Well There's Your Problem Right There

Lost in the aftermath of the Chernobyl-level Iowa caucus debacle is a crucial point: Even if the count were not fucked up, we'd still be questioning Iowa's primacy in and centrality to our process of choosing a president. 

For years, the counts Iowa reported on election night were not raw vote totals but "state delegate equivalents": a somewhat abstracted figure representing the total number of delegates to the state party convention awarded to a particular candidate using a specific mathematical formula. (Not the number of delegates to the national convention, i.e. the count that matters; just the number pledged to the state convention.) In 2016, Hillary Clinton edged Bernie Sanders in this count by 0.3%, giving her a nominal victory. Bernie's crew maintained that he had won more caucus-goers, and if that total were reported, he could reasonably have claimed a victory of his own (or a split decision at worst). But this was ultimately speculative, since the only count reported was the abstracted "state delegate equivalent" number.

This year, as a result of DNC process reforms, Iowa is on the hook to report several numbers:

• The raw vote total per candidate
• The adjusted vote total after "first alignment" (in which supporters of non-viable candidates are reallocated to others)
• The number of "state delegate equivalents" (SDEs) awarded to each candidate

Had everything worked as expected, purely as a function of the way these numbers are constituted it's very likely that multiple candidates could have claimed some version of victory--not just in raw vote or in "state delegate equivalents", but in being a strong second-choice option (and thus acceptable to a broader range of voters). Even a lower-tier candidate could claim "momentum" if they exceeded expectations on just one of these figures. It's thus very likely that the narratives coming out of Iowa would have been completely muddled and impossible to parse, lending the state of play no greater clarity than before the caucus. This alone would have severely damaged the Iowan mythos of being the first great acid test of a real presidential campaign.

Further, with Iowa reporting not just the SDE count but every data point along the way, those of us playing the home game can run the numbers ourselves. Any discrepancy, data entry error, or simple mistype can therefore be audited in near-realtime in a way that was not previously possible. We always just took the Iowa state party at its word that the caucus process and resulting math was being done correctly. Now we know that to be untrue and have the math to prove it. But it's also the case that the instructions to caucus precinct leaders are difficult to follow and the process by which delegates are awarded is convoluted. Perhaps it's always been this way--but now we have greater insight into the precise ways in which the process is convoluted. The above link takes you directly to the Iowa Democrats' precinct leader guide, which I was able to find in about three seconds via Google search. There's no way this document wasn't going to get out and dissected on Twitter. 

If we were designing from scratch a system to nominate Democratic presidential candidates, would we start the process in a small, rural, demographically unrepresentative red state (not expected to be competitive in the general election) using a time-intensive voting system that ultimately relies on 1600+ people to all correctly solve a math problem? I think the answer is self-evident. I can only hope that in future years, this byzantine relic will be relegated to the cornfield where it belongs.