The winner's curse and why cities should prefer not to have Amazon's new HQ

As many readers will know, Amazon has been searching for a location for its second headquarters in the U.S. The Washington Post reported earlier this week (gated, but there is an ungated version on the New Zealand Herald website):

Amazon's search for a site for its second headquarters is now mostly playing out behind closed doors, as officials from 20 finalist locations provide the company with additional materials...

The Amazon search is a serious matter. The chosen city could reap 50,000 jobs and $4 billion in investments from the company. Taxpayers may be asked to foot billions of dollars of subsidies to win the deal. Housing markets and traffic patterns may be dramatically affected by the company's decision. A group backed by the Koch Brothers published a video opposing subsidies for the project. On the other hand, former Virginia governor Terry McAuliffe suggested recently, in an interview, that "whoever wins this thing is going to run for president."

The mayor of the winning city may feel like running for president. But eventually, whichever city wins the affections of Amazon and gets the second headquarters will almost certainly wish they hadn't. Why would I say that? Because of the winner's curse.

Consider a group of cities vying to be the location for the second Amazon headquarters (you can see the shortlist here). Rational city planners (or the city council or mayor or whoever is making the decision to try to woo Amazon) with the same preferences would all have the same valuations (or very similar valuations) for how much economic benefit Amazon will generate for their city (based on jobs growth, for instance). However, not all city planners have the same preferences, and the city planners may make random errors in determining how much value Amazon will provide the local economy, so all of the city planners will have different willingness-to-pay to attract Amazon. In this case, their willingness-to-pay reflects how much in incentives (local tax rebates, subsidies, etc.) they are willing to offer to Amazon in exchange for Amazon locating the second headquarters in their city. For city planners with similar preferences, these differences in willingness-to-pay arise randomly - some will overestimate how much they should be willing to offer to Amazon, and some will underestimate.

Now consider Amazon's decision. They will try to maximise their "economic rent", by choosing the city that will offer them the greatest advantage, which includes the amount of tax incentives and subsidies that are on offer. Cities where the planners have underestimated the value that Amazon will provide will offer Amazon relatively small incentives to locate there, and Amazon won't choose them as a result. The 'winning' city will likely be the city that offers the greatest incentives (the most tax incentives, or largest subsidies, etc.), which will be the city that over-estimates the value that Amazon will provide to the local economy by the most.

So sure, the 'winning' city will get the Amazon headquarters. But they will also win a lot of obligations for tax incentives and subsidies offered to Amazon. Overall, the gain in terms of jobs and tax revenues will likely be less than the subsidies paid to attract Amazon, adding up to a net loss for the city. Sometimes it's best not to win.

Uber and Lyft increase traffic congestion in cities

Many people claim that ridesharing services like Uber will reduce traffic congestion. However, what happens if people start to favour those services over public transport, or over walking or cycling? The New Zealand Herald reported today:

One promise of ride-hailing companies, such as Uber and Lyft, was fewer cars clogging city streets. But studies suggest the opposite: that ride-hailing companies are pulling riders off buses, subways, bicycles and their own feet and putting them in cars instead.

And in what could be a new wrinkle, a service by Uber called Express Pool now is seen as directly competing with mass transit...

One study included surveys of 944 ride-hailing users over four weeks in late 2017 in the Boston area. Nearly six in 10 said they would have used public transportation, walked, biked or skipped the trip if the ride-hailing apps weren't available.

The report also found many riders aren't using hailed rides to connect to a subway or bus line, but instead as a separate mode of transit, said Alison Felix, one of the report's authors.

"Ridesharing is pulling from and not complementing public transportation," she said.

One of the things we are discussing in the first week of my new ECONS101 class this week is economic decision-making. And one of the key aspects of the decision between alternatives is the relative price - the cost of one alternative compared with the cost of another. Public transport may be cheap, but it isn't always convenient. Once you factor in convenience (being able to be picked up from in front of your home, rather than having to walk in the rain to the nearest bus stop or subway station), it shouldn't be surprising that many commuters are choosing Uber or Lyft instead, because in relative terms the cost of Uber or Lyft may be lower than public transport (or walking, or cycling, especially in the rain). The new Uber Express Pool service looks set to make things even worse:

Uber's new Express Pool links riders who want to travel to similar destinations. Riders walk a short distance to be picked up at a common location and are dropped off near their final destinations — essentially, how a bus or subway line functions.

The service was tested in November in San Francisco and Boston and has found enough ridership to support it 24 hours a day. Round-the-clock service was also rolled out last week in Los Angeles, Philadelphia, Washington, Miami, San Diego and Denver, with more cities to follow.

"This could be good for congestion if it causes vehicle occupancy rates to go up, but on the other hand, the Uber Pool rides and I guess these Express rides are really, really cheap, just a couple of dollars, so they're almost certainly going to be pulling people away from public transport options," [Christo Wilson, a professor of computer science at Boston's Northeastern University] said. "Why get on a bus with 50 people when you can get into a car and maybe if you're lucky, you'll be the only person in it?"

Overall, Uber might be a substitute for ambulances, but that isn't going to do you much good if your Uber ambulance is stuck in a traffic jam, mostly made up of other Ubers.

[Update: The irony of this op-ed by Richard Menzies, NZ general manager of Uber, should be obvious]

Paying landowners to change land use

Economists recognise that people respond to incentives. If you increase the cost of doing something, on average people will do less of it. On the other hand, if you increase the benefits of doing something, people will do more of it. So, economists are less surprised than 'normal people' about things like this, as reported in the New York Times at the end of last year:

In environments as different as North America and Africa, new programs are preserving land through short- and-long-term deals that pay people to protect nature on their own land. The innovation makes it possible to transform a binary approach to land use — either devoting it to private development or turning it into a nature reserve — into something in between.

