Since 2007, I have been trying to raise awareness about problems with the Iowa Democratic caucus system: barriers to participation for many who want to have a voice in choosing our president; the fact that not every Democrat's "vote" counts the same toward the delegate numbers; the lack of secrecy and potential for intimidation in caucus rooms; and the distorting effects of caucus math, starting with but not limited to the 15 percent viability threshold.
Again and again and again, I have urged my fellow Democrats to make our caucuses more inclusive and the results more representative of each candidate's actual supporter numbers. I might as well have been pounding on the walls of a sound-proofed chamber, for all the impact my blogging has had on the Iowa Democratic Party's leadership.
The slimmest margin ever between the top two Democratic presidential candidates lends new urgency to the task of cleaning up the caucuses. Yet state party chair Dr. Andy McGuire and others are holding the line against even a full review of the calculations that put Hillary Clinton ahead of Bernie Sanders by 700.59 state delegate equivalents to 696.82. They insist that any minor glitches on Monday night were resolved by 2:30 am, when the party sent out its press release declaring Clinton the winner. The results are final. Nothing to see here, folks.
The longer party leaders drag their feet, the more they will stoke conspiracy theories about the caucuses being stolen for the establishment's favorite.
Consider the logistical challenges with organizing 1,681 precinct caucuses, electing 11,065 county delegates, which must be converted to 1,406 state delegate equivalents.
Now add some inexperienced precinct chairs. A personal friend got roped into doing this job at the last minute and received virtually no training.
In all likelihood, well-intentioned human beings will make some mistakes. We had to count our Clinton supporter group twice in my precinct, because the first time the numbers didn't add up to the right total for our caucus as a whole. This video shows a Des Moines precinct where the Clinton captain appears not to have bothered to do a full count after the second division into preference groups, and the precinct chair let it slide.
Even assuming the supporter groups for Clinton and Sanders were counted properly in every location, we should expect occasional miscalculations by precinct chairs, especially ones who have never done the job before. Caucus math can be complicated.
Sanders told an MSNBC reporter that he hasn't ruled out contesting the caucus results: "We haven’t had the time to analyze it, but our people in Iowa are taking a look at that." The Sanders campaign also wants the supporter numbers from each precinct made public (which I've advocated for years). Iowa Democratic Party spokesperson Sam Lau responded in a statement,
"The Iowa Democratic Caucuses are not a primary – candidates are awarded delegates, not raw votes. In fact, because of our realignment process, raw vote totals are not recorded. As we have always done, and as we have always told the campaigns we would do, we have released state delegate equivalents earned by each candidate."
When I sought comment on the specific irregularities the Sanders campaign has raised in its communications with party leaders, Lau told me,
"On caucus night, each of the campaigns had a representative in our tabulation room, because it was important to us that we were inclusive with our reporting, and that they had the opportunity to bring any concerns to us directly. We worked with all three campaigns on resolving various issues, and we reported verified final results for 100% of our precincts."
McGuire gave the same party line in an interview with the Des Moines Register.
"The answer is that we had all three camps in the tabulation room last night to address any grievances brought forward, and we went over any discrepancies. These are the final results," she said.
You can't expect the whole world to accept "verified final results" that were calculated in a matter of hours, producing a margin of victory well below 1 percent.
I think highly of McGuire and do not believe she tried to fix the caucuses for Clinton. But having been a high-profile Clinton endorser before the 2008 caucuses, she must understand how it looks for her to reject entirely reasonable calls for a precinct by precinct review of results. Tabulation and data entry errors are facts of life. Lambert Strether wrote yesterday at the prominent website Naked Capitalism, "the DNC and the state Democratic Party organizations have a direct, financial interest in the Clinton candidacy. So if Iowa reminded you of Ohio in 2004, perhaps there’s a reason for that." Today, he followed up by commenting, "let us remember that party chair Dr. Andy McGuire has a “known allegiance” to Clinton, and is therefore conflicted; she is in no sense a neutral arbitrer or honest broker and should recuse herself from this issue."
