It’s about that time of year when people who don’t live in Iowa complain about the Iowa caucuses.

Struggling presidential candidate Julian Castro has been the most pointed, charging  that the Iowa caucus system “actively diminishes the voices of African Americans and people of color in the Democratic presidential nominating process.” Joe Biden, who is contending in Iowa but polls worse there than in the other early primary states, has also cast shade on the Iowa caucuses. “Are they representative historically and practically — based on race and creed and color — of the nation?” Biden asked before answering his own question: “No, they’re not.”

It’s not in dispute that the Iowa caucuses don’t give a significant voice to communities of color. In 2016, Iowa caucus-goers were only 9% nonwhite, whereas nationally, Democratic primary and caucus participants were 38% nonwhite. But a more complicated question is: Does the entire Democratic Party presidential nomination process, because Iowa goes first, inflate the role of whites and marginalize the role of people of color?

First, let’s get one debate out of the way: Why shouldn’t we have a 50-state primary, and stop giving preferential treatment for certain states? Simple: A 50-state primary would make it extremely hard for little-known, underfunded candidates to break out. Small-room, up-close, retail campaigning would almost completely cease to exist. Candidates with easy access to early money — including self-funding billionaires — would hold a colossal advantage.

So someone has to go first, and Iowans have eagerly accepted the role for years. However, by 2008 many Democrats recognized a problem with having two nearly all-white states (with the New Hampshire primary following the Iowa caucuses) begin the process, especially for a party that takes pride in its racial diversity.

Since Iowa and New Hampshire are intensely protective of their respective “first in the nation” status, the Democratic National Committee decided to balance them by making Nevada and South Carolina the third and fourth contests. Nevada’s Democratic turnout in 2016 was 41% nonwhite, and 19% Latino, while South Carolina’s was 65% nonwhite, nearly all of that African American.

Did that solve the problem? Has the Democratic primary process become sufficiently racially inclusive? Our data points are few. We have only the 2008 and 2016 Democratic primaries to examine. But so far, under the new calendar, South Carolina has played a more decisive role than Iowa.  In 2008 and 2016, the Iowa and South Carolina winners were the same: Barack Obama in 2008, Hillary Clinton in 2016.

However, Iowa didn’t give Obama or Clinton any momentum for the subsequent New Hampshire primary, even though both states have similar white, liberal and educated Democratic electorates. In stark contrast, their respective South Carolina wins helped them both recover from their New Hampshire defeats, and sweep the Deep South.

Why hasn’t Iowa been more of a springboard? It has been at times. Back-to-back Iowa-New Hampshire victories have happened since Jimmy Carter’s successful underdog campaign infused the Iowa caucus with political relevance in 1976. Al Gore and John Kerry won them both in 2000 and 2004 (and then proceeded to coast to the nomination). But split decisions happened in 1984, 1988 and 1992, as well as those in 2008 and 2016.

While there are specific factors unique to each race, there are two key differences between the Iowa and New Hampshire contests that help explain why one state does not automatically determine the outcome for the other.

One difference is geographic. Iowa has been partial to Midwesterners, such as Minnesota’s Walter Mondale, Missouri’s Dick Gephardt, and Iowa’s own Tom Harkin. Obama, too, came from neighboring Illinois. New Hampshire primary voters often favor their fellow New Englanders, including Massachusetts’ Michael Dukakis and Paul Tsongas, and Vermont’s Bernie Sanders. (This year’s primary features two Midwesterners in Indiana’s Pete Buttigieg and Minnesota’s Amy Klobuchar, and two New Englanders in Sanders and Massachusetts’ Elizabeth Warren — , three, if we count Massachusetts’ Deval Patrick.)

Another is that Iowa is a caucus state, and the time-intensive, low-turnout affairs tend to attract committed party activists. New Hampshire is an open primary in which independents often participate in big numbers. Polls of the 2008 and 2016 Democratic contests show that about 20% of Iowa caucus-goers self-identify as independents, versus at least 40% of New Hampshire primary voters.

