Fixing Our Incentive Structures

Introduction

One of the things we’ve learned over the past eight years is that a lot of what we’ve taken for granted about the way our world works is based on norms, not laws—un-forceable and unenforceable expectations of good behavior. We operate on the honor code far more often than many of us realized.

That’s not likely to change. There may be places where legislation could be helpful, but a lot of what we need are changes to “incentive structures,” the sets of ordinary, informal rewards and punishments that guide our everyday behavior. We can’t legislate every single aspect of our life, and we wouldn’t want to.

Where could we think differently about our incentives and how they drive behavior? Here are a few ideas.

Scenes from a Broken Hand is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

In Education

I wrote about this issue last year in reference to ChatGPT and what its effects on schooling might be. As I said back then:

If students believe that the most important thing is getting an assignment completed and getting a good grade, they will use Chat as they’ve used dozens of other tools throughout the years—as a way to cut corners or cheat. That this is the most predictable result says more about our values than it does about the technology.

The easy-but-wrong response to a new threat is to legislate it away. In schools, that means banning books, devices, tools, etc. Some schools and districts are doing exactly that, but I think it’s always losing a battle. And it doesn’t speak to the real problems. When it comes to plagiarism and cheating, the incentive structure in our schools tells students that scoring matters more than learning. So, how do we change that incentive structure?

The most radical thing we could do is get rid of grades and rankings entirely, but I don’t see that happening, ever. The second most radical thing would be to assign only a single grade, at the very end of the course, representing final mastery of the material. I don’t see that happening, either, although I wish it would. Some students have to work their way towards learning; it takes them a while to “get it” and succeed. Is that really something we should punish or downgrade? Because that’s exactly what we do, while rewarding most lavishly the students who “get it” on Day One and skate easily through the rest of the course. Why should I be punished if I start the course getting poor assignment grades? Isn’t the end result what matters? Why do we have to average all the grades earned during the course? Aren’t we actually punishing effort in this structure? Do we really want to be telling young people that the way they start something will forever hobble and haunt any future progress they manage to make? That doesn’t sound very “American Reinvention” to me.

If we can’t change the entire grading structure, we can mitigate against its harms in smaller ways. One way is to give students opportunities to revise work and retake tests as often as they want to. I would rather tell students that every one of them has the ability to get an A grade on a paper—and that whether they get it or not depends entirely on them and how much work they’re willing to put in. Here’s the scoring rubric: here’s what you’re aiming for. I will give feedback at every attempt, to guide you towards a better score. You can stop at one draft or push through to three or four—as many as it takes to get where you want to go. You decide when to stop. You decide what’s “good enough.” If you stop short of the A grade, that’s your choice, not my semi-arbitrary, one-moment-in-time judgement.

Some people object to this idea, saying it encourages students to be lazy and disorganized, to avoid studying for tests or turning in work on time. Maybe. Maybe, for kids who are already in a position to deliver on-time. But even for them, they’ll eventually figure out that the less up-front work they put in, the more work they’ll have to put in later, to get it right.

Either way, which structure do you think better prepares students for success in life and work—the one I just described, or a year of unannounced pop quizzes? I don’t know about you, but I haven’t had a pop quiz in a very long time.

Another way to incentivize actual learning over corner-cutting and cheating is to make demonstration of skills and knowledge more public than private. Have students stand in front of peers—and sometimes adults from the school and the community— to talk the talk and walk the walk. Let people engage them in conversation about what they know and what they can do. This doesn’t have to be a scary inquisition. If students have done the work, it can be very exciting and give the students a lot of pride over what they’ve accomplished. At the end of high school, where I’ve seen it done, it can be a genuine rite-of-passage, where adults accept young people into their community.

In Social Media

My wife and I were talking about this the other night. How can you incentivize people to behave better online without limiting or chilling their speech? Some people talk about getting rid of the ability to post anonymously, but that’s tricky—there are people out there who can’t be honest without hiding behind a mask, and they may have important things to say.