Consider how Airbnb works. Think of Minneapolis during the coming Super Bowl, when hotel rooms are scarce and residents will be enticed to rent their homes to football fans. Something like that happens in the environmental realm, too: There is a surge in demand for protected land when migratory birds are passing through an area or a threatened species is breeding.

In the United States, the nonprofit Nature Conservancy has been a pioneer in bringing the “sharing economy” business model to conservation. It has been temporarily expanding wetlands for migratory birds in California’s Sacramento Valley since 2014. In early fall, when birds head south for the winter, and again in early spring on their return journey, birds need larger protected areas than the current mix of parks and nature preserves allows, as the website Howstuffworks reported in August.

The big insight was realizing “we could use a rent rather than buy model,” said Mark Reynolds, an ecologist with the Nature Conservancy, which pays rice farmers to flood their fields for the few crucial weeks each fall and spring. Rice growers routinely flood their fields for irrigation and to decompose crop residue after harvest; through the conservation program, named BirdReturns, they do so during periods when the fields would have been dry.

So often, activists and environmental campaigners settle on a 'command and control' model as their preferred policy to ensure positive environmental outcomes ('command and control' policies are policies that say you 'must do' some things, or 'must not do' other things). As the New York Times article makes clear, a market-based approach can be just as effective, if not more effective, in some circumstances. A market-based approach doesn't rely on compelling people to obey, it relies on changing the incentives.

In this case, if you offer landowners a payment for changing their land use for part of the year (e.g. flooding their fields to make wetlands), then landowners can choose whether they want to do so. Landowners who face a low cost of changing their land use (maybe because their land is not highly productive so they won't be giving up much production) will be more likely to do so (because the payment will exceed the costs of land use change). Economists refer to the value of the foregone production for these landowners as an opportunity cost - it is the cost (to those landowners) of choosing to change land use. In contrast, landowners who face a higher cost of changing their land use (maybe their land is more productive, so they would be giving up more production) will be less likely to do so. For these landowners, there is an opportunity cost of not changing their land use - they are giving up the payment they would have received from the environmental group.

How do you work out how much to pay the farmers to get the right number of them to change land use? The article explains:

A team of ecologists and economists figured out how much to compensate the farmers for this change. They ran “reverse auctions” in which landowners specified the lowest payment that would entice them to flood their fields for a given four- to eight-week period.

This auction system adjusts payments to farmers’ costs. For example, flooding during the end of the spring migration season is trickier to fit into an annual rice-growing schedule, so bids — and payments — are higher then. The auction model is also flexible when the weather fluctuates. The early years of the program occurred during California’s prolonged drought, but abundant rainfall in 2017 meant that BirdReturns could dial back the amount of pop-up wetland it procured this year.

Note that the reverse auction is a good way for the environmental group to ensure that they can achieve their desired change in land use at the lowest cost, provided the landowners are genuine in specifying their opportunity cost for changing land use as the lowest payment they would accept. The italicised bit in the last sentence is important. You don't want landowners to simply hold out for higher payments. One way to avoid that problem is to ensure that you invite more landowners (with more land) than would be necessary to achieve your desired amount of land use change.

Finally, it is worth noting that this type of market-based system can't make farmers worse off. Since farmers are not compelled to participate, they will only do so if the benefits to them outweigh the costs. Markets aren't a perfect solution for every problem, but sometimes they can solve problems in a surprisingly simple way.

Book Review: The Secret Life of Money

Sometimes books over-promise and under-deliver. I think I'm pretty good at spotting those and avoiding reading them, but occasionally one slips through. And that was the case with The Secret Life of Money, by Daniel Davies and Tess Read. The promise is big and on pages 2-3 of the book:

The problem is that the kind of economics they teach in universities is all too abstract... What you need to know is the smallest possible set of general principles, but much more about the way in which the various bits fit together...

That's what we're trying to do in this book.

The problem is that Davies and Read have created a sort of straw man argument. The kind of economics I (and many others) teach in universities is very applied (at least, I think so). So, in the very next chapter when they cover lifts (elevators for those in some countries), what they were essentially discussing was customer lock-in, which we cover in ECONS101. In the third chapter, they talk about the economics of trade shows, which are a platform market (also covered in ECONS101). Not exactly an auspicious start, if you're claiming to explain things that aren't taught in university economics. And made worse when, a few chapters later, they discuss the economics of credit cards but completely miss the fact that credit cards are also a type of platform market.

Many of the chapters are exceedingly shallow, such as a chapter on money laundering that could be best summarised as "don't do it". I'm not sure why we needed four pages to tell us that. On top of which, a surprising amount of the book is actually accounting, not economics. We can quibble over where the disciplinary boundary lies between accounting and economics, but most of us would agree that depreciation belongs in the former.

There are some highlights. Despite picking on those first two chapters above, I think they included the most interesting material and examples (perhaps that's why they were at the start of the book), and the chapter on blood diamonds ends with the key observation that there are no 'blood emeralds' (because emeralds are genuinely scarce, but diamonds are not, so emerald miners do not need to create artificial scarcity by shutting out some of the competition). There are also a few examples of epic cynicism, like:

If you want to value a brand, you have to make a discounted cash flow (DCF) model (see your favourite business school textbook - if you don't have one, then mentally substitute 'a magic spreadsheet that gives you a usually rather spurious but toothsomely precise number for the financial value of a company or project').

However, those rare highlights are not enough to redeem the book, which will tell you less about the secret life of money than about the secret life of business consultants with a book deal and too much time on their hands. Give this one a miss.

