There have been a lot of surveys of career and job satisfaction done over the years by both social scientists and marketing companies. It’s an easy thing to fund, because it lets executives quantitatively examine the productivity of their underlings as if they are the mechanical parts of a vending machine while at the same time pretending to care about them.
The data is messy and variable—that’s how SS rolls—but there are some surprising and useful trends. One of the big ones is that when asked “what do you want more of,” the almighty dollar usually comes in at number 2. A huge blow to its ego, I’m sure, but at least it can afford therapy. The number one slot on what people want more of in at work is recognition. Validation. Reward for their efforts and their successes, not in the form of a company Mercedes, but in the form of the boss saying, “Damn, Susan, you rocked the Peterson v Anderson brief! I’d hate to be Anderson’s family right about now!”
(Side note: a disproportionate number of people interviewed in these studies is named Susan. The social sciences have a serious Susan problem.)
As humans we crave validation the way pandas crave artisinal bamboo. Being told we’re a Good Boy worked in our previous incarnation when we were all dogs, and it still works on us today. Most people can remember a job, or a project or phase within a job if they’re less lucky, that was more satisfying than usual simply because they felt we got the recognition they deserved.
To some extent, managers and supervisors understand this. It’s obvious, right? Most positions have some kind of performance metrics that are rewarded or punished accordingly. Hell, a lot of industries base their entire employment model on this principle. So why is it that so many people, even within the fields that give out the most bonuses and commissions, are so unsatisfied with their jobs?
There are a lot of reasons, including the fact that the carrot and stick model is demonstrably a terrible way to motivate people to effectively perform complex tasks. But there’s another one that doesn’t get as much attention, and causes many of these schemes to produce exactly the opposite effect they are intended to.
It is this: a great many structures designed to measure and recognize performance have a hilariously low correlation to actual performance. A lot of people out there, from vacuum salesmen to CEOs, are getting rewarded for accomplishments that they didn’t really achieve, or punished for failures that are only sort of their fault.
In his excellent book Misbehaving, behavioral economist Richard Thaler describes a consulting gig he did at a high powered company. In the room was the CEO and 22 top executives. He proposed the following hypothetical to the execs: Let’s say you have a potential project you could undertake, and it has a 50% chance of success and a 50% chance of failure. If it succeeds, the company makes 2 million dollars. If it fails, the company loses one million dollars. In economics terms this project has an “expected value” of 1 million dollars. That is, the average amount you are likely to gain if you undertake this project is a million dollars. On any given go you could win or lose, but if you did it enough time one million dollars would be the average. He asked the execs if they would undertake this project. Only three of them said that yes, they would. The others said it would be too risky.
Thaler then turned to the CEO and asked him if he would sign off on 22 projects like this. That is to say, 22 projects run by his executives that would collectively early the company, on average, 11 million dollars.
“Of course,” said the CEO. “That’s a no brainer.”
“But you won’t get 22 of these projects,” said Thaler. “You’ll only get 3.” Because indeed only three of the execs said they would take the project on. Thaler turned to one of the people who said they wouldn’t undertake the project and asked why.
“If it makes money, I get a pat on the back, and maybe a bonus worth three months salary,” said the executive. “If it fails I might very well get fired. I like my job. I’m not willing to risk it for three months salary.”
It’s pretty clear that the fault here does not lie with the executives for being unwilling to take the risk. It lies with the CEO, who wants the risk in aggregate but is doing a terrible making his employees want to take it. But it raises another, more fundamental question. Let’s say all 22 projects go through, and some people get bonuses and some people get canned. You might think that’s good strategy. After all, it separates the wheat from the chaff, right? The good executives who made the project work get to stay on, and the bad ones are shown the door.
Maybe not. In the real world, it’s often very, very difficult to tell how much of these kinds of achievements are down to individual skill and merit and how much they are down to chance. We usually assume that people who succeed are good at what they do and people who fail are bad at it. We are a lot less likely to assume this about ourselves.
The hesitancy of these executives in taking the projects suggests that they feel this in their bones. No matter how smart or talented they are, no matter how hard working, they won’t necessarily be able to pull off a win on the project. Someone who had a completely merit-based worldview would assume that they would be in the 50 percent who made the 2 million. Maybe three of them had that kind of confidence. The others know that sometimes, the souffle is going to collapse no matter how carefully you folded in the egg whites.
On a much smaller scale, the morale at my current job is generally fairly low. There are a lot of reasons for this, but more and more I think the low-correlation effect is a big one. We have stats that track a wide variety of metrics. There are about 12 that matter, and each of them is broken down into another dozen or so sub-metrics. Our stats determine how much money we make, our prospects for promotion, and how well we are regarded by our supervisors. Having low stats is depressing.
When our stats as a site are low there are a lot of messages from management about how we need to get this done, how big a deal it is, how badly we are collectively doing. None of it is particularly harsh or punitive, but it is always there, buzzing in the background. Management tries their best to counter the effect of these messages by buying us pizza and giving out prizes and having funny hat day. But none of that addresses the fundamental problem: many of these stats are out of our hands.
Not completely, of course. And before I’m accused of making excuses for my own performance I should say that right now my stats are very good. I had a slump for a while, but even when I did they were still pretty good for the site. So this isn’t about me. At the same time, I’ve been in a situation where individual metrics spiked in the wrong direction purely because of luck. I got a lot of the wrong sorts of calls within a few days of each other, and my stats plummeted.
When I ask my supervisors about this, they always tell me the same thing. “Just do your best on those calls. You won’t get in trouble for a bad result as long as you handled it well and did everything right.” And on an case-by-case level it’s true. If they listen to a call where I give a customer a large credit and they find that I did everything I could to bring it down and save the company money, they’ll tell me I did a good job. That even though it was a large credit, it was still a good call. Hell, I’ve even gotten compliments on calls like that from management for how well I handled them. Compliments, and a nosedive to my stats.
The lesson is clear: even if you do everything right, the only meaningful measurement of your performance can tank due to bad luck. Sure, it’s not totally out of our hands. The people with the consistently worst stats are usually the worst performers, and there are agents whose stats are consistently higher than average. But everyone on every level is fully aware of how much certain of these elements are out of our hands.
In industries across the spectrum this kind of thing is very, very common, and it is extremely dangerous. If morale matters—and study after study shows how much it affects productivity, even if you don’t think businesses have any responsibility to the psychological well being of their employees—then the incentive structure needs a good, hard look, and probably a serious talking to.