As revealed by the Mackinac Center on Wednesday, a school consolidation study by Michigan State University's Education Policy Center senior scholar Sharif M. Shakrani contained what appeared to be unattributed material (about 800 words) lifted from work that was not his own. The report received wide press coverage, yet it is the third study of questionable quality in 20 months from MSU-affiliated researchers, a pattern that has damaged the public policy debate in Michigan.

In February 2009, two economists from MSU's Center for Economic Analysis performed an economic impact analysis of the state's fledgling film subsidy program. Among other things this provides a refundable tax credit reimbursing film makers for up to 42 percent of the expenses they incur shooting here.

The MSU scholars used a software program called "REMI," which is a model to measure the impact of these subsidies. They concluded that in its first nine months the program had created some 1,100 net new jobs. The university issued a press release claiming that the state's film incentive program was a "big hit."

But the researchers left out 100 percent of the program's costs from their model, ignoring the fact that somebody had to provide the cash to cover those subsidy checks and foregone taxes. (Care to guess who? Got a mirror? That would be you.) Including the costs would have dramatically reduced the headline "new jobs" figure, showing the program to be not such a big hit after all.

Then last October, MSU professor Charles Ballard did a little moonlighting, producing a study paid for by the Service Employees International Union (SEIU), although one had to carefully read the fine print to recognize that it was not an official university product. Co-written with graduate student Nicole Funari, the resulting paper purported to show a significant "retrenchment" in Michigan's state government workforce, including pay differentials between public and private employees disadvantageous to the former.

However, the pair didn't bother to include the total compensation cost of government employees. An analysis performed by a Mackinac Center scholar explained the flaw:

The Ballard and Funari paper makes no attempt to account for health care or other benefits in their comparison of wages. The true standard of comparison should be total employee compensation. To make such a comparison, the benefits afforded Michigan state employees would have to be compared with those of private-sector employees with similar educational background and in similar jobs.

Further, we described how the "conclusions drawn by the authors and attributed to the paper in the media are not substantiated by the data presented, and the use of the data is in some cases misleading."

Unfortunately, poor Spartan scholarship often gets more press attention than detailed deconstructions. We'll see if that pattern holds with the latest study, which also looks like flawed methodology.