As the curator of the Daily Kos Elections Polling Wrap, one trend in the 2012 electoral cycle that has not escaped my attention has been the relative paucity of polling in this presidential election cycle, when compared to other cycles.
For example, in 2008, there were 2,730 polls released in either presidential or downballot races, not counting national or primary polls. This year, despite having a much better method of gathering polling data (to say nothing of the assistance of the rest of the Daily Kos Elections team, something I did not have in 2008), this election year has brought just over 1,450 polls to-date. To put it another way, to merely match 2008's polling output, we would need to average a release of about 300-350 polls per week for the rest of the cycle. That is, suffice to say, unlikely.
Unfortunately, we learned this week that the 2012 polling blackout is going to extend to what every losing candidate colloquially refers to as "the only poll that matters":
Breaking from two decades of tradition, this year’s election exit poll is set to include surveys of voters in 31 states, not all 50 as it has for the past five presidential elections, according to multiple people involved in the planning.While mournful, the decision nevertheless had seemed somewhat inevitable. With rising costs, and dwindling resources, we have already seen the major media outlets make slashing cuts to their polling budgets, with few media outlets even polling weekly, with only a month to go until the election.
Dan Merkle, director of elections for ABC News, and a member of the consortium that runs the exit poll, confirmed the shift Thursday. The aim, he said, “is to still deliver a quality product in the most important states,” in the face of mounting survey costs.
The decision by the National Election Pool — a joint venture of the major television networks and The Associated Press — is sure to cause some pain to election watchers across the country.
The criteria for inclusion and exclusion is predicated solely on the presidential race, it would seem. One imagines that the typical consumer of the media outlets only cares about the presidential race, and that gives the NEP the green light to cut states that are generally conceded to be safely in the corner of either President Obama or Mitt Romney.
With this in mind, these were the 20 entities (including the District of Columbia) that were summoned to the polling guillotine by the media outlets:
Alaska, Arkansas, Delaware, District of Columbia, Georgia, Hawaii, Idaho, Kentucky, Louisiana, Nebraska, North Dakota, Oklahoma, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Utah, West Virginia and Wyoming.The casual observer might look at this collection of states and say, "So what? These are all dark-red and dark-blue states. What is the big deal?"
(Continue reading below the fold.)
Here is the big deal:
1. How in the world is the NEP defining a "competitive" or "uncompetitive" state?
Do I think the outcome in these 20 locations is baked in? Perhaps. But the curious thing here is not what was included in the exit poll purge. It was what was not included in the exit poll purge.
Nearly half of the 19 states excluded from the NEP exit poll rotation have something in common: They resulted in a single-digit margin of victory in 2008, or there has been a poll in the 2012 cycle showing that state in single digits. Just last week, a poll out of Louisiana had Mitt Romney's advantage at six points. Do I think that Mitt Romney will lose Louisiana? Well, no, I don't. But Barack Obama's average lead in Pennsylvania over the last month has been greater than that. There have been polls in both Georgia and South Carolina in the last 12 months showing the states within the margin. Does Barack Obama have a respectable shot in either? No. But if the final outcome of a state is liable to be within single digits (which could happen in both, as it did in 2008), it would seem a worthy target for exit polling.
The decision to drop potentially intriguing states is even more maddening when you consider some of the states that the NEP exit poll will keep in the rotation. Why, if you accept the premise that polling all 50 states is cost-prohibitive and some states need to be jettisoned, are Kansas and Mississippi still being polled?
Kansas doesn't even have a Senate or gubernatorial race, and the presidential margin is expected to easily be in double digits. They do have a ballot measure looking to alter the property tax on boats, but that seems to be a pretty damned thin rationale for exit polling a state Mitt Romney is likely to carry with north of 55 percent of the vote.
Meanwhile, in Mississippi, I'll give the exit poll consortium the benefit of the doubt. Maybe they mistakenly thought that freshman Sen. Roger Wicker was squaring off with former Vice President Al Gore, and not this Some Dude of the same name. Otherwise, there is little to no rationale for polling the Magnolia State. Heck, there are one of only 13 states to not even have any compelling ballot measures on the ballot.
2. What about the races downballot?
Just because a state is clearly destined to wind up red or blue in the electoral college does not automatically render that same state meaningless. The consortium's list of banished states ignores, apparently, that we are also electing a Senate whose partisan control may well be on a knife's edge, as well as a dozen governors.
That makes the exclusion of three different states here somewhat surprising. The exclusion of one, North Dakota, is nothing short of outrageous. If the GOP is to claim the Senate, as has been their stated goal since the confetti was swept up in 2010, they are going to need to claim the open seat in North Dakota being abandoned by retiring Democrat Kent Conrad. The early assumption was that this was an automatic pickup for the GOP, given the reddish tendencies of the state. But Democratic nominee Heidi Heitkamp has outhustled Republican congressman Rick Berg, and the willingness of the Democrats to release their polling, and the reticence of the Republicans, are a significant hint that an upset may be afoot. Apparently, though, we are going to have to wait until well into the evening to know for sure.
Two other states have races where there is a pronounced lean in the likely outcome, but are states where a tight race is still eminently possible. In Hawaii, the GOP has thrown a boatload of money at former GOP Gov. Linda Lingle, and while polls show her trailing Democratic Rep. Mazie Hirono in her bid to replace longtime Sen. Daniel Akaka, the sheer size of the GOP effort to make the race competitive makes it a race that merits watching. Meanwhile, in West Virginia, we get a gubernatorial rematch of a race that was decided by three points just one year ago. The assumption is that Democratic Gov. Earl Ray Tomblin has done enough to lock down a full term, but given the anti-Obama sentiment among West Virginians, it might be a race worth getting a read on.
And, again, the exclusion of these three states is maddening given some other states that made the cut. Alabama has no Senate race, a wholly uncompetitive presidential forecast, and not even any reasonably competitive House races. The only race of note may be a state amendment on the ballot gunning for the mandate portion of Obamacare. While a more pressing national issue than the Kansas boat tax, it is still hard to say that it is a more deserving issue for consideration in exit polling than the potential tie-breaking race in the Senate.
3. Who said exit polls are only about Nov. 6?
Even if one were to concede that the outcome in these 20 locations was beyond doubt, and that the trailing presidential candidate has precisely zero chance of scoring the state's electoral votes, there is a big problem with eliminating some of these states wholesale.
Exit polls are not just a predictive measurement of how an electoral outcome will play out. They are also a research and information resource, one which allows us to detect and track electoral trends in states over time.
For this reason alone, the exclusion of Texas is an enormous error in judgment. A number of political observers, and not solely on the left, suspect a sea change in politics is underway in Texas. Changing demographics, it is theorized, could turn the second-largest state in the Union into a purple, if not blue, state within the decade.
But to know if that phenomenon is legitimate, or merely the wishes and hopes of some left-of-center analysts, we need data to support or refute it. Data which, apparently, will not be forthcoming. It would be extraordinarily valuable to not only see how the ethnic composition of the Texas electorate has changed (if it indeed does change) between 2008-2012. Furthermore, it would be of great value to know if, as the Latino population in the state grows, their voting behavior changes, as well.
The same is true for other fast-growing states excluded from the exit polling rotation. Is the rapid growth of Georgia and South Carolina going to reinforce the redness of those two states, or will it edge them closer to parity, as has been the case in states like North Carolina and Virginia? Sadly, it will be harder for us to answer that question, having lost this data of the likeliest of voters.