Imagine a developing country that is relatively new to democracy holds an election with the following characteristics:
The election is held during a relatively short window of time during which many would-be participants are unable to attend due to work commitments. Registered voters are required to re-register before participating, resulting in long lines that force still more would-be participants who took a few hours off to go back to work.
There is no absentee voting, as votes are cast by participants gathering in rooms in their local precincts and physically standing in candidate-specific corners to indicate their choice. Uncommitted voters are able to hear pitches from local representatives for each of the candidates before deciding, but these pitches are presented in front of the rest of the voters in the room, often resulting in shouting matches between the respective camps. This part of the election is administered almost entirely by volunteers, many of whom are completely new to the process themselves, and the proceedings in some rooms (including the choices of late-deciding voters) are broadcast on national television for entertainment purposes.
Raw vote totals are not reported. Instead, members of the bureaucracy administering the election use a proprietary formula to convert vote totals into close-but-not-quite-representative electoral vote equivalents, which are used to decide how electoral votes are apportioned. However, rather than simply count up electoral votes nationwide, they are apportioned on a district-by-district basis. The districts lines are drawn by the dominant political party in the state, skewing representation further.
An American election observer would not recognize the results of such an election as being valid. They would say that the ballot was not accessible too many potential participants, and that there were too many points at which the participation of those who did show up could have potentially been manipulated, diluted or ignored outright. This being the case, it's simply impossible to tell if the results accurately reflect the aggregate preferences of the population in question.
And yet, the election I just described was the Nevada Caucuses.
The Caucuses were an absolute disaster on Saturday. At one point on MSNBC (which really did broadcast an undecided voter making his decision between Bernie Sanders and Hillary Clinton live), Chris Hayes reported that the process had been disorganized, complicated and haphazard. He then talked to a precinct captain, who was caucusing for Clinton, who agreed that she was frustrated with the process, going on to sheepishly disclose that she felt a primary would be better. The next person Hayes talked to, a first-time participant caucusing for Sanders, was similarly frustrated.
In a day marked by spirited political disagreement, everyone involved came together to agree on one thing: this was not what democracy was supposed to look like.
I've yet to see a single person come away from the caucuses with a good (or even bad) argument as to why they are preferable to one person/one vote primaries. Hillary Clinton won, taking 19 delegates to Sanders's 15, but she won in a process that was so busted it's difficult to say whether she or Sanders was more disadvantaged by the caucusing process. In other words, I'm not sure Sanders supporters can make a compelling case that, had Nevada held a primary instead of a caucus, the results would have been different. For all we know, Clinton could have won a primary by even more. We simply don't know.
When the goal is to nominate a candidate for president, that's a huge problem.
An earlier version of this post appeared on AMERICAblog.com.