Skip links

How to Fix the Presidential Primary Process

Dear Readers: In his book A More Perfect Constitution, Crystal Ball Editor in Chief Larry J. Sabato laid out several proposals to modernize the Constitution. One of those was a suggestion about how to better format the presidential primary process, which has evolved over the course of the nation’s history and goes unmentioned in the Constitution. In the wake of the traditional lead-off contests in Iowa and New Hampshire, and growing concerns (particularly among Democrats) about the roles that those two states have staked out at the start of the calendar, we thought we’d offer Sabato’s proposal for a different nomination system. It features a rotating regional primary that would give every state a more realistic chance of voting at the start of the calendar and compresses the bloated presidential nomination calendar.

What follows is an edited excerpt from A More Perfect Constitution dealing with the pitfalls of the presidential nominating process — which are at least as obvious today as they were back in 2007, when the book was released — and suggestions on how to nominate the major party candidates in a fairer, more orderly fashion.

The Editors

Of parties, presidential politics, and the quadrennial orgy

Imagine that a convention of clowns met to design an amusing, crazy-quilt schedule to nominate presidential candidates. The resulting system would probably look much as ours does today. The incoherent organization of primaries and caucuses, and the candidates’ mad-dash attempts to move around the map, would be funny if the goal — electing the leader of the free world — weren’t so serious.

Few want to go back to the bad old days when party “bosses” chose presidential candidates in smoke-filled rooms. Primaries and caucuses are now fundamental to our conception of popular democracy in presidential selection. But there is such a thing as ineffective popular democracy, especially when it is hopelessly disorganized.

A good-size piece of the problem can be labeled “Iowa and New Hampshire.” These two states seem to assume that the Constitution guarantees that they should go first, but a close reading of the text finds no such clause. The New Hampshire lead-off primary was initiated in 1920, and it has arguably been very influential since 1952, when it played a role in both the decision of President Truman not to seek reelection and Dwight Eisenhower’s successful quest for the GOP nomination. New Hampshire reprised its 1952 incumbent-toppling when Minnesota Sen. Eugene McCarthy came within a few percentage points of President Lyndon Johnson in the 1968 Democratic primary, leading in part to LBJ’s decision shortly thereafter not to seek another term.  The Iowa caucus has only played a role since 1968, and its true national debut came in 1972, when George McGovern scored well there on his way to a surprise, ill-fated Democratic nomination. Just four years later, the Iowa caucus propelled a little-known former Georgia governor, Jimmy Carter, to the Democratic nomination and the presidency, assisted also by Carter’s subsequent narrow victory in New Hampshire. In both the Hawkeye and Granite states, Carter received less than 30% of the votes, but in a crowded Democratic field of candidates, this low percentage was enough to prevail.

Why should two small, heavily white, disproportionately rural states have a hammerlock on the making of the president? The truth is that Iowa and New Hampshire have a franchise they are determined to keep at all costs. New Hampshire even has a law that requires its secretary of state to do whatever is necessary to keep its primary first.

Without a constitutional requirement, there is simply no solution to a situation that deteriorates every four years. Try as they might, the national party committees cannot orchestrate a fix. In the end, they can only punish a recalcitrant Iowa and New Hampshire in minor ways, by cutting the size of their convention delegations or giving the delegates bad hotels and seating at the party conclaves. The candidates who campaign in states holding contests earlier than permitted can also be penalized by having the national parties deny them any delegates they may win in those states. But a handful of lost delegates is trivial compared to the whirlwind of positive publicity secured by victory in early states. These penalties, light as they are, may not even materialize. By the time of the national conventions, the parties are unlikely to want to alienate swing states such as Iowa and New Hampshire as the general election campaign begins. Nor will the parties want to aggravate their candidates by penalizing them at a time when they are attempting to fully unify their forces at the convention.