Uber as a substitute for ambulances

When a new, cheaper substitute to an existing good or service becomes available, we can expect the demand for the existing good or service to decline. Importantly though, 'cheaper' refers to the full cost of obtaining the good or service, not just the monetary cost. So in the case of ambulance services, the cost may partly be monetary (such as in the U.S.), but also includes the cost of waiting to receive treatment (which is probably a more important component of the cost in New Zealand, although in some cases ambulance services will charge for a call-out. For example, see here).

Are there cheaper substitutes for ambulances? In a country such as the U.S., where the monetary cost of calling an ambulance could easily be in the thousands of dollars, ride-hailing services like Uber or Lyft could easily be a less costly option overall. However in New Zealand, where the monetary cost of calling an ambulance is not as high, the evaluation is not quite as straightforward. If you call an ambulance for something serious, then you benefit from treatment from the ambulance officers as soon as they arrive at your location. But if you call an Uber, the driver probably arrives at your location quicker than an ambulance would (since they have less far to travel), but then you have to wait until you get to the hospital in order to receive treatment. So, for Uber to be cheaper than an ambulance in New Zealand, the monetary cost savings would have to offset the (likely) longer time to receive treatment.

So, we would expect to see Uber having some effect on ambulance services in the U.S., but less so in New Zealand. But how big an effect? In a recent paper, Leon Moskatel (Scripps Mercy Hospital) and David Slusky (University of Kansas) investigate the impact of UberX on ambulance call-outs. Using U.S. data on the timing of UberX's entry into various cities and the number of call-outs in each city, they find that there is:

...at least a 7% decrease in the ambulance rate from Uber entry into a city.

Moskatel and Slusky's paper is short and not particularly detailed. The analysis is fairly straightforward, and perhaps a little too much so. They claim that their analysis "follows a generalized difference-in-differences framework at the city-quarter level", but I don't think it does because they use only a simple dummy variable to test for the effect of Uber.

A 'difference-in-differences' analysis involves computing the difference between two differences (there's no mystery in the naming of this technique). Essentially, in this case you calculate the difference in ambulance call-outs between the period before UberX became available and the period after UberX became available (treatment cities), and then you calculate the difference in ambulance call-outs between the period before and the period after for control cities (where Uber did not become available). The problem with the analysis in this paper is that there are no control cities - all cities in their analysis had UberX become available. This doesn't bias their results, but it does affect how you interpret them, since they are really only testing for a difference in mean ambulance call-outs between cities with and without UberX.

If you have time series data but no control, you could run an interrupted time series analysis instead, which is very similar to difference-in-differences but simply tests whether the time trend in the data changes between the period before UberX became available and the period after. The results that they present in Figure 2 in the paper suggest to me that their analysis is probably under-stating the impact of UberX, since there appears to be a clear break in the time trends between the period before and the period after UberX became available.

Anyway, that is a fairly technical critique of a paper that tells us something interesting. Although, I wouldn't expect there to be as large an effect of Uber on ambulance call-outs in New Zealand.

[HT: Marginal Revolution]

Justin Wolfers on why we should care about the gender gap in economics

A couple of weeks ago, Justin Wolfers wrote an interesting article in the New York Times on the gender gap. Mostly it covered the latest research, which has been widely discussed already, along with updated data suggesting that the pipeline of female economics PhDs is declining. However, the article was more interesting (to me at least) for answering the important question of why should we care about the gender gap:

New data indicates that the share of women studying the subject in America’s universities has flatlined and the pool of prospective female economists may even be shrinking.

That pattern would be disturbing in any academic field but because economics has an outsize influence on public policy, it means that many important debates are likely to be dominated by men’s voices for years to come...

The scarcity of women economists has already had important consequences.

Consider the stark differences of opinion revealed in a 2014 survey of professional economists. Fully 63 percent of women said income in the United States should be distributed more equally, compared with only 45 percent of men. Female economists were 13 percentage points less likely to say that the United States government is too large; 18 percentage points less likely to say the United States has excessive government regulation; 20 percentage points more likely to say employers should be required to provide workers with health insurance, and 16 percentage points more likely to say current policies excessively favor economic growth over environmental quality.

Perhaps most telling was the question on pay: Only 14 percent of female economists said the gender wage gap is largely explained by differences in education and voluntary occupational choices while 54 percent of male economists agreed with that notion...

Women economists tend to focus on different topics than men. While men dominate macroeconomics, women are more visible among those studying labor markets, health and education. The only majority-female economics conference I’ve ever attended was on the economics of children, a field focused on schooling, family structure and child well-being. If there were more female economists, more attention would surely be paid to these issues.

The most striking statistics of all come from a survey taken over 20 years ago, in which 98 percent of women economists agreed with the proposition that “there is a ‘good-old-boy’ network in the economics profession.” A smaller majority of men agreed. Without more women in the field, that kind of network is likely to persist. 

So, there are several good reasons why we should care about reducing the gender gap in economics, and hopefully interventions like the one I blogged about yesterday can help some of the way.

[HT: Marginal Revolution]

Read more:

• Female representation in economics
• CORE on 'missing women in economics'
• More on the gender gap in economics
• The gender gap in economics

Could role models close the gender gap in economics?

I've written a few posts on the gender gap in economics (the latest one here). If the gender gap is a problem, then it is reasonable to consider how we might best go about closing that gap. First, we would want to know where the gender gap in economics arises, in order to identify where to target an intervention. Is it that female students are less prepared at high school for economics study at university? Some have argued that is the case. Or maybe in their first economics paper at university, we are turning female students off economics (or alternatively, we are doing the job of attracting female students into an economics major less well than other disciplines like accounting or marketing are attracting those same students into their majors)? Initial results from my Summer Research Scholarship student (which I will blog about soon) are suggesting that is the case. Or maybe female economics students are less likely to complete their major? Our results suggest that isn't happening.

