In less than four days, voters in Iowa will head to some location in their precincts and cast the first official votes of the 2016 presidential campaign.  Both because of its small size and because of the unique compositions of the respective parties in Iowa (compared to the national parties), winning in Iowa is not essential to winning either party’s nomination.  What does matter is how Iowa sets up the rest of the race.

The first thing to note about Iowa is two key related issues — turnout and weather.  Most polls for both parties show significant differences between the preferences those who have attended a caucus in the past and those who will be attending their first caucus.  By definition, every election has a group of potential votes who were too young to vote in the last election or lived in another state.  For these voters, this year’s caucus is their first chance to attend a caucus.  But every cycle there are other voters who, even though they could have participated previously, were simply not interested enough to take the time (and it does take time) to attend a caucus, but this time are interested.  Guessing who will win requires making a guess at how many first time attendees will show up to participate.

Part of this guessing process is the weather.  Unlike a primary state where voters can drop by the polling place at any time over a twelve or more hour voting period, a caucus state requires a voter to be able to head to their local caucus in time to get their before 7:00 p.m. and stay for an extended period.   Depending upon the weather and how risky the drive is, voters might just bunker down at home rather than brave the elements and icy roads.  For now, the weather forecasts for Iowa seem to be good with snow scheduled to start after the caucuses are done.  Recognizing some of the problems with making a caucus meeting, the Iowa Democratic Party has authorized satellite caucuses and telecaucuses, but the satellite caucuses will only allocate three state convention delegates and the telecaucuses will only allocate two state convention delegates (pretty much guaranteeing a 2-1 split out of the satellite caucsuses and a 1-1 split out of the telecaucuses and that one extra state convention delegate will probably have no effect on the allocation of national convention delegates).

Putting aside the issues of turnout, the real battle in Iowa is one of perception and media coverage.  The delegates selected out of Iowa represent about 1% of the total delegates to the national conventions of each party, and proportional representation means that no candidate will leave Iowa with a significant lead in the battle for pledged delegates to their national conventions.  As such, what matters is how the media conveys the significance of the results.  The media coverage tends to emphasize the surprising result; so how things change over the next 96 hours will matter as much as or more than who actually finishes first.  Sometimes the candidate who narrowly loses gets more help out the results than the candidate who narrowly wins.  Candidates need results that allow them to claim a “mandate” to continue in the race more than they need to actually finish first.

On the Democratic side, the actual results are unlikely to have any significant impact unless Secretary Clinton or Senator Sanders wins in a landslide or Governor O’Malley shows some unexpected sign of life.  Both Senator Sanders and Secretary Clinton will almost certainly leave Iowa with a result that sends them onward to New Hampshire, Nevada, and South Carolina.  Governor O’Malley is likely to have a result that signals that the end is near.

In discussing the Democratic results, it is important to remember that (except perhaps for the satellite caucuses and telecaucuses) the Iowa Democratic Party will not have a count of the total votes, but rather a count of the delegates to the county conventions that each candidate wins.  Each precinct has a different number of delegates to the county conventions (with the county conventions also having different sizes) ranging from one delegate to twenty-five delegates.   As we learned nationally in 2008, the delegate selection process rewards having additional strength in the right location.  In a precinct that has an odd-number of delegates, finishing in first by one vote gets you an additional delegate.  In a precinct with an even number of delegates, there is no difference between getting 52% of the vote (or higher in smaller precincts) and getting 48% of the vote.  Because the number of delegates that each precinct has to their county convention is already set, it does not matter if the attendance at a precinct’s caucus is 50 Democrats or 100 Democrats.  The information reported to the media will not indicate how many attended any particular precinct caucus.  Additionally, you need 15% support in a precinct to get any delegates.  Governor O’Malley could exceed expectations and have support nearing 10% in the typical precinct and still have 0 county convention delegates at the end of the day.

The Republican contest, on the other hand, will announce raw votes and the media has a lot more ability to make a difference by how they spin the results (in light of the large number of candidates).  Barring a big swing at the last second, it looks like Trump has a narrow (but statistically significant) lead of Senator Cruz.  If either wins in a landslide (10% or more), they would get favorable coverage from Monday’s results, with the other candidate’s ability to win nationally being called into question.  A close win for either (5% or closer) would probably result in converage that treats them both as winners with little impact on the following states.

The bigger story, however, will be about third place.  Currently, it looks like Trump and Cruz will get a combined total of around 60%, leaving the rest to fight over 40%.  If somebody (most like Senator Rubio) can get over 15% and put some distance between himself and fourth place, that candidate would get significant media attention as the third (and last) “real” contender for the nomination.  That media attention, in turn, might allow that candidate to break the current four-way tie for third in New Hampshire, potentially turning this race into a three-person contest.  If there is a close battle for third place with somebody like Carson and the would-be establishment candidate barely breaks 10%, the media attention will almost certainly stay on the top two, and a third place finish would have almost no impact on New Hampshire creating the possibility that none of the wanna-be establishment candidates drops out any time soon, allowing Trump and Cruz to dominate a splintered Republican field for the next several weeks.

Given the closeness of the numbers, I am reluctant to make any big predictions of what happens on Monday or what happens next.  My expectation is that Governor O’Malley will probably not drop out until after New Hampshire.  I think that Governor Huckabee and Senator Santorum (and maybe Fiorina and Gilmore) drop out after Iowa.  For the rest of the marginal Republican candidates, the key state right now is New Hampshire.  None of the candidates in the Republican field need to take third in either Iowa or New Hampshire, but they need to be close enough to third (3-4% in one of the two states) to be able to make an argument that they can move into the top three by mid-March.