IF THE David Duke campaign had any enduring meassage for America, it was this: Competing with demagogues is expensive. Office-seekers who wish to sell a complicated message to an increasingly diffuse electorate must outspend their brassier opponents.

Only a "cheap" message can get through in a "cheap" campaign. It takes more time and money to communicate about complicated issues of governance than to communicate about race. Yet critics are once again calling for reforms that would curb campaign advertising and spending to protect gullible Americans from the spiritual pollution of political snake-oil merchants.

The fact is, our campaigns aren't broken, and don't need that kind of fixing. Voters are not passive victims of mass-media manipulators, and it is dangerous to assume that low-key "politically correct" campaigns would somehow eliminate the power of the visceral image. Restricting television news to the MacNeil/Lehrer format -- and requiring all the candidates to model their speeches on the Lincoln-Douglas debates -- won't solve America's problems.

David Duke, loathesome and frightening though he may be, is neither an argument that campaigns don't work nor that campaign advertising should be restricted. In fact, Louisiana voters knew all about Duke's past and his associations with racist and antisemitic causes; Duke was able to communicate his message just as effectively -- perhaps more effectively -- in interviews and debates.

Reformers say they want to turn down the volume, discuss more important issues and turn out more voters -- worthy goals, but also contradictory. Decorous campaigns will not raise more important issues. Neither will they mobilize more voters nor overcome offstage mutterings about race and other social issues. It was not worthiness and refinement that got 80 percent of Louisiana's voters to turn out.

If government is going to be able to solve our problems, we need bigger and noisier campaigns to rouse voters. It takes bigger, costlier campaigns to sell health insurance than to sell the death penalty; the cheaper the campaign, the cheaper the issue. Big Brother is gaining on the public. Surveys show that voter perceptions about presidential candidates and their positions are more accurate at the end of campaigns than at the beginning; there is no evidence that people learn less from campaigns today than they did in past years. That brilliant 1988 team, Roger Ailes and Robert Teeter, could not recycle Dick Thornburgh; the road to Washington is littered with the geniuses of campaigns past.

Many critics argue that congressional elections do not work because a lack of competition isolates Congress from the electorate; they argue that Democratic control of Congress is based upon incumbency advantage, not the will of the voters. They are wrong. In races for 567 open congressional seats since 1968, the GOP has lost a net of nine. In the 244 open-seat races since 1980, the GOP made no net gains. Democrats won as many previously GOP seats as Republicans won previously Democratic seats.

In fact, the inability of Republicans to capture Congress attests to the limits of voter manipulation. People tend to rate the Democrats higher on issues with which Congress deals, and the GOP higher on issues with which the president deals. Divided government may be slow, cumbersome and confrontational, but it rests upon the divided preferences of the voters -- not slick ads or a lack of competition.

It is also argued that campaigns influence voters to take a "pox on both houses" attitude -- i.e., that informed voters will be less likely to vote. This theory is easy to test: First, take a sample of people across the country and ask what they consider to be the most important issues, where the candidates stand and what they like and dislike about the office-seekers.

Then, after the election, find whether the interviewees, who have been forced to think about the issues, were more or less likely to vote than other people. If they voted less often, there is clear support for the claim that negativism and irrelevancy are turning off American voters. If the people vote more than others, though, the problem is not that people are being turned off but that they are not getting turned on enough.

In fact, there is such an experiment. In every election since 1952, people interviewed in the University of Michigan's benchmark National Election Survey are asked such questions; after the election, actual voting records are checked to see whether the respondents did indeed vote.

The results demolish the trivia-and-negativism hypothesis. Respondents in national studies, after two hours of thinking about the candidates, the issues and the campaign were more likely than other people to actually vote. Indeed, the Duke-Edwards election shows that people will turn out to choose between a Nazi and a crook when the campaign is big enough to keep them mobilized.

The real reason that voter turnout is down is that campaigns are not big enough to keep them tuned in. Changes in government, in society and in the role of the mass media in politics have made campaigns more important today than they were 50 years ago, when modern studies of them began. But the scale of the campaigns have not risen to their larger task.

Campaigns attempt to simplify politics, to achieve a common focus, to make one question and one distinction paramount in voters' minds. But the spread of education has both broadened and segmented the electorate, thereby making it more difficult to assemble a winning coalition. Educated voters pay attention to more problems and are more sensitive to connections between their lives and national and international events. The more divided an electorate, and the more money available to advocates of specific issues or causes, the more time and communication it takes for a candidate to assemble people around a single distinction.

Even as unifying forces in our society -- for example, the proportion of people watching mainstream network programming and news -- have waned, forces tending to fractionalize the electorate have been on the rise. For example, today they include: more government programs -- Medicare, Social Security, welfare and farm supports are obvious examples -- that have a direct impact on certain groups; coalitions organized around policies toward specific countries, such as Israel or Cuba; various conservation and environmental groups; and groups concerned with social issues, such as abortion and gun control.

Furthermore, there are now a great many more specialized radio and TV programs and channels, magazines, newsletters and even computer bulletin boards with which persons can keep in touch with like-minded people outside their immediate neighborhoods or communities.

At the same time, phenomena such as expanded use of primaries have increased the need for unifying mechanisms. Primaries mean that parties have had to deal with the additional task of closing ranks after the campaign has pitted factions against each other. Finally, campaigning under divided government is also more difficult; it is harder to justify a compromise between competing political principles -- the 1990 budget deal is an example -- than to reiterate one's own principles.

What this suggests is that if we really want to increase voter interest and participation -- as well as the capacity of government to tackle our problems -- the best strategy may well be to increase our spending on campaign activities that stimulate voter involvement. In this regard, it is important to note the clear relation that exists between turnout and social stimulation. There is, for example, a large gap between the turnout of educated and uneducated voters; married persons at all ages vote more than people of the same age who live alone; and much of the increase in likelihood of voting seen over one's life is due to increases in church attendance and community involvement.

As for the argument that America already spends too much on elections, the fact is that American elections are not costly by comparison with those in other countries. Comparisons are difficult, especially since most countries have parliamentary systems, but it is worth noting that reelection campaigns to the Japanese Diet, their equivalent to our House of Representatives, cost at least eight times as much per vote as our congressional elections. Indeed scholars estimate that Diet elections cost between $50 and $100 per constituent, while incumbent congressmen here spend an average of $1 per constituent. It is food for thought that a country with a self-image so different from America's spends so much more on campaigning.

Our campaigns are criticized as pointless affairs, full of dirty tricks and mudslinging that ought to be cleaned up, if not eliminated from the system. But the use of sanitary metaphors to condemn politicians and their campaigns says more about the people using the metaphors than it does about the failings of our politics.

Before we attempt to take the passions and stimulation out of politics we ought to be sure that we are not removing the lifeblood as well. Ask not for more sobriety and piety from citizens, for they are voters, not judges; offer them instead cues and signals which connect their world with the world of politics. The challenge to the future of American campaigns -- and hence to American democracy -- is how to bring back the brass bands and excitement in an age of electronic campaigning. Samuel Popkin, professor of political science at the University of California San Diego, is author of "The Reasoning Voter: Communication and Persuasion in Presidential Campaigns," University of Chicago Press, from which this article is adapted.