So, let's say that the problem is in the introductory economics paper that students encounter in their first year at university. How do we better sell economics as a major to female students? Homa Zarghamee (Barnard College) has argued that the way we teach economics needs to change. However, we also need to make it clear to female students that there are career pathways in economics that are suitable (and indeed, attractive) for them. And one way to do that is through exposing female students to role models.

Moneyish reported last month on a new working paper by Catherine Porter (Heriot-Watt University) and Danila Serra (Southern Methodist University), where the authors assessed the impact of bringing past alumna to speak to economics principles classes, on how many female students later chose to major in economics. They first had two female undergraduate students interview (via Skype) and choose speakers from a short-list (which also included male alumni). They then randomised classes, so that some classes had the two selected alumna visit, where:

Each visit consisted of a discussion about the role model's experience as an economics major, a description of their career paths and achievements, and an explanation of how their specific major (economics) contributed to their success on the job.

The randomisation allowed Porter and Serra to test whether students in the classes that had the visits differed in whether they chose economics as a major, compared with students in the classes that didn't have visits. They found that:

...the role model intervention had a significant and large impact on both outcomes for female students, while having no impact on male students. Being in a class that received the role model visits increased the likelihood that a female student would take an intermediate microeconomics class the following academic year by 12 percentage points, and the likelihood that she would express the intention to major in economics by 7.8 points. This corresponds to a 100 percent increase in both variables.

Simply having a female role model visit the class and talk about their experiences and career path was enough to double the probability of female students choosing to major in economics (and had no effect on male students). That's astounding. The effect of the visits was greatest for the top female students (those with a GPA of greater than 3.7). Interestingly (to me), having a female professor had a significant effect on taking the intermediate microeconomics course (their indicator of being an economics major) in the semester immediately after the principles course, but was not significant when they also considered students who took the intermediate microeconomics course the following year. That's actually the measure my Summer Research Scholarship student is using, and we have (tentatively) found that the gender of tutor doesn't affect the choice to major in economics.

So, where to from here? I'm on the lookout for some excellent Waikato alumna to come and speak to first year economics students!

[HT: Development Impact]

Read more:

• Female representation in economics
• CORE on 'missing women in economics'
• More on the gender gap in economics
• The gender gap in economics

Book Review: Misbehaving

I have been looking forward to reading "Misbehaving: The Making of Behavioral Economics", by the 2017 Nobel Prize winner Richard Thaler, for some time. It's been marketed as "As readable as Freakonomics; as provocative as Thinking Fast and Slow" (neither of which I have reviewed as I read both before I started blogging). And for once I think the marketing hype wasn't overplayed and I certainly wasn't disappointed in reading it.

Thaler writes in a very readable style, including many amusing anecdotes to present the story of his research career and how behavioural economics has developed over that time period. There is a lot to learn from the book, even for someone familiar with the key developments in behavioural economics. For instance, I came away with a full page of notes that I will use later this year in my ECONS102 class (where I introduce behavioural economics concepts in the first week). Mental accounting is given a thorough treatment (as you might expect, given that it was one of the early insights that Thaler developed), as well as problems of self-control (ditto) and fairness (based on work with fellow Nobel laureate Daniel Kahneman). On fairness, Thaler questions the fairness (not the economic efficiency) of Uber surge pricing (I have previously discussed Uber's surge pricing here and here).

I was surprised to learn of the depth and breadth of Thaler's contributions to behavioural finance (a field with which I am not familiar at all, so perhaps shouldn't have been surprised) and sports economics (including his consulting for multiple NFL teams). He has even written papers on decision-making in game shows (including Golden Balls, which I blogged about here, interestingly using exactly the same example as Thaler). In reading the book though, I got a very real sense of why Thaler won the Nobel Prize. The book contains a veritable Who's Who of big names in economics, who Thaler has collaborated with or interacted with over many years. Of course, it isn't the calibre of the company you keep that puts you in line for a Nobel Prize, but when the big names are willing to keep you around, this is surely a strong signal of your quality as a researcher.

As you might expect from a book that covers the development of a field over a long period, Thaler ends by looking forward to the future of economics. He notes that:

If I were to pick the field of economics I am most anxious to see adopt behaviorally realistic approaches, it would, alas, be the field where behavioral approaches have had the least impact so far: macroeconomics.

There is good reason that economists mostly failed to anticipate the Global Financial Crisis, and unrealistic macroeconomic models have been rightly criticised for their unrealistic assumptions about human behaviour. Unfortunately, Thaler notes that there has been a lack of success in past efforts towards behavioural macroeconomics, including a short-lived annual meeting organised by George Akerlof and Robert Shiller. On a more positive note, Thaler sees the rise of randomised controlled trials as good news for economics, especially as this type of research is increasingly being published in top economics journals.

There is a lot of depth to this book, but the anecdotes and Thaler's additional thoughts around them were a real highlight for me. I especially liked this bit about Kenneth Arrow at a conference in 1985:

Arrow began by dumping on the idea that rationality is necessary... Arrow noted that there could be many rigorous, formal theories based on behavior that economists would not be willing to call rational... Yet, he noted, one could easily build a theory based on habits. When prices change, the consumer choose the affordable bundle that is closest to what she was consuming before. Arrow could have gone even further. For example, we could have rigorous theories as bizarre as "choose the bundle with brand names in order to maximise the occurrences of the letter K." In other words, formal models need not be rational; they don't even have to be sensible. So we should not defend the rationality assumption on the basis that there are no alternatives.

