Is this any way to pick the nominees? A flurry of early primaries, a barrage of fund-raising, voters who fell voiceless, candidates who feel frozen out. To consider changes in the primary system, Outlook sought the views of two people who have a role in shaping it: the chairman of the Democratic and Republican national committees.

Since the beginning of this presidential election cycle, pols and pundits have been in high dudgeon over the "front loading" of the primary season. Scheduling more than half the state primaries by mid-March drastically increases the importance of money, they say, and tips the scales too heavily in favor of the front-runners.

The critics may have a point. Are we being buried under a blizzard of primaries? And if so, what's to be done?

To answer these questions, we must understand two fundamental changes in the way we choose our presidential nominees that have taken place in the past 30 years. The first is that we have moved from caucuses and state conventions to primaries--and thereby have raised the cost of campaigning. The second change is that we have severely complicated the ability of our candidates to raise money.

These two changes work against each other--like driving a car with one foot on the gas pedal and the other on the brake.

Let me explain:

Primaries were once seen chiefly as little more than a means of demonstrating electoral viability--a sort of dress rehearsal for candidates who needed to prove their stump skills. They were not the basic means of selecting convention delegates; that was done through party caucuses and state conventions. In 1968, only 15 states even held primary elections. Vice President Hubert Humphrey won the Democratic presidential nomination without entering a single primary--and Ronald Reagan actually won more Republican primary votes that year than did Richard Nixon.

But next year, as the result of a process begun by the Democrats following that 1968 election, 45 states and the District of Columbia will hold presidential preference primary elections, and those elections will select almost 90 percent of the delegates.

Certainly primaries have the virtue of involving more voters than do caucuses. But for that very reason they are more expensive, and they increase the difficulty of building an organization and communicating a message. So, by shifting to primaries, we have vastly increased the costs of campaigning. We have put our foot on the gas.

Meanwhile, in 1974, the Federal Election Campaign Act put a strict $1,000 limit on the amount of money an individual could give to a federal candidate. This was an attempt to improve the odds for long-shot, underfunded candidates. It was also an attempt to get rid of corruption--after all, what could a donor really expect for a measly $1,000. So we put our other foot on the brake.

And the result has been exactly the opposite of what was intended. We now have fewer serious candidates and money is more important than ever--because the Law of Unintended Consequences came into play, and, unlike some campaign finance laws, this is one that can't be broken or evaded.

We'll never again see a Eugene McCarthy-type of campaign. McCarthy, you'll recall, sat down with a few rich, liberal donors and made his pitch, and they provided hundreds of thousands of dollars to launch his campaign. But today, a dozen guys each giving the $1,000 limit could barely pay for a single poll.

In other words, candidates going into the 2000 election have to spend far more time raising money than ever before.

Now add front loading to the equation.

States began rescheduling their primaries earlier in the election year because they hoped to give their voters a more powerful voice in the nomination process. But that voice is muted when many states do the same thing. And that is certainly what has happened. As recently as 1980, we Republicans chose just one-fifth of our convention delegates by the second week of March; in 1992, we selected almost half of our delegates by that date; and in 2000, we will have selected two-thirds of our delegates by then. For Democrats, the problem is even greater--while Republicans over the past 20 years have tripled the percentage of delegates chosen by the second week of March, Democrats have actually quadrupled the percentage.

Carry this trend toward its logical conclusion and sooner or later we would end up with a one-day, nationwide primary election sometime in February. And what a pointless exercise that would be: Because financial and organizational advantages would accrue so heavily to the best-known candidate, he or she could easily beat anybody who didn't happen to be rich enough to self-finance a campaign. Do we really want to limit our primary choices to the the most famous or well-connected candidate and some zillionaire?

To a lesser degree, this is the problem we may face in 2000. The result of front loading may be that lesser-known candidates have too little time between state contests to build momentum--to take advantage of an unexpected surge in one primary to raise money for another. We won't know for sure until we have a nominee.

Many thoughtful Republicans are concerned about this phenomenon. So I've put together the RNC's Advisory Commission on the Presidential Nominating Process, led by former party chairman Bill Brock, to examine the problem and make recommendations. Its members are considering a host of potential fixes--each with pros and cons.

The party could create "super delegates," for instance, to represent its long-term interests at the convention. Not tied to any candidate in particular, these super delegates might serve as a check on a potential runaway nominee--or they might unite behind one candidate early on and collapse the contest even sooner.

Bonus delegates--offered to state delegations as an incentive to delay their primaries--came into play after the 1996 Republican convention. But the incentives would obviously have to be made bigger, because a number of states decided it was more advantageous to hold their primaries sooner.

Rotating regional primaries, where the states in a given region would hold their primaries within a specific 30-day window, might stretch out the process again. And offering incentives for states to choose delegates by proportional representation instead of on a winner-take-all basis might let some second-place finishers stay in the race. But each of these fixes might have unintended consequences we can't yet imagine.

It is too early to tell what recommendations the Advisory Commission might make. But one is almost certain--a call for raising the contribution limit from its 1974 level, and indexing it for inflation.

No changes can be enacted until 2004, if then. No changes can occur without either the Congress passing legislation or the national parties directing the state parties to change their ways. That kind of top-down direction has never played well in the GOP--but maybe the problem is serious enough that states will accept it the next time.

One thing is clear, though--a lot of the problems we're dealing with today are the result of reforms we made over the last 30 years. We would be wise to keep that in mind as we examine the system and see if there aren't ways we can make it better. In politics, as in medicine, the Hippocratic oath serves us well--first, do no harm.

Jim Nicholson is the chairman of the Republican National Committee.