On February 3, a bizarre ritual will once more take place across the state of Iowa. Caucusgoers—that is, people whose schedule and abilities allow them to spend hours in a local high school gym, public library, church basement, or American Legion Hall on a weeknight—will try to persuade their neighbors to stand in a corner with them to show their support for their preferred presidential candidate. Once enough Iowans without childcare responsibilities or work obligations have stood in enough corners, the ritual is over. The presidential race will enter a period of fresh clarity.

A few candidates will be crowned winners, either because they actually won or because they lost by less than expected, while others will be declared losers, for the exact opposite reasons. Some candidates will drop out of the race, while others will be the beneficiaries of headlines declaring them front-runners. Even before the caucus, the shape of the race has been determined by what people in Iowa think of the candidates. Iowans like former South Bend, Indiana, mayor Pete Buttigieg, and thus Buttigieg has become a front-runner. Iowa is so influential that a candidate will poll better in the rest of the country because they poll well in Iowa. Every poll released by the Des Moines Register is parsed by the national political press like it’s the I Ching, reporters reading the future in the preferences of 602 likely caucusgoers. Those candidates preferred by the 602 will raise more money, earn more media coverage, enjoy access to debate stages, and find themselves discussed more prominently.

And all of this happens because, eventually, maybe 175,000 people will gather in hours-long meetings around the nation’s thirty-second most populous state, with no secret ballots, with one of America’s least diverse electorates.

Julián Castro was right. This system is a bad one. Before he dropped out of the race in January (after failing to catch on with Iowa voters), he argued that Iowa shouldn’t go first in the presidential primary process. The Hawkeye State’s first-in-the-nation status is a historical fluke; it was picked more or less at random in the early seventies. And given the makeup of Iowa’s electorate, it’s hard to dispute his point: Iowa is more than 85 percent white, while the nation as a whole is only 60 percent Anglo. Nearly a third of the state lives in rural areas, compared with less than 20 percent of the overall U.S. population. And only a tiny subset of the already-limited pool of voters even participates: in 2016, of the 1.9 million registered voters in Iowa, only slightly more than 350,000 of them (16 percent) bothered to show up for either party’s caucuses.

So let’s accept that Castro expressed an accurate sentiment when he said, “Iowa and New Hampshire are wonderful states with wonderful people, but they’re also not reflective of the diversity of our country.” Let’s also accept that the caucus process is a poor way to guide the overall process. Nonetheless, someone has to be first. We humbly submit that it should, in fact, be Texas.

Here’s why. 

Roughly 9 percent of the U.S. population lives in Texas. That’s a hefty chunk of the country, but not such a large percentage that you’d lose the advantages that come with a staggered primary calendar. Texas represents the ideal balance: large enough that candidates would have to prove their mettle among a wide swath of voters from different ethnic and regional backgrounds, but not so large that it would amount to a de facto national primary, which would tend to favor deep-pocketed candidates like Mike Bloomberg and Tom Steyer in the current race, or flavors of the month like Ben Carson in late 2015, or Rick Perry for a brief moment in 2011.

But wouldn’t candidates spend all of the time in Texas’s big cities? After all, Texas is an urban state, and it would be easy to campaign in the urban triangle formed by San Antonio, Austin, Dallas–Fort Worth, and Houston. However, the 19 million or so Texans in the big four metro areas still represent only about 65 percent of the state, which means there are a whole lot of voters to be found elsewhere in Texas. There are large population centers in Iowa, too, and that doesn’t stop candidates from visiting coffee shops in towns like Corydon (pop. 1,585) or working the grill at Shaggy’s Gourmet Burgers in Wapello (pop. 2,067).

Certainly, presidential hopefuls would be common sights in Houston, Dallas, Austin, and San Antonio, but successful candidates would also need to spend a lot of time in El Paso, Lubbock, Corpus Christi, East Texas, the Panhandle, the Permian Basin, and the Rio Grande Valley as they looked to put together a coalition that could win them a primary. Would every single voter in the state get to ask every single candidate a question while they choked down a corn dog? Probably not, but the republic will survive. There simply isn’t much value in a tiny, nonrepresentative slice of the electorate hogging so much of the candidates’ time.

Starting in Texas would force candidates to prove that they could compete in both the grassroots and media environments, while the focus on Iowa mostly tests only the former. Texas has four of the nation’s top forty media markets (and nearly twenty, in total), and 268,597 square miles of land. That is a lot of ground to cover, but that’s a good thing. Primary campaigns are now thirteen-month slogs in a hyperactive media environment. The protracted schedule means that every candidate makes at least three separate visits to each Iowan’s living room, but that kind of hyper-retail politics doesn’t really benefit the rest of America. Even if you accept the proposition that retail politics are still a vital part of campaigning in a nation of 327 million, forcing candidates to spend more than a year courting caucusgoers in Iowa is overkill.

To win in Texas, you have to do both retail and wholesale politics. Take Beto O’Rourke’s 2018 campaign for Senate: he showed that a campaign that travels the state can be much more successful in Texas than campaigns that focus overwhelmingly on big cities and media buys. His attention to smaller counties helped him activate more voters than any statewide Democrat before him, outperforming Hillary Clinton’s 2016 campaign in counties like Nacogdoches and Taylor by several points. In a primary, where there is no Republican to contend with, those are voters that no serious candidate would be foolish enough to ignore.

Which brings us back to demographics. Texas, unlike Iowa, looks a whole lot like America—and certainly like the America of the future. We’re younger than the nation at large—13 percent of our population is over 65, compared to an average of 16 percent among the other 49 states. We’re growing faster than the rest of the country. Texas’s population surged by 15 percent over the past decade, compared with 6 percent nationally. Almost 40 percent of Texans are Hispanic, second only to New Mexico. That’s a strength when you’re talking about a demographic responsible for more than half of the overall growth in the U.S. population. Making Texas the first-in-the-nation state would help correct the longstanding under-representation of Latino voices.

Certainly, there are other states that a person could argue should go first. Nevada is a diverse, fast-growing state. But the Las Vegas metro area accounts for more than 71 percent of the state’s population. The Nevada primary is essentially the Las Vegas primary—hardly a representative sample of the nation. Someone who’s able to speak to the needs of people in Dallas, McAllen, Odessa, Beaumont, and Killeen, though, should cover the bases for most Americans. (Nevada, like Iowa, also awards its delegates through caucuses rather than primary voting.)

Florida, while fast-growing, large, and diverse, is already perhaps the most important swing state. Focusing even more attention on the Sunshine State would mean that America had effectively delegated its presidential election process entirely to one state. Florida Man already has plenty of influence, and couldn’t reasonably be trusted to handle more. Big, diverse New York and California already wield significant influence in national politics, and their residents will never be voiceless in the primary process, no matter where it starts.

Texas makes the most sense. It’s a roughly representative sample of the nation, large enough that candidates would have to prove their mettle to more than just a relative handful of hyper-engaged residents, but still small enough that winning here wouldn’t effectively end the entire campaign in one shot. Candidates would have to be mindful of the concerns of people who live in coastal areas; on the border; in rural regions; in some of the nation’s largest cities; and in sprawling suburbs and exurbs. They would need to address the needs of a growing population, and of diverse groups. The media coverage, and subsequent fund-raising boosts, would no longer be skewed by the priorities of a small collection of people in a state that cares about things (corn subsidies!) that the rest of America does not. The current primary system is broken. Texas can help fix it.