Here’s an idea we came up with. What if the platforms used AI to assess the text of each post for its tone, and then added some kind of flag or marker based on its kindness, supportiveness, or helpfulness vs. its cruelty or mockery? Maybe it’s just positive vs. negative vs. neutral. This might not be meaningful for an individual post, where the tone is pretty obvious on its own—but an overall percentage score tied to the poster’s profile could be really informative and tell you whether or not this is someone you actually want to engage with. It would also be useful and revelatory to see what your own percentage is. (“My stuff is only 62% positive? I thought I was much nicer than that!”)

What would this change? First, it might make each of us a little more thoughtful about what we put out into the universe, whether under our own name or under a pseudonym. We don’t want a system that stops us by force, but we should want a system that makes us pause and think a little before acting. Right now, the algorithms incentivize outrage, because high and hot emotion leads to more engagement. But is that really good for us? I think we would want kindness and helpfulness to be better incentivized in any places where we meet in public—or at least for it to have a fighting chance.

Better incentivized, maybe, but not required. Sometimes we want someone to say something outrageous and mean in public. Sometimes we need to vent, and we want others to hear it (and we enjoy hearing others doing it). That’s healthy. Criticism is important. Figuring out how to evaluate when criticism is healthy and supportive rather than demeaning and cruel would be a challenge, but not an insurmountable one, I think.

Some people might want to ban and utterly prevent the demeaning and the cruel. That’s where we are right now. What constitutes hate speech? What constitutes protected speech? I’m not talking about any of that; I’m simply talking about identifying kinds of speech. Which is tricky enough, I admit—and something many people would object to. But I feel like, if angry venting is you want to do, every post, every day, we you ought to be willing to own that. If I use my social media to spew obscenities at the world, I shouldn’t be coy about it or take to my fainting couch if my profile says something like, “98% Negative.” No one’s going to stop me from saying something nasty, but if I choose to say it, it ought to…count. Somehow. It should be noted.

It might help create a better “buyer beware” environment than we currently have. No one’s speech would be limited, but a little flag next to someone’s name, telling me they’re generally positive and supportive, most of the time (or generally negative and mean), would give me information I could either use or ignore. That would mean a lot more to me as a reader than a blue checkmark that can simply be purchased.

And: if I could see an aggregate score across the entire group of people I follow, and also the group that follows me, I might learn something about the people I like to follow—and about myself: what I’m attracted to, and what kind of people are attracted to what I put out.

Would any of this really change the social media environment and culture? I don’t know. But it feels like maybe it could. At least it might make the incentives for good and bad behavior more balanced.

Obviously, it isn’t a foolproof scheme. The rubric for what constitutes positive vs. negative vs. neutral would need to be public and transparent, and people should have the opportunity to give feedback and even appeal judgments they think are unfair. It wouldn’t work perfectly on Day One, that’s for sure. But over time, with feedback and training, it’s something that AI should be able to do pretty well, I think.

In Politics

I don’t want to go too deeply into this one, because better informed people talk about it all the time. We know our incentive structure in politics is out of whack. The Electoral College privileges empty land over actual population. Gerrymandering pushes us to nominate more and more extreme candidates. First-past-the-post voting enforces our two-party system and keeps diverse voices out of the mix. These are all things that legislation could fix, if we could convince people in power, who have benefitted from the existing system, to change it (good luck with that).

At an individual level, part of the problem is convincing or forcing people to be more virtuous and ethical in political life. And part of the problem is convincing more virtuous and ethical people to get involved in politics.

It’s an old problem. James Madison understood it pretty well:

If Men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary. In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and the next place, oblige it to control itself.

As did John Adams:

Morality and virtue are the foundation of our republic and necessary for a society to be free.

Right now, our incentive structure seems to reward blowhards, con-men, carnival barkers, narcissistic sociopaths, and loons. We could certainly do better. How we get from here to there, I don’t know.

Conclusion

We don’t have to keep doing all the things we’ve done for hundreds of years, just because of inertia and lack of imagination. We may not have the power to make all the changes we want, but at least we can start thinking about what changes we do want. We can start by imagining a better end-product and then working backwards to design a system that would more reliably, more often, deliver that result.

Then it’s just a matter of demanding it and making it happen.

I know, I know…easy to say. But the saying is a start.

Share

Leave a comment