WAL-MART STUDY SHOWS CLEAR BIAS.

Byline: Gary M. Galles Local View

A last-ditch $3 million ad campaign for Proposition 72 features Wal-Mart as a health-insurance miser. The ads even led Wal-Mart to contribute to the ``no'' campaign ``to defend (its) honor.'' Yet unmentioned in the media coverage is that the anti-Wal-Mart research study cited in the pro-72 campaign is so biased that its conclusions are indefensible.

``Hidden Cost of Wal-Mart Jobs'' was published by the University of California at Berkeley Labor Center in August. It purports to document that Wal-Mart imposes burdens on California taxpayers by forcing employees to use state welfare and health-care programs. Its findings have been used as a proof-text by those opposed to the retail giant's plans of opening ``supercenters'' - big-box stores and supermarkets under one roof - throughout Southern California.

Now the report is being recycled to promote Proposition 72 - which would require California businesses to pay 80 percent of the health-care insurance of their employees and their families - while taking another shot at Wal-Mart. Unfortunately, the study offers a misleading set of results engineered from biased assumptions (since manipulating one's premises is the most effective, hidden way to get desired answers), produced by a biased source.

The UC Institute for Labor Employment, the Labor Center's primary sponsor, is not so much a research organization as one that exists to support the union agenda. For instance, in a single newspaper story, the center was described as ``unabashedly pro-labor'' and ``an institution that has been there for the labor movement,'' which was heartily endorsed by the California Labor Federation. Created by Gray Davis in the ``surplus'' year of 2000, the center was so obviously biased that Gov. Arnold Schwarzenegger proposed zeroing out its $4 million budget, before union-backed lobbying restored it to $3.8 million.

The failings of the ``research,'' the conclusions of which are now being touted, are numerous. For example, despite the fact that Wal-Mart employees are far more likely to be retired, secondary workers or teenage members of a family, and to work part time, the study assumed they had the same demographic composition as workers at other employers, thus biasing wage and other comparisons.

Not wanting insurance coverage because a retiree receives Medicare or a worker is already covered (the study erroneously assumes that workers not covered by Wal-Mart's insurance are also not covered by family members' employer-sponsored insurance) hardly indicts Wal-Mart for cruelty. In fact, under Proposition 72, such workers would be harmed because they would have to pay a portion of the cost for mandatory coverage that would duplicate what they already have.

The study's sources are questionable. It relies heavily on newspaper articles, but only cites a single, old, refereed academic journal article. It also fails to show its methodology, referring readers to an earlier study by the same group. However, the earlier study also fails to clearly reveal its methodology, too.

Perhaps the study's least defensible assumptions, however, are economic, based on the flimsiest of cited evidence, which in turn was ambiguous (e.g., the study assumes that lower wages have no effect on employment). However, assuming away the law of demand hardly qualifies as economic research.

Further, the study assumes that the savings people receive from Wal- Mart's lower costs have no further economic effects on employment, incomes, tax revenues, etc. - as if those savings won't be spent for other goods and services directly (added consumption spending) or indirectly (savings funneled to investment spending), as well as making otherwise too costly projects now worthwhile.

The UC Berkeley Labor Center study, as well as the anti-Wal- Mart/pro-Proposition 72 ads built on its conclusions, does not merit much respect. It appears to be little more than a mechanism to conclude what its supporters set out to find, by burying its misleading assumptions and approach deeply enough that no one would look carefully enough to discover them.

The alleged results may swing the outcome of Proposition 72, since it depends on voters whose attention to such questions is very limited, but good policy is seldom based on deliberate distortions.