Some thoughts on intellectual property

This is the classic defense of intellectual property:

Suppose you invested time and money in order to come up with a new idea (a novel, a new method for refining some chemical, whatever). If you didn't have exclusive rights to this idea, competitors could take advantage of your initial investment by using your new idea for free. This would destroy the profits you could make from the idea, and therefore, we should (temporarily) grant exclusive rights on new ideas to their inventors in order to incentivize innovation.

One potential response to this argument is this: why not grant exclusive rights to the idea forever? After all, isn't this what we do with other property? If I invest the time required to grow a carrot, someone can't take it away from me just because he waited a certain amount of time. It's not like I have a monopoly over the carrot business -- anyone can grow his own carrot. Likewise, if you grant me rights on my novel forever, it's not like I control the whole book industry. Write whatever you want, just hands off my book.

The problem with this argument is that carrots and other physical goods differ from novels and other intellectual goods in the following sense: if you take my carrot, I no longer have the ability to enjoy it. On the other hand, you can copy my novel without affecting my ability to read the original.

Put another way, the primary social cost of intellectual property protection is that it allows patent / copyright holders to sell a good for greater than its marginal cost. This is bad because you'll end up with situations in which I value a recording of a song at (say) $5, it would cost society $0 to give it to me (just copy the file), but I'd have to pay $10 to buy it (the price the copyright-holder sets). Under this regime, I won't by the song, and society loses $5 of value ($5 - $0: the surplus from my being able to listen to the song).

Intellectual property policy is a balancing act: You want to make patents last long enough to properly incentivize innovation, but the longer they last the more value society loses to monopoly pricing (e.g., having to buy songs for more than they cost to produce ($0)).

How do you strike a balance? One thing to consider is that the benefit to the creator of the marginal year of exclusivity declines exponentially (in the real sense of the word). Why? Well, how much is $1 delivered in 100 years worth to you today? Here's another way to phrase the same question: how much would you have to put into a bank today such that after 100 years you'd be able to withdraw $1? Suppose the interest rate were 5% -- if you invested x today, after 100 years, you'd have:

(((x * 1.05) * 1.05) * 1.05)...

= x * (1.05)^100

= x * 132

Set that equal to $1 and solve for x and you get:

$1 / 132 = .76 cents

Since money you get 100 years from now is worth so little today, the promise of such money is probably not a significant incentive for today's innovators. E.g., even if holding a copyright confers $100k / year of benefit, you're missing out on less than $1k if its duration goes from 100 years to 99 years (and therefore reducing copyright protection from 100 to 99 years is unlikely to deter you from completing whatever invention you're working on).

Of course existing copyright law is even more absurd: you get exclusivity for your whole life plus 70 years!

While reducing the length of copyright is pretty uncontroversial, there are those who'd like to go even further, perhaps eliminating the concept of intellectual property altogether.

The thought is this: the standard argument in favor of intellectual property makes an empirical claim: without intellectual property protection, innovators will not be sufficiently rewarded for their innovations. To what extent is this claim true?

Consider Starbucks -- it probably took a lot of time and effort to perfect the idea of an upscale coffee shop, figure out a branding / marketing strategy, find ideal locations for branches, etc. Anyone can take advantage of this work for free, yet Starbucks has done just fine without having a patent on the idea of an "upscale coffee shop".

Why? Because their first-mover advantage already rewarded their investors sufficiently. This is the way innovation works outside the intellectual property system -- someone thinks of a good idea and uses it to start a business. Potential competitors notice he's making money, decide to copy him, and gradually competition drives the price down. Eventually no further competitors enter the industry, and prices settle at an equilibrium. Why wouldn't a similar process reward innovation in contexts traditionally policed by intellectual property laws?

Another example: Napster. Napster certainly seems like a slam dunk for proponents of intellectual property rights -- how can musicians make any money if everyone is stealing their music for free? Society certainly benefits by incentivizing musicians, but society would also have benefited by forcing record studios to compete with Napster.

Once a song is recorded, a record label adds essentially zero value by duplicating it (since anyone can duplicate it for free). By letting the record labels make money via a service that adds no value, we didn't give them any incentive to create value via an iTunes-like solution.

The Economics of Guaranteed Sick Days (/ The Economics of Ezra Klein Being a Total Idiot)

Some people have been talking about this chart:

Bumbling idiot Ezra Klein explains:

You're seeing two things here. The light blue line measures paid sick days. This is what you use if you need to take three days off because you have a fever. The dark blue line is paid sick leave. This is what you use if you need to take three months off because you have cancer. Every other country on the list offers at least one. Most offer both. The United States is alone in guaranteeing neither.

