Thursday, March 6, 2014

An SEC Primary in 2016? Not so fast… (Part II)

A couple of weeks ago FHQ examined the likelihood that the states most closely associated with Secretary of State Brian Kemp's (R-GA) southeastern regional primary proposal would be able to implement a presidential primary move. That was more of an internal look at what may affect the calculus in each state. The post touched on outside factors that may affect that decision-making process, but only in passing. Obviously, there are other matters that may intervene to complicate things.

There is some history here. The idea of a southern regional primary is not a new one. Barely a year after the reformed presidential nomination process got its first trial run in 1972, Jimmy Carter was out laying the groundwork for a nomination bid on the Democratic side in 1976 but was also trumpeting the strategic virtues of holding a collective southern regional primary. The benefits seemed clear. The South would speak with one voice and propel a more moderate-to-conservative candidate to the Democratic nomination who could, in turn, better compete in the general election.

As it turned out, it took the states of the South a decade and a half to coordinate this, bringing the idea to fruition. It took some cajoling from the Carter folks ahead of the 1980 renomination run against Ted Kennedy to convince legislators in Florida to hold pat in March and legislators in Alabama and Georgia to move up to coincide with the primary in the Sunshine state. That subregional primary was to serve as a counterweight to the delegate gains Kennedy was likely to win in New Hampshire and Massachusetts.1

Four years later, several southern and border states adopted caucuses for the competitive Democratic nomination race, joining Alabama, Florida and Georgia in March, though not all on the same date. Only the Oklahoma Democratic caucuses were on that same second Tuesday in March date. Caucuses in Arkansas, Kentucky, Mississippi and South Carolina followed later in the month. Collectively the South spoke with something approximating a single voice, but the result was not support for a more moderate candidate.2 Rather, it was support for Walter Mondale.

There was, then, no alignment between the notion of a strong, unified regional voice in the process and a homegrown, southern, moderate-to-conservative candidate. The former seemed more likely with a southern bloc of contests, but that did not happen until the 1988 invisible primary. Even then -- with everything lined up -- the South did not speak with one voice in the 1988 Democratic primary. The unintended consequence was that three Democratic candidates emerged from the Southern Super Tuesday with a claim to victory -- Dukakis in the populous South (Florida and Texas), Gore in the peripheral South and Jackson in the Deep South -- all while George HW Bush used a sweep of the region on Super Tuesday to consolidate his hold on the Republican nomination.

The dynamics of any given nomination race matter and it is difficult to gauge ahead of time -- as a decision-maker on the state level -- what those dynamics will look like in, say, two years time. That is the cautionary tale for those thinking of coordinating primaries in 2016. That past repeated itself to some degree in 2008 on the Democratic side (though not in a regional sense). Barack Obama and Hillary Clinton roughly split the logjam of national contests on Super Tuesday while John McCain significantly stretched the delegate lead he had established during the January contests.

What are the dynamics FHQ is talking about?

The candidates who run combined with the sequence of primaries and caucuses and the rules of delegate allocation are basics. And all are unknown at this point in time to those state-level decision-makers. There is a baseline calendar for 2016, but the question is how state actors view that terrain in light of the national party rules on (national convention) delegate selection. Actually, this constitutes several questions:
  1. Do we want to move our delegate selection contest up (to an earlier point on the calendar)?
  2. Does a new position mean incurring a penalty from one or both national parties?
  3. Does a new position mean conducting an election on the same day as a number of other regional partners?
  4. Does a new position mean conducting an election on the same day as a number of other states with no one dominant region? 
  5. Does moving to a new position to create a regional primary (question #3) mean that other states (or regions -- see question #4) will herd toward that date; typically in the post-reform era, the first date allowed by both national parties (the first Tuesday in March in 2016)?
Now, there is no indication that state-level decision makers actually consider these matters this deeply. Rather, in most cases, state legislators (collectively) see, on its surface, a good idea -- a regional primary -- and run with it. In the process, however, there is little evaluation of the unintended consequences.

None of this is happening in a vacuum. These decision to move a primary or caucuses are not independent of one another. The answer to question #1 depends on the willingness and ability of the state to move based on structural factors. FHQ has already discussed that for the states potentially involved in this retro-southern regional primary concept proposed by Georgia Secretary of State Kemp. Nothing in that proposal suggests that any of the southern go rogue, so the states of the South will avoid penalty so long as the Democratic National Committee retains a similar calendar to the Republican National Committee.

But there is something to questions #3-5 posed above. Partnering with other states in a region has its advantages, but it seems that that exercise has diminishing returns for the states involved as more states sign on. This needs a deeper examination, but one could argue that the most successful regional primaries have been subregional primaries; smaller clusters of contests at a point on the calendar that provides that group of states with the spotlight and is also earlier than the point at which 50% plus one of the delegates has been allocated to one candidate (effectively ending the nomination race). Contrast the 1988 Southern Super Tuesday with the 2008 Potomac Primary (Maryland, Virginia and Washington, DC), for example.