Congress has some power to intervene in the state-based, party-centered nominating process, yet the federal legislature would be highly unlikely to step into that briar patch. Presidential nominating reform has never been a priority for Congress, in part because of the traditional rights of the states and the parties to organize this sector of politics. It is highly doubtful that Congress will generate the will to clean up the nomination mess anytime soon. For one thing, the senators and representatives from Iowa and New Hampshire would be willing to do anything to stop congressionally sponsored reform, quite possibly with assistance from colleagues who would see their own presidential ambitions at stake. A senator who becomes a hero in Iowa and New Hampshire for saving the caucus and primary would be halfway to a presidential nomination! And realizing this, most or all of the senators with presidential aspirations would jump to back the Iowa/New Hampshire status quo. (It’s a rare ambitious senator who doesn’t get up in the morning and see a president in the mirror.)

Thus, the only possible, comprehensive answer is likely to be a constitutional one. In the 21st century we the people need to do what the founders didn’t even perceive as necessary in their pre-party, pre-popular-democracy age. The guiding principle should be one that all citizens, in theory, can readily embrace: Every state and region ought to have essentially an equal chance, over time, to influence the outcome of the parties’ presidential nominations, and thus the selection of presidents. We are one nation, and simple equity demands that all of us, regardless of our state of residence, should have the opportunity at some point to influence the selection of presidential nominees by filling one of the precious, early voting slots.

The nominating process ought also to be moved back into the four months leading up to the party conventions. Presidential politics now takes fully one fourth of a president’s four-year term — whether he is running for reelection or not — and with front-loading accelerating, it soon may consume even more of it. Not only is this bad for the presidency as an institution, but it causes the electorate to tire of the never-ending political campaign. It should be possible to create a system that flows from the first primaries and caucuses beginning in March or April directly into August party conventions, and then into the Labor Day kickoff for the autumn general election. Not only is this not rocket science; it doesn’t even qualify as elementary mathematics. It is easy, if the will and the means are present. The electorate must supply the will, and the Constitution should outline the means.

There have been dozens of proposals to revamp the primary scheme, though none has been offered as a constitutional fix. Clearly, that is both because the Constitution currently ignores the politics of the system almost entirely and because a constitutional insertion — virtually written in stone — would have to be as fair and foolproof as possible. The following plan, the product of much discussion and thought, is proposed in that spirit.

The regional lottery plan for the new Constitution

The Congress should be constitutionally required to designate four regions of contiguous states (with contiguity waived for Alaska and Hawaii, and any other stray territories that may one day become states). The regions would surely look something like the ones in Map 1, with natural boundaries denoting the Northeast, South, Midwest, and West.  All of the states in each region would hold their nominating events in successive months, beginning in April and ending in July. The two major-party conventions would follow in August. This schedule, all by itself, would cut a couple of months off the too-long process currently prevailing in presidential years.

Map 1: National regions for hypothetical primary system

The presidential nominating system would still be state-based, so each state party would be free to choose any date it wished within the region’s month, and further, it would be free, as currently, to choose either the primary or the caucus method of selecting delegates. Of course, all the states in a region might try to front-load their contests on the first possible day, but that actually would make little sense, except perhaps for the first region in the series. Even in that first region, a state might have more influence coming later in the month, perhaps standing alone on a particular day — a situation that would encourage presidential candidates to spend time and money in the stand-alone state. After all, the post-primary headlines would belong solely to the candidate who won that stand-alone state. If there were 10 states on a particular day, the headlines as well as the candidates’ time and money would be split 10 ways. Note, too, that the regional system would concentrate the candidates within a single region for a month. They would have a better opportunity to get to know the problems and peoples of the region and its states, and the geographic proximity of the campaigning would cut down on the wear and tear on the candidates, to some degree anyway.

