The voting population that matters gets smaller


In a previous post, I wrote about how voter volatility has dropped sharply beginning with the 1996 elections, resulting in much greater stability in voter patterns. This has resulted in the cementing of party preferences in about 80% of the states leading to a semi-permanent red state-blue state map of the US, with just a handful of about 10 states (Ohio, Florida, Pennsylvania, Virginia, and Wisconsin being the largest of this small group) considered swing states whose outcome is now considered to be still up for grabs.

Because the presidential election is determined by the electoral college system that assigns votes by states, campaigning and resources are now largely targeted to just those states that are seen as competitive, with the rest of the nation being pretty much bystanders. Ritchie King has compiled a chart beginning in 1972 comparing the margin of victory in each state with the overall margin of the national vote and finds that the gaps have increased steadily since 1992. As King states, “from 1992 to 2012, blue states grew a bit bluer, red states a bit redder and tossup states a bit fewer in number.” However, the current election has seen a narrowing of the spread, at least as far as opinion polls go.

This year, in addition to having less of a spread, has more states with closer absolute margins than other recent elections. In 2008, for example, only six states had electoral margins of 5 percentage points or less; this year, our polling averages show 11 states as being that close.

Part of the reason there are more competitive states this year is that Trump is nudging some red states into the tossup zone while outperforming past GOP nominees in some blue states.

Professional political strategists advise candidates to pour almost all their resources into just these swing states. But while this may make sense tactically, it is not good for democracy in general. Donald Trump, defying convention as usual, has been holding rallies in the states of Connecticut and Washington that are widely believed to be definitely voting Democratic. He has been ridiculed for this but I applaud him. The president is supposed to represent the nation as a whole. What does it say if the candidates appeal to less than 20% of the states?

With the advent of data mining and micro-targeting, the situation is getting steadily worse as campaigns are now able to predict how smaller communities and even individuals will vote. Steven Rosenfeld reports on a new study David Schultz that suggests that as few as 20 counties in the nation, comprising fewer than 500,000 voters or just 4% of the total vote, can tilt the outcome one way or the other.

Where are 2016’s deciders? In Ohio, it’s Hamilton County, home to Cincinnati. In Pennsylvania, it’s Bucks and Chester Counties, to the north and south of Philadelphia, and also Lackawanna and Luzerne counties. In Florida, it’s Hillsborough and Pinellas counties, surrounding Tampa and St. Petersburg. In Wisconsin, it’s Brown County, where Green Bay is; nearby Winnebago County, further inland; and Racine County to the south near Chicago suburbs. In Iowa, it’s rural tiny Bremer County, and in New Hampshire, it’s Hillsborough township, inland on the Massachusetts border.

There are a few more: population epicenters such as Nevada’s Clark County, home of Las Vegas; Virginia’s Prince William County, outside Washington D.C., North Carolina’s Wake County, with Raleigh and Durham; New Mexico’s Bernalillo County, containing Albuquerque; and surprisingly, Dona Ana County near Las Cruces, which has a big state university.

“These seem to be the counties within the swing states where the candidates go,” said Schultz. “They view them as battlegrounds. They seem to be pretty good bellwethers, in the sense of predictors of how that state is going to vote… Even if they appear blue or red, there’s a question of how great the turnout will be.”

This is another reason why election blowouts are becoming increasingly rare and Trump still remains competitive while alienating many demographic subgroups.

Indeed, when Schultz identified these swing counties, he looked at their 2012 votes, and realized that had Mitt Romney been a little more successful in various combinations of these locales, he would have been elected.

“People don’t realize that the Romney-Obama race was actually far closer than most people think,” Schultz said. “There was an Electoral College blow-out, but if you had a shift of just a few hundred thousand votes across a few states, Romney would be running for re-election this year. We’re really looking at no more than a half-million votes shifting, depending on how you actually define it, and that’s a very small number of votes.”

How far can this process go? Will we end up with a handful of competitive neighborhoods? Streets? Individuals? Will a future presidential election focus on just whom Thelma Brown, of 240 Main Street, Cincinnati, Ohio will vote for as she will be the last remaining swing voter?

