That’s the thrust of John Judis’ new article in TNR on the crisis of the labor movement. Recently, the labor movement has been embroiled in an acrimonious debate about whether and how much to reform the AFL’s approach to organizing and collective bargaining.

On one side are SEIU’s Andy Stern and his allies in unions like the Teamsters, UNITE HERE, the Laborers and the UFCW, who advocate a dramatically stepped-up emphasis on union organizing. Stern, in particular, has advocated rebating as much as 50 percent of AFL dues to member unions to be used for organizing, as well as essentially forcing smaller unions to merge into larger unions to promote more effective and concentrated bargaining. And he has said a number of times that he will consider taking SEIU out of the AFL if adequate progress is not made toward his goals.

On the other side are John Sweeney, the rest of the current leadership of the AFL, unions like AFSCME, AFT, CWA and almost all the smaller unions who argue that such a large rebate would gut the AFL's successful political mobilization program and that forcing smaller unions to merge is both unnecessary and fundamentally undemocratic.

At the recent semiannual executive meeting of the AFL in Las Vegas, this conflict sharpened when the Teamsters moved a motion to implement the 50 percent rebate to support organizing (though they pointedly dropped the more-controversial proposal to merge smaller unions into larger ones). In the end, it was defeated easily, 15-7, though that margin of defeat probably conceals considerable support for a new direction at the AFL and, particularly, replacing John Sweeney himself. As Judis puts it:

Sweeney's performance has privately aroused the ire of many union presidents, but, over the last year, his opposition has become chiefly identified with Andrew Stern, the lean, gray-haired president of the SEIU....Stern and his allies have not launched an open challenge to Sweeney's reelection, which will be decided this July. Instead, Stern has launched a complicated set of structural proposals that will not necessarily solve any of the Federation's problems. And he has done so in a manner that has focused the debate on his own contentiousness rather than on Sweeney's deficiencies....

Altogether, Stern's actions may have actually strengthened Sweeney. Many union officials interpreted Stern's threat to bolt as a gesture of contempt toward them and the AFL-CIO. One union official described Stern as a "millstone" around the neck of the unions that want Sweeney replaced. Unions that had previously railed against Sweeney and his staff, another official explained, "are supporting Sweeney because they hate Stern more."

Judis argues that's a shame, because the need to replace Sweeney is so pressing. As he puts it:

Since its founding in 1955, the AFL-CIO has provided the largest counterweight to business interests, both in the workplace and in Washington. But, as the labor movement's share of the workforce has declined, the Federation has lost power. The union presidents desperately want to halt that decline, but disagreement about how to do so has been acute, reaching new heights at the Las Vegas meeting, where the issue of structural reforms dominated the agenda. That's too bad, because the solution to the Federation's problem lies elsewhere: in finding an effective replacement for John Sweeney.

I am more sympathetic than Judis to the structural reforms Stern and others are proposing, but it is hard to argue with the proposition that the AFL is in urgent need of a leadership change. Consider the sorry record of the Sweeney era, as highlighted by these data pulled together by Judis (comparison is of Kirkland's last full eight years with Sweeney's last full eight years [neither was in control in 1995]):

Kirkland era

1986: 16,975,000 union members; 17.5 percent of workforce unionized
1994: 16,740,000 union members; 15.5 percent of workforce unionized

Change, 1986-1994: 235,000 members lost, 2 percentage point decline in union density

Sweeney era

1996: 16,269,000 union members; 14.5 percent of workforce unionized
2004: 15,472,000 union members; 12.5 percent of workforce unionized

Change, 1996-2004: 797,000 members lost, 2 percentage point decline in union density

It isn't that Sweeney's record isn't much better than Kirkland's in the all-important area of union members and unionization. It's that it's been no better at all. Sweeney's leadership, in this sense, has been an utter failure.

But wait, say Sweeney's defenders, what about labor's big success in political mobilization, boosting union household turnout from 18 percent of voters in the 1992 election and 14 percent in the 1994 election, to 24 percent in 1996, 23 percent in 1998, 26 percent in 2000 and 24 percent in 2004? This would indeed be impressive if it happened. But it almost certainly did not.

The reason is simple: the figures above are all from exit polls from the respective years and the low apparent union turnout in the 1992 and 1994 exit polls was primarily driven by a change in question wording in those years. In 1988, when the union household membership was measured by a separate yes/no question, reported union household membership among voters was 25 percent. But in 1992 and 1994, union household was included in a lengthy "grab-bag" list of items that respondents could check as applying to them, rather than as a separate yes/no question. That format depressed the number of respondents reporting membership in union households, accounting for most of the falloff in union household voters to 18 percent in 1992 and 14 percent in 1994.

As soon as the format was changed in 1996 to a yes/no question, reported union household membership among voters shot back up to 24 percent and has stayed in that range ever since. Contrary to the claims of the Sweeney defenders, that question wording change is a more plausible explanation for the big jump in reported union turnout than a fabulously successful political mobilization program.

In reality, there has probably been no big jump in union turnout. Indeed, labor economist Richard Freeman, in his exhaustive study of union turnout and voting data, "What Do Unions Do....to Voting?", concludes:

[T]he share of voters in unions fell by about 1 percentage point and the share of voters in union households fell by about 2 percentage points from 1990 to 2000. The [exit poll]-based estimated increase in the union share is erroneous.

Of course, just keeping the union share of voters from falling much in the face of the steady decline of unionization can be reckoned a modest accomplishment. But the Sweeney regime's claims of big success in the political mobilization area are based on faulty data and should not be taken seriously.

So the reality is that Sweeney's leadership has produced precious little real gain for the labor movement anywhere. Clearly new leadership is needed and the sooner the better. The problem is, as Judis points out, the current faction-fighting in the labor movement may make it harder, not easier, to produce that leadership change. Let's hope Stern and his allies learn from their defeat in Las Vegas and start focusing more realistically on what it will take to get Sweeney and his ineffectual regime out of office.