There is a battle going on for the soul of the Democratic Party, a battle between its new ideas and its old constituencies. It is a battle that has been going on quietly these past two years, when few Democrats had to attend to the messy everyday business of governing. It will rage more heatedly in 1983 and 1984, not only in the presidential campaign, but also in the House and the big state capitals where Democrats in the 1982 elections won the responsibility for governing.

The course of the battle is not plain to see. Many ridicule the Democrats' new ideas as "Atari liberalism" -- putting trendy labels on old-fashioned Democratic programs. Instead of calling for public works programs, Democrats now call for rebuilding the nation's "infrastructure" -- which turns out to mean the roads, bridges and sewers government has been building for 140 years. Instead of calling for more money for education, Democrats are calling for upgrading "human capital"-- which turns out to mean more training and education for young people.

This new nomenclature is important, however. It shows that the Democrats are turning from a politics of distribution to a politics of production, from a politics that focuses on distributing society's wealth to the unfortunate to a politics that focuses on producing more wealth.

The immediate beneficiaries of the politics of production, as advocates of distributionist politics point out, are not always needy: students capable of absorbing higher education and adults capable of holding high-skill jobs are not usually from the poorest income groups. But in the long run, everyone in society will be richer for their work. In embracing these policies, Democrats have abandoned, for a time, the distributionist impulse. In the 1970s it led them, to take just one example, to oppose decontrol of gas and oil prices because of the effect higher prices would have on the poor. Over time the Democrats came to accept decontrol on the ground that it maximizes production.

This is what Democrats have done in the past: the First New Deal and the Kennedy- Johnson 1961-64 policies were designed to stimulate production. Only afterward, in the Second New Deal and the Great Society, did the Democrats get around to the politics of distribution. This makes sense if you believe that only a society with sustained economic growth can afford a politics of redistribution, and that a society without growth must concentrate on production first.

But will policies of production prevail in a Democratic government? We heard little about them on the campaign trail, and there are few infrastructure, human capital, and national investment bills being pushed on Capitol Hill the way federal aid to education and Medicare were pushed in the late 1950s (an exception, the gas tax/highway bill, is likely to be passed easily with bipartisan support in the next few weeks). The lobbying forces behind most of these new ideas seem weak. How will such policies prevail when they come into conflict with major Democratic constituencies that favor what I would call not policies of distribution but policies of protection, policies that spend rather than create national wealth?

Labor unions, for example, have memberships concentrated in industries and occupations that seem likely either to contract or be stagnant in the 1980s: autos, steel, building trades, government. Each union has its own agenda for protecting its members' interests: the UAW wants the local content bill, the Steelworkers want import quotas, the public employees' unions want generous pensions. They want to maintain, through subsidies or protectionism, an economy and job structure like that of the America of 1975 or 1965 or 1955. The unions have enormous political clout with Democrats, given labor's active role in the 1982 campaign and the AFL-CIO's decision to consider endorsement of a presidential candidate in November 1983.

Other constituency groups lead the Democrats toward protectionism. Geographically, the party is strongest in the Northeast and Great Lakes industrial rything is involstates where growth has been low and outmoded industries seek to maintain a status quo the free market will not support. Incoming Democratic governors support domestic protectionism: Mario Cuomo brags that he will see that New York buys subway cars in New York, not Montreal or France; Richard Celeste makes as one of his five major economic stimuli a state "buy Ohio" policy. Job quota programs sought by blacks and other minorities are, after they have been in effect awhile, a form of protecting a status quo rather than producing a new and different future. So are policies maintaining the size of city bureaucracies in cities whose population has dropped.

So while the politics of production is in the policy paper stage, the politics of protection, with its strong constituencies, has drafted bills, signed up cosponsors, and gotten pledges of support. Get almost any Democratic politician off in a corner, off the record, and you'll find someone who recognizes the long-term stagnating effect of the politics of protection. But get a Democrat in office or on the presidential primary campaign trail, and you may hear something else again.

The best clue as to how the Democrats would perform in office if they win in 1984 will come not in the Congress, where Democrats stoll have to respond to Republican initiatives and in the awareness that anything must be approved by Republicans to pass, but in the big states. In New York, Ohio, Michigan, Minnesota, Wisconsin, Massachusetts, Democrats are in power and face difficult problems. Governors like Cuomo and Celeste are eloquent supporters of the politics of production. But they also support many protectionist policies. In the crunch, which side will they come down on? The battle for the soul of the Democratic Party is going on right now, and the outcome is by no means as certain as the purveyors of the new ideas of politics of production would have you believe.