Showing posts with label counterfactuals. Show all posts
Showing posts with label counterfactuals. Show all posts

Saturday, March 24, 2012

The Myth of Proportionality's Impact is Dead

FHQ and John Sides put together a short op-ed piece for Bloomberg View over the last couple of weeks. The premise was pretty simple: Much is being made of the changes to the Republican delegate selection rules in 2012, but what impact are those changes truly having on the race? FHQ has argued since the summer of 2011 that the true measure of change would be felt through the more evenly dispersed calendar of primaries and caucuses and that the new proportionality requirement would have minimal effects on the accrual of delegates throughout that calendar.

To test this John and I took the 2008 delegate selection rules -- a mix of proportional, winner-take-all, and other hybrid rules in between -- from the states that have held primaries or caucuses thus far in 2012 and simulated a reallocation of the delegates under those rules.  The results were revealing in the face of so many complaints from within some Republican circles that the new proportionality requirement is drawing out the current nomination process (...with the implication that the protracted fight is not helpful to the Republican Party or its standard bearer in the fall). Through the same point in the process -- post-Illinois -- Mitt Romney, Newt Gingrich and Ron Paul would all have lost delegates and Rick Santorum would have gained a handful. The net change to the current delegate margin between Romney and Santorum would  have been to have tightened it by 65 delegates. Instead of accelerating the process -- pushing Romney closer to the 1144 delegates necessary to clinch the nomination -- the 2008 delegate rules would actually have slowed things down (...while still providing Romney with a better than two to one delegate advantage over his nearest competitor).

--
From this came a rather strange series of critiques from Dana Houle (via Twitter). I'm not entirely sure what Mr. Houle was reading into or in our op-ed. But he did at one point seemingly attempt to make the piece about inserting the full 2008 Republican rules -- hyper-frontloaded calendar and all -- into the 2012 context as opposed to the intended isolation of the new proportionality requirement. [And FHQ has taken issue with Mr. Houle's interpretation of the Republican delegate selection rules in the past.] Again, FHQ has argued that the rules-based changes in how the process has progressed are almost solely due to the calendar.1

But Mr. Houle stretched this out into a discussion of the impact the new proportionality requirement has had, contending that changes to delegate allocation rules begets changes to candidate strategy or campaign tactics. Has the proportionality requirement changed how candidates have amassed delegates in 2012? It has, but that impact is minimal; negligible even.2 Has this had an impact on candidate strategy/campaign tactics? Perhaps, but FHQ would argue that the nature of that change, too, has been or would have been minimal as compared to a 2012 race under 2008 delegate selection rules.

Particularly interesting was Mr. Houle's contention that Ron Paul would have been "irrelevant" and that Rick Santorum and Newt Gingrich would have been apt to "blow off a state rather than try to get some delegates here and there". Let's examine this. I enjoy the criticism, and the chance to clarify the bottom line here is a good thing.

January
3
Iowa:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)
Strategic change: none

10
New Hampshire:
2008: proportional primary
2012: proportional primary
Strategic change: none

21
South Carolina:
2008: winner-take-all primary statewide and by congressional district
2012: winner-take-all primary statewide and by congressional district
Strategic change: none

31
Florida:
2008: winner-take-all primary
2012: winner-take-all primary
Strategic change: none

February
4
Nevada:
2008: non-binding caucuses (unbound delegates)
2012: binding, proportional caucuses
Strategic change: Ah, well here's something. If Nevada had been non-binding in 2012 -- as the 2008 rules outlined -- perhaps the candidates would have "blown off" the Silver state and simply ceded it to Romney as was the case in 2008.3 It could also have been the case that "irrelevant" Ron Paul would have contested the caucuses in an effort to influence actual delegate selection process rather than focus on the non-binding straw poll as he has attempted in other similar 2012 contests. As for Gingrich and Santorum, maybe they skip ahead to Colorado or Minnesota, or maybe they treat Nevada as a non-binding caucus on par with those next contests and see no harm in the potential momentum a surprise win (or performing better than expected) would bring.

7
Colorado:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)
Strategic change: The presidential nomination process is sequential, so changes at one point create changes at subsequent sequential events. FHQ will grant Mr. Houle that. But that was never anything that we were arguing. The likely strategic change would have been minimal. Do I know? ...for sure? No, but I invite comments on how things -- outcomes -- would have changed.