Another anecdote, from the Preface, stuck with me throughout the book:

[In a phone call with a New York Times Magazine reporter] ...I heard Danny [Kahneman] say: "Oh, the best thing about Thaler, what really makes him special, is that he is lazy."...

To this day, Danny insists that it was a high compliment. My laziness, he claims, means I only work on questions that are intriguing enough to overcome this default tendency of avoiding work.

I, for one, am glad that Thaler was not too lazy to complete this book. It was an excellent read, and I recommend it for anyone interested in how behavioural economics has developed into one of the most interesting fields within economics.

Why honey thefts are on the rise

Last week, the New Zealand Herald reported:

"Theft of beehives has become a growing issue over the last few years. Where once there was the odd, isolated regional incident, today we're seeing theft occur more often," [Apiculture New Zealand chief executive Karin] Kos said.

"A lot of hives are in isolated areas and it seems to be small groups or individuals who have some knowledge of bees and how to transport them.

"We believe most of the activity happens at night," she said. That's also the time when the bees are in the hives and quiet.

An increase in honey value, particularly manuka honey, appears to be a key factor behind the rise in crime throughout New Zealand, but most occurs in central North Island, Bay of Plenty and Northland, Kos said.

Why the increase in honey thefts? The 1992 Nobel Prize winner Gary Becker identified that rational criminals would weigh up the benefits and costs of their actions, in his economic theory of crime (see the first chapter in this pdf).

A similar way of thinking about it is represented in the diagram below. Marginal benefit (MB) is the additional benefit of engaging in one more honey theft. In the diagram, the marginal benefit of honey thefts is downward sloping - the more honey thefts a criminal engages in, the more likely they are to get caught and the harder it is to 'fence' their stolen honey the less they can sell their stolen honey for (because it is harder to 'fence' greater quantities of stolen honey). Marginal cost (MC) is the additional cost of engaging in one more honey theft. The marginal cost of honey theft is upward sloping - the more honey thefts a criminal engages in, the higher the opportunity costs (they have to give up more valuable alternative activities they could be engaging in, and as well, they are more likely to get caught and it becomes harder to 'fence' their stolen honey). The 'optimal quantity' of honey thefts (from the perspective of the thief!) occurs where MB meets MC, at Q* honey thefts. If the criminal engages in more than Q* thefts (e.g. at Q2), then the extra benefit (MB) is less than the extra cost (MC), making them worse off. If the criminal engages in fewer than Q* thefts (e.g. at Q1), then the extra benefit (MB) is more than the extra cost (MC), so conducting one more theft would make them better off.

Now consider what happens in this model when the value of honey increases (because of increased demand from China or elsewhere). The benefits of honey crime increase. As shown in the diagram below, this shifts the MB curve to the right (from MB0 to MB1), and increase the optimal quantity of honey thefts by criminals from Q0 to Q1. Honey thefts increase.

So, how do you combat honey crime? Becker's model suggests that you can reduce crime by either increasing the costs of crime (e.g. by either increasing the probability that criminals are caught, or increasing the penalties for criminals who are caught, or both), or by decreasing the benefits of crime (e.g. by making it difficult for stolen honey to be traded, such as by having some sort of registration and tracking system for legally-traded honey). If you increase the costs of crime by increasing policing and enforcement or increasing punishments, this shifts the MC curve up and to the left, decreasing the optimal quantity of honey thefts. If you decrease the benefits of crime by making it difficult to trade in stolen honey, this shifts the MB curve down and to the left, also decreasing the optimal quantity of honey thefts.

Which is better? Both alternatives are not without cost - more policing or longer prison sentences are both costly, as is the setup and maintenance of a system of registration and tracking of honey. To really work out which is more cost-effective, we'd need to know their costs, as well as how responsive criminals are to changes in the costs and/or benefits of crime. Given police resources can become rather stretched though, perhaps the honey industry should be looking at developing a solution themselves (and at their own cost)?

The challenges for producing ethnic population projections

I spent the last two days at the Pathways Conference at the Albany campus of Massey University. The Pathways conference is run by the research team of the CADDANZ (Capturing the Diversity Dividend of Aotearoa New Zealand) project, of which I am one of the team members. Usually, the Pathways conference focuses on migrants and immigration, but this time there was a more direct focus on diversity (you can find links to videos of the keynote speakers' presentations here).

My presentation at the conference was on ethnic population projections, especially for small ethnic groups. I won't post today on the full details of that presentation (which will be a forthcoming working paper that I will talk about then), but I did want to discuss three challenges for producing ethnic population projections.

Statistics New Zealand produces ethnic population projections only for the main 'Level 1' ethnic groups in New Zealand (European or Other, Maori, Pacific, Asian, and the omnibus groups Middle Eastern/Latin American/African), as well as for the three largest 'Level 2' ethnic groups (Samoan, Chinese, and Indian). There are good reasons why they don't produce projections for smaller groups such as Dutch, Fijian, or Vietnamese (being the three groups that I presented projections for at the Pathways conference).

The first challenge is lack of data. If you want to produce population projections using the traditional 'cohort component model', you need to be able to project births, deaths, and migration for the population groups you want to project. To project future births, deaths, and migration, you create a model based on observed numbers and rates of births, deaths, and migration in the past. This is very difficult for small population groups, because the numbers of observed births, deaths, and migration events is smaller and noisier. In some (or many) years, there might be no events of that type. For example, there might be no births to Vietnamese mothers aged 15-19 in some years, which makes it difficult to project.

The second challenge is that, in addition to projecting births, deaths, and migration, you also need to project inter-ethnic mobility. Inter-ethnic mobility occurs when a person's ethnicity changes. You might think that ethnicity is static, but that isn't true at all, as this article by Carolyn Liebler and others explains:

Add something else to the list of things that seem simple but are actually complicated – the way someone reports their race or ethnicity... With over 160 million cases [from the U.S. Census] covering all U.S. race and ethnicity groups we found that 6.1% of people in the (not-nationally-representative) data had a different race or ethnic response in 2010 than they did in 2000.

