Facts on public sector wages

The National Institute on Retirement Security in the U.S. produces some really excellent reports which should be more widely read, and not just on pensions or retirement income. 

Last week they published a good report, Out of Balance?  comparing public and private sector compensation over the past 20 years, written by two professors at the University of Wisconsin. 

This report found that when adjusted for education levels and other factors that have an impact on wages, state and local government employment typically earn about 11 per cent less than comparable workers in the private sector.  These calculations are based on regression analysis accounting for different variables and wage differences related to gender and race are also fairly dramatic.   This report got some good pick-up in the US media, including on CNN.

In contrast, the Canadian media continue to perpetuate some of the misinformation that emanates from the Fraser Institute and the CFIB on this and other issues.  I usually don’t take the time to respond to all the gross misinformation coming from these sources, as it would be more than a full-time job, but feel compelled to do so on some instances.

In a column in today’s National Post Herb Grubel perpetuates and amplifies the misinformation from the CFIB’s Wage Watch report of last year as evidence for his argument that public sector unions should have their right to strike eliminated.  He claims that if the incomes of public sector workers were equal to those in the private sector, the deficits of governments would be lowered by $19 billion.

In fact, the CFIB study engaged in a lot of manipulation and distortion of Statscan census data in order to come up with their conclusions that canada’s public sector workers are paid more.  What is telling about it is that they didn’t report on wages for individual occupations but instead constructed an average after a considerable amount of manipulation of the data.

A number of people have published critiques of the CFIB report, including Sylvain Schetagne at the CLC, myself and David MacDonald for NUPGE, but only a few in the Canadian media have reorted  on the real facts.

The truth is that overall average levels of pay between the public and private sector are very similar.  There’s much more discrepancy in pay levels within the private sector, with higher paid occupations paid considerably more in the private sector and lower paid occupations, especially those for women, considerably less.   When adjusted for education level, pay in the public sector is often less.

Grubel claims that the pay differentials in the public sector are entirely due to higher levels of unionization in the public sector and somewhat disingenuously calls for efforts to ensure “equal pay for equal work”.

In fact, relatively higher wages at the lower levels are due in large part to pay-equity laws, implicit government policies to provide more equitable pay scales and, yes, union efforts to redress the large income and pay inequalities that exist in our society.

We just don’t all share the same blind faith that private markets always deliver equal pay for work of equal value.


  • After reading this article, I cannot believe Grubel was once a professor. How can someone who has responsibility to teach within economics could actually make the following statement. Isn’t this one of the first things that a fascist does when put within a place of power. It surely is an embarrassment within the Canadian academic community for somebody such as this to shoot off at the mouth with such anti-worker sentiment. I wonder if a survey was done of economists in Canada, how many would recommend his first cours eof action, or any of his other brilliantly dimwitted ideas. I wonder if the man even lifted the hood on this “study”. The methods are what I would classify as propaganda and silly but dangerous ideas such as promoting the restriction of free trade unions can get a country in trouble real fast. Very irresponsible, adn I hope his next cheque from the Fraser Institute is a big one cause I think his conference tour drawing abilities just took a sharp nose dive. However, their is always openings at the circus for proven clowns.

    Something I think right leaning economists do not understand, and it is a key point in nation building- a nation state’s cohesion is directly causal in its wealth and success. This cohesion is based upon drivers of equality and inclusion, like free unions, social safety nets, public healthcare etc.

    Here is his the quate that I cannot believe the FP printed

    “What measures would help reduce the existing fiscal deficits and restore equity in the compensation of public-and private-sector workers? The most obvious measure would be the elimination of public-sector unions, which would probably appeal to the vast majority of Canadians who are not members of unions.”

  • and on a day that 3 people in Greece are killed in a protest and display of quite hectic social protest that are being driven by anti-worker sentiments as displayed.

    Does anything like these incidents ever click within the minds of the apparent informed.

    I just can’t believe such statements are made based upon such disgusting studies as that which the CFIB conducted.

    If you look closely at the methods within that study, you will see that the occupational control that the CFIB stated they maintained within the study was in fact its problem.

    In the study, as mentioned by this article, the method controlled and compared occupations based on similar occupational codes (NOCS). If you look within the methods section, its states very clearly that occupations with at least 20 workers within both the public and private sector were compared.

    If one takes a look at publicly accessible data from the 2006 census that was used in this study on the statcan site, one can easily see that many private and public sector occupations vary in size. (again the method made explicit on how private and public sectors are defined, is a rudimentary the brute force method)

    The end result is the CFIB were comparing in a very large proportion of occupations some that had a very low amount of public sector workers in many cases less than 500 against some that had hundreds of thousands of private sector workers. Yet somehow this was allowed. Method wise it introduces a bias into the process that makes the reports findings completely useless. One cannot make conclusions based on such data errors. This especially occurred in such lower waged occupations as retail and food services. So somehow the CFIB allowed hundreds of thousands of low wage workers to be factored into the occupational mix where only a very small amount of public sector workers actually exist. That my friends is not a method but simple trickery to bring in a whole pile of bias that nullifies the “studies” outcome.

    This on top of the fact that never once do they mention data quality of any of their estimates again makes one seriously think how such a newspaper article with grandiose claims designed to promote anti-worker beat up on public sector unions theme is totally and categorically academically criminal.

    If you have any questions on the above feel free to contact me for the public data located online.


  • “There’s much more discrepancy in pay levels within the private sector, with higher paid occupations paid considerably more in the private sector and lower paid occupations, especially those for women, considerably less”

    Women also make up a higher share as a percentage of workers because its cheaper to hire them right or wrong legally then men. If there pay was equal, that percentage would fall fast. You make the assumption more women would be hired, I think women as percentage of the workforce would decrease, for right or wrong reasons. Its the same debate if we were to remove all unlegal workers off the books, that suddenly business would hire legal canadians workers, they might just closeshop, and or offshore etc.

Leave a Reply

Your email address will not be published. Required fields are marked *