In the wake of almost every Democratic defeat since 1972, liberals can be found insisting that, if their candidate had adhered to the party's core economic beliefs and steered clear of social issues, he would have done much better, if not won. If Democrats were to return to "the liberalism this country once heard from Woodrow Wilson, Franklin Roosevelt and John F. Kennedy," Princeton University sociologist Paul Starr declared in The New York Times last month, it would "give the Democratic party back its majority." But this electoral advice--whatever its merits--sidesteps a much more basic and disturbing question: Is it possible any longer to enact the kind of liberal program that Roosevelt and his successors did? In other words, even if a Democrat were elected in 2008 on a liberal platform, would he or she be able to put it into effect?

If you look at the history of liberalism, what you discover is not reassuring. From 1932 through 1974--even when Republicans Dwight Eisenhower and Richard Nixon were president--liberals got much of their program enacted, but, since then, they have failed abysmally. In 1977, Jimmy Carter championed bills that matched almost perfectly what Starr includes in his liberal agenda--"progressive taxation, affordable health care ... environmental [and] labor protection"--but Carter failed to get any of them passed even though he had a sizable Democratic majority in Congress; in 1993 and 1994, Bill Clinton couldn't enact his signature health care measure with an almost equally large congressional majority.

It is convenient to blame these failures on incompetence, but the truth is that structural factors were more important. Liberalism's success from the '30s through the 1960s was based primarily upon certain special economic and political conditions: popular pressure from below, business' acquiescence in reform, and the conviction of the nation's opinion-makers that reform was good for America. Since then, dramatic changes in the international economy have turned business against reform and weakened the other forces supporting reform. Liberalism is by no means defunct, but it has been put on the defensive--most particularly, in this second Bush term. If Democrats want to revive liberalism, and not merely win office for themselves, they will have to address--and, where possible, rectify--the conditions that have undermined it....

To revive liberalism fully--to enjoy a period not only of liberal agitation, but of substantial reform--would probably require a national upheaval similar to what happened in the '30s and '60s. That could happen, but it doesn't appear imminent. What is more probable is a gradual move back toward the center, where older programs would be protected from assault (although not from refinement), where incremental change could be made, and where the stage could be set for a fuller revival if circumstances warranted. This could result, ironically, from the same causes that initially turned the United States away from liberalism.

In the past, the world has overcome industrial overcapacity through world wars and depressions, but with neither likely (one hopes!), the world economy is unlikely to recover its earlier brisk postwar pace. The extended U.S. boom of the late '90s, it is now clear, was a momentary spike brought about by speculation in information technology. If the economy does continue to grow sluggishly--creating greater insecurity even among the so-called investor class--it is likely that public discontent with business will rise again, as it did in the early '90s, and provide an opening for Democrats, and, through them, for liberal reform. But liberals will have to take advantage of this opening in a way that they failed to during the '90s.

Taking advantage would mean devising new approaches to domestic and international policy that are fair and efficient--and that don't allow the opposition to raise the specter of big government. A bloated national health insurance system could eventually be worse than none at all, but, designed properly, a national health system could widen access and keep costs under control, benefiting many businesses as well as workers. The perils of globalization can't be effectively addressed through trade protection, as some industrial unions still insist, but, by improving education, by encouraging foreign manufacturing firms to locate here, and, as economists like Rodrik have begun to argue, by international reforms that will protect workers from the vicissitudes of monetary instability.

Liberals would also have to rebuild the infrastructure of democratic pluralism through encouraging, subsidizing, and defending unions and whatever other form of countervailing social organization is feasible--from community groups to Internet-based virtual communities. The Republicans took this lesson from the older New Deal movement and have built a political infrastructure of their own while attempting to destroy what the Democrats have constructed. Liberals would have to do whatever is necessary--including, above all, tightening labor law--to rebuild their movement from below.

Liberals are not without recourse. Just as the liberals of the '30s could draw upon three decades of progressive reform plans, today's liberals can draw upon 30 years of discussion about national health insurance, environmental and consumer protection, pension reform, international trade and investment, and worker training. These policies have been featured in many campaigns, although they have frequently been overshadowed by abortion or gun control. Starr and other liberals are right to insist that they be front and center--not because, by doing so, Democrats will be guaranteed electoral success, but because, without them, the Democrats are simply a collection of heterogeneous interests and identities united primarily by distaste for Republican conservatism. They may win elections, but they will be totally unprepared to change America for the better. And that, after all, is what politics should be all about.