With regards to the post below on "open-source unionism" and the catch-22 involved with labor law, a reader asks a good question: How did unions get so strong the first time around? After all, it's not like business is any more hostile to organized labor today than it was back when Carnegie's Pinkertons were machine-gunning steel workers. So why should it be any more daunting for unions today?

My understanding is that the growth of unions in the twentieth century was mostly done in "spurts," and to a good degree based on a lot of contingent factors. During World War I, for instance, the Allied countries needed to make nice with labor in order to fight their little war, and so, for instance, things like the War Labor Board were created. The 1920s were a darker time for labor, but the Depression and World War II once again happily steered the government back to labor-friendly territory. Labor density also rose so dramatically in the 1930s and 1940s in part because a new form of union had suddenly appeared—the industrial union. The same thing had happened in the 1880s: a new form of organizing, the "local assembly," had given labor its big boost.

Looked at his way, the usual prescriptions for bolstering union strength—labor-friendly legislation and better organizing—sometimes seem to fall a bit short. After all, labor density still declined during the Clinton years, and while Clinton was no FDR, he was about as labor-friendly as any Democrat likely to gain the presidency in the near future. It's also worth looking soberly at the limits of regular old organizing. Richard Freeman has estimated that if it costs about $2,000 to organize a worker, then adding 1 million new members would require 40 percent of all union dues—about what the SEIU wanted to spend—but that would add only a point to labor density in the United States. A million new numbers would be a phenomenal gain—as would card checks and other labor-friendly measures—but not nearly enough on their own to reverse the decline.

Freeman and Rogers' proposal for "open-source unionism" points to at least one way forward—a "new form" akin to local assemblies and industrial unions—for an age where a shrinking manufacturing base is decimating union strength, and the government has been trampling on labor since the Reagan years. (Although it's worth pointing out that their proposal will only work if companies are dissuaded from cracking down on pro-union workers.) I don't know how one would turn their open-source proposal into practice—I'm way too far down on the labor food chain for that—but maybe if everyone claps their hands and says they believe in fairies or something.