Adapt (book review)

Several years ago one of the keynote speakers at the New Zealand Association of Economists conference was Tim Harford. He had a new book at the time, Adapt, with the by-line "Why success always starts with failure". I regret the book has been sitting on my shelf ever since, but I've finally gotten around to reading it.

I've really enjoyed most of Harford's books (see other reviews here and here), and this one was at least as good as any of the others. Harford has a talent for pulling together appropriate anecdotes to illustrate the points he is making. And the overall point of this book is the importance of using trial and error to adapt to challenges. In other words, it is increasingly important to identify new ways of dealing with challenges as (if not before) they arise, if we want to be successful.

Towards the end of the book, Harford identifies three essential steps for adaptation (the 'Palchinsky principles', named for a Russian engineer who identified problems in Russian coal mines in the early 20th Century):

First, try new things, expecting that some will fail. Second, make failure survivable: create safe spaces for failure or move forward in small steps... And third, make sure you know when you've failed, or you will never learn.

The first principle makes a case for constant innovation, experimentation, and rigorous testing, even when things are going well. If it isn't broken, it can still be fixed to work better. The second principle is about managing risk. If you fail, do so early before it becomes catastrophic. Which is where the third principle comes in. Identifying when something you try has failed and not feeling beholden to continue it is challenging, but essential.

These are good principles for business, and in the final chapter Harford illustrates how they can be applied for individuals. I highly recommend this book as a good read, not because of the economics content (of which there is a fair amount), but because of its more general applicability. Enjoy!

Big brother is watching your drunken tweets, and using them for research into alcohol use

Around Easter, I spent the last bit of my study leave at La Trobe University in Melbourne, working with Michael Livingston. We've been working on the relationship between alcohol outlet density and crime for a number of years, but in this case we were looking at developing a unified theory of the relationship. The necessity for a unified theory arises because there are a number of competing (but, as it turns out, quite complementary) theories for why having more alcohol outlets in an area would lead to more violence (I'll post more on this later).

One of the earliest theories for this relationship is called 'availability theory'. The theory is reasonably simple, but is underlined by a lot of stuff we teach in introductory economics. If you have more alcohol outlets in an area, then the 'full cost' of alcohol falls, and because alcohol is now less costly people will drink more. And when people drink more, more bad things (violence, property damage, accidents, etc.) happen.

When we say the 'full cost' of alcohol, we mean not only the price of the alcohol itself, but also the cost of travelling to and from the location of purchase. So, there are two mechanisms where having more outlets in an area leads to a lower 'full cost' of alcohol. First, having more outlets probably means that people have to travel less far to obtain alcohol, so the travel cost is less. Second, having more outlets probably means an increase in competition between outlets, and we know that competition tends to result in lower prices.

Unfortunately, the empirical support for availability theory is pretty patchy. Some studies find that greater alcohol outlet density is associated with more consumption of alcohol, while others find no effect (e.g. see here or here). Added to that, unpublished work that Bill Cochrane and I have done shows that having more outlets in an area is not associated with lower pricing (using cross-sectional data - we've been collecting longitudinal data now for a number of years, so it's getting time for us to revisit the analysis with better data). Which is why other theories explaining why alcohol outlet density is associated with violence (for example) have arisen.

Which brings me to this recent paper by Nabil Hossain, Tianran Hu, Roghayeh Feizi, Ann Marie White, Jiebo Luo, and Henry Kautz (all from University of Rochester). The authors use a machine learning algorithm to identify drunk tweets. MIT Technology Review explains:

The team began by collecting geotagged tweets sent during the year up to July 2014 from New York City and from Monroe County on the northern border of the state, which includes the city of Rochester. From this set, they filter all the tweets that mention alcohol or alcohol-related words, such as drunk, beer, party, and so on.

They then used workers on Amazon’s Mechanical Turk crowdsourcing service to analyze the tweets in more detail. For each tweet, they asked three Turkers to decide whether the message referred to alcohol and if so whether it referred to the tweeter drinking alcohol. Finally, they asked whether the tweet was sent at the same time the tweeter was imbibing.

They then used the geolocated Twitter data and asked the mTurk volunteers to identify tweets that were sent from home. They then used that data to train a machine learning algorithm to identify other tweets that were sent from home. Now, armed with a dataset of users' homes (to within about 100m) and their tweets while drinking, the researchers are able to establish a correlation between drinking and alcohol outlet density. Again, MIT Technology Review explains:

...Hossain and co point out that a higher proportion of tweets in New York City are associated with alcohol than in Monroe County. “One possible explanation is that a crowded city such as NYC with highly dense alcohol outlets and many people socializing is likely to have a higher rate of drinking,” they say.