The lead editorial in today's Des Moines Register rightly notes,
What happened Monday night at the Democratic caucuses was a debacle, period. Democracy, particularly at the local party level, can be slow, messy and obscure. But the refusal to undergo scrutiny or allow for an appeal reeks of autocracy.
The Iowa Democratic Party must act quickly to assure the accuracy of the caucus results, beyond a shadow of a doubt.
First of all, the results were too close not to do a complete audit of results. Two-tenths of 1 percent separated Bernie Sanders and Hillary Clinton. A caucus should not be confused with an election, but it’s worth noting that much larger margins trigger automatic recounts in other states.
Second, too many questions have been raised. Too many accounts have arisen of inconsistent counts, untrained and overwhelmed volunteers, confused voters, cramped precinct locations, a lack of voter registration forms and other problems. Too many of us, including members of the Register editorial board who were observing caucuses, saw opportunities for error amid Monday night’s chaos.
The Sanders campaign is rechecking results on its own, going precinct by precinct, and is already finding inconsistencies, said Rania Batrice, a Sanders spokeswoman. The campaign seeks the math sheets or other paperwork that precinct chairs filled out and were supposed to return to the state party. They want to compare those documents to the results entered into a Microsoft app and sent to the party.
I don't believe there was any grand scheme to rig the caucus results. The vast majority of precinct chairs did a thankless job well, despite sometimes getting shouted down by Sanders fans who didn't understand how supporter numbers convert into county delegates. By the way, caucus math distortions didn't favor any one candidate across the board; in this West Des Moines precinct, twelve extra bodies in the room for Sanders gave him twice as many county delegates as Clinton.
That said, questionable actions by precinct chairs are not unheard of. In this video shot by Matthew Palevsky during a 2008 caucus, the precinct chair and secretary cut off efforts by Joe Biden supporters to find the one additional person they need to become viable. The incredibly close delegate counts for Clinton and Sanders magnify the significance of any similar mistakes by precinct leaders this year.
The sooner the better, should agree to a full recount of this year's caucus results. In the longer term, they should think seriously about reforms that could improve the integrity and fairness of the caucus process.
I understand that a caucus is not a primary, but I don't understand why Iowa Democratic Party leaders are so opposed to running a straw poll like the Republican Party of Iowa does at its caucuses. Theoretically, a candidate can win a plurality of delegates from a Democratic precinct even if a larger number of caucus-goers walked in the door intending to support someone else. What's wrong with telling the public how many Democrats stood in each candidate's corner? When I have asked that question over the years, I have repeatedly been told that New Hampshire's secretary of state will not allow the Iowa Democratic Party to release raw vote totals, because that would look too much like a primary. Why not follow the Iowa GOP's path, and release numbers from a straw poll that is not tied to county delegate selection? Clearly New Hampshire does not object to that procedure.
Unprecedented situations call for unprecedented actions. Otherwise Iowans may have just experienced our last first-in-the-nation caucuses.
P.S.- The Facebook status posted today by former Johnson County Auditor Tom Slockett is a good illustration of how many Sanders supporters view the Iowa caucuses now:
The Clintonites are trying to make a purse out of a sow's ear. Forget the coin flips, they are irrelevant. The problem is that significantly more Democrats supported Bernie than Hillary at the Iowa caucuses. This information is readily available and was accepted by each caucus as it was necessary to use in calculating the number of delegates apportioned to each delegate. It is the numerator in the fraction comprising the (number supporting each viable candidate) / (number supporting all viable candidates) X (number of delegates assigned for the precinct). So the number supporting each candidate in each precinct exists. This is far and away the single most important piece of information produced by each precinct. Add this number together with the same number from each precinct for each candidate and you have the total number of Democrats who caucused for each viable candidate. But this number is being suppressed by the State Party. Why? They say it is because that is the way it has always been done, the oldest excuse in the book for covering up a mistake. The real reason, because the State Party is controlled by the Clintonites and the billionaire donator class and they are covering up the fact that more Democrats in Iowa supported Bernie at the caucuses than supported Hillary. This is a fact they know full well the public deserves to know and as Clinton supporters they are determined to cover it up. This is the truth and those are the sad facts about the state of affairs of our Iowa Democratic State Party.
