Showing posts with label Incentives. Show all posts
Showing posts with label Incentives. Show all posts

Saturday, November 16, 2019

What, Exactly, Do You Want?

(The inspiration for this post comes from here, and from a coffee conversation with The Grinch, who asked me the subject line bluntly during a political discussion.)

It seems prudent to break the subject into three discrete questions, all of which might on their own be considered variations on the subject line, but which are, on reflection, very different.

1. What are the problems of modernity?

If you take people on the right, either outer right or mainstream right, and ask them what are the specific problems with modernity, you will get a surprisingly uniform and well-agreed-upon list of problems. The most ardent reactionary and the most mainstream Republican will probably not actually disagree very much on what are the particularly wretched aspects of life in contemporary America. Open borders. Welfare moochers. Criminals not being punished. Endless propaganda against straight white Christian males. You get the idea.

This is worth pondering. The things which trouble us are a substantial cause of agreement.

2. What would be your ideal solution to fix those problems?

If you took the same range of people, and asked them what kind of governing arrangements they would prefer, they might give you quite different answers. Normie conservatives want a return to the constitution as understood in 1950, or 1920, or 1850, or what have you. Further right republicans might want the authoritarian capitalism of Lee Kuan Yew, or the colonial administration of British Hong Kong. Reactionaries might want an absolute monarch, or Moldbug's sovereign corporations. 

And yet, among the more thoughtful ones, even if they disagreed on exactly what was their preferred model, they would probably also take any other right-winger's governing arrangements over the status quo, if it could actually be achieved. In other words, we might haggle over the ordering, but any of them would likely be an improvement. If agreeing to this were all it took to get it to actually happen, we'd have a reasonable shot at all being on board.

3. How do we get to there from here?

A.K.A. What should we actually do about all this? And on this question, people come to all sorts of answers. I think the reason is that nobody really knows. I tend to believe the Moldbug point that activist strategies are a disaster for the right (see, for instance, Charlottesville). But this is very different from knowing what will work, or even might work. I don't honestly have a great idea, but I'm working on it. When it comes to things like voting, we can't even agree on which direction the sign goes.  

To see how wild the divergence can be between the first point (agreeing on current problems) and the third (what to do to concretely improve things), consider the following description of leftism. See if it rings true:
Almost everyone will agree that we live in a deeply troubled society. One of the most widespread manifestations of the craziness of our world is leftism, so a discussion of the psychology of leftism can serve as an introduction to the discussion of the problems of modern society in general.
But what is leftism? During the first half of the 20th century leftism could have been practically identified with socialism. Today the movement is fragmented and it is not clear who can properly be called a leftist. When we speak of leftists in this article we have in mind mainly socialists, collectivists, “politically correct” types, feminists, gay and disability activists, animal rights activists and the like. But not everyone who is associated with one of these movements is a leftist. What we are trying to get at in discussing leftism is not so much movement or an ideology as a psychological type, or rather a collection of related types.
Leftists tend to hate anything that has an image of being strong, good and successful. They hate America, they hate Western civilization, they hate white males, they hate rationality. The reasons that leftists give for hating the West, etc. clearly do not correspond with their real motives. They SAY they hate the West because it is warlike, imperialistic, sexist, ethnocentric and so forth, but where these same faults appear in socialist countries or in primitive cultures, the leftist finds excuses for them, or at best he GRUDGINGLY admits that they exist; whereas he ENTHUSIASTICALLY points out (and often greatly exaggerates) these faults where they appear in Western civilization. Thus it is clear that these faults are not the leftist’s real motive for hating America and the West. He hates America and the West because they are strong and successful.
Sounds pretty reasonable, right?

So what do you think the guy who wrote this proposed to do to fix this problem?

