The decline of unions everywhere is plain; many people question whether we need them at all. But unions are necessary -- for economic reasons that smart capitalists will appreciate and because the basic function unions perform remains absolutely essential to democracy

Among friends and foes alike, it seems, unions are widely regarded as dinosaurs. Appropriate, perhaps, to an age of manufacturing, mass production, and class struggle. But not to a service economy allegedly heading toward a "workplace of the future" in which all workers are "empowered" and management is just another job.

Few outside the labor movement see unions as part of the solution to current economic and social problems. And if you're not seen as part of the solution -- well, you're in very deep weeds.

In their heyday unions played a crucial and overwhelmingly positive role in the economy. Unions served the needs of both management and workers -- and the economy at large -- by helping make mass production work. Unions standardized jobs, instituted grievance procedures, and continually raised their members' wages so as to boost economywide demand for consumer goods and services. That created a new class of customers for the companies, enlarging their markets and giving them reasons for renewed investment. And that, in turn, increased productivity and lowered the cost of goods, which was good for everyone.

Today, however, a number of factors have led capitalists to question the need for unions at all. Among those factors are the devolution of mass production, the inability to significantly stimulate demand by raising the wages of only one group in an increasingly global economy, and the disintegration of large vertically integrated companies in need of stable blocks of workers.

But unions are necessary -- for economic reasons that smart capitalists will appreciate and because the basic function unions perform remains absolutely essential to democracy. For working-class Americans to enjoy the status of equal citizenship, they need financial clout -- which means they need independent collective organizations of their own.

That unions are necessary does not, of course, ensure their viability. To survive and prosper in the new industrial system, unions will have to do three things simultaneously: secure material benefits for members or potential members, solve a problem capitalists cannot solve on their own, and chiefly by doing those things, win widespread support and approval from nonmembers. If unions can do that, they will be as productive a force in today's economy as they were in the big-business model they helped bring about. The question is how.

Take first the fundamental union role of serving members' interests. In today's economy, unions can no longer guarantee job security. The alternative is to promote career security. That is, instead of trying to define a worker's place in a fixed structure, the new type of union would seek to ensure workers' power in a fluid structure. The way to do that is to provide all workers with advanced technical (and, increasingly, "business") training so they can participate in the design of work teams and find a new job easily if their current employer goes under.

Today there is little likelihood that individual companies will provide the appropriate training and counseling services; they are too focused on immediate returns to make the long-term investment. State and local governments will not make the investment, because they cannot determine the proper sorts of services needed. In fact, the only institution qualified to invest in such services is the unions.

Unions cannot, however, equip their members for careers independent of a particular company if compensation and work rules vary so much from company to company that no one dares to move. That problem, apparent today from the way people cling to jobs to preserve pension or health benefits, can be solved only by standardizing compensation and organizational practices for all companies. Workers who gain the skills needed to survive in the free-lance economy will be barred from participating in it otherwise. Just as unions can serve their members by making their skills more versatile, they can serve them by pressing both companies and government for greater uniformity in the conditions of compensation and employment.

Note that unions alone have the capacity to play that role. They are much better than government officials or companies are at knowing just what underlying standards need to be respected in establishing "comparable" work settings. Because unions extend across many companies, they can gather wide knowledge of the different jobs that cluster into careers. Companies and groups of employees trying to reconcile differences could thus well look to the unions when conflicts arose over the definition and application of rules. In fact, as institutions of workers in the economy itself, unions are indispensable vehicles for enforcing standards -- with the local knowledge and the capacity for disruption needed to play that role.

The interests of companies, too, would be advanced by the new model of unions. Most companies would like to wave a magic wand over their current work force and suddenly, costlessly, have employees with precisely the skills they need. If unions can help companies figure out how to make effective use of the vast public funds available for training, they can be the magic wand.

Employers likewise have an interest in unions' working to achieve comparability across companies in conditions of work and compensation. As companies continue to become more decentralized and to cooperate more closely with other companies, there is less connection between who employs a person and where and with whom that person works. Without some uniformity in work conditions, rules, and compensation, advanced forms of cooperation are far more difficult to enter into and manage, and then fold when the task is done.

So unions can, as they once did, play an economic role that both advances their members' interests and solves economywide problems. Yet their transformation would prove a difficult task.

To become innovative agents of flexibility and productivity would, for most unions, require considerable change in how they think about their role; and change is always hard. The strategy outlined here would also require unions to vastly increase their reach -- implying substantially greater commitments to organizing, and willingness to experiment with new forms of representation in sites where unions might not have majority support. Unions would also have to make a vast effort to layer into their own operations the technical competence needed to perform this new role.

Finally, the strategy outlined here would require forcing U.S. business to accept -- as in general it currently does not -- the legitimacy of unions. As during the 1930s, today the basic principles of workplace democracy, and of democracy in economic policy-making generally, need to be established over intense opposition. Doing that would mean a fight.

We suggest that virtually everyone in America has a clear interest that the fight be won.

* * *

Joel Rogers is a professor at the University of Wisconsin; Charles F. Sabel is a professor at Massachusetts Institute of Technology.