UPDATE: McGuire wrote an op-ed for Friday's Des Moines Register. She sounds determined to keep her head in the sand.
These are not contests of popular raw votes — nor do we think they should be. If that was the case, our candidates would spend all of their time in Des Moines and Cedar Rapids, and not in small towns like Manchester and Carroll, where Sen. Sanders and Secretary Clinton traveled, respectively, in the final days before our caucus, to confront the diverse concerns of rural Iowa.
[...] Asking for raw vote totals demonstrates a misunderstanding of our process. As does asking for a recount. [...]
That being said, we continue to work with the campaigns, as well as our local party leadership, on any questions they have on our reported results on a case-by-case basis. [...]
These few questions they are raising should not obscure the fact that we took more steps than ever before to ensure the accuracy of our results.
And Iowa Democrats will remain adaptable to any changes that can improve on the process while preserving what makes the caucuses special. That’s why we worked to develop the new reporting app. That’s why we hired our largest caucus team ever who conducted the most widespread trainings in our party’s history. And that’s why we listened to the criticisms we have heard in the past on barriers to participation and pro-actively responded this year by launching our first-ever tele-caucus and satellite caucuses. [...]
In that same spirit, we will thoroughly review the concerns we heard this year on space issues and night-of difficulties and work to form appropriate solutions. We believe it is always important to improve, and we’re listening.
But we believe it is important to recognize the intricacies of a Democratic caucus, and to celebrate the enthusiasm we saw from impassioned Democrats all across our state who showed once again why Iowa is first in the nation. The party is stronger because of Monday’s caucus, and we look forward to continuing to build upon this strength from the grassroots on up.
First, the satellite and tele-caucuses were a nice gesture, but with fewer than 300 people signed up in advance and fewer than 200 people participating on February 1, they are hardly a real solution to the problems keeping thousands of Democrats from having a voice in the presidential nominating process. I could probably find 200 politically engaged Democrats who weren't able to caucus in my little suburb of Windsor Heights alone. Exactly one retirement community in this state was a satellite caucus location. How many similar facilities and nursing homes does Iowa have?
Second, I reject the premise that being interested in raw supporter numbers means a person doesn't understand the caucus process. That sounds like a cover for the Iowa Democratic Party not wanting to expose the distortions that can happen when the views of scores or hundreds of people are reduced to a small number of county delegates, and Democrats in low-turnout areas have more influence over state delegate equivalents than their counterparts in political hot spots. A full review may indicate that the party correctly calculated the number of state delegate equivalents for Clinton, Sanders, and Martin O'Malley, and that more Democrats came to their precinct caucuses intending to stand for Sanders. (Contrary to what Slockett wrote, we can't jump to that conclusion yet, since Sanders did better in smaller counties than many people expected.) Such information would be embarrassing for the state party but instructive as leaders consider reforms that would give every Democrat equal voice in future caucuses.
Third, despite "widespread trainings," some precinct chairs were unprepared for the task. Many people from different precincts related various problems to me. This report by Ryan Foley for the Associated Press includes more anecdotes.
Fourth, and more troubling from my perspective, counting large groups of people in over-crowded rooms is a difficult process to manage even for an experienced and competent precinct chair, like the one in my precinct. The state party may not want to hear about it, but many Democrats were discouraged by the chaos they witnessed at their caucuses and have been complaining on social media. Here are a few examples of comments I have seen posted since Monday night:
• "I know that Iowans are in love with the caucus. It was explained to me to be this great, wonderful thing, maybe a little crazy, but still awesome. But frankly, I am still in a daze and trying to process what the hell happened. I don't know that I want to participate again."
• "we had 2 exits out of an elementary school cafeteria with 400 people crowded around 10 rows of lunch tables. Cooperation in counting was challenging enough, getting that many people out without trampling one another? There was also another precinct crammed into the library. I realize we have no idea how many people will show up. It just seems to be another reason to change this process."
• "The caucuses have always been a mess, but with high turnouts they only get worse."