Rates of inter-ethnic mobility in New Zealand are similar (see the report from Statistics New Zealand here). This challenge arises because people can self-identify with any ethnicity, and their self-identification can change over time. These changes in self-identity are not common, but they provide yet another rare event that needs to be projected as part of an ethnic population projections model.

The third challenge is that people can hold more than one ethnicity. That might not sound like much of a challenge, but traditional models assume that each population group (by age, sex, location, etc.) is mutually exclusive. That is, a person cannot simultaneously belong to more than one group. But, if people can hold more than one ethnicity then they will belong to more than one group, which means that ethnic population projections models must run in a different way to traditional models.

Those challenges are the main reasons why Statistics New Zealand provides ethnic population projections for only a limited number of ethnic groups. In a future post, I'll discuss a method that Jacques Poot and I have been applying that allows us to go a little further and produce projections that are complementary to Statistics New Zealand's projections, and cover a wider number of much-smaller ethnic groups at both the national and regional levels.

Loss aversion, ideology and the peer review process

Last month, Andrew Gelman wrote an interesting blog post about peer review, or more accurately about how researchers steer their papers through the peer review process:

...researchers are taught to be open to new ideas, research is all about finding new things and being aware of flaws in existing paradigms—but researchers can be sooooo reluctant to abandon their own pet ideas...

My story goes like this. As scientists, we put a lot of effort into writing articles, typically with collaborators: we work hard on each article, try to get everything right, then we submit to a journal.

What happens next? Sometimes the article is rejected outright, but, if not, we’ll get back some review reports which can have some sharp criticisms: What about X? Have you considered Y? Could Z be biasing your results? Did you consider papers U, V, and W?

The next step is to respond to the review reports, and typically this takes the form of, We considered X, and the result remained significant. Or, We added Y to the model, and the result was in the same direction, marginally significant, so the claim still holds. Or, We adjusted for Z and everything changed . . . hmmmm . . . we then also though about factors P, Q, and R. After including these, as well as Z, our finding still holds. And so on.

The point is: each of the remarks from the reviewers is potentially a sign that our paper is completely wrong, that everything we thought we found is just an artifact of the analysis, that maybe the effect even goes in the opposite direction! But that’s typically not how we take these remarks. Instead, almost invariably, we think of the reviewers’ comments as a set of hoops to jump through: We need to address all the criticisms in order to get the paper published. 

Gelman argues that there is a problem with the way that researchers are trained to deal with peer review. Researchers deal with peer review by making the minimal number of changes necessary in order to ensure publication, Gelman argues that that needs to change. I don't disagree. However, I think he is missing something important about researchers, as real people.

People are loss averse. We value losses much more than equivalent gains (in other words, we like to avoid losses much more than we like to capture equivalent gains). Loss aversion makes people subject to the endowment effect - we are unwilling to give up something that we already have, because then we would face a loss (and we are loss averse). Or at least, there would have to be a big offsetting gain in order to convince us to give something up that we already have. The endowment effect applies to objects (the original Richard Thaler experiment that demonstrated endowment effects gave people coffee mugs), but it also applies to ideas.

I've thought for a long time that ideology was simply an extreme example of the endowment effect and loss aversion in practice. Haven't you ever wondered why it's so difficult to convince some people of the rightness of your way of thinking? It's because, in order for them to agree with you, that other person would have to give up their own way of thinking, and that would be a loss (and they are loss averse). It seems unlikely that the benefits of agreeing with you are enough to offset the loss they feel from giving up their prior beliefs, at least for some people. Once you consider loss aversion, it's easy to see how ideologies can become entrenched. An ideology is simply lots of people suffering from loss aversion and the endowment effect.

Now, back to peer review. Researchers are also loss averse, and when they submit an article for publication, the article includes their own ideas, analysis, and conclusions. They don't want to give those up, since that would be a loss (and researchers, like everyone else, want to avoid losses). So, it is natural for researchers to deal with reviewers' comments in a way that minimises the sense of loss that they feel from making the necessary changes (while preserving the gains from getting the article finally published). Making small, incremental changes to a research paper minimises the loss the researchers feel, and so that is the way that researchers deal with peer review.

How to solve the problem? Gelman's solution appears to be better training for researchers. However, I can't see how you can train someone out of being loss averse, which means that we are stuck with researchers who will, as Gelman says, "respond to legitimate scientific criticism in an angry, defensive, closed, non-scientific way". Maybe instead it's time to reconsider peer review? Maybe peer review should exist only to screen out the most egregious errors and not to quibble over minor details (most of which can easily be quibbled over later in online commentary on the paper)?

[HT: Marginal Revolution]

Book Review: Beauty Pays

I've written a couple of times about the effects of beauty on economic outcomes (see here and here). So, I was quite interested to read Daniel Hamermesh's 2011 book Beauty Pays: Why Attractive People are More Successful. Hamermesh is the godfather of this strand of the labour economics literature, having been involved in many of the key studies. This is a book where he lays out the state of the literature (as it was in 2011, although it might be fair to say that it hasn't moved on a whole lot since).

The book covers some key themes. First among these is expressed in the title of the book: beauty pays. Those who are more attractive earn a wage premium compared with those who are less attractive. By now, this finding is incontrovertible, although still controversial. How big is the premium? Hamermesh notes that the effect of good looks on earnings for men is about the same as an additional one and a half years of education, or five years of work experience. The premium for women is about two-thirds as large. When expressed over a lifetime, the difference between a good-looking and a bad-looking worker is about US$230,000 in lifetime earnings. So, it's not a small effect, but again it is less for women.

