DYNOMIGHT ABOUT RSS SUBSTACK

Litter and circles of internalization

Litter and circles of internalization

Jul 2022

Say you’re a dictator fed up with your citizens’ littering. What can you do?

Option 1: Make littering illegal.

This is obvious, but it’s clearly no magic bullet since everyone’s already done it and yet littering hasn’t perished from the earth.

Why? Primarily because it’s hard to catch people. You could put police everywhere, but this would cost more than it’s worth.

Option 2: Cultural norm

In some places, dropping trash on the street will result in someone picking it up and running up to helpfully return the object you “accidentally dropped”. A strong social norm like this is very effective and largely self-enforcing.

But say your citizens don’t have a norm. You can run a bunch of ads to try to change the norm, like the US did this the 1970s using this charismatic Native-American Italian-American actor:

crying "indian"

Probably this helps. I think these ads shifted the norm away from everyone littering all the time for no reason. But it wasn’t enough to get people to punish defectors, so some people still litter.

What’s next?

Option 3: Game theory

The average penalty for littering is

   (Expected cost) = (Probability of getting caught) × (Penalty if caught).

If your citizens are rational, all you need to do to stop littering is make this expected cost larger than the benefit/convenience of littering. If it’s hard to increase the chance of catching people, the other option is to increase the penalty. For example, if people get $1 of value from littering and get caught 0.1% of the time, then the penalty needs to be at least $1000.

This idea traces back to Becker in 1968, but it’s rarely used. Some people doubt if people are “rational” in this way. I’m sure it’s not a perfect model but come on—in Singapore you can get a $3600 SGD fine and 12 hours of coercive labor for discarding a single cigarette butt. Guess how much litter Singapore has?

(In many US states you can theoretically get hefty fines for littering, but this is seldom enforced—in many places you can throw trash on the ground in front of police with impunity.)

The bigger objection to this policy seems to be that people are repelled by random huge penalties getting handed out for minor misdeeds.

That objection doesn’t particularly resonate with me. Life is kind of tragic and high variance by nature—stock markets are uncertain, you could have a brain aneurysm at any time, and we accept firefighters risking death for the greater good. If we had 100 policies that each increase average welfare but add a bit of variance, implementing all of them would make almost everyone better off. And you don’t have to litter.

But I still don’t endorse this solution, for three reasons.

First, it’s unpopular. An eagerness to die on tiny hills is essential for any blogger (split-brain experiments are overstated; fight me) but this isn’t worth it.

Second, we don’t want to eradicate littering—only reduce it to the socially optimal level. If you’d have to risk dropping a baby to pick up a scrap of paper, you should leave it. This means that this policy will sometimes hit people with huge penalties for doing things we want them to do, which seems weird.

Most importantly, I worry that random harsh punishment could harm relations between the police and the public, making it harder to stop robberies or whatever. Even a small effect here would make the cure worse than the disease here, I wonder if this is part of why it’s so hard to get police to enforce existing littering statutes.

Option 4: Tax litter creation

Some cities tax businesses for creating litter. For example, Oakland California, “to maintain safe, clean and sanitary streets, sidewalks, and public spaces”, taxes fast food restaurants and convenience stores according to their gross receipts:

Annual gross receipts Excess litter fee
≤$4,999 $0
$5,000 - $499,999 $230
$500,000 - $999,999 $910
≥$1,000,000 $3,815

I think this is pretty idiotic. You can sell 250 $20 sandwiches wrapped in rapidly-dissolving wax paper that your customers recycle while I sell 5000 $1 cups of coffee in huge styrofoam cups that my customers crumble and throw in the ocean. As far as this fee is concerned, we’re equivalent!

This fee creates no incentive for anyone to solve the litter problem. The only effects are to slightly increase prices at these businesses (and slightly decrease the number of such businesses). That has some effect on litter, but it’s tiny.

Collective internalization

To review: In an ideal world, everyone who litters would pay a small fine equal to the cost they impose on society—perhaps on the order of a dollar. But that’s impossible without a panopticon surveillance state, so we’re left with poor solutions:

  • Try to shift cultural norms.
  • Create super-high penalties that are rarely enforced.
  • Crudely tax businesses based on what they sell.

