The election of a president of the United States is not just democracy in circuslike action. It also may be, at its core, the most elaborate and expensive recruitment and hiring process that mankind has ever created.
You can think of a presidential election as being like a particularly large company’s search for a chief executive. In this case, the search costs a couple of billion dollars (the amount that will be expended on campaigns), has a hiring committee of 127 million people (the number of voters last cycle) and is covered at every turn by virtually every media organization on earth.
What, then, can the latest evidence about best hiring practices tell us about the election, in which hiring the best employee has particularly high stakes? Good news: The answers might just make you feel a little better about U.S. democracy.
In the last several years, there has been a lot of evidence, both from academic work and from companies that approach recruitment analytically, that traditional job interviews are not particularly good tools for identifying the best employees.
One conclusion: It is a bad idea to hire someone primarily based on a job interview, or on a manager’s gut instinct. Some people perform better when being interviewed, but that seems to be a self-contained skill. It does not tell you much about whether a person would be a good software engineer or accountant.
“Interviewers are likely to feel they are getting useful information from unstructured interviews, even when they are useless,” concluded a 2013 article in the academic journal Judgment and Decision-Making. “Our simple recommendation for those who make screening decisions is not to use them,” wrote Jason Dana, Robyn Dawes and Nathanial Peterson.
Major companies have come to the same conclusion. Google long had a reputation for putting job candidates through multiple interviews and asking potential employees brain-teaser-type questions to see how they responded. Not anymore.
“We looked at tens of thousands of interviews, and everyone who had done the interviews and what they scored the candidate, and how that person ultimately performed in their job,” Google’s chief human resources executive, Laszlo Bock, told the Times in 2013. “We found zero relationship. It’s a complete random mess.”
Maybe the search committee for the next president – that is, all U.S. voters – should not put too much weight on what they see in unstructured press interviews with candidates. I argue that speeches and debate performances fit the same pattern, in that they are performances that do not tell you a ton about how a person would perform in the vast portion of the presidency that takes place off-camera.
So what does work? Google found that structured interviews, in which applicants are asked uniform, concrete questions, were helpful.
And a large body of work suggests that tests and simulations can give particularly useful information about how a person will perform on the job.
Economists looked at data from firms employing low-skill workers, in which some managers had the discretion to hire the employees they thought were the best fit, while others relied on a performance test. Those hired with tests alone stayed in the jobs 15 percent longer than those chosen by managers, a sign they were a better match. And in an experiment with British doctors, a daylong series of simulations, such as interaction with an actor playing a sick patient, more accurately predicted their future performance than a 90-minute test.
Plenty of companies in a wide range of fields have embraced the idea. Consulting firms like McKinsey have long hired bright young MBAs, in part, by giving them a case study of a business challenge and asking them a series of questions about how they would approach the problem.
If tests and simulations are the way to go in picking an employee, what does that mean for picking a president?
Perhaps the ideal scenario would be to put Ted Cruz and Donald Trump and Hillary Clinton and Bernie Sanders through a simulation in which they must jump between persuading a member of Congress to vote for a highway bill and conducting an arms control negotiation.
That will not happen anytime soon. But in a weird way, we might be seeing a version of exactly that simulation.
What are the qualities it takes to be a successful president? He or she needs to be good at hiring and trusting the right people; making constant big decisions with limited information and often while exhausted; setting the right big-picture strategy; and knowing when to stick with it as circumstances change and when to make tactical adjustments.
If you look at it that way, running a presidential campaign starts to look like exactly the kind of simulation of being president that our search committee needs to pick a president!
The leading candidates in this campaign come from very different backgrounds: two senators, a former secretary of state and a businessman, which makes it hard to judge them against one another. But right now we are seeing an apples-to-apples comparison in how they run a campaign.
Is the campaign disciplined and well run, displaying the aforementioned mix of smart big-picture strategy and clever tactical choices? Does the candidate hire skilled staffers and put trust in people who are competent? Does the candidate have the kind of stamina to put in the long hours of campaigning, or does he or she get tired and say the wrong thing or make bad decisions?
It is easy to criticize the length of U.S. campaigns, and the complexity of the system by which delegates are selected to choose each party’s nominee. But in a weird way, the duration and complexity of the task make campaigns an even better hiring test.
Just as a daylong simulation was a better way to judge the competence of British doctors than a 90-minute test, a long campaign gives more time for a person to flunk the simulation by making bad strategic decisions, listening to bad advice or saying something stupid.
In early 2008, a person would have had plenty of reason to be skeptical about whether Barack Obama, who had served in the U.S. Senate for about three years and had never run anything larger than his Senate office, had the leadership chops to run the entire federal government. The fact that he ran his primary campaign more effectively than Clinton was one piece of evidence that he did.
The U.S. campaign system has evolved over the years without any master plan. But whenever you groan about yet another state primary or long debate, look at it this way: If you were on the search committee to hire a CEO, you would want the longest, hardest simulation you could get. And in your role on this very large search committee, that’s exactly what you’re getting.