I know some of you will be wondering why the beauty premium is lower for women (as with many of you, I would have thought it would be larger). Hamermesh explains that:

...one explanation for the surprisingly larger effect of looks on men's than on women's earnings is that women have much more latitude than men in choosing whether or not to work for pay, and that beauty affects that choice. Part of the reason for the gender difference in the effects of beauty on earnings is that beauty alters the mix of female workers, so that the distribution of workers contains proportionately fewer below-average looking women. That is less true for men.

In other words, the less-good-looking women are less likely to work at all (compared with less-good-looking men), which means the difference between good-looking and less-good-looking women in terms of observed wages is reduced compared with the difference between good-looking and less-good-looking men.

Hamermesh then gives an excellent discussion of reasons why the beauty premium may arise due to the actions of employers. The overall discussion is difficult to excerpt, but this bit in particular caught my attention:

If we think of looks as part of a product or service, and if we assume that potential customers value looks, then it is clear how better-looking employees can raise a competitive company's sales. At the same average cost of all the other inputs into the product and at the same price charged, customers will be more likely to buy the product and/or will be willing to buy more of it. More will be sold; and the company will expand at the expense of its competitors.

It is a similar argue to one my wife is making in her PhD thesis about the work of baristas and how consumers consume the product (coffee) as well as the experience, which includes the emotional labour of the barista. Both aspects are valued by the customer.

However, there are a couple of parts of the book where I must disagree with Hamermesh. He argues that the beauty premium is socially productive, i.e. that it is associated with greater economic output and production, since good looking workers are more likely to sell things than less-good-looking workers. I'm less convinced on this point, since it ignores the fact that the beauty premium dissuades some workers from working entirely (see the paragraph on gender differences above). So, while the beauty premium might increase sales, it also has an offsetting negative impact on economic output and it isn't clear which effect (positive or negative) would be larger.

Most of the time, Hamermesh addresses any concerns I have with the book in the later chapters. He finishes the book with arguments for why ugly workers should be protected, which relies on an assumption that the cost of such a policy (through reducing the beauty premium, which is socially productive) would be outweighed by benefits to ugly workers. Again, I'm unconvinced, but there is plenty of scope for future research to address that question as there is little evidence in either direction. Hamermesh does note that helping ugly workers may come with unintended opportunity costs:

Put in stark terms, aiding workers in one disadvantaged group tends to reduce wages and take jobs away from those in other disadvantaged groups.

In that case, maybe it is better not to help ugly workers?

Overall, if you are interested in the economics of beauty, this is a great place to start. Hamermesh does an excellent job of covering the literature. Regardless of whether you agree or disagree with the normative conclusions he draws, this book is a good read.

The overstated impact of the Super Bowl

Super Bowl LII will be played in a couple of days, between the Eagles and Patriots (and with my Panthers out of the running, I'll be backing the Eagles - the Patriots have had their fill of success in recent years). Tens of thousands of fans have descended on Minneapolis for the game, and accommodation providers, hospitality businesses, and retailers will no doubt see increased sales as a result. However, there is another neglected group that benefits greatly from large sporting events - the consultants who write economic impact studies, and charge thousands of dollars for them. As I've written about before (most recently in terms of the America's Cup), most economic impact studies are junk, or at the very least heavily overstate the benefits. Kevin Draper made the case for recently in the New York Times, related to the economic impact of the Super Bowl:

Depending on what, exactly, constitutes a public dollar, taxpayers contribute an average of about $250 million to build N.F.L. stadiums, according to the advisory firm Conventions, Sports & Leisure International. For U.S. Bank Stadium in Minneapolis — which opened in 2016 and will host the Super Bowl on Sunday — the state of Minnesota spent $348 million and the city kicked in an additional $150 million, a bit less than 50 percent of the stadium’s total cost.

An economic impact report commissioned by the Minneapolis Super Bowl Host Committee stated that much of the taxpayer investment in the stadium would be recouped by the region during the event. It estimated that the Super Bowl would contribute $343 million to the region, including $29 million in tax revenue.

“We are taking a conservative approach with the numbers,” Michael Langley, the chief executive of the Minneapolis-St. Paul economic development agency, said. “But even if you are only talking about $350 million to $400 million, that’s a huge benefit to the community, just in terms of dollars spent in February.”

Sports economists don’t view the situation quite the same way. They said the economic impact study for the Minneapolis Super Bowl began by saying all the right things about how past estimates had “been criticized as extremely overinflated, inaccurate, even purposely misrepresented.” In the end, though, it did the same thing.

The main problem, usually, with economic impact studies is the incorrect specification of a counterfactual - what would have happened if the event (e.g. the Super Bowl) hadn't happened? In most cases, there wouldn't have been no visitors to the area, so it's inappropriate to suggest that all economic activity during or around the event represents extra economic activity. Draper notes:

Take hotel rooms, for example. To host the Super Bowl, Minneapolis had to show that there were at least 24,000 of them within 60 minutes of the stadium, capable of accommodating visitors during the entire 10-day Super Bowl celebration. Accordingly, the economic impact report estimates the Super Bowl will generate 230,000 nights of hotel stays.

But if the Super Bowl were not in town, many of those hotel rooms would have been filled anyway, by business travelers, conventiongoers and — yes, even in Minnesota in the dead of winter — tourists.

A related example, and closer to home. The World Sevens Series is being played in my city (Hamilton) today. My wife and I were thinking about going out for dinner in town last night, but wisely thought better of it when we considered how busy it was likely to be. The restaurant probably gained custom from the rugby fans here for the weekend, but they also lost the custom from my wife and I.