Minnesota:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)
Strategic change: See Colorado

Missouri:
2008: winner-take-all primary
2012: non-binding primary/binding caucuses (in March)
Strategic change: This is a tough one. The process that produced the switch to a caucus system in Missouri was a chaotic one; rife with legislative division over what to do and a general misunderstanding of the primary system. One thing that drove the division was the threat of losing half of the delegation for holding a primary too early. That is a function of the calendar rule change (not the proportionality requirement) and was not what we were testing. We withheld Missouri from the reallocation simulation, or treated it as similarly unbound (as it was in 2012 -- unbound until the delegates are bound at the congressional district conventions and the state convention). But if we want to play of the game of Missouri holds a binding, winner-take-all primary on February 7, we can. Missouri would have lost half of the party's delegation -- cutting it to 26 total delegates -- and the winner-take-all allocation would have been under a similar level of threat of challenge (on par with Florida above). That would have added 26 delegates to Santorum's total. Would Santorum or even Gingrich have known that Missouri was going to be a good state for one or both of them beforehand? That is dubious. It would have been unclear ahead of time -- just as it was in reality in 2012 -- that was the case. One of the patterns that we have come to recognize as predictable -- evangelicals for Santorum -- had yet to emerge.

Now, I don't think that the Missouri GOP would have challenged the RNC rules and held a binding primary even in the absence of the proportionality requirement. However, if it had, that delegate boost would have helped Santorum push back on the chorus of "He won, but didn't win any delegates" charge that was lobbed at his campaign in the time after his February 7 sweep. And that may have put some additional pressure on Gingrich to drop out of the race.4 Whether that would have actually moved Gingrich, well...

11
Maine:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)
Strategic change: There is an issue of timing with the Maine caucuses in 2008 compared to 2012. Four years ago Maine Republicans held their caucuses across one weekend, but in 2012 the bulk of the caucuses stretched across two weekends. Most of the attention paid to Maine in 2012 occurred in the time after the February 7 contests and that likely would not have changed under 2008 rules. Attention would not have increased nor would other strategy have changed much in regard to the non-binding caucuses in the Pine Tree state.

28
Arizona:
2008: winner-take-all primary
2012: winner-take-all primary
Strategic change: none

Michigan:
2008: winner-take-all by congressional district/proportional statewide primary
2012: winner-take-all primary statewide and by congressional district
Strategic change: This is another tricky one. The Michigan GOP apportioned 1 delegate per each district in 2008 and let the remainder be proportionally allocated at-large delegates (based on the statewide result). In 2012, however, the state party opted to apportion two delegates per congressional district with just two leftover delegates to be at-large. The bottom line is that candidates would have had opportunities to win delegates in each plan whether by congressional districts in the 2012 plan or by proportional allocation of bigger pool of at-large delegates in 2008. Given the that the difference is just one delegate in the true allocation (16-14 in favor of Romney) versus the simulated allocation of delegates under 2008 rules (15-15 tie), the change was small. Again, there were opportunities in both plans for the same candidates who did well in that contest in 2012. Very little would have changed, particularly with Michigan sharing the spotlight with Arizona.

March
3
Washington:
2008: non-binding caucuses (unbound delegates) & winner-take-all by congressional district/proportional statewide primary
2012: non-binding caucuses (unbound delegates)
Strategic change: We could not simulate a primary that did not/does not exist in Washington in 2012. Since the Washington legislature eliminated it anyway, it would be impossible to include it in the simulation. The assumption, then, is that Washington would have held the same non-binding caucus under 2008 or 2012 rules and that there would have been no change in tactics as result.

6
Alaska:
2008: proportional caucuses
2012: proportional caucuses
Strategic change: none

Georgia:
2008: winner-take-all primary statewide and by congressional district
2012: proportional primary statewide, top two by congressional district (or winner-take-all if majority winner in district)
Strategic change: It would still have been a Gingrich state, but Romney and Santorum would have had opportunities in congressional districts. Would they have "blown off" Georgia to focus elsewhere? Maybe, but the Romney/Santorum's focus was mainly on Ohio anyway. Gingrich would have ended up with a greater number of delegates under the 2008 rules.

Idaho:
2008: proportional primary
2012: conditionally winner-take-all/proportional caucuses
Strategic change: There were larger delegate prizes on Super Tuesday that would have drawn the candidates away from Idaho regardless of the rules in a state that favored Mitt Romney anyway. Though it should be noted that Romney would have lost delegates in the Gem state under different rules due to the proportional primary in 2008. Ron Paul, perhaps, would not have gone to Idaho as it would not have been a caucus state under the 2008 rules.