South Carolina’s power rests in its linkage to the other Deep South states. Together with Louisiana, Mississippi, Alabama and Georgia, South Carolina is part of the “Black Belt.” Each state’s Democratic electorate is majority African American or close to it. All five states hold primaries, not caucuses. And since 1988, they have voted for the same candidate, with the exception of 2004 when South Carolina voted for its native son John Edwards.

In 2008 and 2016, winning the Black Belt has made a big difference in what matters most: the delegate count. In those five states, Obama won a net of 65 delegates over Hillary Clinton, which was about two-thirds of his final margin of victory in pledged delegates . Eight years later, Clinton won 153 more delegates than Bernie Sanders in the Deep South, and another 147 in five states where African Americans constitute a significant minority: Arkansas, North Carolina, Tennessee, Texas and Virginia. The delegates from those 10 states accounted for more than 80% of her margin of victory.

The Black Belt didn’t just matter in the final delegate count. It also helped direct the momentum of the race. For example, on “Super Tuesday” in 2008, which involved contests in 22 states 10 days after South Carolina, California was thought to be the biggest prize, as it offered the most delegates of any one state: 370. But Clinton’s eight-percentage-point win in the Golden State popular vote netted her a 38-delegate advantage over Obama’s share. On the same day, Obama won Georgia popular vote by 35 percentage points. That netted him 33 delegates, almost fully negating Clinton’s California win, and more than accounting for Obama’s slight delegate edge from all the Super Tuesday contests. Without winning South Carolina, Obama doesn’t win Super Tuesday, or the nomination.

In 2016, Super Tuesday was a smaller affair of 11 states. But Clinton won seven of them, five of which were in the South, widening a delegate gap that Sanders would never come close to bridging. A week later, Sanders pulled off his upset in Michigan’s primary, which he hoped would shift the race’s momentum. But Michigan was a close contest, and he netted only four delegates. On the same day, Clinton won Mississippi in a rout, and netted 26. Even on Sanders’ best day, the black vote gave Clinton more delegates.

Because South Carolina has functioned as a Southern bellwether, it has had more political relevance than has Nevada. The Silver State caucuses happen before South Carolina, and have given Latinos a larger voice in the process. But Latinos, who made up about one-fifth of the Nevada turnout in 2016, don’t dominate the state in the same way that African Americans do in South Carolina. And they didn’t serve as any sort of bellwether. In 2016, Sanders won the Latino vote in Nevada (at least, according to the entrance poll), but still lost the state to Clinton. Soon after, on Super Tuesday, Latinos helped power Sanders’ victory in the Colorado caucuses, but Clinton won the Latino vote, and the overall vote, handily in delegate-rich Texas.

This time around, Super Tuesday is just three days after South Carolina, giving campaigns little chance to breathe before competing in 14 contests accounting for 40% of the available delegates. The two biggest states, California and Texas, will be on tap, as well the Southern states Alabama, Arkansas, North Carolina, Tennessee and Virginia.

Granted, there’s no certainty that South Carolina will prove again to be the pivot point of the primary storyline. The field is too unsettled, and the African American vote, which is showing signs of division along generational lines, is not monolithic.

But considering that New Hampshire voters do not always follow the dictates of Iowans, and that South Carolina Democrats — since holding a stand-alone early state primary — have yet to follow New Hampshire, there’s no reason to assume that black South Carolina voters will blindly follow what white voters in Iowa and New Hampshire do.

The winner of Iowa may well go on to win New Hampshire, Nevada and South Carolina and then run the table. But if that happens, it would not be because black voters in the Palmetto State lost their voice. It would be because they liked what they saw up north and out west, and made up their own minds.

A candidate certainly doesn’t want to fizzle in Iowa, and risk being winnowed out before reaching South Carolina. That chronological reality does give Hawkeye whites a significant role in the process that others do not have. But if there is one early state that is more important than all the others, I would still say it is South Carolina. The black vote remains critical to the outcome, and their voices will not be diminished.