How close was Iowa? Florida 2000 close

Dan Guild walks through the math at the precinct caucus he attended, to show how small shifts can alter delegate counts. -promoted by Laura Belin

At this writing, with 100 percent of Iowa precincts reporting but an unknown number of precincts to be recanvassed, the difference between Bernie Sanders and Pete Buttigieg is 1.5 state delegate equivalents (564.01 to 562.497).

I don’t think any account I’ve read has adequately explained how close this was.

I should stress here that it makes far more sense to report the Iowa results by popular vote, rather than by state delegate equivalents. (Sanders led Buttigieg by 43,671 caucus-goers to 37,557 in the first alignment and by 45,826 to 43,195 after the final alignment.)

But most of the media are reporting the winner based on state delegate equivalents, a practice the Iowa Democratic Party has encouraged.

On caucus night I worked at Urbandale 9, a precinct in a Des Moines suburb with eleven Polk County convention delegates. The numbers looked like this:

Total attendance: 313

Needed for viability: 47

Total County Delegates: 11

First Allocation Final Allocation Delegates before Number of State
Delegate Equivalents
rounding Delegates
Sanders 81 85 2.987220447 3 0.84
Buttigieg 60 88 3.092651757 3 0.84
Warren 65 77 2.706070288 3 0.84
Biden 47 63 2.214057508 2 0.56
Klobuchar 35


Remember Sanders trails by only 1.5 state delegate equivalents across the entire state.

So what happens if we take seventeen votes from Buttigieg and give them to Sanders? Notice Sanders gets four delegates now versus only two for Buttigieg. The results in an additional 0.56 SDE’s for Sanders.

The result: Buttigieg’s statewide lead is cut by a third, to under 1.

First Allocation Final Allocation Delegates before
Delegates State
Delegate Equivalents
Sanders 81 102 3.584664537 4 1.12
Buttigieg 60 69 2.424920128 2 0.56
Warren 65 77 2.706070288 3 0.84
Biden 47 47 1.651757188 2 0.56


Another scenario: let’s take one vote away from Biden.  In this case, he is not viable.  We know from polling that the vast majority of these supporters would have gone to Buttigieg. This is a pretty good guess based on that polling of what would have happened.

The result: Buttigieg’s lead goes up by 0.56 state delegate equivalents.

First Allocation Final Allocation Delegates before
Delegates State Equivalent
Sanders 81 85 2.987220447 3 0.84
Buttigieg 60 125 4.392971246 5 1.4
Warren 65 87 3.057507987 3 0.84
Biden 47 0 0 2 0.56
Klobuchar 35
Others 25


Again, my point here is not complicated.  If Buttigieg loses the state delegate equivalent count, it is very likely that one vote in Urbandale 9 might have made the difference.

Iowa in 2020 was really close.  As in Florida 2000 close.

Editor’s note from Laura Belin: here’s another example of how minute shifts in voter preferences can affect delegate counts. In my precinct (Windsor Heights 2), total attendance was 254, and the viability threshold was 38. Numbers after realignment were 62 for Warren, 61 for Buttigieg, 48 for Klobuchar, 43 for Sanders, and 40 for Biden. Because our precinct allocated seven Polk County delegates, they split two each for Warren and Buttigieg and one for each other viable candidate.

In past years, our precinct had only six county delegates. If that were still the case, each of the five viable candidates would have received one delegate and Warren would have gotten the sixth for having the largest group. Buttigieg would have been down by one county delegate (0.28 state delegate equivalents), despite having only one fewer caucus-goer than Warren.

Or, assuming my precinct has seven delegates, let’s say three fewer people had gone to Biden, making him not viable. Assuming most of his support had realigned to Buttigieg, the mayor could have received three of the Windsor Heights 2 delegates instead of two–adding 0.28 state delegate equivalents to his total.

Now consider what would have happened if Sanders had six fewer supporters in my neighborhood, falling just short of viability. Whether his supporters went mostly to Warren or mostly home, that’s 0.28 of a state delegate equivalent he would not have received from my precinct.

Top image: Photo of the February 3, 2020 caucus in Urbandale precinct 9 and digital photo of the caucus math worksheet provided by Dan Guild and published with permission

Maintenance Notice - As of November 14, 2023 we are still seeing issues with replying to comments...Thanks for your patience, this will be restored.

  • Caucus or Primary

    This comment is not necessarily tied to the specific theme of the post, but I need to ask, What is the objective for Iowa Democrats? If all we want is to cast a vote for a preferred presidential nominee, then go to a primary. But what is done with the convention, then? Who will decide delegates? How will the process for platform proposals change? I remember an article elsewhere about the Republican straw poll in which my opinion was confirmed, i.e., the people who participated in the straw poll were those people who believed that by being involved they could make a difference. The opponents of a straw poll felt it was a useless exercise. Selecting a nominee of a political party is not the role of government, it is the role of the party – and of the people who believe they can make a difference in who the nominee will be. Do we return to the smoke-filled backroom deal making for our nominees? Most people will say no. Do you want a primary where you cast your ballot 29 days before election day and your candidate drops out the next week? Do you want a caucus where, even if your candidate is not supported by other people in your precinct, you can still have another opportunity to influence the nomination? Set aside for a moment all the other items, such as reporting results, the national party, the international media, the specifics of any given precinct caucus. Just look at all the pros and cons of primary v. caucus, including how one or the other may affect who runs for a state-level office, then decide which is preferred. Then worry about being first in the nation and all those other issues. Thank you.

  • This post hurts Iowa

    This post is counterproductive. Breathless analysis of math and margins feeds the critics who forget that in a primary all voters who backed a candidate drawing under 15% have wasted their vote. Come November it gets worse–all people who back losing candidates are effectively disenfranchised in a winner-take-all system.

    Critis compare our caucus to the electoral college because we round off those fractional delegates. They forget that we are giving credit to all the viable candidates whereas the electoral college gives all its credit to just one candidate.

    The only benefit for Pete to claim victory is to help his fundraising and to focus more cameras on him. But he has done no better than Sanders. The Iowa caucus produces leaders, not winners. That is how it should be. .Celebrate our inclusive caucus.with its very popular instant run-off wrinkle for fringe candidate supporters.

    There are never winners. But there were clear losers: Yang, Steyer, and those who merely pretended to be running like Gabbard and Patrick.

  • Florida no, Iowa yes

    Remember how Florida turned out? They needed a “winner” so badly that Republicans staged their mini-riot at the office that was recounting votes. The state Supreme Court got involved only to be overruled by the US Supreme Court. The ballot counting was halted. A dubious winner was installed over the whole country. He soon took us to war needlessly.

    But here we went through sleepless nights because news junkies and talking heads need “winners” even when only a tiny fraction of the delegates to months-distant convention was at stake. That’s the real scandal. Our proportional result vindicates our caucus approach. We now personally know which Democrats in my county are most interested in this election. We will be better off for it even if Bloomberg becomes the nominee.