Massachusetts:
2008: proportional primary
2012: proportional primary
Strategic change: None. The rules didn't change and neither did the fact that Romney was the former Bay state governor.

North Dakota:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)
Strategic change: none

Ohio:
2008: winner-take-all primary statewide and by congressional district
2012: winner-take-all primary by congressional district with conditionally winner-take-all/proportional allocation (majority winner takes all at-large delegates)
Strategic change: There are two ways of thinking about a change in the rules in Ohio. First of all, if the 2008 rules had been in place, Santorum still would have had opportunities to win delegates in congressional districts. Second, that race was competitive enough that proportional or winner-take-all, it would have brought the candidates' attention. Romney would have come out of Ohio with many more delegates under the 2008 rules.

Oklahoma:
2008: winner-take-all primary statewide and by congressional district
2012: conditionally winner-take-all/proportional primary statewide, top two/three in congressional districts
Strategic change: Like Ohio above for Romney, Santorum would have been the beneficiary in the Sooner state. The two states largely offset each other.

Tennessee:
2008: conditionally winner-take-all/proportional primary statewide, top two or proportional in congressional districts
2012: conditionally winner-take-all/proportional primary statewide, top two or proportional in congressional districts
Strategic change: none

Vermont:
2008: winner-take-all primary
2012: conditionally winner-take-all/proportional primary, winner-take-all by congressional district
Strategic change: There were too many other large prizes on Super Tuesday for things to have changed much from the campaigns' perspectives in Vermont. Romney would have gained delegates under the 2008 rules.

Virginia:
2008: winner-take-all primary
2012: conditionally winner-take-all primary statewide and winner-take-all by congressional district
Strategic change: Very little would have changed in Virginia due to the ballot situation. Romney would have gained three delegates and the automatic delegates under the 2008 rules, but little else would have changed.

10
Guam
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)5
Strategic change: none

Kansas:
2008: winner-take-all caucuses statewide and by congressional district
2012: proportional caucuses statewide and winner-take-all by congressional district
Strategic change: Likely none. With Alabama and Mississippi on the horizon -- good states demographically by this point in the race for Santorum and/or Gingrich -- the focus was there already and would have been under the 2008 rules. Romney was also focused on those states. That would not have changed.

Northern Mariana Islands:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)5
Strategic change: none

Virgin Islands:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)5
Strategic change: none

Wyoming:
2008: non-binding county conventions
2012: non-binding county conventions
Strategic change: none

13
Alabama:
2008: conditionally winner-take-all/proportional primary statewide, conditionally winner-take-all/top two by congressional district
2012: conditionally winner-take-all/proportional primary statewide, conditionally winner-take-all/top two by congressional district
Strategic change: none

American Samoa:
2008: non-binding caucuses (unbound delegates)
2012: non-binding caucuses (unbound delegates)5
Strategic change: none

Hawaii:
2008: convention (unbound delegates)
2012: proportional caucuses
Strategic change: The big change here -- as was the case with Nevada above -- is that under the 2008 rules, the Hawaii delegates would have been unbound instead of proportionally allocated. That may have affected the (Paul, Romney and Santorum) campaigns' decisions to send one of their children each on a vacation/campaigning trip. But that's probably about it.

Mississippi:
2008: conditionally winner-take-all/proportional statewide and winner-take-all by congressional district
2012: proportional primary statewide and conditionally winner-take-all/top two or three by congressional district
Strategic change: With no change in Alabama rules-wise from 2008 to 2012, there may have been some shift in the focus to Mississippi under 2008 rules with the potential for a wider delegate margin from the state. Santorum would have gained from the switch to 2008 rules.

17
Missouri:
2008: winner-take-all primary
2012: non-binding primary/binding caucuses (in March)
Strategic change: See Missouri entry above. A non-binding caucus with no reporting of results is difficult to plan for. But if a winner-take-all primary had been scheduled for this date instead of a non-compliant February date, Missouri Republicans would have had a full apportionment of delegates. That would have changed the strategy around the contest, but the proportionality rule was not what led to switch to the caucuses.

