Striking Out.(The Talk of the Town)

In the spring of 1988, television and movie writers went on strike. The strike, which lasted for twenty-two weeks, was rhetorically bitter and economically destructive: it cost an estimated half billion dollars in lost revenues and wages and sent network ratings down by nine per cent. But the walkout had only limited impact at the negotiating table: when an agreement was finally reached, it looked very much like a deal that could have been made five months earlier. The strike almost certainly cost both sides more than the sums they had been fighting over.

Twenty years later, entertainment writers are on the picket lines once again. They may do better this time, but history is against them. Walkouts may call to mind labor triumphs like the Flint sitdown strike of 1936-37, which gained union recognition for the United Automobile Workers at G.M., but most don't end that well--nor do they generally end as badly as the 1981 air-traffic controllers' strike, in which everyone got fired. Instead, strikes often end tepidly, with no major gains or rollbacks, and economists have found that, on average, strikes these days have little, if any, impact on what workers get paid. (Paradoxically, unions raise worker wages, but strikes generally don't.) Given the negative economic consequences--lost paychecks for workers and lost business for employers--the economically rational thing for both sides is usually to settle before the walkout starts. So why don't they?

One obvious hurdle to a settlement is that neither side knows what the other's true position is. In economists' terms, strikes happen as a result of "asymmetric information"--when one side knows more than the other about the real economics of the situation. Entertainment writers, for instance, want a share of the revenue generated from their work in new media, including programs streamed on the Internet. Producers insist that they need flexibility with regard to new technologies and that it's too early to know how much they can afford to pay for streaming programs. This may be just a bluff, or it may contain some truth--it's hard for the writers to know the difference. Going on strike is one way to find out. If a company concedes quickly, that's a sign that it was just bluffing. If it's willing to endure a long strike, that may be a sign that it meant what it said. That's why the longer a strike lasts, the less likely it is to produce a big victory for either side: you're willing to cut a deal after a long strike that you wouldn't have been willing to cut before in part because the strike has told you that the other side wasn't just bluffing.

Even if negotiators are acting in good faith, it's still hard to settle. Both sides, to begin with, are likely overestimating their chances of victory, thanks to the well-documented tendency people have toward overconfidence. A strike isn't always a mistake: sometimes workers do win big. But if both sides think a strike will help ...