Comments

  1. sonofrojblake says

    The only people who have the power to change how elections work are, by definition, the people who have the least motivation to do so, since the system that put them in the position of power is the one that “worked” for them. Why would they change it? It gave the “right” answer.
    The problem, such as it is, is the availability of information. If all the votes in all the country were stuck in a single big pile and just counted – fully anonymised not merely by voter name but also by location – these analyses would not be possible. But since votes are identifiable by location, even if the electoral college system was eliminated and the election was decided by popular vote, it would still be possible in principle to identify where it made most sense to concentrate electoral attention. Only by deliberately hobbling the statisticians could you avoid this.

  2. Brother Ogvorbis, Fully Defenestrated Emperor of Steam, Fire and Absurdity says

    Lackawanna and Luzerne counties.

    I live in Luzerne County and work in Lackawanna County. And I have seen some heavy flow of political advertising. And more than 90% of the adverts I see are Clinton campaign adverts, or private groups raising awareness on her behalf (they don’t say to vote for her, but the conclusion leaves no doubt). Four years ago, we were inundated with Romney and Obama adverts.

    I wonder if this has more to do with Trump’s inability to raise massive amounts of money? or is the Trump campaign focusing only on the Pittsburgh/Philly suburbs?

    And thank you. This is yet another data point showing just how modern communications (such as polling data and data mining) can be antithetical to actual democracy.

  3. OlliP says

    But without the electoral college every vote would matter. It would make sense to try to win in a state or county by a bigger margin or lose by a smaller margin. As things stand, why would either candidate make any effort to shift the margin in a state where realistically the winner is certain and only the margin can be influenced by campaigning and policy.

  4. says

    How far can this process go? Will we end up with a handful of competitive neighborhoods? Streets? Individuals?

    The Koch Brothers are already working on that.

  5. sonofrojblake says

    @OlliP, 3:

    without the electoral college every vote would matter

    True. But if you have access to detailed breakdowns of voter habits by area, you will still end up targetting your efforts on the areas that the stats show have the largest numbers of undecideds.
    Imagine you have just three states, where history shows the west coast reliably comes out Democrat by a margin of 70:30, the east coast is reliably Republican by the same margin, but the land between is 51:49 and wavers. (Assume all three regions have the same population). The Democrat would be wasting their money spending in the west because it’s in the bag and the people voting against them are likely unpersuadable, and similarly wasting their money spending in the east because it’s a lost cause. Same logic reversed for the Republican. Even if every vote counts, there are simply more waverers up for grabs in the central area, so that’s where rational planners would concentrate the campaign spend.
    Only by preventing said planners from knowing where votes had been cast could you eliminate that behaviour and force candidates to campaign equally to everyone. Never gonna happen.

  6. hyphenman says

    Mano,

    A critical subplot in Shonda Rhimes’ TV show is just this kind of shenanigans happening in Defiance County ,Ohio.

    Jeff Hess
    Have Coffee Will Write

  7. KG says

    The Democrat would be wasting their money spending in the west because it’s in the bag and the people voting against them are likely unpersuadable, and similarly wasting their money spending in the east because it’s a lost cause. Same logic reversed for the Republican. Even if every vote counts, there are simply more waverers up for grabs in the central area – sonofrojblake@9

    Remind me never to hire you as my election agent. The balance between the two parties last time tells you little or nothing about how many people are potentially persuadable either way. Saying “because it’s in the bag” indicates that you haven’t actually taken on board that it’s the total vote that counts – there’s no sense in which any of the three states can be either in or out of the bag for either candidate. You give no reason – there isn’t one – why “the people voting against them are likely unpersuadable” in the state where they get more votes. The information that would be useful is how volatile the electorate is in each state – for which how much the result last time differed from that the time before is a reasonable although by no means perfect proxy. So you’re correct that information about past elections would rationally lead the candidates to focus their efforts, but wrong about what information is actually of use.

Leave a Reply

Your email address will not be published. Required fields are marked *