We have one year to go before the US presidential election, and this seems as good a time as any to start talking about manipulation in the context of elections. Not to say that you’ll get 24 posts about elections, but we’ll try and make that a recurring theme over the coming year, especially since so many of us have been involved in election strategy and tactics over the years.
So, let’s say that you’re the campaign manager of a US presidential candidate, and you’re tasked with delivering a state – let’s go with Ohio, which gets a lot of press. What do you do first?
It’s important to first understand the scope of the problem. Ohio has around 11M people, and convincing them to all vote your way seems like a lot of work. But wait! There are several things that can help you here:
1. First, most people don’t vote. Yes, it’s strange to say, but most of these 11M people won’t vote. First off, several are the too young to vote, so let’s just take the voting-age population: 8.8M people. Take out the prison population, and the non-citizens, and you’re left with 8.5M people that can vote. Out of those, a huge percentage will not bother to vote: in 2008, for example, only 5.2M citizens voted for president in Ohio. So you don’t need to get the majority of 11M people to vote for your boss, just the majority of 5.2M! Your workload is already halved! Well done!
2. Many of these votes are already spoken for. Okay, so you need at least 2.6M voters to vote for your boss, not an easy week’s work. But wait! It turns out that you can predict, fairly accurately, how some districts will vote based on who lives there. For example, let’s take two counties, Warren and Cuyahoga. Warren is a relatively upscale county, with a median income of $58,000. Warren is solidly Republican – in 2008, it voted two for one in favor of McCain. Cuyahoga is poorer, with a median income of just $39,000. It is also heavily Democratic, with a two-for-one vote for Obama in 2008. The thing is, for all the talk of Ohio as ‘swing state’, most counties in Ohio vote fairly consistently:
Notice that some counties remain stubbornly red or blue, election after election? You can actually see that interactively here, but the point remains that you are unlikely to convince the voters in Trumbull, for example, that have voted Democratic for the last 5 elections, to switch to a Republican candidate. But you can count on Fulton County, proudly (and exclusively) Republican since 1960. So, realistically, before you really start your job, you can reasonably count on support from around 1.2-1.4M people who are solidly Republican, no matter what. They will vote for you because their father voted Republican, and his father before him, and on all the way back to Secession. So really, you only need to convince around 1M people or so to vote for your boss to win the state. Well done, again – you’ve halved your load again!
3. You have a lot of money. Let’s summarize, here. For your boss to sweep Ohio, a state of 11M people, it turns out that you only need 2.5M votes or so, and most of these are already ‘yours’ because of party affiliation. Now, convincing the remaining 1M or so people to vote your way can still seem like a lot of work, but your candidate did give you some cash to work with. McCain, for example, spent around $16M in Ohio during the 2008 campaign. $16M is a fair bit of change – around $16/person, in fact – but of course, what matters is how you use it: do you for (expensive) TV ads, and if so what do you say? Do you do a direct mail campaign? Pay for staff to man the phones and try to get more of your voters out on election day than the Democrats? That’s where your skill comes in (that’s why you make the big bucks!) but you could run ads like these (touting your commitment to renewable energy, ironically enough):
Even with $16M, convincing a pile of people to vote for you is still going to be a lot of work. What about if you could, instead, convince some of these 1M people to NOT vote for your opponent?
Meet Mr. Russel Pierce, from Alabama, who just lost an election despite having some very creative thinkers on his side. Mr. Pierce was the force behind the controversial Alabama immigration bill SB 1070, which essentially made it mandatory for people who ‘looked like’ immigrants to carry their papers with them at all times or face arrest. Many people took offense at this, and ran a campaign to recall Mr. Pierce (the first of its kind which has ever succeeded, by the way). That’s not the manipulative part here.
Leading the effort to recall Mr. Pierce early on was, ironically enough, Ms. Cortez, a Latino woman who wanted to recall Mr. Pierce and… well, it was never very clear. Ms Cortez’s campaign was, to say the least, weak. She rarely spoke to the press. When she did, she gave inarticulate and feeble answers.She had very little marketing, and did little advertising, other than Spanish street signs…
All in all, then, it was shaping up to be a comfortable race for Mr. Pierce.Ms Cortez was Latino and would likely appeal to Latino voters, but she was weak enough that most of her support would be weak, and unlikely to materialize come election day. And, as mentioned above, it doesn’t take a lot of voters to stay home to change the outcome of an election.
This was exactly the plan hatched by Greg Western, an Ally of Mr. Pierce and Tea Party strategist. Greg had a simple idea: since getting more votes for Mr. Pierce was hard, how about creating his own opponent? An opponent that would be the face of the opposition to Mr. Pierce, but weak enough that supporters couldn’t really rally around her? How about a Latino woman who had no experience in politics, and was not very articulate, for example?
So Greg Western, a close Pierce ally, recruited and became the campaign manager to Ms Cortez – to run against Mr. Pierce! It became somewhat surreal, in fact: Mr. Pierce’s nieces worked to gather the signatures needed to put Ms. Cortez on the ballot, and his own brother drove them around to help gather enough support to put Ms. Cortez on the ballot.
It was an interesting manipulation – creating your own opposition, strong enough to register (i.e. get her name on the ballot), but weak enough to create tepid support and entice voters who wanted to recall Mr. Pierce to stay home on election day. The only reason the manipulation failed is that another challenger stepped forward, with a lot more political experience, and eventually found out about the sham candidacy and sued to remove Ms. Cortez’s name from the ballots (which was not done, by the way).
This kind of trick is hard to pull off in national presidential elections, of course, but it does illustrate quite well one of the core rules of election mathematics: you can actually change the outcome of elections – even presidential elections – by shifting a very small number of votes. Obama only won Ohio by less than 300,000 votes in 2008, and that was enough to give him a very strong lead over McCain. In fact, it’s an even smaller margin than this – given the strong party support in various counties, the Obama campaign only really had to ‘switch’ maybe around 100,000 voters in some key counties to win in Ohio overall. 100,000 voters is not a small number, but that’s just a little over the capacity of the Cleveland Stadium..
* * *
As we gear up for a year of electioneering, keep in mind that winning elections, at the tactical level, is all about getting a relatively small number of people to vote your way, or, as Mr. Pierce showed, to at least NOT vote for your opponent. A lot of the manipulation that we will see will not be focused on broad policies or large promises – but on the nitty gritty work of shifting small packet of votes here and there. Remember Joseph Stalin’s words:
“Those who vote decide nothing; those who count the votes decide everything.”