As it turns out, the completely bonkers answer was to send home made bombs to try to blow up a random computer store owner, geneticist, biologist, computer scientist, and a bunch of others. The quote is from the Unabomber manifesto. He's actually one of the few exceptions to point 2 - I wouldn't actually prefer to live in his entirely pre-industrial-revolution society, even if I can agree on some of what has been lost. It's actually worth reading, if for no other reason than the whiplash disconnect between the quite incisive observations he makes about the problems of modernity, and the absolutely insane things he did that made him famous:

Step 1. Send home-made bombs to random people who do things somehow associated with modernity
Step 2. ????
Step 3. Profit!
Step 4. Worldwide revolution where technology is abolished. 

That said, I think the widespread agreement on point 2 is, in many ways, equally surprising. Because it's not obvious that we should necessarily find some kind of consensus improvement just because we dislike the status quo. And moreover, it suggests that there actually might be something common to all these alternative governing arrangements that would make them an improvement. Why exactly, do we all instinctively support a lot of each others' solutions?

I think there is a value in trying to condense down a description of what it is at a high level we want. Partly, this is because even if one dislikes voting, one can't get around having to convince people somehow. Maybe not the general public, maybe just the elites. Fine. But also as an important intellectual exercise of understanding what the actual problem with the status quo is. What is it that unites Singaporean authoritarianism, British Hong Kong, absolute monarchy, and sovereign corporations, but isn't present in modern democratic states?

The shortest answer I have, which I think can be given in a lot of different settings without offending people, is this.

Well-defined and secure property rights in the state itself.

You want to know what I want? That's what I want.

Who owns the state? And in whose interest is it to be run? Not some b.s. answer like "the people". Which people, specifically, and in which proportion? What exactly is the decision-making process? How are the cash flows going to be distributed?

That's what's so appealing about sovereign corporations. Not only is there one answer, but everybody knows that answer, and everybody pretty much has incentives to continue to respect that answer. Who owns Microsoft? Not "the people", that's for sure. And in a sense, does it matter who specifically owns Microsoft? Not really. What matters is the structure and the incentives. If the shares all changed hands tomorrow, would you expect Microsoft to act very differently? No. The shares will change hands tomorrow, and you won't even bother to think about it when you load up Windows.

But if we all agree that Lee Kuan Yew owns the state, or that the British Civil Service with their well-defined system of promotions and personnel owns the state, or that Louis XIV owns the state, that's probably fine as well. If they really truly own it, they probably will have sufficient incentives to run it decently well. They might screw it up - this is why I generally prefer sovereign corporations to absolute monarchy, because regression to the mean and bad genetic draws periodically produces fools at the top. But if they do, they'll be hurt themselves, which probably gives them incentives to hire or consult someone who understands statecraft properly.

A completely well-defined and secure set of state property rights is probably itself a Platonic ideal that can't actually be achieved. Specifying exactly why the individual soldier or policeman carries out the order in every situation, or where exactly the King/CEO's advisors come from, will always be thorny. But we can probably still rank order different arrangements in terms of approximate stability and agreement on state ownership.

Why do we want well-defined and secure property rights in the state?

Because a stationary bandit is better than roving bandits. 

When you have one bandit, you get expropriated efficiently. When you get multiple bandits, it turns out like the descriptions in one of the War Nerd podcasts of what it was like to have armies coming over your area during the Thirty Years War. Even if they were just passing through, they had to resupply. So they'd come by, steal all your grain, use your furniture as firewood, perhaps rape a daughter or two, and then camp for the night. Then a few months later, another army would do the same, except they'd be annoyed that the first guys had taken most of the good stuff.

Strikingly, in Brecher's recount, it made surprisingly little difference whether the army was notionally friendly to the country in question, or notionally hostile. Either way, you were getting everything taken. It didn't matter if they were taking next year's seed grain, or you didn't have enough to survive the winter. That's your problem.

Such accounts also show the risk of being hyperbolic. In modernity, we aren't getting randomly mobbed as a society in anything like the way peasants were in the Thirty Years War. Why not?