• "I was at the [...] site where we had to stand outside for an hour. There were people with babies and young children, elderly people, people on crutches or walkers. Once we went back inside, we were crammed in so tight that I saw one woman have an anxiety attack."
Fifth, to suggest the Democratic Party is stronger now than last week is wishful thinking. Any lingering questions about the caucus process or results can only hurt efforts to energize the grassroots for the general election.
SECOND UPDATE: I forgot to mention that statewide candidates routinely campaign all over Iowa before general elections, where the winner is determined by popular vote. Republican presidential candidates don't stick to a few high-population cities either, even though the GOP caucuses are are straw poll rather than a battle for delegates.
THIRD UPDATE: Democratic volunteer John McCormally defends the caucus system against "misplaced" criticism at Iowa Starting Line. Several of his points don't apply to me; for instance, I have been pushing back against the coin toss conspiracy theorists on social media. But a few are relevant to this discussion.
A caucus does not record "votes" for each candidate. The "accounting" some people are demanding is the one used for delegate apportionment and it cannot provide an accurate summary of the caucus results. That is because caucuses involve alignments, and then realignments, before delegates can be selected. For example, most precincts recorded no "voters" for Martin O’Malley. However, O'Malley had some initial support nearly everywhere–just not the 15% needed for viability. O'Malley's numbers were never recorded because only viable preference groups were counted. O'Malley voters simply folded into other groups or left. In the end, delegates are awarded not based on raw votes, but with a delegate formula tied to county convention delegate allotment. In a caucus, it is delegate numbers that truly matter, and those are publicly available.
Precinct chairs should have record the size of all preference groups after realignment, which were used to calculate the county delegates assigned. Although those numbers will not show how many Iowans came to caucus for O'Malley, they will show whether precinct chairs did the math correctly. I assume that nearly all of them did, but some honest mistakes may have occurred. That's one reason I support releasing the numbers. If the Iowa Democratic Party does not have the numbers for every precinct (because some chairs dropped the ball on submitting them by paper or electronically using the new app), officials should attempt to get those numbers by contacting the chair or secretary from the precincts with missing data.
McCormally also notes the difficulty of securing adequately-sized rooms, especially since some school districts denied the parties' requests to use their buildings. My purpose is not to assign blame for scheduling, but to point out that it's unwieldy to count hundreds of people crowded in a small space. It would be easier to count ballots, as Republicans use during their caucuses.
[T]he 325 people in that room [in his precinct] caucused for their preferred candidate. Of those, over 200 had to fill out voter registration forms, indicating they were new caucus-goers. The turnout was twice what both the Sanders and Clinton campaigns expected in our precinct. That new-voter enthusiasm tells you that people are hungry for democratic participation—but also explains why there are so many criticisms of the process. Caucuses are not primaries—they are party building events rife with idiosyncrasies. If you’ve never participated, they seem weird. They don’t always lend themselves to clear results, and this one in particular doesn’t fit into our cultural obsession with picking a clear winner and loser. Caucuses are nuanced, and nuance doesn’t translate to 10 second soundbites and 140 character tweets.
The reality is that Democrats had two very strong candidates who head out of Iowa with virtually equal support. That’s not something to be investigated, its something to be celebrated.
The Iowa Caucuses are special—but they were never built for the horse race journalism that pervades today’s media. At the end of the day, the caucus is for assigning delegates and building a party. We had two strong candidates who were assigned virtually equal delegates. Over 170,000 Iowa Democrats debated the future of our country by standing with their neighbors in 1,681 neighborhood precincts spread over 99 counties. That’s pretty amazing. Kudos to the entire Iowa Democratic Party staff–I’m proud to have been part of it.
Let's not pretend there is some virtue in "idiosyncrasies" that sometimes mean Democratic caucuses don't produce "clear results." Clinton and Sanders fought a hard race and finished in a virtual tie. My problem is not the closeness of that contest, it's the fact that the county delegate apportionment sometimes distorts the real preferences of people in the room. "Nuanced" caucuses seem less valuable to me than the Iowa Republican approach to caucusing. At least we know how many Republicans showed up in every precinct intending to support each candidate.