What’s more, the geolocation data reveals that a higher proportion of people drink at home (or within 100 meters of home) in New York City than in Monroe County, where a high proportion of people drink further than a kilometer from home...

They also found a correlation between the density of alcohol outlets in a region and the number of tweets indicating that somebody is drinking now.

The latter result is of course correlation, not necessarily causation. Perhaps having more alcohol outlets in an area causes people to drink more, but on the other hand perhaps people who drink more are more likely to choose to live where there is readier access to alcohol. However, it is an interesting approach, and worth following up, perhaps if an appropriate instrument for alcohol outlet density can be found (to try and overcome the issue of potential reverse causality). Maybe availability theory isn't dead after all?

[HT: Marginal Revolution]

Try this: Bazinganomics

Dirk Mateer has done it again. While browsing the latest issue of the Journal of Economic Education, I found this short paper by James Tierney (Penn State), Dirk Mateer (Arizona University), Ben Smith (University of Nebraska-Omaha), Jadrian Wooten (Penn State), and Wayne Geerling (Penn State), referring the to website Bazinganomics. The website uses scenes from the comedy show The Big Bang Theory to illustrate economic concepts - everything from markets, elasticity, and externalities, to unemployment and economic growth. Some of the examples are a little strained, but if you're a fan of the show (and I know many of my students are), then there is a lot in there that is of value.

Enjoy!

New arguments against free trade

Earlier in the month, The Economist reported on the rise of anti-dumping tariffs:

Dumping is the practice of selling goods in foreign markets at an unfairly low price—typically, one lower than the going rate in the exporter’s home market. Anti-dumping measures are intended to prevent a company from selling goods below cost in order to drive competitors out of business, before using the resulting market power to gouge customers.

So countries create additional tariffs to apply when foreign firms are thought to be using otherwise free trade to 'dump' their products into the domestic market. However, as Gary Becker has noted (for example, in this book), this type of predatory pricing is probably not sustainable. If you drive competitors out of the market and raise your prices, then new competitors will enter, assuming that the fixed costs of production aren't high enough to constitute a barrier to entry (which is arguable for some industries, such as steel production). However, even if there are high fixed costs, if you have relatively free trade then competitors from other countries can also step in when the predatory pricing stops.

There are other arguments put forward in favour of anti-dumping tariffs (from the same Economist article):

In Britain, where rock-bottom global steel prices now threaten Tata Steel, the owner of the country’s biggest surviving mill, proponents of tariffs argue that it is important to preserve domestic steelmaking to ensure supplies for the defence industry, among others.

The 'national security' argument is an old favourite of some anti-free-trade groups, particularly those that stand to benefit directly from trade barriers. The Economist points out the folly of the argument:

...it is hard to see how the use of French steel in British submarines harms Britain’s security (its pride is another matter). For manufacturers of all sorts, most notably carmakers, cheap steel is a boon.

In other words, the gains (to domestic consumers, and to domestic manufacturers using steel in production) from free trade in steel probably outweigh the losses (to domestic steel producers). And national security isn't much of a consideration.

But is it always the case that the gains from freer trade outweigh the losses? In another interesting blog piece from earlier this month, Tim Harford writes:

Fifteen years ago, the conventional economic wisdom was that free trade was almost unambiguously a good idea. Here’s the basic logic. There are two ways for the British to get hold of wine. We can grow and press our own grapes, or we can make something that the French want and trade with them. If we’re good at making, say, computer games and the French are good at making wine, then trading is the better way to get what we want...

I’ve been phrasing all this “conventional wisdom” in the past tense but, for the most part, it stands up. However, it is acquiring an important and depressing footnote. A new research paper, “The China Shock”, from David Autor, David Dorn and Gordon Hanson, is part of a rethink under way in the economics profession...

Autor, Dorn and Hanson conclude that the American workers who have been hurt by competition with China have been hurt more deeply, and for a longer period, than many economists predicted. Employment has fallen in industries exposed to trade competition, as expected. But it has not shown much signs of rising in export-oriented sectors.

The research paper by Autor et al. is here (ungated here). Economists recognise that there are gains from trade. In fact, it is one of the things that economists most agree on. However, although there are gains from trade, and therefore gains from making trade freer, those gains are not obtained without cost. When we open our markets to international trade, our export industries (the industries in which we have a comparative advantage) will produce more, while other industries (in which we have a comparative disadvantage) will shut down. In New Zealand we experienced this - which is why we no longer have much in the way of car assembly plants, whereas there were previously several large plants around the country. The conventional wisdom is that there is some short-term pain (workers losing jobs in the industries with comparative disadvantage) for long-term gain (additional jobs in the industries with comparative advantage).

