What is the science of winning an election, and why don’t politicians use that knowledge?

What’s the science of winning an election? We can know what the winners do, but that’s never made for good science, since you can’t isolate one variable at a time. (It’s a bit like looking at the list of the wealthiest Americans and trying to write a list of how to get rich. The winners tend to leave out the part about their rich dad who started everything for them.)

What we do know is this: most political scientists speculate that campaigns can influence the outcome of an election in the neighborhood of 2-5%. Last month I interviewed Christoper Wlezien, author of The Timeline of Presidential Elections. After decades of research, he told me that the best and most reliable poll is the one taken right after the conventions, which has decided 14 of the last 15 elections. Every other change is merely a “temporary bounce,” which disappears after a few weeks. Permanent bumps, which have long-term effects on the campaign, are much more rare.

We have mountains of data about what actually works in campaigns. For example, the quality of interaction matters a great deal; mobilization techniques like canvassing and personal phone calls are much more effective than robo-calls, emails, or snail mail. TV ads are effective, but short-lived. People remember negative ads for a long time, but they don’t necessarily draw people to the other candidate. Because candidates are so convinced that everything on the list is necessary–from travel to TV ads–campaigns often end up engaged in counterproductive behavior. Candidate A decides to blanket West Virginia with a certain message? Candidate B steps in and targets the same territory. The net effect is zero.

So why don’t candidates ever seem to follow any of this advice? Old habits die hard.

1.) Campaign organizers need the work. They’ve built an entire industry dissecting and building up each step. The more complex and unwieldy they can make the process of a campaign seem, the more they can charge, and the better job security they have. Candidates are far more likely to listen to the smooth-talking guy who’s gotten them this far than a disinterested professor who’s carried out tons of research.

2.) The other reason why campaigns engage in frustratingly inefficient behavior is a little more perverse. In the U.S., we expect campaigns to run this way. Yes, they’re inefficient, insanely expensive, and generally annoy everyone within 20 feet of cable TV or a loud-mouthed relative for months on end. But having elections any other way would feel weird, even if it were a better way.

The take-away: there might be mountains of data telling you what you need to do to win, but don’t underestimate other people’s desire to stick to the status quo. Of course, you might be the one secretly wanting to stick to the status quo. Or maybe your high school friends don’t want to you grow up. Perhaps your spouse or partner is comfy-but-not-incredibly-happy. No matter how many tricks, lifehacks or secrets you know, your ability to change your own behavior will ultimately make the difference between whether or not you’re able to live the life you want.