Part of the changes in the DNC Call for the 2020 Convention and National Delegate Selection Rules were provisions governing the caucus states.  In past cycles, the results in states which used caucuses as their delegate selection process but also used a later non-binding primary showed two things.  First, significantly more people participated in the non-binding primary.  Second, the voters in the non-binding primary had different preferences than those who attended the caucuses.  Additionally, the rules in some of the caucus states created an opportunity for “mischief” at the later levels of the delegate selection process permitting a well-organized campaign to win additional delegates at those later levels and costing a poorly-organized campaigns delegates that they had apparently won on caucus nights.  The new rules attempted to address these “problems”  In particular, Rule 2.K of the Delegate Selection Rules includes requirements that caucus state have a procedure for early or absentee votes in the caucus, have a mechanism to allow participation by those who are unable to attend their local caucus at the time and location set for the local caucus, a means for reporting the “statewide and district level results for each candidate based on the first expression of preference by the participants” in the first level of caucuses; and require that “the allocation of all national delegates, be locked in at the final expression of preference” in the first level of caucuses.  However, Rule 14.B and Rule 14.E seem to suggest that caucus states might still be able a later level as the determining step.  (In primary states, these same rules require using the primary vote.)

In 2016, the Iowa Caucus (held under the old rules) did have a process by which voters could vote absentee via satellite and tele-caucuses but the satellite caucuses only elected three state convention delegates and the tele-caucuses only selected two state convention delegates.  Voters participating in either of these alternative caucuses had no role in the selecting delegates at the district level.  For those who could attend the precinct caucuses, at the precinct caucus, attendees would divide into an initial preference and determine which groups were viable (with a general 15% threshold unless the precinct was electing three or fewer delegates to the county convention).  After the initial count, attendees would have the opportunity (based on which preferences were viable or close to viable) to change their preference.  The precinct chair would report the results of this second count to the state party in terms of “state delegate equivalents” and would not result raw votes.  The delegates selected at the precinct caucuses would attend the county conventions where a similar process would occur to select the delegates who would be attending the congressional district convention and the state convention.  A similar process would again occur at the congressional district conventions and the state convention to determine the allocation of the national convention delegates selected at those conventions.  For multiple reasons (the possibility of delegates elected at precinct caucuses and county conventions not attending later conventions, the possibility of changes in preference of such delegates, delegates pledged to withdrawn candidates choosing between the remaining candidates, and the fact that each delegate chosen at a precinct meeting was a fraction of a state delegate and those fractions would be converted to whole numbers at the county convention), the report of the state delegate equivalent only provided a rough estimate of the national delegates that each candidate was likely to receive from Iowa.

We now have a draft of the 2020 Delegate Selection Rules for Iowa.  (Of course, these rules still have to go through a public comment period, be finally approved by the Iowa Democratic Party, and by approved by the Rules and By-laws Committee of the Democratic National Committee before becoming final.)  This draft gives us a first look at how the caucus states might change their state rules to comply with the new national rules. 

First, the satellite caucus and the tele-caucus have been replaced by virtual caucuses.  There is still no absentee voting in the precinct caucus and individuals must actually attend their precinct caucus to participate in that caucus.

Second, the precinct caucuses will now record the vote from the first vote of the attendees and report that result to the state party.  For the purposes of electing delegates to the county convention, however, the same threshold rules still apply and attendees will have the opportunity to change their preferences after the first vote.  However, only those in “non-viable” preference groups can change their vote after the first round.  (In other words, if your candidate got 14%, you can either move to another candidate or try to recruit supporters from other candidates who missed 15%.  However, if your candidate got 40%, you can’t change your preference to help another candidate become viable.  If, in a small precinct, there are more viable preferences than delegates, the smallest of the preferences is deemed non-viable and those attendees must switch to the viable preferences.)  These final preferences will also be reported to the state party. 