Well, in the modern west what we have isn't several groups of roving bandits. That indeed would be worse. When the bandits are entirely an outside force, and one that will only be here temporarily, you get absolute catastrophe, and they'll take everything. Rather, we have internal banditry by unstable coalition. None of the bandits are individually strong enough to rob you, but collectively they are. The good news is, the bandits are part of the society themselves, so they're here for a while at least. The other good news is that they (so far, and on average) view their coalition as kinda sorta stable  - enough that they can exploit semi efficiently the various groups they eating. And, so far at least, they're smart enough to figure out a way to do that. This is why modern America is grim in many respects, but you'd never confuse it with Mugabe's Zimbabwe, or a the Holy Roman Empire when the Swedish army paid a visit. In a modern context, neither stable coalitions not semi-intelligent leaders exist, for instance, in modern South Africa. Unsurprisingly, it's rapidly going down the toilet. My predictions on that have been correct for as long as I've been writing. If any of my readers are working for the South African tourism board, I suggest the slogan "See it before it's gone", or if you're more optimistic, "See it before it gets worse".

It's still approximately the same group on top over time, or at least there's a fair degree of overlap, which is why the modern west isn't a total disaster. But who exactly is on top, and who is getting more of the spoils, and who is at risk of being cast out altogether - those things are much less stable.

Hence the need for endless political propaganda, because the coalition needs to signal out-group hatred to keep itself together. Hence the ever shifting potpourri of fashionable causes competing for attention, because the relative ownership stakes in the coalition aren't well defined. Hence the endless insistence on public assent to obvious lies, because you need credible litmus tests to determine loyalty. Hence the need to keep bringing in foreign electoral ringers, because some of the important organs of power are still laundered through our four year American Idol contest, and there's the risk that some of the yokels might get uppity.

The best summary of the theoretical solution to all this is still Moldbug's first post on Formalism. Find out who already owns the state, and give it to them officially. That way you only have one hard problem to solve (find out who owns the state), not two (expropriate the existing owners and give it to some new set of stable owners)

The problem is that we have no idea how to define properly who owns the state, because the answer keeps changing. There's also a problem that people who think they should own the state but currently don't may not assent to formalising the current ownership arrangements. And overconfidence makes everyone think they'll get more in the next round of coalition rearranging.

But a whole lot of us instinctively seem to see the appeal in systems where that problem has been solved somehow. Anyhow. King Charles II, Emperor Bill Gates, or the dispersed shareholders in the Disney Company. It doesn't matter much to me. I'll take any of them.

So we're still somewhat back at the problem of point 3. How do we move towards a system where property rights in the state are well-defined?

I don't know, exactly.

But at least I have a more concrete idea of what I'm aiming at, and what might achieve it, rather than having to split the difference between Catholic Integralism and Sovereign Corporations.

And it's phrased at a sufficiently high level of generality that you can say it at a dinner party and not immediately get thrown out.

Maybe that's something. Maybe it's not.

Nobody said point 3 would be easy, or even solvable at all.

Monday, April 8, 2013

Why school group work sucks

If you, like me, were a nerdy type-A personality at school, you probably loathed getting put into groups for assignments. Inevitably, you'd be stuck with some bunch of lazy idiots who could credibly commit to either:
a) not caring if they got a terrible grade, and hence being uninterested in working, or
b) not being capable of getting a good grade even if they did work.

Both of these would get you to the most common outcome - the smart kid does all the work, usually ostentatiously announcing beforehand that the dumb kid is going to screw it up and thus insisting that he leave it alone, all the while still resenting the dumb kid for his idleness. The dumb kid laughs at the smart conscientious kid slaving away like a sucker.

Teachers would always spin you a bunch of junk about this being useful preparation for the real world, and how it was important for you to learn to work with people you didn't necessarily pick.

Looking back now, I realise that this was all a crock of crap - school group assignments prepare you for nothing useful, and all the irritation you felt was in fact completely justified.

