 3 shows what the calculations are. For the study's conclusions to be wrong, it would have to be under-counting public benefits or over-counting private benefits.

The regression adjustments potentially raise the biggest red flag BUT the author gets the same results with simpler comparisons in Table 2 that just consider educational differences. The adjustments just aren't making that much of a difference. It would be more concerning if the results in Table 4 were radically different from the results in Table 2.

The results tell a very consistent story. Once you account for educational differences, compensation for Wisconsin's full-time public employees is lower than compensation for its private employees. On average, public employees receive more compensation, but this advantage is due to their higher levels of education. Within education brackets, public employees are paid less.

You can compare the EPI analysis with a recent one done by the conservative Cato Institute, which concludes that public employees are underpaid. You'll notice, however, that the Cato analysis does not mention the education levels of public employees even once.

Should the Cato Institute have controlled for education? Well, consider what its researchers do when they want to argue that women aren't underpaid--they control for every job characteristic under the sun, including occupation, experience, and "non-standard" work schedules.

Preston said...

Thanks for your thoughtful response. The educational level issue was one that puzzled me. I'd say that in private industry educational level is less significant to pay level than it is in government service, particularly in education. I'm thinking that in private industry it serves more as a gate-keeper issue than as a measure of compensation issue. I don't know how this might affect the comparison of public vs. private compensation.

Preston said...

Dave wrote: "If you can find a non-partisan study that finds huge compensation advantages for public workers, I'd love to see it."

It's not non-partisan, but I did see this conservative critique of the EPI study from yesterday's Washington Post. It begins "The Center for Union Facts is out with a new, compelling study that makes clear how shoddy is the analysis of the most widely cited pro-union study arguing that public-sector employees are actually underpaid."

Dave Ribar said...


The comparison is shoddier
- it only reports outcomes for hourly and not total compensation;
- it replaces a questionable adjustment in the EPI analysis (the control for firm size) with an even more questionable adjustment (predicted firm size); the most reasonable thing to have done is to omit the firm size control; and
- it adds in part-year workers.

The main gripe is with the predicted adjustment for firm size. Because predictions are only made for the public workers and not for the private workers, the meaning of the public coefficient changes. In the new paper, it is effectively the public differential at a firm with fewer than 100 employees.

If you look closely, the use of the predicted firm size adjustment accounts for the lion's share of the change in results.

Preston said...

I see there was more criticism of the EPI report in the Wall Street Journal today:

"Consider a study released last October by the Center on Wage and Employment Dynamics at the University of California, Berkeley, which concluded that Golden State public employees "are neither overpaid nor overcompensated." The Economic Policy Institute has generated reports arguing that government workers are underpaid.

These studies are misleading. Public-private pay comparisons vary from state to state, but a full accounting shows clearly that large, union-dominated states tend to overpay their workers. California is a good example.