The Autor et al. paper demonstrates that the short-term pain can last much longer than previously thought (the paper is mostly very readable - I encourage you to do so). Although the 'national security' argument is weak, perhaps the so-called 'jobs argument' against free trade has some truth to it after all? Which might give us reason to pause on free trade agreements, particularly when the gains are marginal at best, or occur far in the future (while the costs are incurred in the short-term).

Free market environmentalism gone wrong

I've been writing for some time the problems of trying to save endangered species (see here, here, here, here, and here). One potentially effective (and controversial, compared with orthodox environmental values) solutions is to grant property rights over endangered species and allow them to be farmed. Of course, this would also entail legalising the trade in endangered species. Then the farmers have incentives (backed by legal rights) to protect their herd of rhinos from encroachment by others (including poachers), and will be much more likely to sustainably manage their herd (otherwise, their livelihoods will suffer). You could argue this is one of the reasons why rhinos are endangered, but cows are not.

Now, it seems that South Africa has moved one step closer to this ideal, as Science Alert reports (emphasis added):

A high court judge has upheld a decision to lift the ban on buying and selling rhino horn within South Africa. Rhino horn trade has been outlawed in the country since 2009, and internationally since 1977, but if the high court's decision holds up against an appeal, that could all be about to change...

The push to lift the ban on selling rhino horn came from game breeders, John Hume and Johan Kruger, who claim that legalising the trade within the country will reduce rhino deaths - rhino horn is similar to our fingernails, and can actually be harvested without harming the animal. Hume also argued that if the ban on rhino trade continued, he'd no longer be able to afford to keep his 1,200 farmed rhinos.

However, this doesn't solve the overall problem, because the international trade in rhino horn is still illegal:

But there's a big problem with their argument, and that's the fact that pretty much all of the demand for rhino horn comes out of Asia - and lifting the ban on being able to sell within South Africa isn't going to do anything to stop that...

In fact, if the ban is lifted, it'll probably just create a new market for rhino horn, and put more pressure on the already endangered species.

So, rather than reducing the price of rhino horn by flooding the international market with supply of farmed rhino horn, this measure likely increases the demand for rhino horn, raising the price, and increasing the incentives for poachers!

It'll take something more coordinated internationally to create the right conditions to eliminate (or at least, minimise) poaching of endangered rhinos.

[HT: Marginal Revolution, back in January]

Read more:

• Can technology save the rhinos?
• Because crushing ivory is so much better than burning it
• Poisoning rhino horns to save the rhinos
• This couldn't backfire, could it?... Ivory burning edition
• Lions are dying to be saved

The deadweight loss of rent control

Caleb Malik pointed me to this blog piece he wrote recently about rent control. Given that rent control is covered in the ECON100 test this week, I thought it would be timely to discuss it again (I previously discussed rent control in the short run and long run here).

Caleb's piece covers the ground well, but I want to look specifically at the economic welfare effects of rent control, and who gains and who loses from controlled rents. Remember that governments (local or central) typically enact rent controls in order to help low-income tenants who may be struggling to pay high rents (often in inner-city neighbourhoods).

Consider the market in the diagram below. If the market is at equilibrium, the market rent is R0, and the quantity of housing rented is Q0. Now say instead that there is a binding rent control at R1 (below the market rent). The lower rent makes renting more attractive relative to owning your own home. Some people would find it cheaper or more convenient to be a renter at this lower rent, so the quantity of rental housing demanded increases (to QD). However, the lower rents make rental housing a less attractive investment for landlords. Perhaps they convert that rental housing into commercial rentals instead (e.g. offices) or maybe they choose to live there themselves (the opportunity cost of living in the house is now lower). Either way, the quantity of rental housing decreases (to QS). The difference between QD and QS represents the excess demand for rental housing at the controlled rent – there are fewer houses available than the quantity people want to rent.

It should be obvious that landlords are made worse off by rent controls - their rental housing attracts lower rents (R1 instead of R0), and they provide fewer rental housing units at that lower rent (QS instead of Q0). The producer (landlord) surplus is a measure of the landlords' economic welfare - it is the difference between the rent they receive (either the market rent or the rent controlled rent), and their costs (which are shown by the supply curve). Without the rent control, the landlord surplus is the area R0EF on the diagram, but with the rent control, the landlord surplus falls to the area R1CF. Landlords are made unambiguously worse off by the rent control.