The standard complaint is that you're being allocated into groups you didn't pick, and with hugely varying levels of skill. Neither of these really describes the real world. You get to pick the company you work for, even if you don't get to choose who is on every project with you. That said, it's highly unlikely that any semi-competent manager would lump together one guy who knows what he's doing and a bunch of morons who don't. Hopefully, there's a minimum level of competence required to maintain gainful employment, and you're unlikely to be stuck with someone truly awful.

That said though, it's become increasingly obvious that this isn't the real problem that makes school projects uniquely worse than real-life group projects.

No, the real problem with school projects is the following:

Everyone has accountability, but nobody has authority.

In other words, everyone is responsible for the performance of the group, but nobody has the authority to actually order anyone to do anything. If someone does a bad job, or hands things in late, or generally is so clueless that you'd rather do it yourself, there's not much you can do. If it gets really bad, you can complain to the teacher. But they generally don't want to deal with your whinging.

The assumption is that general social sanction for shirking, combined with the fact that everyone needs to work together to get the marks, should be enough to make it work. But isn't it obvious that setting up a mini-communist state for mark allocation is always going to produce a free-rider problem? And that the equilibrium is going to be that the guy who cares about it more does all the work?

This is like some Frankenstein version of real world group tasks. In most corporate settings, you're going to have a boss or team leader who is directly responsible for the team's performance, and can order people to do certain things on pain of getting fired. Hand the report in on Friday at 12pm or you get canned. Simple enough. Your boss may be an idiot, in which case it's a huge pain (of a very different sort). But at least there is a single person with the incentives to see the group succeed, and the authority to make it happen and solve the co-ordination problem.

If people get to pick their own groups and there's multiple assignments, the repeated game aspect can deter shirking somewhat.

But in general, teachers create a horrible system for assignments that simply teaches smart kids that the world is full of moochers, and that you'll end up doing a disproportionate share of the work only to see some slacker enjoy the fruits of your labour.

You might argue that this lesson is crucial for teaching them about the operation of the tax system and pork-barrel public-sector employment, and I wouldn't necessarily disagree.

I'd like to think that this was the well-thought-out plan all along, but somehow I doubt it. The only way to fund these taxes is to have businesses whose internal team dynamics are so different that productivity results and there's a surplus to be stolen in the first place.

Sunday, December 11, 2011

Charging By The Hour

I'm always amazed by the number of jobs and services that are paid or charged by the hour.

Charging by the hour makes the most sense for jobs where the main requirement is actually just to be there for a given amount of time - working at a front desk, or some sort of customer service position that doesn't have a big up-selling component. If you're serving ice cream, the store really needs you to be there for those eight hours, and it makes sense to pay accordingly.

But the most mysterious are those when the job is complicated and based around a particular outcome, not a length of work. Take something like legal work. At least in Australia, the vast majority of it is charged by the hour, with different lawyers having different hourly rates.

This creates exactly one good incentive effect - it means that your lawyer does not have incentives to scrimp and save on the level of work they're putting in. Incentivising your lawyer to spend a few more hours at the office to look up other possible defenses to your embezzlement charge is probably something you actually want to do. This becomes particularly important when the amount of work required is not clear up-front, which may well be true in legal cases. 

But balance this against all the other bad incentives this creates. First, the moral incentives - it punishes honesty. Their inputs are typically unobserved, and thus you create huge incentives to overbill. The honest ones won't, but they're going to be hurt for it. And  I never, never want to incentivise dishonesty. It's not as clear as outright fraud either - since lawyer time is billed out in six-minute increments, there's always a question of exactly how much to deduct. If you do 30 minutes of work, have a 2 minute toilet break, do another 30 minutes of work, grab a coffee and check your work email  for 4 minutes somewhere in the interim, is that 10 units or 11? What if you don't remember exactly when you started? Do you just round it up to some other number? I don't imagine the average lawyer worries about these questions much, but they've come to some accommodation on this, and are getting rewarded or punished accordingly.