Instead, here’s what I propose: Tax businesses based on the amount of litter their customers create. Something like this:

  • Each business has a QR code that they put on everything they sell.
  • After litter is picked up in parks or whatever, the codes are scanned and businesses are charged to social cost of litter.
  • (You could also scan just 1% of litter and charge businesses 100x as much per item if that’s easier.)

The idea is: OK, maybe we can’t internalize the costs of littering on the single actors who do it. But we can still internalize costs in a coarse-grained way.

How would businesses respond to this new incentive? I’m unsure. Perhaps they would reduce packaging, or have some kind of deposit, or put inspirational text on their products, or band together to create a private litter-removal army, or something else.

I have no idea what would work, but that’s the point. (Markets create unforeseeable emergent structures and all that.) At a minimum, surely businesses could find some more effective solutions than the government has now.

Now, perhaps this kind of “collective internalization” is unfair. Say you never litter but you live somewhere where lots of others do. Under this scheme, your local businesses will price in littering fees or deterrence schemes, and you’ll have to pay them even though you aren’t responsible.

That doesn’t seem fair. But I guess we have lots of situations like this already?

  • Ever taken an international flight and wondered why the airline was so fussy about your visa situation? The reason is that if airlines carry someone to a country without a valid visa, most governments make the airline pay a fine of thousands of dollars.
  • In the US, health insurance is often provided by employers. To save money, these employers will often encourage their employees to have healthier lifestyles.

But I think these understates things—they make it look like collective internalization is a policy choice, whereas it’s usually just the natural order of things.

For example, in some cities police do little to stop shoplifting. So, in areas where people shoplift a lot, stores must raise prices to either eat the losses or pay for private security.

But really, nothing is free. Even if the police do enforce shoplifting laws, someone is paying the taxes for them to do that. So people are still collectively punished for living in a city with shoplifters.

And even for litter, what happens now? If you live somewhere where people litter, you’re already being collectively punished by having to look at everyone else’s garbage!

In the end, collective internalization is almost impossible to avoid, it’s just an issue of how big we draw the circles, and how efficiently we design the incentives.

Or… perhaps businesses would respond to this tax by building models for individual customers’ disposition to litter and charging them accordingly. That would eliminate the collective punishment aspect, but it’s pretty horrifying from a privacy perspective.

Comments at substack.

new dynomight every thursday
except when not

(or try substack or rss) ×
Historical analogies for large language models

ice, chess, tractors, railroads, swords, sword fighting, mass production, pre-fab homes, photography, Segways, scissor doors, pornography, calculators

How will large language models (LLMs) change the world? No one knows. With such uncertainty, a good exercise is to look for historical analogies—to think about other technologies and ask what would happen if LLMs played out the same way....

Do economies tend to converge or diverge?

History suggests that poor countries have the same growth as rich countries on average, just with much higher variance.

I used to have a mental model that economic growth was about: 1. Figuring out clever ways to do stuff. 2. Doing it. Rich countries are at the technological frontier, so they have to do both of these things at...

Sales tax creates more unnecessary pain than value added tax

Explains through a model of selling coconuts while value added taxes distort the economy less than sales taxes.

It turns out that sales tax has a huge, gigantic, terrible flaw: It punishes specialized businesses. A value added tax (VAT) has no such problems.

Why I'm skeptical of universal basic income

Common arguments for universal basic income are about liberty or automation of jobs. The "real" argument is much more mundane.

Universal basic income (UBI) is an odd duck. Proponents range from futurists to libertarians to social democrats. Why this weird range of people?

Comparative advantage and when to blow up your island

Explains comparative advantage through trades of coconuts and bananas. Also, why it might be smart to blow up your island.

Economists say free trade is good because of “comparative advantage”. But what is comparative advantage? Why is it good? This is sometimes considered an arcane part of economics. (Wikipedia defines it using “autarky”.) But it’s really a very simple idea....