Why the Iowa Caucus Vote Needs To Be Audited From Raw Data & the Audit and Data Published

by Thomas Neuburger

The new show-your-work requirements for Iowa seem to have revealed that not every precinct follows the rules and many struggle with the math of calculating delegate equivalents, there has probably never been a clean caucus.
—Matt Bruenig here (emphasis added)

For the Democratic Party to undermine faith in the fairness of its own process will have devastating consequences, just the first of which is the re-election of Donald Trump. It’s crossroads time, folks.
—Yours truly here

When the Iowa caucus results were originally withheld by the state party, one of the reasons given was "inconsistencies" in the reported data. That was Monday night. It's now many days later, and those inconsistencies have not only not been ironed out, they're still popping up. And being reported by caucus managers via Twitter and noticed by news outlets everywhere.

I'll present just a few of the "inconsistencies" below, but first, a larger point. If the Democratic Party in general, and the Iowa Democratic Party in particular, want to undo the damage they have done, not only their own reputations but to the electoral process as a whole, they will hire independent auditors to reverify the results from raw data — and then publish those results and the data that supports them.

At this point, I don't think anything less will repair the harm done by the #IowaCF that, as I write, is still ongoing. Consider just the few examples below, then ask yourself, no matter whom you prefer as a candidate, will you believe that whatever they say about the results is accurate, absent the raw data and an auditor?

Example 1 — Polk County caucus. From Sean Bagniewski, chair of the Polk County Democrats, writing on Feb 5, two days after the election:

We’re aware that some of the Polk County precincts reported by the Iowa Democratic Party this afternoon were incorrect. Our precinct chairs and our Executive Director Judy Downs have notified the IDP [Iowa Democratic Party] and we’re helping them get it corrected.

Note that he's not saying the county data incorrect; he's saying that the record of that data at the state Democratic Party is wrong.  

Example 2 — Black Hawk County caucus. From Chris Schwartz, Black Hawk County supervisor:

Click to enlarge

Mr. Schwartz added this on February 5:

We have known for over 24 hours as verified by our county party that @BernieSanders won the #iacaucuses in Black Hawk County with 2,149 votes, 155 County Delegates. #NotMeUs #IowaCaucuses

Schwartz two hours later:

The state party is now being forced to walk back their error of giving @BernieSanders delegates to @DevalPatrick who received zero votes in Black Hawk County. Press can dm me.

The "error" he refers to was that the state party transferred, perhaps inadvertently but who knows, a number of votes from Bernie Sanders to Deval Patrick, who was out of the race and had no adherents.

Example 3 — Walnut/Grimes/Grant caucus. The transfer of votes from Sanders to Patrick, and perhaps from Warren to Steyer, appears to have occurred elsewhere.

The highlighted part of the following graphic appears to be from the Walnut/Grimes/Grant caucus results as reported by the state party:

Click to enlarge

If the graphic is accurate, Deval Patrick got or was given 20 and 23 votes on the first and second alignments respectively, while Sanders got or was given 2 and 0 votes on the first and second alignments.  Also in the same caucus, Steyer got 20 and 23 votes while Warren got 0 and 0, if I read this correctly.

The Patrick-Sanders data swap seems an obvious problem, but to my knowledge, it hasn't been addressed. And the Steyer-Warren data could be fishy as well, but without the raw data we'll never know. 

We'll also never know if errors like these have been fixed without an independent and published audit.

Example 4 — Polk County-Des Moines Precinct 14 in Merle Hay. This is from James Payne who said he attended the caucus:

They fucking misreported my precinct. Bernie won 2 delegates in Polk County - Des Moines Precinct 14 in Merle Hay and we fought like hell for them. And @iowademocrats put one of our SDEs in Warren's column but correctly reported vote

Payne later adds, "Here is the raw misreporting on the failing @iowademocrats's website. That .27 should be a .52 SDE. Instead, they gave it to Warren."

Here is the raw misreporting on the failing @iowademocrats's website. That .27 should be a .52 SDE. Instead, they gave it to Warren. They're trying to steal this. #IowaCaucuses #iowa #IowaCaucusDisaster pic.twitter.com/ciTvBI4mTH

— James Payne (@Banalization) February 5, 2020

Click through for the images. 

Of his own experience, Payne writes, "I flew to Iowa from NYC, slept on a blow-up for a week, knocked in the cold, called, was trained & I met tons of great people & had an awesome time but no one I was with was being paid, everyone was taking a hit to do this, I will not stand for this shit. #IowaCaucusDisaster"

Example 5 — Statewide data analysis. Finally, let's consider this from Michael Culshaw-Maurer, a PhD candidate in Ecology at UC Davis.

He writes, "Been working with @SethYefrican (who is now famous) on some analysis of weird #iowacaucases results. Here I outline a bunch of cases where a candidate went from VIABLE in the first round to NOT VIABLE in the final round, which isn't supposed to happen".

This image accompanied his tweet. Click to enlarge it.

If I read this right, the candidate reports boxed in red were viable on the first round but not viable on the second round after realignment (see candidate key on right of graphic for color coding). According to Iowa Party rules, a viable candidate cannot lose viability.

Also as I read this, there doesn't seem to be a pattern as to which candidates are disenfranchised. Even so, this is quite a number of errors. If this data is taken from the state's reporting site as seems likely, it indicates a broad serious problem with the validity of the published results.

It also begs the question, have all Iowa caucus been run this sloppily? 

Bottom Line: Audit the Caucus Results and Publish the Audit

None of this is proof of malfeasance, but it's not proof of innocence either. What it is proof of is that the Iowa caucus results contain a mass of data errors that, deliberate or not, throw the entire results report into question.

If the Democratic Party of Iowa wants to restore its reputation as a reliable broker in the business of democratic elections, it will turn over all data to independent auditors who will publish all of the results — even if it proves Matt Bruenig right, that "there has probably never been a clean [Iowa] caucus."

The alternative is an election, barely begun, that's already tarred beyond recognition, and a public that sees these facts in shock and responds, understandably, in disbelief.