18
Puerto Rico:
2008: conditionally winner-take-all/top two or three caucuses
2012: conditionally winner-take-all/proportional primary
Strategic change: The 2008 to 2012 change would have perhaps altered Santorum's decision to spend any time in Puerto Rico, but that depends upon whether you think the trip down there was vacation or not. The reality is that Romney won by enough to win all of the delegates under either delegate selection plan.

20
Illinois:
2008: loophole primary
2012: loophole primary
Strategic change: none

--
Would there have been strategic/tactical changes if the 2008 delegate allocation rules had been in place in 2012 instead of the current rules? Of course there would have been changes. Would those changes have affected where this race is now or the outcome of certain races? Probably not or at least I am hard pressed to think of a primary or caucus that would have changed or how the course of the race would have changed (...other than, say, at the margins). Over time, it may have caused the campaigns to shift more focus onto states that would have been favorable to one candidate or another, but that would have a canceling out effect. Santorum would have focused more on the South/prairie and won more delegates there while Romney would have focused minmal effort there and more in his areas of strength. The effect in the aggregate is very similar to the delegate situation the race is in at the moment.

What this exercise does shed even more light on is that the new proportionality requirement has had very little influence over the course of the 2012 Republican nomination race and minimal effects on the resultant strategy. The myth is dead. If the attempt is to pinpoint a rules change that is drawing the process out, then the finger should be pointed at the calendar rules changes (and state-level reactions to them) that produced the 2012 primary calendar.

FHQ gets the point Mr. Houle is trying to make. There are factors that cannot be accounted for in all of this. I won't argue that point because it is true. However, there is very little evidence that the presence/absence of the proportionality requirement would have fundamentally altered this race (or even altered it at the margins in ways that add up to even a small change in the course of the race). If anything, the delegate picture would look largely the same while the candidates would have potentially more greatly emphasize areas/regions/states of strength. And that would tend to reinforce the current (demographic to delegate) dynamics while ever so slightly closing the delegate gap.

[NOTE: FHQ will have more on this data throughout the weekend and into next week.]

--
1 There are obviously much different dynamics (outside of the rules) in this race than in 2008 or any other cycle. That is the nature of the presidential nomination process. You never know what you're going to get. Others will point to the impact Super PAC spending has had on the 2012 race as well. Both have more of an impact on all of this than the proportionality requirement.

2 Does the switch alter the delegate count? Yes. Does it change the state of the race as it is now? No, it doesn't. Romney would still have a healthy lead in the delegate count; just not quite as healthy.

3 Of course, Nevada shared the same date in 2008 as the South Carolina primary and that dynamic was what drove Romney's victory in the state. Had the Nevada caucuses been the only event on its caucus date in 2008 -- binding or not -- the candidates likely all would have been there.

4 And IF Missouri was binding and winner-take-all that may have caused the candidates to focus more of their efforts there instead of in either Minnesota or Colorado. But again, Missouri VERY likely would not have gone in that direction; opting instead to shift to a later caucus. The Missouri GOP was very wary of losing delegates and would have made the move regardless.

5 The island territories held conventions that both started and completed the delegate selection process. In each case, decisions were made to pledge/bind those delegates (in contrast with the other non-binding caucus states where the delegate allocation has yet to take place).


Recent Posts:
2012 Republican Delegate Allocation: Louisiana

August Presidential Primary Resurrected in Kentucky Legislation

Divining the Meaning of Illinois


Are you following FHQ on TwitterGoogle+ and Facebook? Click on the links to join in.

Wednesday, January 18, 2012

Of Course Rick Santorum Won Iowa...

...but Mitt Romney did, too.

Both were able to beat the medium term expectations that had developed around their campaigns relative to the Iowa caucuses.1 Part of the story coming out of the Hawkeye state was the closeness of the top two, and while that is a fun footnote -- or will be in the history books -- to the caucuses, the main stories from FHQ's perspective were that Santorum was able to become the top not-Romney in the state and that Romney, despite the underlying demographics of caucusgoers, was able to finish in the top tier. Santorum exceeded expectations and Romney -- even in the worst case scenario -- either met, by being in the top tier, or exceeded expectations.

