The Nevada caucuses — the first major vote of the Democratic presidential nominating contests in the western United States — will be held Saturday. The results could give some candidates significant momentum. Or, if the caucuses play out like those in Iowa earlier this month, the results could lead to prolonged confusion and calls for recounts.
The Nevada Democratic Party has been scrambling to put in effect safeguards and redundancies in its caucuses to avoid the myriad issues that plagued Iowa.
Here is how the caucus process works in Nevada, along with the lessons from Iowa that show us what could potentially go wrong.
Early Voting and ‘Strip’ Caucuses
Some 75,000 Nevadans have already cast ballots, ranking their top three to five candidates in order of preference. Those choices will come into play in every step of the caucus process.
The volunteers who run the caucuses Saturday will be given data from the early votes and will essentially have to pretend the early voters are present. (More on this later.)
In addition to early voting, some of the caucuses will be held in hotels on the Las Vegas Strip to accommodate the city’s many shift workers.
At the Caucus Sites
Hundreds are expected at some caucus sites in Nevada. Keeping track of the people, combining their choices with those of early voters and accurately converting those votes into county delegates could prove difficult.
Here is how the basic process works.
Captains assemble: As a caucus begins, a designated precinct captain from each campaign will stand in an assigned spot, holding a sign to show where a candidate’s supporters should gather.
Caucusgoers arrive: They sign in and receive a preference card, which lists the candidates they can vote for. Volunteers add up the number of participants and early voters to get the total number of caucusgoers for that precinct.
First alignment: Caucusgoers congregate around the group representing their first-choice candidate.
Initial count: The caucus chairman and volunteers count the supporters for each candidate, both in-person supporters and early voters. Groups with support from at least 15% of all participants are considered “viable” and can earn delegates.
In Iowa, some viability miscalculations meant that candidates who did not have enough supporters still received delegates.
Shuffle: After results of the first round are announced, the viable groups can try to persuade those in nonviable groups to join them. If an early voter’s first choice is not viable, they are automatically moved to their next ranked candidate with a viable group. (Caucusgoers in viable groups remain with their first choice.)
Final alignment: The resulting groups are counted again, and each group’s share of the total vote is calculated.
In Iowa, some groups that did not meet the minimum viability threshold made it to this stage.
Volunteers must record their counts two ways: on a “caucus math poster,” a large worksheet that is to be hung on a wall, and in a “caucus calculator,” a Google Forms application on an iPad provided by the state party. The app will check the numbers for errors and will help the volunteers calculate viability as they go through the alignment process. In the next stage, it will also help to calculate delegates.
When the final alignment is finished, the vote share of each group must be converted into a number of county delegates: people who will attend the upcoming county conventions.
In Iowa, some precincts made calculation errors that led them to award the wrong number of delegates to candidates in some cases. This year, for the first time, both Iowa and Nevada must release raw vote totals — the number of people in each candidate’s group in the first and final alignments. This means that caucus math can be checked by the public.
To understand the math, imagine a precinct where there are 100 caucus participants, including early voters and those physically present. Let’s say this precinct is responsible for electing eight delegates to the county convention. If there are four candidate groups in the final alignment, each with exactly 25 members, then they each get a quarter of the eight delegates — two delegates per candidate.
Of course, the numbers usually aren’t so even. Let’s say one of the groups had 16 members, or 16% of the 100 participants. Their share of the eight county delegates, 16% of eight, would be 1.28.
County delegates are real people, and precincts cannot send fractions of people to the county conventions, so the resulting numbers have to be rounded. In this example, 1.28 would round down to 1, and the group would receive one delegate.
Because of this rounding, precincts can end up with too many or too few total delegates awarded to candidates. In those cases, the math that determines which candidates should gain or lose delegates can result in a tie.
What if There’s a Tie?
In Iowa, these situations were handled with a coin toss. In Nevada, they are handled with a card game.
The state party provides each caucus site with a new, unopened deck of cards. In the event of a tie, jokers and instructional cards are removed and the deck “should be shuffled by a precinct chair or site lead at least seven times before use,” according to a memo released by the Nevada Democratic Party. “The high card determines the winner, and aces are high.”
When the counting and calculating is complete, volunteers will submit the results to the Nevada Democratic Party, where they will be tabulated and released to the public. Volunteers will submit results four ways: via the Google Forms app on the iPad, by phone to a hotline, through a text message to the state party with a photo of the paper worksheet and with a paper backup of that worksheet collected at the caucus sites.
In the Iowa caucuses, the first signs of trouble came when volunteers tried to submit the results of their precincts. The plan was for most precincts to send their results via an app, but many either chose not to or found that the app did not work. Volunteers could use a hotline to call in results, but it was quickly overwhelmed.
The people taking the calls were writing down the results on paper and handing them off to be entered into a spreadsheet. In that process, various errors occurred, and some of the results that were initially released to the public had to be corrected.
The four reporting methods in Nevada are intended to create a robust paper trail and to avoid many of the missteps made in Iowa. Nevada Democrats had planned to use the same reporting app but quickly came up with a new system after the app proved faulty.