Third, there are six virtual caucuses (scheduled at different times and dates in the last week leading up to the caucus, including one that is simultaneous to the actual precinct caucuses).  Participants in the virtual caucuses must pre-register to participate and will not be eligible to participate in a precinct caucus.  The caucuses will either be tele-caucuses or other on-line mechanism.   Each of the virtual caucuses will use preferential voting (i.e. the participants will rank their candidate preference — giving at least her top five candidates — including uncommitted).  These preferences will be used to allocate an additional pool of delegates (representing 10% of the delegate totals at the district and state conventions).  The first choice preference for each candidate will be counted and (starting at the bottom) the votes for non-viable candidates will be reallocated until only viable candidates remain.  At that point, the pool of delegates for the district and state conventions will be allocated with the campaigns designating individuals to serve in those delegate positions.  Viability will be determined by congressional district.  (Based on the 2016 allocation of state convention delegates, the number of delegates selected using the six virtual caucuses will range from 28 or 29 in the Fourth District to 38 or 39 in the First District with a total of 140 delegates state-wide.)

Fourth, while the precinct caucuses will report both the raw vote total from the first expression of support and the final expression of support, they will still also report the state delegate equivalents from their caucuses.  The allocation of national convention delegates will be based on the total state delegate equivalents (both at the district conventions and at the state conventions) from the precinct caucuses and the virtual caucuses. 

Fifth, candidates can ask for a review/recount of the results from an individual precinct (i.e. for miscalculation of the allocation of delegates or mistakenly determining viability threshold) or at the district or state-wide levels (misadding the results from the precincts).  Any request must be made shortly after the caucuses (by February 7) with the final result to be issued by the end of the month.   

Sixth, at the district and state conventions, delegates chosen by the county conventions who support candidates who did not win any national convention delegates at that level can opt to switch to a viable candidate (and thereby have a vote in choosing the national convention delegates).  Such changes in support will not have an impact on the total number of national convention delegates won by each candidate.  (Similarly, if a delegate supports a candidate who is not viable at the county convention can switch to a viable candidate for the purpose of electing district and state convention delegates.)

Taking a first glance at these rules, three things stand out.  Most significantly, the rules are still based on equivalent delegates won on caucus night (or in the virtual caucuses.)  Raw vote (either initial preference or final preference) still does not matter.  Thus, voters in precincts that have low attendance per state delegate equivalent will have a bigger impact than voters in precincts with high attendance per state delegate equivalent.    

Additionally, while better than in 2016, the allocation of delegate equivalents is still weighted heavily in favor of the precinct caucuses over the virtual caucuses.  (Another reason for favoring raw vote over delegate equivalents.)  While a virtual caucus is not quite the same as a primary (as it still takes more time to vote in a virtual caucus than at a primary), in 2016, the vote totals from the primary were slightly more than double the vote totals from the caucus.  Giving the virtual caucuses only 10% of the delegate slots at the state convention still seems to give the participants in the virtual caucuses less weight than they should have.

Third, by using state delegate equivalents, the rules still seems to be slanted against marginal candidates and in favor of the leading candidate.  A candidate who receives 15-17% of the first preference raw vote in a district (or state-wide) will probably get 20% in some precincts (including potentially in the district-wide vote after second preferences are included) but 12% in other precincts.  As a result, such candidates who would win delegates based on the raw vote numbers, may find themselves with only 10-11% of the state delegate equivalents and thereby win no delegates.  In past caucuses, there have been some candidates who were strong enough to win delegates in some precincts with several notable candidates breaking 10% in state delegate equivalents but not the 15% needed to actually win national convention delegates (Birch Bayh in 1976, George McGovern in 1984, and Dick Gephardt in 2004).  Particularly, if there were a 10+ candidate field, I could see the top two candidates getting in the  low 20s by raw vote but getting near 40% in state delegate equivalents as the third, fourth, and fifth place candidates only manage to break the 15% threshold in some precincts even while breaking or nearing that level in the state-wide raw vote.

I am not sure that the current round of changes will alter the actual results that much.  If the national party were to require Iowa to use the first preference votes (or even have the attendees at the precincts cast their first vote using preferential voting and then use preferential voting to determine who wins the delegates), that could make a big difference in results (as well as helping the middle of the pack candidates survive longer by more accurately reflecting the gap between the first tier and the second tier).  The current rules changes, however, should not make a big difference in the final results but will lock them in after caucus night.  Since Iowa only has 41 pledged delegates to the national convention, it is unlikely that this change will impact more than 4 or 5 delegates in the final count.