Whether the two flip flop their positions in the Iowa GOP-certified results this week will do little to change the dynamic that has developed in this race: Romney is the frontrunner and Santorum's name is on more lips and in more minds post-Iowa than they would have been if he had finished behind Newt Gingrich or Rick Perry on the night of January 3. Going back and trying to rewrite the story based on the shifting of less than 100 votes or so in the margin will have very little effect on what's going on now. Rick Santorum would still face the same sort of questions Mike Huckabee faced four years ago (Specifically, can insta-organization compete with the well-oiled machine of a well-financed frontrunner?), and Mitt Romney would still have -- at the very worst -- met expectations in the first two states while his rivals, with the exceptions of Paul in both Iowa and New Hampshire and Santorum in Iowa, underperformed.

This would have been a fun question -- the type Public Policy Polling likes to throw into their surveys from time to time -- to include in a South Carolina or Florida poll. My strong hunch is that it would make very little difference in vote choice in either the Palmetto or Sunshine state.2 The only time that this might have mattered was in the early morning hours of January 4. Good luck constructing that counterfactual. If anyone is able to, please let me know. I want to check out your time machine.

"Stop the inevitability narrative in its tracks"? Eh, probably not. It may be a speed bump, but more like one of those varmints Mitt Romney once hunted than an elk or moose in the headlights.

--
1 By medium term I mean something akin to a rolling average of expectations over time; something that is not susceptible to an outlier survey's snapshot of the race.

2 Granted, I think it would be difficult to determine whether that was actually part of a voter's decision-making calculus anyway. ...but that's a whole different can of worms from the political science/public opinion literature.




Are you following FHQ on TwitterGoogle+ and Facebook? Click on the links to join in.

Monday, May 12, 2008

Would McCain Have Won Under the Ohio Plan?

I spent the weekend (at least in terms of this blog) looking at the potential presidential nomination reform plans that are being considered by both of the parties. The one plan that has some traction for the moment, is the Ohio Plan that the Republican Party Rules Committee passed last month. Next stop? St. Paul, where the plan will face increased scrutiny this summer at the GOP convention. And it may even face a hostile group of delegates if it reaches the floor for broad consideration. No, not because of those Ron Paul delegates. The Ohio Plan could encounter resistance from McCain delegates. Why, you ask? Well, victorious nominees rarely back plans that change the rules under which they won their nomination, especially if that means they (or someone similar to them) wouldn't have won. If it ain't broke, don't fix it. And at a McCain-centered convention, delegates may be more willing to yield to their standard bearer on the issue.

Plus, let's remember that the GOP won't have the benefit of hindsight. Unlike the Democrats, the GOP can only alter its rules for presidential nomination during the preceding national convention. The Republicans then, won't know if having McCain wrap up their nomination three months prior to the Democrats will have an effect on the general election outcome (though they'll likely have a good idea whether the Democrats are indeed divided as a result of their longer process).

The question then, is would McCain have won the Republican nomination under the rules outlined in the Ohio Plan? First, let's glance at that map again:
Next, we'll have to willingly suspend our disbelief that such a plan could ever be put in place (and if you've been reading, you know FHQ has a laundry list of disbeliefs that will have to be suspended here). Let's also assume that if a candidate won a contest in 2008 under the current rules, they would have won the same contest under the Ohio Plan rules (at least among the competitive group of Republican contests--anything from Iowa to Texas/Ohio). According to the Ohio Plan rules, the same four states that led off the process under the current rules, would have kicked off primary season (Sorry Wyoming, Michigan and Florida). Iowa would have been followed by New Hampshire which would have been followed by South Carolina and Nevada. That wouldn't change any of the results we've seen under the current rules (but it would change the timing of the contests. None of these states would have been allowed to go prior to the beginning of February). Huckabee would have won Iowa. McCain still would have won New Hampshire and South Carolina and Romney would have won the Nevada caucuses.

Instead of Florida coming next, though, the process would shift from the "early 4" to a grouping of the least populous states (in teal above). In the process, McCain would have lost the advantage of that Florida win; a win that propelled him to the showing he had the next week on Super Tuesday. By the same token, Romney would have lost the influence of Michigan and Wyoming; wins that kept him viable heading into the next, contest-heavy week. With three different winners of four contests, there would be no clear favorite heading into the small state primary (Dare I call it Tiny Tuesday?) during the third week of February.

Among these 12 states' contests, McCain would have won one (Vermont), Huckabee would have won one (West Virginia) and Romney would have won five (Alaska, Montana, North Dakota, Maine and Wyoming). The remaining five states (Hawaii, Idaho, Nebraska, New Mexico and South Dakota) are states that have yet to hold contests during this cycle, so we don't know how those outcomes would have looked in a competitive environment.

