You’re probably already familiar with some of the basic arguments for the elimination of caucuses from America’s byzantine presidential primary system:
• they exclude people who have to work on weekends (or evenings)
• they exclude people who are elderly or disabled and don’t have the mobility or stamina to go out and participate in something that takes several hours
• they exclude people with young children, who no matter how much organizers say “oh children are welcome” don’t want to spend two hours getting the stink-eye from fellow caucus-goers
• they exclude people who believe in a secret ballot or just want to avoid being antagonized by supporters of opposing candidates, and
• they’re a deterrent to people who simply think it’s a waste of time to spend two hours on something that should only take a few minutes.
All of that adds up to remarkably low turnout in caucuses. Even the vaunted Iowa caucus had considerably lower turnout, percentage-wise, than the least-attended primary (in Texas). That’s not what we as a party are about—we want as many people to vote as possible and have their voices heard. We want to remove barriers to voting, not construct additional barriers like requiring voter ID or reducing the number of polling places. We certainly don’t want to restrict the vote to only the most motivated, most fervent voters with the most time on their hands.
And now we have serious evidence of one other consideration: Caucus results don’t even accurately represent the will of the larger party electorate when they tend to sample only the party’s most motivated members. In 2016, we’ve now had two instances of states that, on the Democratic side, had a caucus (that was binding, for purposes of allocating delegates) and then a primary (that was a “beauty contest” and didn’t count toward delegates). In both cases, a different candidate won the caucus than won the primary.
Nebraska held its Democratic caucus on March 5; Bernie Sanders received 19,120 votes and Hillary Clinton received 14,340 votes. This was a 57-43 margin, giving Sanders 15 pledged delegates to Clinton’s 10. However, on May 10, Nebraska held its non-binding Democratic primary. Clinton got 42,665 votes to Sanders’ 37,705, a 53-47 edge for Clinton.
And Washington held its Democratic caucus on March 26. Sanders received 19,159 precinct-level delegates to 7,140 for Clinton; this became a 74-27 advantage for Sanders in pledged delegates, which was in fact his biggest net gain of any state in the entire primary process. (The only formally reported numbers from Washington are delegates, not popular votes. However, the Washington Post estimated, based on the party’s reports of 230,000 participants, that the actual popular vote broke down 167,200 to 62,330.)
At the end of Tuesday night, however, the results from the Washington primary were 354,688 for Clinton and 306,715 for Sanders, a 54-46 edge for Clinton, a reversal of nearly 30 points in each direction. One noteworthy caveat is that Washington is a vote-by-mail state, where ballots will be counted so long as they are postmarked by Election Day; the estimate is that the Tuesday night count reflects only 74 percent reporting. There’s also a trend in Washington where younger Seattle-area voters tend to wait until the last minute to submit their ballots, so that close statewide races tend to break in the more liberal direction as the count unfolds. So it’s possible that Sanders will close the gap somewhat as votes are counted in coming days. The likelihood of overcoming a nearly 50,000 vote gap, though, seems slight, and the Associated Press, hopefully aware of this pattern in Washington, has already called the race for Clinton.
This would be less alarming if it only happened in one state, or if the same candidate won both the caucus and the primary but won the primary by a reduced margin (as happened with Barack Obama in Washington in 2008, winning the caucus by a 68-31 margin and then winning the primary only 51-46). But in 2016, it happened in both states that use this system—and in each case, the broader sample of the electorate chose an entirely different candidate than the smaller, self-selected group who attends the caucuses did.
Granted, this wouldn’t pass muster as a controlled experiment where the external conditions remain the same throughout. It’s entirely likely that, if the primaries were the only events in those states that counted, the candidates would have contested them much harder (or at all, seeing as how there was no advertising in the lead-up to the Washington primary). Sanders, for instance, would have had his widely-attended Washington rallies prior to the primary, rather than prior to the caucus, which might have motivated more people to turn out for him in the primary. Perhaps his supporters didn’t bother with the primary, figuring that they’d already done their duty and didn’t need to spend 48 cents rubber-stamping the results.
The timing within the primary cycle may have had an effect as well. It’s likely that Sanders supporters have less expectation of an overall victory now than they did in late March, and might vote with less intensity as a result. Nevertheless, it’s hard to argue that those differences in motivation alone could be responsible for a 55-point swing in the margins in Washington, especially given the minimal amount of effort involved in voting by mail in the Evergreen State.
This might still be leaving you with one big question: Why do these states have such redundant systems? In both cases, the primary is required by state law. In Nebraska, the presidential primary coincides with the primaries for downballot congressional and legislative races, though in Washington, the downballot primaries are in August and the presidential race is literally the only thing on the May primary ballot.
The Republican parties in both states simply rely on the primaries to allocate delegates—which makes sense, since there’s already a perfectly fine primary that they can use for free. The Democratic parties in those states, however, insist on using a separately-scheduled caucus anyway to allocate delegates. In Washington, the rationale is “party building,” which means expecting that people will enjoy the caucus so much that they keep going to other district-level meetings. But what it probably really means is a way to harvest email addresses for fundraising solicitations.
(In Washington, there’s also the rationale offered that there’s no registration by party, so caucuses are a way to limit the presidential pick to the true believers. But the Washington caucuses are “open” caucuses anyway—anyone can show up without having to flash their Democrat card or pay dues, and certainly lots of self-described independents participated in Washington’s caucus this year.)
In Washington, we already have the single best tool for ensuring high turnout, and for ensuring that government-imposed barriers (like voter ID requirements at the polls, or long lines caused by having inadequate polling places) won't stop people from voting: Vote-by-mail. (Vote-by-mail states Washington and Oregon regularly vie with Minnesota and Wisconsin for the highest turnout in the nation.) It makes absolutely no sense to spurn that system in the presidential nominating process in favor of a different system that is unnecessarily time-consuming, intimidating, serves to exclude broad swaths of the population—and as we saw in Iowa and Nevada this year, can lead to uncertain, vague results that just infuriate losing candidates' supporters.
Minnesota, the second-largest state (after Washington) that still used a caucus this year, has already realized this and will not use the caucus starting in 2020, opting to switch to a primary. It’s time for the remaining caucus states to join them, now. Kill the caucus.