What about tenants? The consumer (tenant) surplus is the difference between the amount that tenants are willing to pay (shown by the demand curve), and the amount they actually pay (the market rent, or the rent controlled rent). Without the rent control, the tenant surplus is the area AER0, but with the rent control this increases to the area ABCR1 [*]. Tenants as a group are made better off. However, this ignores that many prospective tenants are missing out on rental accommodation (remember the amount of rental housing has fallen from Q0 to QS). Because of the shortage of rental housing, landlords have choice over who they rent their housing units to. Given the choice between, say, a couple who both have professional jobs earning relatively high income, and a low income family who have precarious jobs, the landlord is likely to choose the former, who they probably assume to be lower risk. So, it is likely that the tenants that the rent control policy is designed to help most are likely to be one of the groups that loses from it - an unintended consequence of the policy.

Finally, the sum of the tenant and landlord surpluses in this market is what we term total welfare - a measure of the benefit to society arising from this market. Without the rent control, total welfare is the area AEF, but with the rent control this falls to the area ABCF. There is a loss of total welfare of the area BEC - the deadweight loss of the rent control. This deadweight loss arises because the gains to tenants (in terms of more tenant surplus) from the rent control are much less than the losses to landlords (in terms of lost landlord surplus).

Moreover, there are other negative effects such as landlords' incentives for keeping their housing units in good repair are reduced (because there is a shortage of housing - tenants can't afford to be choosy about the unit they rent), and the incentives to build new developments are reduced (because they will be much less profitable than if the market rent prevailed). The effects are also worse in the long run (including a larger deadweight loss). So, as Caleb Malik concludes:

...rent control is often marketed by politicians as a way to help the lower class. They paint the landlord as an old miser who is simply out to get the poor, and rent control as a way to stop this antagonist. The public often idealizes such a scenario. When not critically analyzed, rent control sounds utopian. One of our largest expenses has now been limited? Who wouldn’t sign up for that? Yet, the effects are rather tragic, as both liberal and conservative economists agree.

On the one hand, tenants find extremely limited housing, and the housing they do find is run down or comes with a price tag that is nowhere near affordable. On the other hand, landlords can no longer profit from their investments. This results in both a loss of new developments and a reduction in the quality of current housing.

*****

[*] However, it is by no means certain that tenant surplus would increase. If the controlled rent is too far below the market rent, then tenant surplus starts to decrease. To see why, consider the extreme example where the controlled rent is zero - at that rent no landlords would offer their properties for rent, and the tenant surplus falls to zero!

Read more:

• Why rent controls have worse effects over longer time horizons

The advantages of creating a market for live kidney donations

In the ECON100 tutorials, one of my favourite questions covers the welfare effects of creating a market for live kidney donations. And it's topical too - there was a front page story on the issue on the New Zealand Herald on 8 April (though it is no longer available online?).

The premise for the market approach is fairly basic. At the moment in most countries (including the U.S. and New Zealand) there is no market for kidneys, and kidney donors are not directly compensated for their kidney. Effectively this means that, in the market for kidneys, the price is fixed at zero, as in the diagram below. At this price, the quantity of kidneys demanded (QD) is greater than the quantity of kidneys supplied (QS). So, there is a shortage of kidneys - not every patient who has a need for a kidney transplant is able to receive one. The waiting list for a kidney transplant is about 450 in New Zealand, and about 93,000 in the U.S.

If live kidney donors (or deceased donors' estates) were compensated for giving up their kidneys, this would remove the effective price control from this market. The market price for a kidney would rise from zero to P1 (in the diagram above). The quantity of kidneys supplied would increase from QS to Q1 (as at least some people who wouldn't have given up kidneys for nothing would choose to do so in return for compensation), while the quantity of kidneys demanded would fall slightly from QD to Q1 (the slight decline here is because demand for kidneys is very inelastic - there are few substitutes for a kidney [*]). There would no longer be a shortage of kidneys.

The welfare impacts are likely to be large. Without the market for kidneys, the consumer surplus (a measure the amount by which transplant recipients benefit from kidneys) is the area ABCO (plus the triangle above AB which extends above the top of the diagram), which is also the measure of total welfare because there is no producer surplus. If the market was allowed to operate (no pun intended!), consumer surplus would be ABEP1 (plus the area above AB), and producer surplus (a measure of the amount that donors benefit from the market) would be P1ECO. Total welfare would now be ABECO (plus the area above AB) - an increase of BEC.