But how would the order of the regions be determined? In many cases, there would still be a bonus in going first. The establishment of a U.S. Election Lottery, to be held on New Year’s Day of the presidential election year, would yield fairness and also add an element of drama to the beginning of a presidential year. One of the nation’s famous lottery machines with the popup ping-pong balls would finally find a purpose beyond bestowing untold riches on people who can’t handle it. Four color-coded balls, each representing one of the regions, would be loaded into the machine, and in short order — the length of a 10-second lottery TV drawing — the regional primary order would be set. Since none of the candidates would know in advance where the political season would begin, part of the permanent presidential campaign would be dismantled. After all, even a very wealthy candidate wouldn’t waste the money necessary to organize all 50 states in advance, and the four-year-long homesteading in Iowa and New Hampshire would be gone forever (a note from the perspective of 2020 — OK, maybe Michael Bloomberg would or could). Much more important, the “law” of averages would give every state and each region, over time, the precious opportunity of going first. Clearly, there is no guarantee that a particularly lucky region would not be repeatedly chosen to start the process, but the equal-access principle is key to the fairness in this plan. These new constitutional provisions would “repeal” the nonexistent constitutional right to go first that Iowa and New Hampshire have appropriated for themselves.

Another benefit of the Regional Lottery Plan would be the reasonable spacing between contests, allowing candidates potentially to recover from setbacks in one region and to regroup prior to the next set of contests. The news media and voters in each region would certainly be on the same page on this critical matter, demanding their fair share of attention.

One additional facet should be added to the plan in order to enhance its effectiveness. The best argument made for Iowa and New Hampshire is that their small populations allow for highly personalized campaigning. The candidates are able to meet individual citizens for lengthy and sometimes repeated conversations about the issues, and these voters are able to size up potential presidents at eye level, without the candidates having the protection of the usual large retinue of image makers and staffers. In that sense, lightly populated states can serve as a useful screening committee for the rest of us. The United States is a continental country, after all, and each large region is still enormous in size.

There is a way to combine the advantages of small-state scrutiny of candidates with the inherent fairness of round-robin regional primaries. We can achieve the best of both worlds by adding a second lottery on Jan. 1. The names of all states with four or fewer members in the U.S. House of Representatives (at present, 21 states) would be placed in a lottery machine, and two balls would be selected. This plan excludes the island territories, which are far-flung and don’t influence the November presidential outcome because they have no electoral votes. The District of Columbia should be included, however, and this would mean 22 jurisdictions would have a chance to be selected in the second lottery. With a larger population than Wyoming, and with three electoral votes assigned in November, the District’s citizens — currently and shamefully without full voting representation in either house of Congress — would no doubt relish and deserve this opportunity, should Lady Luck in the lottery deliver it to them.

The two small states (or D.C.) with relatively low populations would lead off the regional contests, and they would be held on or about March 15 — at least two full weeks before the initial contests would begin in the first region. These two states would be free to stage a primary or a caucus, and the candidates would be free to participate in none, one, or both. As a practical matter, most candidates would choose both, unless a prominent candidate hailed from one of the lead-off states. Traditionally, a home candidate gets deference and is sometimes unopposed for the state’s delegate votes. Of course, the other party can still have a full-fledged fight in the state’s primary or caucus.

No doubt, all the candidates would rush to these lead-off states right after the lottery on January 1, and they would have two and a half months to campaign. But there would be no permanent, four-year campaigns there, and personalized, one-to-one campaigning would be a large part of the effort. In other words, the two states would offer all the advantages of Iowa and New Hampshire, without having to always be Iowa and New Hampshire. Additionally, the guarantee of at least two weeks of decompression after the leadoff states make their choices gives voters in the first region a chance to evaluate the results and reevaluate the winners — and possibly to make different choices.

In sum, the Regional Lottery Plan would achieve many good things simultaneously for a selection process that currently makes little sense. The election campaign would be shortened and focused, a relief to both candidates and voters. All regions and states would get an opportunity to have a substantial impact on the making of the presidential nominees. A rational, nicely arranged schedule would build excitement and citizen involvement in every corner of the country, without sacrificing the personalized scrutiny of candidates for which Iowa and New Hampshire have become justly known. And all of this can only come about by putting the politics of nominations and elections in its proper place — the United States Constitution.