Draper also raises another good point:

The rooms cost more than they otherwise would, generating about $28 million in additional revenue. But Stephenson cautioned against assuming that the money had stayed in Indianapolis; in fact, there was heavy leakage.

“They don’t give it to the housekeeper or bellboy or front-desk person,” he said. “A lot of it just flows to whoever owns the hotel.”

To suggest that all of the benefit of additional spending in the city is a benefit to the city is to ignore where the profits will go. That's an even more important consideration for events in New Zealand, where a significant proportion of profits may be siphoned off by international hotel chains, travel providers, and so on.

In any case, it's a good reminder that we should be careful what we believe when we are presented with big numbers that are labelled 'economic impact'. Including when we read the forthcoming "independent economic report" for the Hamilton Sevens (mentioned here).

Some further notes on rent increases

Rent increases have been in the news a lot recently, especially in Wellington (see my post from earlier in the week). Based on some discussions with students, I thought I would make a few notes.

First, some have suggested that rents have increased by $50 because student allowances and Student Loan living cost payments have increased by that much (e.g. Grant Robertson's Facebook post on this). This is possible, but unlikely. If our incomes increase, we will be willing to pay more for rental housing (rental housing is what economists call a 'normal good'). That shifts our demand curve for housing up and to the right, as in the diagram below (from D0 to D1). The equilibrium rent will increase from R0 to R1. How much more will be willing to pay, and how much will the equilibrium rent increase by? Certainly our willingness-to-pay will not increase by the whole additional $50 of our income, because when our incomes increase we are also willing to pay more for food, clothing, entertainment, and all of the other normal goods that we like to buy. Our willingness-to-pay for rent will increase by some proportion of the $50 increase in income. So, I think it's extraordinarily unlikely that an increase in demand, driven by the increase in incomes, is solely behind the increase in rent.

Second, prices increase when there is excess demand. Is excess demand driving the increase in rents by $50? I covered this in my post on Monday and Eric Crampton also covered it. It's possible that rents are currently below the equilibrium rent, as shown in the diagram below. At the rent R0, there are Qd tenants looking for a house, but only Qs houses available to rent. There is excess demand for housing. Some tenants are missing out on housing. Landlords and tenants both recognise the excess demand, and tenants might start approaching landlords and offering a bit more in order not to miss out, or landlords might increase rents knowing that tenants will be willing to pay a little bit more in order not to miss out. Either way, rents start to rise and eventually the market ends up at the higher equilibrium rent R1. However, we seem to be in a perpetual state of excess demand in the rental market (if you doubt that, look at any of my past posts on housing). I've mused that maybe landlords are offering 'efficiency rents' - rents that are deliberately below the equilibrium rent, because that allows them to have the pick of applicants. So, I think it unlikely that excess demand is a big factor in increasing rents.

Third, and something that I haven't seen anybody considering, is that the accommodation supplement will increase on 1 April. Landlords signing rental agreements now must know that the accommodation supplement will increase during the term of the tenancy agreement, and that this increase in subsidy works sort of like the increase in demand shown above. However, in the case of the accommodation supplement, we can be fairly sure that it does increase willingness-to-pay for housing by $50, because it can't be used for anything else. [*] Why adjust rents now, when the accommodation supplement doesn't change for two more months? Because the rent that is agreed now can't be changed for several months after April - it makes sense to build this increase in now. This is shown in the diagram below. The accommodation supplement is a subsidy, paid to the tenant, so we show this with the D+subsidy curve, which is above the demand curve (for simplicity, the diagram doesn't show an increase in an existing accommodation supplement, which is actually what is happening, but making the diagram a bit more complex doesn't change the story at all). Another way of thinking about this is that, once the accommodation supplement increases by $50, households are willing to pay the same as before (shown by the demand curve), plus the extra $50 of the increased accommodation supplement. The rent before the increase in the accommodation supplement is R0. After the increase in the accommodation supplement, the rent that households pay to the landlords increases to RL, and the effective rent (after subtracting the increase in the accommodation supplement) falls to RT. Notice that most of the benefit of the increased accommodation supplement is captured by landlords. This is because of the very steep (inelastic) supply curve in the rental market - when rents increase, there aren't a lot of additional landlords rushing to make their houses available for renting.

So, my feeling is that the increase in rents we are observing now is the effect of landlords recognising that the accommodation supplement will soon increase, and they are trying to capture that increase early. And not the effect of the increase in student allowances, or excess demand for rental housing. Or maybe it is a little bit of all three? Of course, the easiest way to find out whether this explanation is the right one would be to ask landlords. But given the potential for a media beat-up, would you really expect them to come clean about this?

Finally, some dimwits have been advocating for rent controls (see Anna Mooney of Renters United quoted here). No. Just no.

Read more:

• That January tradition... rent increases, Wellington edition
• This is what happens when you have shortages of rental property
• Students, rental shortages, and renting over the summer
• January tradition continued... Rent increases again
• A January tradition - complaints about rising rents
• Out-of-control rent increases in Auckland - why?

*****

[*] You may be thinking here that income is fungible (it can be used for other things besides housing). That means that, when the accommodation supplement increases by $50, households' willingness-to-pay for housing might not increase by $50, because they might re-direct some of their other intended spending away from housing. However, in a world where people are quasi-rational and affected by mental accounting, it turns out that fungibility may actually be reasonably low (I'm currently reading Richard Thaler's excellent book Misbehaving: The Making of Behavioral Economics - more on that when I review it soon). Quasi-rational decision-makers have a household budget for rent, and when the accommodation supplement increases by $50, they may be thinking that means they can spend $50 more on rent (rather than other things) and act accordingly.