There are other ways of demonstrating the gains from compensating kidney donors. A recent paper in the American Journal of Transplantation by Philip Held (Stanford), Frank McCormick (Bank of America - retired), Akinlolu Ojo (University of Michigan Health Systems), and John Roberts (University of California San Francisco Transplant Service) evaluates the costs and benefits of government compensation of kidney donors. The paper is very readable, and I encourage you to read it (it is ungated). It's also notable for having twelve supplements of additional material to the paper, supporting the analysis and conducting sensitivity testing. The abstract provides the best summary of the results:

From 5000 to 10 000 kidney patients die prematurely in the United States each year, and about 100 000 more suffer the debilitating effects of dialysis, because of a shortage of transplant kidneys. To reduce this shortage, many advocate having the government compensate kidney donors. This paper presents a comprehensive cost-benefit analysis of such a change. It considers not only the substantial savings to society because kidney recipients would no longer need expensive dialysis treatments—$1.45 million per kidney recipient—but also estimates the monetary value of the longer and healthier lives that kidney recipients enjoy—about $1.3 million per recipient. These numbers dwarf the proposed $45 000-per-kidney compensation that might be needed to end the kidney shortage and eliminate the kidney transplant waiting list. From the viewpoint of society, the net benefit from saving thousands of lives each year and reducing the suffering of 100 000 more receiving dialysis would be about $46 billion per year, with the benefits exceeding the costs by a factor of 3. In addition, it would save taxpayers about $12 billion each year.

There's a lot that my ECON110 students can gain from reading this paper, as it makes extensive use of techniques we develop in a much simpler way in that paper. Among other things, Held et al. demonstrate that kidney transplantation is more cost-effective than kidney dialysis ($49,000 vs. $186,000 per Quality-Adjusted Life Year gained) for end-stage renal disease.

The overall conclusion though is again, that compensation for live donors of kidneys makes economic sense. Held et al. have a final comment for those who still hold a dissenting view:

Finally, we encourage those who oppose compensating kidney donors to place a monetary value on their concerns and to show how they outweigh the very large net benefits demonstrated by this analysis. If they do, they may discover—as we did in Supplement 6—that many of the arguments usually made against compensation of kidney donors turn out instead to be arguments in favor.

[HT: Marginal Revolution for the Held et al. paper]

*****

[*] Although, it could be argued that the demand for kidneys would actually increase, because a kidney transplant can be expected to last only 10-15 years, after which the transplant recipient would require another kidney transplant.

Are robots the cure for the cost disease?

I just finished reading William Baumol's book "The Cost Disease", which is based on his long-standing research explaining why costs in education and health care are spiralling out of control. In short, the premise of the cost disease can be fairly easily summarised, as follows.

Baumol contends that there are two sectors in the economy: (1) a "stagnant sector" where labour productivity growth (the growth in the amount of 'stuff' that can be 'produced' in each hour of a worker's time) is slower than average; and (2) a "progressive sector" where labour productivity growth is faster than average. Activities that fall into the "stagnant sector" are those that require substantial human input that cannot be easily reduced (at least, not without a consequent reduction in quality). For example, it is difficult for a doctor to see more patients per hour, without the quality of care being reduced. In contrast, activities that fall into the "progressive sector" are those where technology is making workers much more productive, such as in manufacturing.

Now, because these two sectors have diverging productivity trends, Baumol argues that this has implications for costs. Because workers in the "progressive sector" are becoming more productive over time, their wages increase. This should be familiar to students of economics, as we expect that in competitive labour markets the wage will be equal to the value of the marginal product of labour (VMPL; the value of stuff that the next worker employed will produce). If productivity increases, the VMPL increases, and so will wages. However, the situation in the "stagnant sector" is different. In that sector, wages increase not because of productivity gains, but because if wages didn't increase the labour force would increasingly move into the "progressive sector" to take advantage of the higher wages (which would reduce labour supply in the "stagnant sector", increasing wages there).

Now, think about the implications for costs. In the "progressive sector", wages are increasing, but the workers are producing more, so the cost of producing each unit of output are stable or declining. In contrast, in the "stagnant sector", wages are increasing but productivity gains are low, so the cost of 'producing' each unit of output are increasing.

Which brings us to the central conclusion of the cost disease: that costs (and prices of things produced) in the "progressive sector" will tend to decline over time, while costs (and prices of what are typically services) in the "stagnant sector" will tend in increase over time.