I'm working at a serious newspaper now and so I'm going to try to avoid words like "barbaric" to describe policy decisions I don't like. But this is certainly unnecessary. CEPR ends with the economic argument: "Each year millions of American workers go to work sick, lowering their own productivity and that of their coworkers and potentially spreading illness to their coworkers and customers." I'm willing to cut employers some slack: Many don't offer paid sick days because they don't think doing so will make them money. That is to say, they make marginally more money by letting their workers fall ill. That may be a good decision for the employer. But it's not good for the worker. And it's an appalling state of affairs. Residents of the world's richest nation should be able to stay home when they have the flu. (emphasis mine!)

"Employers = bad, workers = good" -- well done, Ezra. Your improbable career rolls on.

Instead, let's try some analysis.

Your salary consists not only of the wages you receive, but also of the non-monetary things of value you get from your job. For example:

• Free lunch
• Free Metrocard
  
• Health insurance
• Vacation / sick days

Your "real" salary is your wage combined with the value you receive from your non-monetary compensation (from now on I'll say "salary" when I mean your "real salary" (including benefits) and "wage" when I'm talking about the dollar amount of your paycheck). Free lunch is worth $4k/year to you? That means giving you free lunch is equivalent to $4k/year extra in wages.

What happens to real salaries if you force employers to give their employees more days off? They stay the same -- employers won't suddenly start paying more for equally productive employees when they have the option to keep real salaries constant by lowering wages or other benefits. Employers who did lower real salaries would get outcompeted those who didn't.

So the real question here is: "Given that salaries are going to remain constant, is it a good idea to force employers to provide more days off and less of other types of compensation (wages, free lunches, Aeron chairs)." Or, put another way, "should the government regulate how salaries are packaged?"

So what's the answer?

First, consider that people have different preferences for days off. As an employee's vacation days increase, his wage and other benefits decrease, and at a certain point he will prefer the extra wage to the marginal day of vacation. By setting a minimum number of vacation days, you will, in some cases, be effectively forcing employees to purchase vacation days from their employer against their will.

The state doesn't force you to buy a certain amount of food (even though it sucks to starve) because the assumption is that you'll do a better job optimizing the amount of food you have on hand versus your other goods / savings. Likewise, else equal, employees will do a better job negotiating with their employers / picking a job that gives them their preferred mix of vacation versus other benefits than the state will on their behalf.

But this analysis won't hold if there are externalities. You could argue that there are two:

• Sick employees can infect other employees. However, the chance that you will get sick figures (negatively) into your salary. Thus, if an employer -- via a stingy sick leave policies -- is getting his other employees sick, they will view it as an effective pay cut, and demand higher wages / other benefits. Therefore, the employer has the incentive to appropriately consider the chances of employees getting each other sick.
  
  One potential problem here is that there's often imperfect information -- it's often difficult for potential employees to evaluate the health risks of a potential jobs.
  
  
• Sick employees infect customers. Again, it seems like employers have the correct incentives here. The price a customer pays for a good includes both monetary and non-monetary costs, one of which is the chance he gets sick. Employers have the incentive to consider total cost to the customer when making decisions.
  
  One potential problem here is that customers don't themselves internalize the entire cost of their getting sick (they can always go to the emergency room for free). (Of course one response is that they should have to internalize these costs, but whatever).

Anyway, my main point isn't that the US should or shouldn't force employers to provide sick leave. The caveats I mention in my discussion of externalities might be dispositive, or there might be some additional factor my model misses (maybe real salaries will go up somehow). I don't know.

My point is that if you do feel strongly one way or another about this issue, you need to engage these arguments. Ezra might be correct in his outrage, but if he wants to contribute something to the debate, his argument needs to be a bit more subtle than "Residents of the world's richest nation should be able to stay home when they have the flu."

Abortion: the last word

Abortion is -- along with "death penalty: cool or no?", and "can I smother my crying baby at the movies?" -- the lamest, most played out moral dilemma. But maybe a part of you still wonders "wait, for the record, what is the actual right answer there?"

To get us warmed up, let's look at a famous example of moral "reasoning" that should remind you why this topic is generally so painful:

You wake up in the morning and find yourself back to back in bed with an unconscious violinist. A famous unconscious violinist. He has been found to have a fatal kidney ailment, and the Society of Music Lovers has canvassed all the available medical records and found that you alone have the right blood type to help. They have therefore kidnapped you, and last night the violinist's circulatory system was plugged into yours, so that your kidneys can be used to extract poisons from his blood as well as your own. [If he is unplugged from you now, he will die; but] in nine months he will have recovered from his ailment, and can safely be unplugged from you.^[4]

Well fuck, the kidnapping wasn't cool, but that doesn't mean you can unplug the guy! Imagine you're the violinist in this thought experiment -- you'd be all "WTF are you doing?! This wasn't even my idea! Seriously, just chill out -- the inconvenience to you is minor in comparison to my dea--arghhhhhh!". You'd also probably offer the woman a lot of money to save your life, a not irrelevant fact (see below!).

