It is happening again: Iowa, New Hampshire and South Carolina are all but deciding the Republican Party’s nominee, with GOP voters in the remaining 47 states left to rubber-stamp the judgment of the early trio.

Every four years, Republicans outside of the “holy trinity” complain that they’re effectively shut out of their party’s most important decision in most cycles; every four years, there are rumblings about changing the system, and most cycles, nothing changes other than the primary contests beginning earlier and earlier.

We should set out to organize a primary calendar that permits as many Republicans as possible to have a meaningful effect on the nominee selection process by voting. (I speak as a Republican to the lament of my party’s voters and candidate selection process, but much of this diagnosis and potential solution also applies to the Democrats.)

Some have called for a national primary, but that would bring its own set of problems. If you think the current system is prohibitively expensive and crowds out promising candidates, imagine having to run a national television advertising campaign to have any chance at all at the nomination. It would be just a larger version of our current Super Tuesday–style mega-primary days, which require candidates to campaign in as many as a dozen states in the same short period.

Super-primary days pretty much ensure that the candidate with the most funds will win because those candidates are the only ones who can afford to run ads in all of those states voting on the same day.

A state can present value to a presidential candidate in one of three ways: the number of delegates at stake (often tied to population size); the state’s chronological order in the primary calendar (a chance to make an early splash and create momentum for later contests); or ease or cost-effectiveness for campaigning (small size and short travel distances, cheap television advertising rates, etc.).

Iowa and New Hampshire are not particularly populated states — this year only 40 delegates are at stake in these two states out of 2,286. But because they’re first, they’re worth immense investments in time and money on the part of candidates.

If we wanted to balance this out so that each state offers value to a candidate, we would begin with the least-populated states, which have the fewest delegates to the GOP convention, and work our way up to the largest and most delegate-rich states. To prevent a prohibitively long and expensive 50-week primary, we would cluster the states into geographic groupings.

Under this system, the earliest states would still get enormous attention, but they would have the least consequence in terms of number of delegates. A candidate could stumble and still rebound in the next week’s contest, at least for a little while. Because of their small population (and, often, geographic size), the early process would still involve retail politicking with a lot of personal interaction with primary voters. Early state victories in relatively inexpensive places such as Delaware and Montana might translate to influxes of funding and momentum, so the idea of an underdog rising to the top would be more plausible. But the steady week-by-week drumbeat of increasingly larger midsized states (Indiana, Colorado, Arkansas) would require candidates to demonstrate that they’re more than just a flash in the pan.

The biggest states, California and Texas, might lament that under this system they’re destined to go last. But one or both could very well end up with the “kingmaker” role, putting one of the leading final contenders over the top. They would be more than just ATMs that are visited by candidates for fundraisers but largely ignored in terms of actual candidate campaign stops.

Under this system, the 2016 Republican primary process would look something like this:

Jan. 26: Maine, New Hampshire, Vermont (65 delegates). Notice that everyone in the political world gets to enjoy their holiday season and New Year’s, with nearly the entire month of January to prepare for the first contests.

Feb. 2: Delaware, the District of Columbia (36 delegates). Contests can’t get much smaller or cheaper than these. If an underfunded but talented candidate can’t break out through retail campaigning in contests that consist of a city and three counties, they really can’t expect to break out anywhere.

Feb. 9: South Carolina (50 delegates). It retains its first-in-the-South status under this system.

Feb. 16: Montana, North Dakota, South Dakota, Wyoming (111 delegates).

Feb. 23: Iowa, Minnesota (68 delegates).

March 1: Nevada, Utah (68 delegates).

March 8: Oregon, Idaho, Washington (103 delegates).

March 15: Connecticut, Rhode Island (47 delegates). These delegate-poor states were moved back several weeks to prevent five of the first 13 states being in New England; also note that because of the cost of the New York and Boston media markets, these would be the first expensive primaries.

March 22: Arizona, New Mexico (81 delegates). This would be the end of the “small state” portion of the contest, with every remaining state having at least 30 delegates and more expensive media markets, with a pair of exceptions.

March 29: Michigan, Indiana (106 delegates).

April 5: West Virginia, Kentucky (76 delegates).

April 12: Colorado, Nebraska, Kansas (111 delegates).

April 19: Arkansas, Mississippi, Louisiana, Alabama (172 delegates). This is about as close as it gets to Super Tuesday on this calendar, and the contest consists of four contiguous, relatively inexpensive states with similar cultures and economic concerns.

April 26: Maryland, Virginia (86 delegates).

May 3: Massachusetts, New Jersey (91 delegates).

May 10: Wisconsin, Illinois (111 delegates).

May 17: Florida (99 delegates).

May 24: Missouri, Oklahoma (95 delegates).

May 31: Hawaii, Alaska (47 delegates). These are the exceptions. Geography will make these states challenging for candidates at any point in the calendar; some candidates may choose to skip these contests entirely. The Tuesday after Memorial Day weekend seems like as good a time as any to have the candidates travel outside the 48 contiguous states.

June 7: North Carolina, Tennessee (113 delegates). Notice that the number of delegates at stake each week is irregularly but gradually increasing.

June 14: Ohio, Pennsylvania (138 delegates).

June 21: Georgia (76 delegates).

June 28: New York (95 delegates).

July 5: Texas (155 delegates).

July 12: California (172 delegates).

This would still leave about six weeks until the national party conventions, if they stick with their traditional late August or early September scheduling.

At some point, we would have to fit in primaries for the U.S. territories of American Samoa, Guam, the Northern Mariana Islands, Puerto Rico and the Virgin Islands. They have few delegates at stake and distance makes them difficult for candidates to reach, but perhaps they should be squeezed in early to give candidates a chance for a late winter campaign stop on a warm beach, or added to the day for Alaska and Hawaii.

Some might see this as a formula for a long, drawn-out, hard-fought primary; of course, Barack Obama and Hillary Rodham Clinton campaigned in one of the longest and most contentious primaries in U.S. history in 2008 and that turned out pretty well for Obama. Under this system, it’s possible to have multiple winners most weeks. Or perhaps an exceptional candidate would sweep the early contests and make the rest of the calendar moot. Much of this is in the hands of the primary voters, which is how it should be.

You can’t guarantee a competitive primary race that will go on for many weeks, allowing millions of primary voters in many states to weigh in. But you can arrange the schedule to avoid any quick slam-dunks that allow the frontrunner to lock up the nomination quickly. Our parties, and the country, are ill-served by a system that often amounts to a three-state coronation.

Jim Geraghty, a contributing editor at National Review, writes the Campaign Spot blog on NRO and appears regularly on CNN, MSNBC and Fox News.