I think there are a number of interesting implications that flow from this realisation of the cost disease. First for me is that this has implications for the "throw-away society". Because manufacturing goods (from the "progressive sector") become cheaper over time, while most repairs must be conducted hands-on (in the "stagnant sector"), then it will get increasingly less expensive to throw away and replace items that cease to function rather than repair them. So for example, when my watch strap broke the other week, it was much cheaper for me to buy a new watch than it was to pay to have the strap repaired (helped of course by the watch being an inexpensive Warehouse watch).

The second implication of interest is the implications for economic growth. Over the long term, economic growth is essentially based on increases in productivity. However, if the "progressive sector" is becoming more productive (through labour-saving technology), it seems to me that an increasing proportion of the labour force will end up working in the "stagnant sector" (e.g. low paid service work). Since the "stagnant sector" is not increasing in productivity as fast, then the implication is that economic growth will slow down as a result.

Finally, it was interesting to read this book while the Labour Party's Future of Work Commission was in the news. In particular, we heard a lot about the 'rise of the robots'. No, not this:

But this: "Robots could do 46 percent of NZ jobs". Many of the jobs cited in that article (drivers, clerical workers, technicians, accountants) would probably be considered to be in Baumol's "stagnant sector". Even baristas might be at risk. So, that suggests to me that the "stagnant sector" may not remain stagnant for much longer. As labour-saving robots (and drones, and machine-learning algorithms, and other new technologies) increasingly take over these jobs, productivity growth in those industries will increase and costs should eventually fall. So, perhaps robots will cure the cost disease?

But in the process put a lot of people out of work. As Brian Fallow notes in this interesting New Zealand Herald article:

You might have every confidence people will come up with plenty of new ways of earning a living that do require the intelligence of the human brain, the dexterity of the human hand or a sympathetic heart.

And that the time and effort digital technology saves will get distributed as more leisure for everybody.

Eventually.

In the meantime, however, it is clear that the scale and pace of the technological revolution we are in the midst of is overwhelming the ability for those good things to happen.

Though technology boosts some people's productivity, it collapses others' to zero as they become redundant or struggle to find their first job.

Fallow concludes that a universal basic income might be necessary. However, the cost is large, but Fallow argues that this could be covered by broadening the tax base - starting with taxing capital gains. However, I'll leave that discussion for a future post.

Ice cream and intelligence

The Economist's daily chart on April Fools' Day demonstrated the correlation between per capita ice cream consumption and average score on the PISA reading scale (which measures reading for 15-year-old students in each country):

Clearly New Zealand students are underperforming in reading relative to expectations based on our ice cream consumption. But hold on before you start to wonder if Chick-fil-A is going to boost the intelligence of Americans. Of course, while the data is serious and the correlation is real, there is unlikely to be any causal relationship here (which was of course the point of The Economist's data team posting this on April 1).

Countries that are richer tend to consume greater quantities of animal protein. These countries also tend to spend more on education, which should in theory lead to better reading scores for richer countries (maybe that's the case in this graph?). Of course, there's a lot of other stuff going on, but those two contemporaneous relationships with national income are enough to ensure that there is a correlation between ice cream consumption and reading scores. That is, because ice cream consumption and national income are positively related, and reading scores and national income are positively related, then it would appear to a casual observer that ice cream consumption and reading scores are positively correlated.

Inequality, redistribution, and fairness

Yesterday's post about preferences for redistribution got me thinking, and reminded me of a couple of articles I read last year (and had filed for future reference, but not come back to until now). As I noted yesterday, preferences towards redistribution on average haven't changed over the last 30+ years in the U.S., but New Zealanders view redistribution less favourably than earlier. In the U.S., certain subgroups (older, richer, better educated, and Republican voters) have grown less in favour of redistribution over time than other subgroups.

But, preferences for redistribution don't really tell us about preferences for inequality. Redistribution is one solution to reducing inequality, but it probably matters how the redistribution is undertaken. I would argue that people (rich and poor alike) want to feel like the system treats them fairly.

Fairness is something we rarely consider in economics. However, I think it was one of the more interesting aspects of Daniel Kahneman's book Thinking, Fast and Slow. In Kahneman's research, he showed that fairness was important to how people perceive economic decisions. For instance, price increases make buyers unambiguously worse off (they have to pay more for the same good or service). However, buyers are less concerned with their loss if the price increase is perceived as fair (such as the seller passing on genuine cost increases) than when the price increase is perceived as unfair (such as the seller raising the price of umbrellas on rainy days, to take advantage of increased demand).