What are the most likely directions those contests would have gone, though? Given Romney's success in the Prairie and Rocky Mountain states, the temptation is there to allocate him the wins in both Idaho and South Dakota. The difference is that the wins he actually got in Montana, North Dakota and Wyoming were all caucus victories (Romney seems to have approached the 2008 primary season in much the same way that Barack Obama and his campaign did. They both focused on organization in those small, caucus states.). Idaho and South Dakota have primary systems in place and it is unclear how Romney would have fared in such a scenario in those states. Southern Idaho looks a lot like Utah, where Romney did very will on Super Tuesday and it is likely he would have won Gem state. South Dakota is bordered by six states, four of which Romney won, so I'll lump it in with the other Romney wins. That gives him seven wins out of those 12 small states.

As for the other three unknown states, McCain's home state advantage would likely have stretched into New Mexico, but Hawaii and Nebraska are tough ones to figure out. Nebraska likely would have been a Romney/Huckabee battle given the strength both had in the region. I'll be generous and throw Huckabee a bone on this one. Hawaii would have been a far away caucus; a set of circumstances that would have favored Romney. Of the 12 small states, Romney would have been a winner (or in good shape) in eight, while McCain and Huckabee each would have managed two wins.

That's a pretty significant win for Romney heading into the big states. Do those eight wins equal what Florida did for McCain, though? That's a tough question to answer. Romney looked good in Iowa, New Hampshire and Nevada before the actual primary season got underway in 2008, and though he had several foreseeable paths to the GOP nomination, just couldn't come through in the end. Projecting Romney's potential success in the early states of the Ohio Plan to the rest of the country, then, is not automatic, but nine wins in sixteen early contests is nothing to sneeze at either. He clearly would have had an advantage in money and delegates prior to the three "pods" of big state contests. That could have pitted McCain and Huckabee in a one-on-one battle for the number two position behind Romney instead of the Romney/Huckabee fight for the same distinction behind front-runner McCain that we actually witnessed. And that is a position Romney did not find himself in in 2008.

Under the Ohio Plan then, McCain would have been in trouble instead of being in control after the first month of the process. Would McCain give his blessing at (or before) the convention to a plan that would have potentially cost him the 2008 nomination had it been in place? That, too, is a tough one to answer. McCain is a maverick (or so they tell me), so he may be willing to buck conventional wisdom. The only recent precedent is the Bush convention in 2000. That convention discussed the Delaware Plan (minus the New Hampshire/Iowa exemption) but it ultimately failed. And even in a season of change, the Ohio Plan will likely face similar resistance at the St. Paul convention (whether McCain endorses it or not).


Recent Posts:
Tales from the Kennedy School Symposium on Presidential Primaries

The Delegate Race: Is Obama There Yet?

ABC News: Obama Now Leads in Superdelegates

Tuesday, April 8, 2008

What if the General Election was Run like the Primary System?

Yesterday's post on the GOP's presidential nomination reform plan and the possibility of a national primary triggered an interesting discussion (between readers Rob and Bill) and question. Rob decried the pitfalls of a national primary and so doing indirectly pointed out how the primary system in the post-reform era has conditioned the American electorate. The first two experiences following the McGovern-Fraser reforms (1972 and 1976)) witnessed two "long shots" win the Democratic nomination. The result has been that some within the electorate think of this system in terms of its ability to nurture competition and allow for seemingly unknown (yet potentially well-qualified) candidates emerge to vie for the nation's highest office. As the primary system has become more frontloaded, that conception has been threatened. The compression of the calendar, the conventional wisdom holds, creates an easier road to the nomination (Unless you consider 2008. But that's something completely different.).

Something about Rob's rejection of the national primary elicited a dig of sorts from Bill who questioned (in so many words) that if the a national primary is so bad, how come it works in November every four years? And that raises a tangential and counterfactual question: What if the general election was conducted in the same way that primaries have been conducted in the post-reform era? How would that potentially affected the outcome of those general elections? Let's assume then, that states can hold their presidential elections anywhere from the Tuesday after Labor Day to the first Tuesday after the first Monday in November. And let's go just one step beyond that to assume that states have yet to frontload their contests to the earliest possible point (Tuesday after Labor Day)--the motivation to do so would be the same in that context as it has been in primary elections from 1972 onward. While we're assuming, we have to also assume that the Constitution provided no guidelines on how this process was to be conducted (though we'll leave the electoral college in place). Just for the fun of it, let's use the primary calendar of 1980 as a guide. There was a fairly even dispersion of contests through the March to June window that year.