Anyway, the only two tenable positions on abortion are:

Abortion is cool, but so is infanticide (up until the baby develops a personality / "becomes a person" / has preferences / is self-conscious / whatever, work it out yourself (Peter Singer is more or less correct)).

Let's be real -- there's just no way birth is a morally significant event. If two humans are exactly similar in every way except that one is in a womb and the other is in an incubator, it can't possibly be immoral to kill one and fine to kill the other.

Imagine that in 50 years technology is at the point where humans can gestate entirely in incubators. Women love it -- drink and smoke all you want, and minimal risk that your vagina rips straight through to your anus (which happens a lot during birth, I recently heard). At what point is killing the fetus-thing okay? If birth is morally significant, what happen when birth goes obsolete?

Clearly you can't believe that abortion is okay and infanticide is wrong -- what reason is there to believe that both are okay?

The intution here is look: at some point we're literally just talking about a bunch of cells. The thing we're "killing" has no preferences, no consciousness, isn't self-aware, has no capacity to feel pleasure or pain; what does it have in common with all those other things that I feel really bad about killing (humans, dogs, etc)?

The fetus / newborn child is "alive" in the same way the brain dead person is "alive". Perhaps technically true, but irrelevant -- functionally speaking, brain dead = dead, and fetus / newborn = not yet alive.

Of course one difference between a newborn and a brain dead 50 year-old is that, while the 50 year-old will likely stay brain dead and therefore never be a person again, the newborn has the potential to become a person, and likely will if you don't fuck with it.

This brings us to the second tenable position:

Abortion isn't cool, but neither is forgoing an opportunity to have a kid -- i.e., we're all morally obligated to have dozens and dozens of children (until the marginal child would be very unhappy or the inconvenience to us would be extremely large).

What the first position ignores (obviously) is that "potential person-hood" is a morally relevant characteristic. It's bad to "kill" things that will someday become people.

Well, not really -- it's not like the fact that this ball of cells can someday become a person makes the ball of cells special in some way, it's just that people are good to have around, and the ball of cells can become a person.

What do I mean "people are good to have around"? Suppose you could push a button and a 25 year-old person would pop into existence. He would be perfectly normal, have all these memories of growing up, etc. He would be just like you. (If this strikes you as an outlandish thought experiment, I'm curious why you're so confident that you didn't pop into existence 20 seconds ago with fake memories.)

Are you morally obligated to push the button? Obviously! What if pressing the button debited your bank account $1, or you had to swim a really long way to reach it, etc. It still seems like you should be trying to bring more happy people into the world.

How costly does pushing the button have to be before you're allowed to forgo doing so? Well, suppose you were just created by the push of a button -- seems like a pretty good outcome. Wouldn't the consequences of not pushing the button have to be pretty great to outweigh the consequences of pushing it (i.e., your being created)?

And so the thought is that having a baby and raising it until it's 25 is like pushing the button. You have to keep doing it until it's a total pain (which, come on, has to be at least 20+ babies)

What's the right answer? The sluts are going to hate me for this, but unfortunately position 2 is correct -- abortion's not only prohibited, but procreation is mandatory. I don't like it any more than you do, but I can't think of a way to rebut the argument.

Take solace in this though: just as I would gladly pay whoever just created me with all these fake memories for his trouble, so too should real 25 year-olds be willing to pay their parents for bringing them into existence. Therefore, parents should have the right to a (large) percentage of their child's lifetime earnings.

Wrongful Imprisonment = Eminent Domain

What, if anything, does the government owe you when it imprisons you for a crime you didn't commit? Apparently not much:

With more than 140 exonerated prisoners released since 2000, 22 states and the District of Columbia now compensate them using formulas ranging from lump sums to calculations of lost wages.

But the amounts vary widely. Wisconsin provides $5,000 a year up to a maximum of $25,000 total. California offers $100 a day. Tennessee provides up to $1 million total.

Twenty-eight states offer nothing — including states with multiple cases of discredited convictions — forcing former inmates to sue in state or federal court. There they have the difficult task of proving bad faith or intentional misconduct by authorities.

Contrast this with what happens when your house stands in the way of a state's plan to build a highway. In this case, the state can force you to do something you don't want to do (namely give up your house), but it's required (5th Amendment) to compensate you to the tune of the cost of the house (really it's supposed to 'make you whole' -- i.e., indifferent to losing your house, which requires above-market value compensation to account for your subjective value).

