Every presidential election seems to produce a great, new strategic theory on how candidates can woo and win the hearts of vital elements of the electorate. If only they can reach the "forgotten American," the "silent majority," the "alienated voter," the "young voter," even the "nonvoter," candidates dream, they will sweep to victory.

Nearly all of these strategems prove illusory. Invariably, the electorate fails to perform as anticipated; targeted blocs of voters do not vote as blocs. First-time 18-year-old voters, supposedly more liberal, either don't participate or wind up voting for such conservative candidates as Richard M. Nixon and Ronald Reagan.

Probably the best example of the foolhardiness of banking on voting blocs came in the last presidential election after much talk about the so-called "gender gap." According to this theory, women were more inclined toward Democratic candidates. Thus, putting a woman on the ticket supposedly would benefit the Democrats.

Geraldine A. Ferraro turned out to be no help to Walter F. Mondale as greater percentages of women voted for Reagan that time than four years earlier. So much for the gender gap.

With these considerable caveats in mind, one approaches a new study again suggesting that the "women's vote" might help Democrats in the 1988 election. This work by the Washington-based Democratic polling firm of Hamilton, Frederick & Schneiders (HFS) seems worthy of consideration and might even be important.

"The gender gap is alive and well but not widely recognized," the pollsters report. "Women differ from men in their orientation toward the political world. Women certainly vote differently than men; nine Senate races in 1986 went to the Democratic candidate because of support from women (CBS News exit polls). In fact, if women voted the same way men did in 1986, the U.S. Senate would still be firmly in Republican hands.

"Recent HFS research on the subject of why the gender gap exists shows stark differences in perceptions of the economy, of the parties and the fundamental role of government. While the subject is complex, in simple terms, women tend to be more Democratic because they want a compassionate government that directly helps people improve their lives. Key issue areas are family finances, education and the elderly."

These conclusions are based on random sampling of 1,500 voters in July for the American Medical Political Action Committee (AMPAC) and a focus-group study of middle-class women under 50 with family incomes of less than $35,000. No female executives were included. This was conducted in Seattle, Nashville and Philadelphia.

Some of the findings about voter opinions on the direction of the country are intriguing, especially in the wake of the Iran-contra affair and scandals involving Wall Street and electronic preachers. They appear to reflect concerns expressed during earlier grass-roots reporting by Washington Post reporters and show a marked difference in attitudes among men and women voters.

For instance, asked for opinions on the direction of the country, 47 percent of women under 50 said it is going in the right direction while 60 percent of men in the same age group agreed. Only 36 percent of men thought the country was headed in the wrong direction, compared with 46 percent of women. Among voters over 50, both sexes held more strongly pessimistic views on that general question, but women's responses again differed from those of men.

Feelings about Reagan reflected the same pattern.

"Two major factors appear to be driving the gender differences in perception of the country's direction -- the economy and trust in politicians," the pollsters concluded. "Focus groups with middle-class women under 50 reveal a strong level of 'disappointment' with players and events in the political world. Scandals with the Iran-contra dealings, Gary Hart, defense-contractor abuse and PTL drove these perceptions."

The pollsters offer more gender differences of political significance. They said women are more worried about the economy, more opposed to contra aid and funding the Strategic Defense Initiative and more supportive of education and "help" programs.

This might make a difference next year, but don't bet on it. In the end, the quality of the candidate counts more than the demographic composition of the electorate.

Haynes Johnson's column will appear on Friday, beginning next week.