The former was a mega-primary that allowed candidates to pick and choose their spots (as on the Democratic side in 1988). One could also just as easily see such a contest giving advantage to an unintended beneficiary (as on the Republican side in 1988). That is, someone of the party opposite the dominant partisanship of the region or a front-running candidate with the resources to compete in such a large number of states. Alternatively, the latter, if shrewdly scheduled (in this case a week after a rush of more than 25 contests in 2008), can draw candidates into a small area of competition with similar issues. Again, that was true in 2008 with the Potomac Primary, but one could also consider the Alabama/Mississippi cluster the week after Super Tuesday in 2012 another of these. Many have argued that those contests were evidence of Romney's poor showings in the South, but while the former Massachusetts governor lost in both, he emerged at near parity with Santorum and Gingrich in the delegate count in each. In other words, it was competitive; something a subregional cluster would desire.

This is actually an idea that the DNC attempted to nurture in 2012: clusters of primaries. Neighboring groups of three or more states that held concurrent primaries in or after April on the calendar a 15% delegate bonus. That was viewed as a way of matching up state and candidate interests but also for giving incentive to later primary and caucuses dates.

Broadly speaking, though, this is an hypothesis that needs some additional research. Is there at point of diminishing returns in terms of what states and candidates get out of a Super Tuesday pile up of contests. Smaller, distinct (date and regional proximity) clusters may be better able to accomplish this. That seemed to be part of the lesson that states seemed to have learned after 2008. Part of the motivation many states had in moving back was a change in national party rules (the February to March transition of the post-carve-out window), but the other part was that a number of states herded to Super Tuesday in 2008 and got nothing out of it.

Those are the competing interests facing those states willing to move around for the 2016 cycle: 1) Learn the lesson of 2008 and attempt to pick and choose a spot on the calendar (either alone or as a small cluster of subregional states) or 2) Move en masse to the earliest date allowed by the parties -- the first Tuesday in March.3 Those two options are not mutually exclusive. It could be that a group of southern states, for instance, cluster on March 1 (fulfilling the first option with the exception maybe of the small cluster) and that has the effect of triggering a rush on the date by other states. That reactionary group of states would be operating under the rationale -- as was the case before 2008 -- that if they do not move they will run the risk of falling after the point in the races where enough of the delegates have been allocated to have singled out a presumptive nominee.

There may also be the added layer of indirect involvement from the national parties as well; coaxing some states to move around. And this goes both ways; not Democratic and Republican so much as moving up and back. In 2012, there was some talk about national Democrats urging some states to move back to negatively affect the Republican process. Northeastern states would move back, making the front half of the calendar more southern and conservative. That would, in turn, hypothetically hurt Romney. The result was something of a mixed bag. Romney had a somewhat rough path to the nomination, but that was not a function of a conservative first half of the calendar. There was a good regional mix of early contests even if a group of mid-Atlantic/northeastern states moved back into late April. The real culprit for the drawn out Republican contest was the dispersion of contests across the entire calendar.

Assuming we witness some movement on Secretary Kemp's southern regional primary on March 1, we could see Republicans (nationally or the RNC quietly) urge just the opposite of what Democrats wanted in 2012. The idea of a southern regional primary isn't new as discussed above, but neither is the idea of a regional primary in this cycle. It was just last November that RNC Chairman Reince Priebus was talking about a midwestern regional primary. If contests in the South begin moving up to March 1, there could very quickly be a quiet yet concerted effort to find a group of contests to serve as a counterbalance on the Republican calendar either on March 1 or not long after. Ohio is already scheduled for that week after Super Tuesday.

States once were slow to react to primary/caucuses movement in other states. A move in one cycle was met with a move in a the next subsequent cycle (if a state was compelled to move at all). That process has sped up over the last several presidential election cycles and reaction time had decreased. Since California moved its presidential primary from June to March in the 1996 cycle -- shifting with it the center of gravity in terms of the balance of delegates allocated over time across the primary calendar -- states have begun reacting within cycle. In other words, moving to a date that looks ripe for the taking now does not necessarily mean that that same date will not be jam packed with a number of other contests in the near future.

This hypothesis fits well in the policy diffusion literature. It also is something that FHQ has explored to some extent in a regional context. If one state moves its primary or caucuses, does that increase the likelihood that a neighboring state moves as well? What we found across a limited dataset -- the 2004 and 2008 cycles -- was the exact opposite: That if a neighboring state moved up, it decreased movement in surrounding states. At this point, FHQ is willing to chalk that up to a limited number of observations in just primary states across just a couple of cycles. It bears further research.

Again, it is easy to look at the surface issues here and move on if you are a state-level actor. Move up, bring along some regional partners, get more attention and affect the nomination. Under that surface, though, there is a lot to think through. It can quickly become a complicated series of unknowns. The changes to the Republican delegate selection rules have limited the world of possibilities by adding some penalties with teeth, but that does not mean that there are not 50 states -- some with multiple actors involved -- that are attempting to reduce uncertainty, game the system and gain an advantage for themselves (in terms of gaining attention and influencing the process). One move by a state or a series of states can set off any number of possibilities in reaction.

That's the take home message in this jumbled mess: unintended consequences. One move begets another move that may negate your original move. And there usually is not a rejoinder to the response. There isn't time.

1 As it turned out, Carter won New Hampshire and all three southern states in the 1980 primaries and it was not until later in the calendar that Kennedy began to close the delegate gap. Even that was too little too late.

2 Jesse Jackson's win in South Carolina and Gary Hart's in Oklahoma were the only two holes in an otherwise unified South. Those exceptions were early (March) contests and undercut the idea of the South collectively influencing the process by backing one candidate.

3 There is a third option as well. States could simply hold their ground and stay where current state law has the primary scheduled. Many states will do this as well.

No comments: