Confucius: The Economist?

Compare the Analects 2.3...

The Master said: Guide them with policies and align them with punishments and the people will evade them and have no shame. Guide them with virtue (de) and align them with li and the people will have a sense of shame and fulfill their roles.

...with the daycare experiment made famous by Freakonomics:

Imagine for a moment that you are the manager of a day-care center. You have a clearly stated policy that children are supposed to be picked up by 4 P.M. But very often parents are late. The result: at day's end, you have some anxious children and at least one teacher who must wait around for the parents to arrive. What to do? 

A pair of economists who heard of this dilemma... offered a solution: fine the tardy parents... The economists decided to test their solution by conducting a study of ten day-care centers in Haifa, Israel. The study lasted twenty weeks, but the fine was not introduced immediately. For the first four weeks, the economists simply kept track of the number of parents who came late; there were, on average, eight late pickups per week per day-care center. In the fifth week, the fine was enacted. It was announced that any parent arriving more than ten minutes late would pay $3 per child for each incident. The fee would be added to the parents' monthly bill, which was roughly $380. 

After the fine was enacted, the number of late pickups promptly went . . . up. Before long there were twenty late pickups per week, more than double the original average. The incentive had plainly backfired.

The Microeconomics of Child Support

I was recently fortunate enough to do some work experience for a judge. As well as seeing some of the dramatic debates that one imagines taking place in court, I saw some of the smaller scale judgments that are made at tribunals. Primarily these were in relation to child support; either a parent with custody appealing that the specified level of child support was too low, or a parent without custody appealing that the specified level of custody was too high. In the cases that I saw all the parents with custody were mothers.

The first thing that struck me was the extent to which fathers went out of their way to avoid paying child support. Of course, this was clearly partially a result of the selection effect; the people at the tribunal were the select few trying to avoid paying. Nevertheless, there was an example of a mother suing a deceased father's estate for unpaid child support; their daughter was not a beneficiary of the estate at all. Surely, however much parents may come to hate each other, they would not go so far as to prevent their children from receiving money after their death?

However, once I thought about it, I realised that they might reasonably see child support going not primarily to their child but to their ex-spouse. Consider an imaginary example in which a father is ordered to pay £200 weekly to a mother with custody. It is essentially a transfer from the father to the mother to compensate her for the child's expense. One might reasonably expect that the mother will spend some of the money on the child, but it seems unlikely that it will be spent exclusively on the child. To think of it in another way, if the mother was made £200 poorer she would not reduce spending on the child by £200; she would seek to cushion the drop in income by reducing some of her own expenses. In fact, it seems likely that, given that in most cases the mother will already have bought the child the things which she deems necessary for their happiness, a significant portion of the money will essentially recompense her for this expense rather than be used for additional spending on the child.

Now, it is quite likely that the child support will not pay for half of the costs of bringing up the child (especially if one values the mother's time, too). As a transfer, it seems more than fair. Nevertheless, from the perspective of a father who hates the mother, one can see why he might be reluctant to pay child support even if they wish their child well. This doesn't excuse their behaviour, but it does begin to better explain some of it.

The Economics of Elections

In his latest New York Times column, Paul Krugman alleges that the Republican party has been deliberately obstructing measures that might help the American economy. I don't know enough about Republican strategy to be able to guess whether or not this is true, but if it were it would make excellent tactical sense.

As Ezra Klein of the Washington Post points out, incumbent politicians in growing economies almost always win elections. Incumbency and whether or not the economy is growing are by far the two most important factors that determine elections. The Republicans did their best to prevent the former in 2008. In 2012, the only one which they have any control over is the latter, the economy. Add to this calculus the fact that most measures that boost the economy significantly today have some future cost, and it becomes easier to understand why an anti-growth strategy could feel justified.

Leaving the Eurozone

Much speculation has gone on about countries leaving the euro. Having their own currency would be of immense advantage to struggling European countries like Spain and Greece, insofar as they could use monetary policy to attempt to revitalize their economies. On the other hand, these countries are exactly the ones which can least afford to bear the costs of abandoning the euro. The issue's urgency was recently highlighted by the revelation that some companies are making very specific preparations for a Greek exit.

In response to this, some have suggested that it may be Germany that exits the euro. It can better afford the associated cost (even just printing the appropriate quantity of paper money is a logistical nightmare), and this would cause less trauma to the weaker economies than in a scenario where one or many of them exited. However, such speculation misses an important point, beyond the political implausibility of this suggestion.

Germany has no need of its own currency - it has already got one, and it's called the euro. As Evan Soltas points out, the German economy is a decisive factor in the European Central Bank's policy making, getting twice as much consideration as the size of its economy would intuitively suggest.

Was China's Stimulus Worth It?

Tyler Cowen of Marginal Revolution forwards a challenge:

[Nick] emails me:

I have greatly enjoyed the recent pieces on China as I currently live in REDACTED…Fwiw, my take is that yes, the 2009 Chinese stimulus was a great test of large-scale govt stimulus and that we are seeing the results of what this looks like *in practice* rather than *in theory*. In practice, large-scale government stimulus is an invitation for corruption and a diversion of resources that builds up knowledge and capital in unproductive areas. As Obama said after the US stimulus package: there’s “no such thing as shovel ready projects.” Even in China.

Matt Yglesias at Slate responds:

But look at the numbers. In 2007, China's PPP-adjusted GDP per capita was $5,562. In 2008 it was $6,201. In 2009 it was $6,810. In 2010 it was $7,571. In 2011 it was $8,407. The 2012 growth numbers look certain to show a drastic slowing down of growth. But let's get really pessimistic about China and assume 0% growth across 2012 followed by a five percent contraction in real GDP per capita in 2013. That would still leave China's per capita GDP level about 50 percent higher than it was when the world starting going into recession. That seems to me like an enormous success story.

The big knock on the Chinese economy is that there's all this malinvestment and tons of white elephant projects. And perhaps there are. But there's no sector of the economy with lower productivity than the unemployment sector, and no public investment more wasteful than having a human being spend years scanning help wanted ads and feeling depressed. Does someone seriously want to argue that none of these projects have any value? It doesn't seem to me that anyone in China is debating whether or not the typical household is better off than it was four years ago. Here in the USA we've had about 2 to 3 percent of our workforce persistently engaged in doing nothing for the past several years, and the forecasts all show that trend continuing for a while. That, to me, looks like intolerable waste.

High Impact Investment

If you want to support a charity, should you invest the money so that your donation grows in size before it is given? The typical answer is no; helping a charity immediately will usually have a future 'return' in the form of its beneficial impact at least comparable to any pecuniary return on your investment. Perhaps you've decided that investment is worthwhile, however - maybe you're a budding Warren Buffet.

How should you invest your money to maximise your altruistic impact in the future?

Risk Averse Investors

Most people are risk averse. To see this, we need only look at insurance, an example of people paying money to reduce the risk of extremely bad outcomes. If somebody insures their house against fire, they will pay more money on average than they will receive on average in insurance if their house is burnt down. Insurance companies are after profits; if they are paying out (to recompense those whose houses are burnt down) more than they are taking in (from everybody insuring against fire) then they will stop offering fire insurance. Assuming that insurance is a zero sum game, this means that people taking out fire insurance cannot be making money out of it on average.

So on average, fire insurance is not a good bet if all you are interested in is maximising your average bank account balance. However, most people quite reasonably care more about preserving a minimum bank account balance than maximising its average, so they are happy to buy fire insurance. A similar rationale can be applied to investment. Investors pay risk premia to avoid risk. This is why safe assets like gilts have lower returns than more risky assets.

In microeconomic theory, the standard explanation for this is that investors have a diminishing marginal utility of money¹. Whether or not you consider this explanation valid, being more tolerant of risk in your investment will clearly earn you a better return. If you care more about saving more lives on average than you do about guaranteeing that a certain number of lives will be saved, it may be worth taking riskier investments. There are fewer diminishing marginal returns to charitable donations; so long as people are in areas at high risk of malaria, purchasing bed nets at a rate of $5 will be worth doing.


^{1. To illustrate this, imagine somebody who has no money (and no other possessions). We'll call them Max, since that happens to be my name and I've got very little money! There are a lot of things that I want to buy, but can't. I am more eager to buy some things than others, however; a car may seem attractive, but if you gave me £20,000 I would be more likely to spend it on a place to sleep than a car. Simplifying for the purpose of example, imagine that cars and cheap flats both cost £20,000. If I am given £20,000 I will buy a flat, increasing my welfare by a certain amount. If I am given another £20,000 I will buy a car, increasing my welfare again. This second increase must be less than the first because I wanted a flat more than I wanted a car. Therefore the second £20,000 is of less value to me than the first. Therefore there is a diminishing value to each unit of money that I am given. This is the diminishing marginal utility of money.↩}

Radical Moral Ideas: Speciesism

Who Are These 'People' You Care About?

In my previous blogposts I've discussed the implications that radical altruism has for ethics. Yesterday I considered what altruism towards future people might imply. Today I'm going to have a go at considering what happens when we examine our definition of people.

Most people who are likely to read this post won't believe that race is reasonable grounds upon which to care more or less about somebody. If you're truly egalitarian, nation probably shouldn't make a difference either, although the widespread preference of national/local charities over international ones suggests that many give a priority to the welfare of people in their own nation. Why might this not be reasonable?

For me, it seems that it's the quality of subjective experiences that I care about when thinking altruistically. It doesn't matter where you're from, everybody can experience pleasure and pain and all the other complex emotions that make up life. At least, you can experience these things unless you're a human vegetable, at which point (provided we're certain you're not just locked in) cutting off life support isn't doing much harm. Certainly discriminating on the basis of other mental qualities seems unfair to the mentally disabled and young infants. No matter how unintelligent someone is, so long as it is possible for their subjective experience of the world to be made better or worse I find helping them an appealing idea.

If this is the case, then why not care about the suffering of animals? That they're "not human" seems to me to be no more convincing an argument than that other humans are "not English" or "not white". Such an attitude is, to use an ugly word, speciesist. In fact, it seems likely to me that some mentally advanced animals like elephants, dolphins and dogs are probably significantly more capable of meaningful subjective experience than young infants. We should, therefore, care deeply about their welfare.

If you accept this argument, then a few actions seem imperative. Firstly, you should become vegan. The food looks a lot less tasty, I agree. If I were to go vegan, I'd have to learn how to get cooking tofu right, which seems quite a challenge. But all the same, something like 60 billion animals a year are killed to produce food, living lives of great suffering. If you find this argument convincing, you'll also want to check out effective animal charities and donate to them.

Radical Moral Ideas: X-Risks

How much do you care about future people?

In my previous post it looked like the extreme ease of helping people through certain charities suggests that we should give them most of our money. However, surely some consideration should be given to the fate of future people. You might think that their lives should be given somewhat less moral weight than those of currently living people, although to me the prevention of an entire lifetime's worth of experiences seems as bad as killing someone alive today, but either way a future with lots of happy people seems better than one with fewer sadder ones.

If you give a non-trivial moral weight to the lives of future persons, however, you might find that you have to take into account a new set of concerns: X-risks, or extermination risks. In the past, many species have gone extinct. When carbon dioxide munching plants first filled the atmosphere with oxygen, the pre-existing species of bacteria had to adapt or die. Notoriously, the dinosaurs were (probably) put to rest by a deadly asteroid impact, only the mammals' smaller size allowing them to survive. Unfortunately, mammals are no longer tiny, and the risk of catastrophic asteroid impact has not gone away.

If humans were wiped out by an asteroid, several billion people currently alive on Earth would die. However, this is not the end of the catastrophe, if we value the lives of future persons. All of humanity's potential future descendants would be prevented from living. This could run into the trillions. In fact, if you think that humanity would otherwise survive for billions of years (potentially colonizing other planets), then the number of beings whose existence would be prevented by human extinction is unfathomably large.

Any other moral considerations surely pale by comparison. Ensuring that the human race is not wiped out in the foreseeable future becomes a moral imperative, even if you can only contribute to this cause by the tiniest of amounts. A number of potential factors that might exterminate the human race come to mind. A weaponised virus, for example, might be manufactured by a genocidal lunatic that could be exceptionally infectious and deadly. At the extreme end of the probability curve of predictions of the impact of global warming are temperatures that might render Earth uninhabitable to humans. A particularly interesting possibility that I hope to discuss in the future is artificial intelligence.

In short, when the future is at stake, nothing else matters. Read Nick Bostrom's essay for more.

Radical Moral Ideas: Cost-Effective Charities

This is the first of a series of posts I'm planning to write on moral ideas that would radically change our ethical priorities if we were to accept their premises. Each have a lot of extremely clever people convinced, although they tend to make an unconventional first impression.

How much do you care about other people?

Most people agree that killing is wrong. It doesn't seem to matter much whether this is by killing somebody in front of you with a gun, by deliberately failing to hit the brakes on your car, or by launching an inter-continental ballistic missile at their house. Somewhat more contentious is the division between active and passive killing. Is shooting somebody worse than deliberately failing to hit your brakes or failing to tell them that a particular item of food is lethally contaminated? To some degree, at the very least, all of this seems bad.

How much cost can a person be expected to bear to avoid killing? If your car's brakes are broken, and you must swerve your car into a ditch to avoid hitting a person, are you morally obligated to bear the cost of repairing your car? Peter Singer gives a powerful example that appears to decide the argument: if you must ruin your expensive shoes to rescue a child from a pond, you should surely do so.

If we accept these arguments then we've established that it is right to bear a cost of up to several hundred pounds in order to save the life of someone else. In fact, the most cost-effective charities can save lives for roughly this value¹. It seems imperative, therefore, to give much more money to these most cost effective charities.

A corollary of this is that you should not give money to less effective charities. Charities fighting cancer and helping the disabled have laudable aims, but their operating costs are high. Donations to them save fewer lives than donations to more cost-effective charities, thereby causing more people to die.

<sup>1. The NHS guidelines approve drugs that prolong lives2 at a cost of less than £20,000 annually.↩
2. Actually the relevant criterion is the quality adjusted life-year or QALY.↩</sup>

Free trade doesn't always benefit everyone, but it's worth it

In the interests of blowing the dust off my attempt at communication, here's an essay my economics tutor liked. I'd apologise, internet, for not to talking to you, but my friends will tell you that I don't communicate with them enough either. Ignore the terrifying title, please.

Consider the 2x2 production model with constant returns to scale, where the two factors are immobile internationally and the two consumer goods are mobile internationally. Using this model, explain the statement “free trade does not benefit everyone”.

To explain why free trade between two countries might not benefit all involved, I shall first lay out the 2x2 production model, before describing how expanded markets for a particular good will reduce incomes for those whose factors of production are scarce in their country. I will then relax the assumption of factors' free movement between industries, and examine the potential short term effects of this. I shall finally briefly argue that the negative impact of free trade can be compensated for if lump sum redistributions are possible in the favour of those who would otherwise be losers from free trade.

The standard 2x2 production model includes two goods, which are produced by different ratios of two factors of production. For the purposes of this essay, these two factors of production will be called labour and capital, abbreviated to L and K respectively, while the goods will be called guns and butter, abbreviated to G and B respectively. Butter is a labour intensive good and guns is a capital intensive good, so for a given set of factor prices the ratio of labour to capital used in the production of butter will be higher than the ratio of labour to capital used in the production of guns. The 2x2 production model also includes two countries with competitive economies, which have different ratios of the factors of production. For the purposes of this essay, these countries will be called Australia and China. Australia is a capital abundant country and China is a labour abundant country, since China has a higher ratio of labour to capital than Australia.

Given that butter is a labour intensive good compared to guns, and China is labour abundant compared to Australia, the price ratio between butter and guns will be higher in Australia than in China, if trade is not free. Furthermore, when trade is not free, production of butter and consumption of butter must be equal within (but not between) each country, as must production and consumption of guns. Assuming that demand curves for each good are the same for each country, and the technology is the same, owners of scarce factors of production in each country will be relatively better off than owners of the same good overseas. Capital owners in China will get more purchasing power from each unit of capital than capital owners in Australia, because the ratio between returns to investment and wages will be higher in China.

When trade is free, however, only the value of production in each country will have to be equal to the value of consumption.

P_BxD_B + P_GxD_G = P_BxQ_B +P_GxQ_G

which can be rearranged to the following budget constraint:

D_G – Q_G = (P_B/P_G)x(Q_B – D_B)

Guns imports = Price ratio x Butter exports

Free trade between Australia and China will lead to a convergence of goods' prices, as free markets clear. In China, this will lead to a rise in the price of butter and therefore, since prices are relative, a reduction in the price of guns. In Australia the reverse will occur. Each country will export the good that it can make more efficiently than the other, leading to an equalisation of factor prices, despite the international immobility of the factors of production themselves. In China, owners of labour (workers) will have their purchasing power increased, because the rising price of butter will increase the price of the factor proportionately more used in its production; wages will increase. In Australia, for similar reasons, owners of capital will be better rewarded. However, as the relative price of butter rises in China, owners of capital in China will be made worse off as the relative price of guns falls. In Australia, owners of labour will be made worse off as the relative price of butter in the country falls. Therefore, in this situation, “free trade does not benefit everyone”.

By relaxing the assumption that factors of production can be freely moved between industries, 'specific factors' become possible. For example, a specialist in gun manufacture may not be as efficient at milking cows as she is at making guns. Factors of production, labour in this instance, can in the short term be paid different prices to make different goods. Only in the long run will earnings equalise, as factors of production gradually shift (by, for example, retraining). In 2x2 production model used above, a fall in the relative price of butter in Australia, though in the long term improving the purchasing power of owners of capital, may in the short run reduce the earnings of owners of capital whose capital is currently invested in herds producing butter, until they can reinvest their capital in gun factories.

Finally, I shall briefly note that according to the second fundamental theorem of welfare efficiency, lump sum redistribution of goods could potentially redistribute the gains from trade. A Hicksian compensating variation (the amount of money, and therefore value of goods, required to put a consumer back on their initial utility curve) would restore Australian workers (and/or Chinese capitalists) to their original welfare, allowing free trade to produce at least a weakly preferred allocation. If the gains from trade are greater than this, as they can potentially be because trade expands the consumption possibilities of an economy, then it is possible that everyone can strongly benefit from free trade.

In conclusion, because according to the 2x2 production model free trade causes the equalisation of factor prices, those who own factors whose relative price decreases will lose purchasing power, and (if standard assumptions are made about utility maximising consumers) will therefore be worse of after the trade. Some owners of the abundant factor of production may be worse off in the short term after free trade if they are specific factors. Despite these results, if lump sum transfers are possible, it should be at least potentially possible for everyone to gain from free trade.

Communication is hard

Getting ideas across to other people is difficult. That's a nice optimistic way to start a blog, isn't it?

Here's a post on Less Wrong that's pretty useful. I sometimes find it hard to remember that my beliefs have been formed over the 20 years I've had to think with. Trying to talk to people about ideas which have been stewing for that long can be tricky. And since tricky things should be practiced, I've decided to start doing so.