Over at Baseline Scenario, James Kwak raises the issue of whether unions are actually bad for business. That's certainly what you'd assume, given the scorched-earth tactics many companies employ to block collective bargaining rights. The problem with researching this questions, writes James, is that though it's easy enough to tally up differences between companies that have unions and companies that don't, "there is a huge problem of selection bias: since companies with unions are unlike companies without unions in many ways, you can’t say whether any differences in outcomes are due to the effect of the unions themselves, or due to the effect of other factors that would be there regardless of the unions." James points to an effort by John DiNardo and David Lee to get around that selection problem in their 2004 paper, “Economic Impacts of New Unionization on Private Sector Employers: 1984-2001.” Their insight was to compare essentially like firms that faced a unionization drive that either barely won or barely lost -- but where it could have gone either way. In theory, that should give the clearest idea of unionization's impacts: It's unionization on the margin. The results don't give comfort to supporters of unions or opponents. What DiNardo and Lee essentially found was...nothing. "The evidence suggests that—at least in recent decades—the legal mandate that requires the employer to bargain with a certified union has had little economic impact on employers, because unions have been somewhat unsuccessful at securing significant wage gains." The unions, in other words, achieved collective bargaining agreements, but nothing really changed on either side. Differences in wages, employment, productivity, and output all proved insignificant. There are a couple problems with the data set: Among them, it only assesses the manufacturing sector, which is fairly heavily unionized and so has higher wage norms. Service sector jobs, which are now at the center of most unionization battles, weren't examined. The study also focused on average wages, and unions may have a greater effect on wage distribution within a firm. But it does leave us with the question of why employers would prefer to grenade their firm than let the workers bargain if the actual impact is so minimal. One theory is that the results in this paper aren't applicable to firms where unionization wins in a walk. The effort to correct for sample bias simply introduced a different sort of sample bias. Another is that most firms aren't much affected by unionization, but some are terrifically affected, and it's hard to say which is which in advance. And a third is that it's a simple control issue: Even at a firm where collective bargaining doesn't result in large wage gains, it still results in a larger voice for workers and a reduction of control for management.