What is the result? Well, which state goes first is huge. Does that state lean one way or the other along the partisan spectrum? Is their an incumbent involved? Interestingly, Iowa and New Hampshire have been competitive in the last two general elections, so they are ideal (in some ways) first states in this scenario. A state that is solidly red or solidly blue is more likely to be discounted by one candidate, but a state that is evenly divided between the parties becomes an interesting battleground, especially for a candidate challenging an incumbent. Winning or posting a close finish would be a boon to a challenger heading into subsequent states (And simply winning a "state you're suppose to win" would even be beneficial to a challenger as well; more so than it would be to an incumbent.). Winning early then, for a challenger, is important in swing states. Voters in those states who might ordinarily opt for the incumbent (or the status quo) when they are undecided, may be more apt to consider a challenger if that candidate has done well early. Such a system would allow a challenger to potentially cast doubt on an incumbent's ability to win with early victories.

The big question is, how all of this could have affected the outcomes of past races? The closer elections are the ones most likely to see a shift (at least in who the winner was). More comfortable victories or landslides may have seen some states change at the margins but without affecting the outcome (Remember we still have the electoral college here and those electors end up serving essentially the same purpose as delegates to national conventions in this sort of system.). So Mondale or McGovern could potentially have been able to win more than one state each or Gore could have bested Bush (He certainly would have focused more resources on Tennessee.).

How else would having the primary system in place for the general election have affected things? What say you, loyal readers of FHQ? This is a fun one for discussion.

I'll be back late tomorrow with a new set of electoral college maps accounting for the new polls that have come out this week.

Tuesday, February 26, 2008

What Could Have Been: New Jersey

February 26 wasn't always this lonely. It wasn't always about Cleveland debates and looking forward to primaries in Texas and Ohio. No, for a couple of years (between 2005 and 2007), New Jersey's presidential primary was situated on this, the fourth Tuesday in February. Traditionally one of the states to bring up the rear in early June, New Jersey's legislature opted to position the state's presidential primary three weeks behind Arkansas, Missouri and Oklahoma, then the only three states scheduled for February 5. Once the momentum began to build behind the February 5 Super Tuesday in late 2006 and into 2007 though, the New Jersey legislature moved forward again, joining what even then in April 2007 was shaping up as a logjam in early February.

The intent of the move was to give New Jersey voters an opportunity to weigh in on who the two parties' nominees were to be for the 2008 cycle. And you can't blame the legislature for assuming that the Super Tuesday model would hold for this cycle as it has since basically the 1992 phase. To move and still not be consequential would have looked bad. But the nominations were not wrapped up on Super Tuesday, and that assumption and subsequent gamble may not have paid off as it could have if New Jersey had remained on February 26. Yes, the intent of the move was fulfilled, but residents of the Garden state (at least the Democratic and independent ones) could have been far more important to the Democratic nomination had the brakes been put on the second move.

Just for the heck of it, let's play out this bit of counterfactual history. If New Jersey had kept its presidential primary on February 26, it would have been the only event on that date; the only game in town. Typically that means a ton of media coverage and candidate attention. In 2008 though, that attention would have grown exponentially. Let's call it New Hampshire, part II. In addition, think about the current race for the Democratic nomination. Obama has rolled off eleven straight victories since Super Tuesday (Yes, the Virgin Islands and Democrats Abroad count. They do provide delegates after all.). Clinton may have still faced those same eleven defeats if New Jersey had been on February 26, but at least a contest in some naturally hospitable territory would have been on the horizon. In addition, it could have served as a nice springboard into the contests of next week; possibly throwing the outcome of the nomination into further doubt (Clinton wins next week could still do that, but with a New Jersey win, it would have been easier.). Now sure, a Clinton win in a February 26 New Jersey primary could have been spun by the media as a contest she should have won, thereby shifting the focus to the margin of victory and delegate totals. However, you can't underestimate how important potentially breaking Obama's winning streak ahead of March 4 could have been to the Clinton camp. It could have fundamentally altered the course of the race.

Finally, and most importantly, a February 26 New Jersey primary would have meant that political junkies wouldn't have to sit idly by waiting two weeks for the next round of contests.