So when the state takes your $1M house it has to pay you $1M, but when the state takes 20 years of your life it doesn't have to pay you anything? Sheeeeeet -- you couldn't pay me any amount of money to go to jail for 20 years.

When the state wrongfully imprisons you, it should have to compensate you for both the opportunity cost of the time and the hardship of prison -- i.e., it should have to pay you enough money to make you indifferent to going to prison in the first place.

Cost Internalization

Any system of criminal justice is going to have costs and benefits. Benefits include general public safety, the minimization of inefficient property transfers (i.e., theft), encouragement of investment (I'm more likely to build that house I want if I'm confident it won't get stolen), etc.

Costs include both administrative costs -- paying judges, air conditioning courthouses, etc -- and the costs incurred by those wrongly imprisoned.

Just as we expect society as a whole to pay for the administrative costs of the criminal justice system -- we wouldn't allow states to force random lawyers to be judges against their will -- we should expect society as a whole to pay the non-administrative costs as well.

After all, what would happen if the state could force lawyers to be judges without paying them? The state would get the benefits without paying the costs -- creating a classic externality. If the state didn't have to pay judges, it would "employ" too many of them.

Likewise, when the state is insulated from the costs of wrongful imprisonment, we should expect an inefficiently high rate of wrongful conviction. This fact is exacerbated by the fact that legislators and the majorities that keep them in power are unlikely to be wrongfully convicted (i.e., even if the state could conscript judges, lawyers' political power might curb the state's overconsumption a bit -- not so in the case of wrongful imprisonment).

Insurance

Compensating the wrongfully imprisoned is equivalent to forcing everyone to buy insurance against their own wrongful conviction.

Why? Well, what would happen if the state had to compensate the wrongfully imprisoned? Everyone's taxes would go up a little (i.e., everyone would have to pay an insurance premium), and when one of us was wrongly convicted, he would get a huge payout (i.e., would have a huge insurance claim).

Is this desirable? Since being wrongfully convicted exacts a huge cost (more than any amount of money for most people), everyone who's risk averse should want insurance. The question is whether the state or the private sector should provide this insurance.

On this point you'll find all the usual arguments -- a private market for insurance might not develop because of adverse selection, or maybe you can't count on people having the foresight to buy this kind of insurance on their own, whatever.

Whatever the reason, the reality is that you can't buy this kind of insurance in the private sector and it doesn't look like you'll be able to any time soon. Given that everyone wants to be insured against the risk of such an enormous loss, the government should provide it in the form of compensation for the wrongfully convicted.

Conclusion

These are the same arguments that people use to defend the requirement to compensate you when the state builds a highway over your house. They are even more convincing here because it's harder to game the system -- i.e., you don't want to compensate too liberally in the highway case because if you did people might intentionally buy property that is in the way of a highway just to get the compensation money. There's no analogous risk in the wrongful imprisonment case (it's quite tricky to engineer your own wrongful conviction).

Stimulus package -- Woo!

One thing I'm trying to stop doing in 2009 is saying "what does that even mean" or "I don't understand", when I actually do understand / know what it means, and I'm just trying to emphasize that I think it's a bad point.

That said, I truly don't understand the way stimulus packages are supposed to work.

Let's examine this in the context of the classic "spend the money on infrastructure" stimulus package meme. The thought here must be that, as part of the stimulus package, the government should spend money to upgrade infrastructure when it's not cost effective to do so. (If it were cost effective to upgrade the infrastructure in question, the effort wouldn't be a "stimulus package" in that the government should do it regardless of the current economic climate).

Suppose there's some bridge in Michigan that could benefit from repair. The repair costs $1000, and confers $800 worth of social value in terms of reduced maintenance costs, fewer accidents, etc. I assume that repairing the bridge (for a net social loss of $200) is the sort of project that the government might take up as part of a stimulus package, and I don't understand how it could be a good idea.

Any explanation of why this is a good idea has to address not only the causal mechanism by which this move increases overall social welfare, but also why such "stimulus" is not also a good idea in prosperous economic times. I.e., not only do you have to explain how the state's buying something for more than its worth could somehow be a net good, but also why it's a net good now and not (say) 6 months / 1 year ago.

One thought might be that it "creates jobs" to repair the bridge. The idea being that the workers who would be repairing the bridge would be sitting around idle, and thus its good to give them a job by repairing it. But why is employing people to do unproductive tasks socially beneficial? If you want the bridge workers to spend more money, why not just give them the money you were going to use to repair the bridge? This way they could use their time more effectively (since every hour they spend working on the bridge gives society less benefit than it costs them).

