In the presidential nominating process, Democrats usually put on a better show. In 1972, for instance, Edmund Muskie had his dramatic cry on the steps of the Manchester Union-Leader in New Hampshire; and when he faltered, Hubert Humphrey fought George McGovern with a surprising amount of venom. Among the Republicans, meanwhile, John Ashbrook had strong ideas but could not break double-digits in his fight against Richard Nixon.

This year--for the first time since 1964--the Democrats do not have a serious contest for their party’s presidential nomination. All the action belongs to the Republicans, so this is a good time to examine how they choose their candidate.

The Republican National Convention will take place August 12-15 in San Diego. A total of 1,990 delegates will meet, so a candidate will need 996 to win. Each state gets at-large delegates according to a formula that includes (among other things) the share of votes won by all its Republican presidential, senatorial and gubernatorial candidates between 1992 to 1995. And every state receives three district delegates for each of its seats in the U.S. House. (The District of Columbia and the territories also have a small number of delegates.) Ohio has 57 district delegates and 10 at-large delegates.

Most states choose Republican convention delegates by primary, in which voters express their preferences in the voting booth. Ten states use the caucus method, in which voters in each precinct gather to choose delegates to conventions at the county, congressional-district, or state level.

Because the caucus method is indirect, and because national Republican rules do not require proportional representation of voters’ candidate preferences, the outcome of precinct caucuses bear only a loose relationship to the actual distribution of delegates.

Likewise, primary results do not translate neatly into delegate totals. In states such as Ohio, the candidate who wins the statewide vote gets all the at-large delegates, while the winner in each congressional district gets the delegates for that district. So in a two person race, a candidate could receive 49 percent in each district and get no delegates. The phenomenon is even more pronounced in California, where all 165 delegates go to the statewide winner.

In about half the states, participation in GOP primaries and caucuses is not restricted to Republicans. Some states do not have party registration at all, while others (e.g.,Ohio) allow Republicans and Democrats to cross over to the other party’s primary. Without a serious Democratic contest, liberal non-Republicans might try to cause some mischief by voting the GOP candidate they think is most likely to lose to President Clinton.

The states that vote early tend to get the greatest attention from the media and the political community. Over the years, therefore, many states have moved up the dates of the primaries and caucuses. In 1996, about two-thirds of the states will vote by the end of March: Ohio will hold its primary on March 19.

Most observers think that this "front-loading" will enable one candidate to clinch the nomination by early springtime. But there is another possibility. New volunteers and contributors always show up after a candidate wins an early victory; but now the schedule is so tight that the new help might arrive too late to do much good. Conversely, an early loser may mount a comeback before disappointment drives away money and grassroots support. The customary "winnowing" may not take place, and several candidates still might be running after the snow melts.

In laying your bets on who will win, beware of placing too much faith in polls. This early, most people simply have given little thought to their presidential preferences, so poll numbers can shift like clouds in your coffee. And polls that purport to show the electorate’s "underlying feelings" can be even worse. Consider the following example.

In October, a CBS/New York Times survey showed a large fraction of GOP voters harboring liberal views on issues. The poll sampled 1,269 adults, 38 percent of whom said that they usually voted in Republican primary elections or planned to do so in 1996. If this result elected the electorate, then 74 million Americans would be getting ready to vote in GOP primaries.

In 1992, however, the Republican and Democratic primaries together generated only 33 million votes. Add the 600,000 or so who voted in caucuses (turnout figures are estimates), and you get less then half of the New York Times number. Add any fudge you like and you still cannot come close to a turnout of 74 million.

What happened? Nonvoters often feel embarrassed, so when pollsters ask them if they vote, they simply lie. And in this case, many others may have misunderstood the screening question, thinking that the poll was asking them if they would vote Republican in November. Whatever the case, the Time’s snapshot of the GOP primary electorate was so distorted that its results were meaningless.

History, a better guide, suggests that conservatives will predominate among primary and caucus voters. No wonder Pete Wilson and Arlen Specter dropped out so early--and why the rest of the field is running to the right.

John J. Pitney, Jr. is associate professor of government at Claremont McKenna College. Having co-authored a book on the House Republicans, he is currently writing one on the ideas of the modern Republican Party.