Coming back to inequality and redistribution, fairness probably matters a great deal here too. I think Paul Bloom nailed it with the title of this article in the Atlantic last October: "People don't actually want equality - They want fairness". Bloom writes:

But in his just-published book, On Inequality, the philosopher Harry Frankfurt argues that economic equality has no intrinsic value. This is a moral claim, but it’s also a psychological one: Frankfurt suggests that if people take the time to reflect, they’ll realize that inequality isn’t really what’s bothering them.

People might be troubled by what they see as unjust causes of economic inequality, a perfectly reasonable concern given how much your income and wealth are determined by accidents of birth, including how much money your parents had, your sex, and the color of your skin. We are troubled as well by potential consequences of economic inequality. We may think it corrodes democracy, or increases crime, or diminishes overall happiness. Most of all, people worry about poverty—not that some have less, but rather “that those with less have too little.”

Frankfurt argues, though, that we aren’t really bothered by inequality for its own sake. He points out that few worry about inequalities between the very rich and the very well off, even though these might be greater, both absolutely and proportionately, than inequalities between the moderately well-off and the poor. A world in which everyone suffered from horrible poverty would be a perfectly equal one, he says, but few would prefer that to the world in which we now live. Therefore, “equality” can’t be what we really value.

What we really value, according to Bloom, is fairness. We want to know that the inequality we observe is not grossly unfair. We want people who work hard (noting that this need not be paid work) to be rewarded with a larger share of the wealth (or income), but equally we want those who would have worked but were unable to do so, to not be left with nothing.

We must also want redistributions to be fair as well. Consider the Kenyan experiment described in this Economist article, also from last October:

The Busara Centre for Behavioural Economics in Nairobi, Kenya, runs experiments with participants from slums and rural areas. Its researchers looked at the results of a lottery-like scheme in rural Kenya, in which a random sample of 503 households spread over 120 villages was chosen to receive cash transfers of up to $1,525. The average transfer, $357, was almost enough to double the wealth of a typical villager. The researchers measured the well-being of villagers before and after the transfer, using a range of different methods: questionnaires about people’s life satisfaction, screening for clinical depression and saliva tests for cortisol, a hormone associated with stress.

Since not all the villagers received a transfer, the experiment sheds no light on what would happen if everyone’s wealth increased equally. But the study does mimic the distributional results of economic growth, which tends to allot gains unevenly. As expected, those who received transfers reported greater satisfaction with their lot after the money arrived. Cortisol levels and the incidence of depression fell too.

However, the satisfaction of those who did not receive anything fell sharply as their neighbours’ fortunes improved. The decline in satisfaction prompted by seeing one’s peers get $100 richer was bigger than the increase of satisfaction from getting a handout of the same size. The bigger the handouts to others in their village, the greater the dissatisfaction of non-recipients. (The handouts did not seem to have any impact cortisol levels or the prevalence of depression among non-recipients.)

One problem with the experiment is that people are very concerned about relative comparisons - we like to be doing better than our peers. So while making some people in the village better off than others will be good for those that received the transfer, it will be bad for everyone else who didn't receive the transfer. These transfers might also be seen as unfair. Although the cash transfers were distributed randomly, I (and probably many others) wouldn't consider it fair to double the wealth of some people and leave their equally-deserving neighbours poor - especially if we could have given a smaller transfer to everyone instead without much loss in efficiency.

So, if we want fairness rather than equality, and we are concerned about our own (income or wealth) status relative to others, then that has implications for our views about redistribution. If a person is most concerned that redistribution will lead to a distribution of wealth (or income) that substantially removes the relative rewards for hard work, then they are less likely to support such redistribution. More so if they are likely to be one of those giving up wealth as a result of the redistribution (notwithstanding any altruistic 'warm glow' one might feel from giving to others). On the other hand, if a person believes that redistribution is the best way to provide for those who cannot do so for themselves (or to raise incomes for the less fortunate more generally), then they are more likely to support such redistribution. And more so if they are likely to be a net recipient of any redistribution. And this might go some way towards explaining the results in the U.S. noted above, especially the growing polarisation in preferences between Democrats and Republicans. As I've noted before, the results for New Zealand still require some further analysis.

Read more:

• Changes over time in attitudes towards redistribution in the U.S.
• Who cares about income inequality?

Changes over time in attitudes towards redistribution in the U.S.

In November last year, I wrote a post about how New Zealanders' attitudes towards redistribution had changed over time (based on this Philip Morrison article in Policy Quarterly). A new article published in the journal Economic Modelling (ungated version here), by Maria Grazia Pittau, Alessio Farcomeni, and Roberto Zelli (all Sapienza University of Rome), takes a more thorough look at similar data on preferences for redistribution from the U.S. General Social Survey covering the period 1978 to 2010.