Moreover, I cannot disagree more strongly with the notion that the caucuses "were never built for the horse race journalism that pervades today's media." As the Des Moines Register's Jason Noble discussed at length in his "Three Tickets" podcast series, the caucuses became significant as a media event and would never be what they have become without the intense focus of national political reporters.
As for the enthusiasm we saw all over Iowa on February 1, I'm always happy to see new people engaged in the political process. I would celebrate more if I didn't know that thousands and thousands more Iowans who wanted to participate were not able to attend their caucuses. Here's a comment someone posted on the Bleeding Heartland Facebook page.
My wife and I were all set to caucus, but our 17 month old needed urgent care and we were unable to make it in order to take her to the doctor. With a primary, we could have easily voted much earlier in the day instead of not participating at all.
Caucus was cool when it was small. But it's too big for what it was designed for. With 350,000 voters total (across Dem and GOP) - it's time for a primary.
Another person who has volunteered extensively with Democrats in her county added on the same thread, "It has been my experience that caucuses do not build up local parties. The professional politicians move on, but the locals carry grudges for a long time."
FOURTH UPDATE: Radio Iowa's O.Kay Henderson reported after the February 5 taping of Iowa Public Television's "Iowa Press",
The man who managed President Obama’s 2012 reelection effort in Iowa expects Iowa Democratic Party leaders will convene a commission to propose possible changes in the party’s presidential caucus procedures. Brad Anderson says Democrats should consider having straw poll ballots, just like Republicans do at their Caucuses.
“The one thing that is troubling is in a close race there should be some kind of recount procedure and right now because of the way it’s done on the Democratic side, it’s just not possible,” Anderson says. “We don’t have ballots.” [...]
“We need to form a commission that takes a look at it in the same way that Republicans did in 2012,” Anderson says. “I think there are very fair questions, but I think at the end of the day this process is one that makes Iowa the center of attention and we have always handled the spotlight fairly well.”
Anderson was the 2014 Democratic nominee for Iowa secretary of state, so he has put a lot of thought into fair election procedures.
FIFTH UPDATE: Clinton's Iowa campaign director Matt Paul unloaded in a blog post against efforts by the Sanders campaign "to poison an otherwise historic night for Hillary Clinton, our supporters, and volunteers." He notes "conspiracy theories" floated by campaign officials before the caucuses and false claims that ran wild on social media the night of February 1. Paul then points out,
For the Sanders campaign's challenges, there are 7 total precincts they are questioning (out of a whopping total 1,681 precincts!) and the Iowa Democratic Party is checking the results on a few more.
And they will follow the same standard procedure for these questions as they did for precincts we asked them to look at — they'll look at the numbers reported that night, check the math and figure out if there are any discrepancies or misallocation of delegates.
Here's the kicker.
Even if the Sanders campaign were to win all of their challenges — the marginal bump they receive in support would not be enough to overcome Hillary Clinton's win margin. Those are just the facts.
Here's the math.
Total Precincts in Contest: 11 (7 from Sanders and 4 from the Iowa Democratic Party)
Total State Delegates being Contested: 1.7947 (not enough to change outcome)
If every contest were to go in the favor of Sanders:
HRC = -.775
Sanders = +.758
MOM = +.0166
To be clear, my goal is not to poison Clinton's narrow win on state delegate equivalents. I just want the Iowa caucuses to be more inclusive and representative going forward.
SIXTH UPDATE: Jennifer Jacobs reported for the Des Moines Register on February 5 that results from one precinct in Grinnell were reported as 18 delegates for Sanders and 8 for Clinton, when it should have been 19 to 7. When the Iowa Democratic Party incorporates those changes, Sanders will gain .072 state delegate equivalents--not enough to alter the outcome.
The case shows that 1) small tabulation errors are to be expected across more than 1,000 precincts, and 2) systematic fraud or misreporting of county delegate totals would not be possible, because there are so many witnesses to what happens in Democratic caucus rooms. Several people who attended in that Grinnell precinct noticed the discrepancy on the Iowa Democratic Party's results website.