Merely having people working cannot be a social good -- if it were, why shouldn't the state actively break the bridge once its done being repaired! If the state kept doing this, it could employ the bridge workers indefinitely.

But even if you agree that giving the bridge workers money is a good idea, that money has to come from somewhere. Perhaps giving it to the workers will make them spend more, but the people you took it from might spend less. Granted, this is borrowed money (so no one's tax burden is going up today), but in theory people should take into account their expected tax burden when they make purchasing decisions (which goes up whenever the government borrows).

Possible explanations / further thoughts:

• People don't think about their future tax burdens. The government can borrow (and thereby commit to raising taxes in the future) without disincentivizing people from spending money today due to their anticipated tax increase.
  
  Were this true, deficit spending would be free. The government could borrow indefinitely without crowding out private investment since people only pay attention to their current tax levels when spending money. This seems implausible.
  
• People are spending money very inefficiently right now. Repairing the bridge will cost $1000 for an $800 benefit, but that $1000 would otherwise be conferring even less than $800 in social benefit. For example, the original owner might have over-cautiously invested it entirely in Tbills.
  
  Were this true, it seems like there would be arbitrage opportunities for people who were able to assess risks correctly. Shouldn't these market forces put pressure on the people who are being overly risk averse?
• Repairing the bridge actually is cost effective. However, it's politically difficult to repair the bridge unless you call it a "stimulus". Right.
• Pork! All Pork! Perhaps these infrastructure projects are pure pork / rent seeking.
  

New Year's Resolution for all my friends: Broaden the base, lower the rate

So New York is having a budget crisis, blah blah blah, we need $4 billion, cuts, taxes, etc.

Gov. Paterson is planning to raise the money by instituting sales taxes on consumer goods instead of raising income taxes.

Here's a plausible thing you might think about this proposal:

Taxes on consumer goods like furs and boats are more fair than income taxes because you have control over how much tax you pay -- just modulate your purchase of the goods being taxed. On the other hand, raising the income tax is less fair since you can't avoid paying it (or at least you can't avoid it as easily).

A female newscaster made this argument to me while I was getting my hair cut last week. Unfortunately, as I told Mahbod and Katie as soon as the barber finished shaving my face, precisely the opposite is true. The best tax would be totally unavoidable, and conversely the worst taxes are those that are easily avoidable. If you're considering a tax policy in 2009, my advice is to broaden the base and lower the rate.

Why are avoidable taxes bad?

From the perspective of a consumer, increasing the tax on a particular good is equivalent to increasing its price. Since consumers will purchase a good only when it gives them more utility than it costs, increasing a good's price (via tax) causes some consumers not to buy it.

From the perspective of society though, we only want a consumer to forgo buying a good when its social cost is greater than the utility it gives him. But the price added by the tax does not represent actual social cost (unlike the part of the price that accounts for, say, the cost of shipping the good to your house). Therefore, taxing a good that one can choose not to buy risks causing some consumers to inefficiently choose not to purchase it.

For example, suppose a certain good costs society $10 to produce and gives a certain consumer $30 worth of pleasure. Hopefully the price of the good is $10, and the consumer can purchase it at a private (and social) surplus of $20.

If we add a $5 good to the tax, and the consumer cannot avoid purchasing it (maybe it's insulin), the overall surplus is the same -- the consumer pays $15, gets $30, and the government gets $5, for a total gain of $20.

Now suppose the consumer avoids purchasing the good and instead buys a different untaxed good for $14. Now the consumer pays $14 and gets $30 for a total gain of $16. The consumer is better off, but society is worse off since the government didn't get any tax revenue.

Now, it happens that some of Paterson's taxes might actually deter behavior that was inefficient pre-tax. How? Well, if a good's price doesn't reflect its entire social cost (e.g., the price of cigarettes doesn't include potential future medical bills), then a tax can improve incentives by bringing the price the consumer faces closer to the price society faces.

This is why his proposed tax on non-diet sodas is a good idea. (Though the exception for drinks that contain more than 70% fruit juice is BS).

I'm not quite sure whether the name is offensive

So as anyone who reads my half-assed twitter feed knows, I'm currently involved in an I Am Legend situation here in New York. (Though not, as Ariel informed me in a completely neutral tone, a Legend situation).

The City is deserted and freezing, and I have no one to confide besides my work computer's humming fan. On the other hand, I can throw my hoodie on the living room couch without worrying about inconveniencing someone (i.e., this week has been only slightly better than average).

Anyway, the only way to keep busy is to manically start blogs, so... Here's another blog!