Importantly, Pittau et al. use multi-level modelling to disentangle the cohort effects from the changes over time (a solution to the age-old problem of distinguishing between age, cohort and period effects). I hadn't considered multi-level models as a solution to this problem before, but apparently it is increasingly common (with repeated cross-sectional data).

In terms of results, the overall picture is fairly uninteresting. Attitudes to redistribution have barely changed over time, as shown in the figure below (when the proportion of respondents supporting redistribution is on the y-axis, the solid line is the average, the dotted line is the linear time trend, and each observation is actually the average of a five-year birth cohort).

The more interesting results are summarised in the conclusion of the paper, and probably confirm a lot of priors:

– Aging influences redistributive attitudes. However, support for redistribution among older people substantially decreased in the last four decades.

– Personal income has a strong performance as a predictor over the whole period, and rich people tend to oppose redistribution more strongly over time.

– There are two different time patterns for education: a downward trend for less-educated American citizens and an upward trend for the highest education level. University or college graduates increase their probability to be pro-redistribution constantly and significantly over time, while non-high school graduates reduce their likelihood persistently.

– Systematic differences between Democratic and Republican voters have enlarged in the past thirty years. Americans are much more polarized on redistributive issues by self-declared party affiliation than they were in the past.

– Ethnicity is generally regarded as a driving factor in mapping preferences towards redistribution. Our findings however show that ethnicity matters at least until the 1990s but ethnic group preferences gradually move closer over time and in the 2000s the gap seems to close.

– Further investigation confirms that in the late 1970s the racial gap was much more important than the political gap in shaping preferences for redistribution, but it was the reverse in the 2000s.

It would be interesting to see what a similar analysis for New Zealand would reveal, especially given Morrison's finding that New Zealanders' attitudes have been shifted away from a preference for redistribution. And it would be also interesting to look at home ownership as an important variable in terms of attitudes to home ownership (following Phil Morrison's subsequent suggestion that home ownership is a neglected variable in the inequality debate in general).

Read more:

• Who cares about income inequality?

Dealing with squealing children, NSW edition

I've written a couple of posts in the past about dealing with the problem of squealing children (see here and here). When I read this article last month, I thought it was time to write another. From the article:

A SYDNEY mum is furious after receiving a letter from her apartment building strata company threatening her with legal action unless she can stop her toddler from creating “excessive noise”...

In the letter, which Ms Mayer posted to Facebook over the weekend, the strata company says it has received reports from her neighbours of “shouting and screaming”, disturbing other residents and putting her in breach of the strata scheme by-laws.

Squealing (or shouting and screaming) children is a classic negative externality - an uncompensated impact of the actions of one party on a bystander. The poor residents of the apartment block face a cost that is imposed on them by the actions of the child (shouting and screaming create noise pollution). Since the child has no incentive to take into account the costs that they are imposing on the apartment residents, they generate too much noise compared to the socially efficient optimum.

How can the externality problem be solved? One solution is proposed by The Coase Theorem, which tells us that, if private parties can bargain without cost over the allocation of resources, they can solve the problem of externalities on their own (i.e. without government intervention, or the intervention of the building strata company in this case).

However, a bargaining solution is unlikely to work for the apartment building, because it would require the child (or rather their mother) to enter into an arrangement with each of the other residents of the apartment (separately or all together). We know that bargaining solutions break down (or fail to arise) when there are many parties to the bargaining - either because of coordination problems, or because one or more parties may try to hold out against a solution, in order to get a better deal for themselves (what we refer to as a 'hold-out minority').

Instead, the apartment in the story uses a command-and-control policy - a rule against excessive noise, which if breached results in a penalty of $550 for the perpetrator (or in this case, their parent). This solution is based on the "polluter pays principle". Under this principle, the party that is responsible for the pollution is solely responsible for making restitution for the damage they cause.

However, the polluter pays principle is not always the best solution to problems of negative externalities. That is because there may be other ways of solving the problem that involve a lower cost (as I have argued before). Following this 'least cost principle', instead of imposing fines on parents for their noisy children (which would be an ongoing cost to the parents), perhaps the apartments could be better sound-proofed. That would only entail a one-off cost, and although that cost might be high initially, it would also reduce the problems of externalities from neighbours who enjoy loud dinner parties or other loud activities. Avoiding those other activities entails an ongoing cost that may be more costly overall.

Read more:

• Solutions to the problem of squealing children, Japan edition
• Dealing with squealing children at least cost