The second is that charging by the hour actively punishes anybody who finds a more efficient way to solve a problem. If you find a way to solve a client's legal problems in a tenth as much time, you're going to earn less than the guy who just chugs through hours of sort-of-relevant research. A lawyer who can demonstrate an ability to solve problems quickly in several situations might eventually be able to charge a higher rate. But clients who don't know their reputation might shy away from someone who charges a lot. And what if the client has very little idea how long the task should take? Then they won't even appreciate the fact that you've been efficient.

The law isn't even the worst example of hourly contracting - having web designers being charged out by the hour is even more barmy. Surely you can have a rough fixed-price estimate of how much to charge for a basic web page?

Generally speaking, wherever I have a choice I'd rather pay people a flat fee for whatever the service is, and then build in explicit incentives where necessary ($50K upfront, another $50K if I'm found innocent, another $25K if my sentence is less than 3 years, etc.). Of course, it's hard to negotiate this with your lawyer, but when I'm writing the contract I'd rather do it this way.

More practically, I do have one choice available to me, and it is this - I don't want to work in any industry where people are charged out by the hour. And I especially don't want to be someone else's employee in that industry where I'm being paid a fixed salary but being charged out by the hour. It means that it's going to be very hard to have any sort of work-life balance and be successful, because the incentives of the firm are just to have you spend more time at the office.I also don't want to worry if I'm doing work too quickly from the perspective of the firm. Having to do a worse and slower job because the firm effectively demands it would be soul-destroying.

Tuesday, September 20, 2011

What's Italian for 'Incentives'?

From Boston.com, comes this absurd story from Italy:
Seven scientists and other experts went on trial on manslaughter charges Tuesday for allegedly failing to sufficiently warn residents before a devastating earthquake that killed more than 300 people in central Italy in 2009.
Hindsight bias! Get yer hindsight bias, piping hot and fresh from Italia! Cheaper when purchased by the gallon!

Let's put aside for a minute any question of actual moral culpability, and whether it's fair to hold people criminally liable for scientific predictions. Instead, let's assume that we're just interested in saving as many lives as possible and increasing the accuracy of earthquake forecasts (although astute readers will note that even this phrasing involves tradeoffs between the two goals). Does arresting seismologists who fail to predict earthquakes sound like good policy?

The trouble with incentives is that you need to think very carefully about whether you're actually creating the incentives you think you're creating.

I imagine the Italian authorities have the following mental model:

1. Massively increase the penalties for wrong predictions
2. Lazy and corrupt seismologists put more effort into getting the right predictions
3. Save lives!

But this is assuming the continued existence of seismologists making predictions in the first place. If you don't do this, then you have the following:

1. Massively increase the penalties for wrong predictions, while not increasing the payoff from correct predictions
2. Result is massively reduced payoffs from making any seismological predictions at all.
3. Seismologists quit en masse, or refuse to provide any guidance.
4. No more predictions.
5. Earthquake deaths.

Not quite the same thing, is it?

Okay, let's put Mussolini in charge! He'll chain those lazy seismologists to their desks and force them to keep making forecasts! Surely we'll be okay then, right?

Or maybe you've then set up the following incentives:

1. Massively increase the penalties for type II errors (i.e. false negatives: saying there won't be an earthquake when there will be one), but not increasing the penalties for type I errors (i.e. false positives: saying there will be an earthquake when there won't be one).
2. Seismologists predict an earthquake every single day of the year
3. People quickly learn to ignore all the warnings
4. Signal to noise ratio in earthquake predictions goes to zero, meaning that there are effectively no more meaningful earthquake predictions
5. Earthquake deaths.

Bureaucrats seem to think that ramping up penalties will create scientific knowledge where there was none. But it won't. All it will do is shift around the incentives and behaviour of agents in ways that are completely predictable, even if they are almost certainly not predicted by the bureaucrats.

In the mean time, you can expect displays of righteous indignation by prosecutors, giddy with hindsight bias, exactly sure that they knew there was going to be an earthquake, so why didn't those lazy seismologists? What you won't be able to get from them is any useful guidance on when the next earthquake is actually going to happen, notwithstanding their amazing hindsight powers.