The FAA understands and applies marginal analysis

So often, I'm disappointed at how economically illiterate governments are, both among policy officials and elected representatives. So, I was a little surprised and happy to read this post by Eli Dourado on personal aviation. It was this bit in particular that caught my attention:

It’s possible by some lights that the market will not produce the optimal amount of safety at all relevant margins, and so regulation persists. Yet everyone should recognize that there is such a thing as overregulation.

FAA does recognize this. The graphic below comes from a document in which they explain the safety continuum doctrine. “If certifcation [sic] requirements and oversight are overly stringent,” they write, “safety can be jeopardized because the burden of certification will prevent the adoption of safety enhancing technologies.” It’s hard to get more clear than that.

Here's the graphic that Dourado refers to:

Students from my ECONS102 class should immediately recognise this as a variant of the classic marginal analysis graph:

Can the FAA (Federal Aviation Administration) regulate personal aviation too heavily? Yes, they can. They can also regulate personal aviation too loosely. But there is a 'Goldilocks zone' where the level of safety regulation is just right. Consider the graph above, with Q representing the level of safety of personal aviation. Marginal benefit (MB) is the additional benefit of an additional safety measure (more regulation). The marginal benefit of safety regulation is downward sloping. The first and most obvious safety regulations provide the greatest marginal benefits, but after FAA has enacted the obvious regulations, additional safety regulations provide much less additional benefit, as personal aviation will already be quite safe. Marginal cost (MC) is the additional cost of a safety regulation. The marginal cost of safety regulation is upward sloping - the more regulation is added, the higher the opportunity costs of additional regulation. Think about the labour involved. The more workers are diverted to regulating safety, the more society is giving up in other production from those workers. Or, to attract more workers to s, we safety regulation, we would have to offer higher wages. Either way, the marginal cost of safety regulation increases as the FAA increases safety regulation. The 'optimal quantity' of safety regulation occurs at the level of safety where MB meets MC, at Q* in the diagram. At Q1 level of safety, we would be better off with more safety, as the marginal benefit of additional safety is greater than the marginal cost of additional safety. At Q2 level of safety, we would be better off with less safety, as the marginal cost of the last unit of additional safety is greater than the marginal benefit that additional safety provides.

The FAA's diagram works similarly. At low levels of safety effort, there is too little rigour. The marginal benefit of additional safety effort is greater than the marginal cost. At high levels of safety effort, there is too little safety innovation. The marginal cost of the additional safety effort is greater than the marginal benefit.

It is good to see that there is at least one government agency that is able to apply some basic marginal analysis to their decision-making.

[HT: Marginal Revolution, back in October]

This week in research #3

After a weekend off last week for the holidays, here's a double dose of what caught my eye in research over the past two weeks:

• Lan, Clements, and Chai look at the research productivity of economics and finance PhD graduates at Australian universities (open access)
• Pugatch and Schroeder show that a message with basic information about the economics major can increase socioeconomic and racial diversity in economics (ungated earlier version here) - treat with caution, as it looks like a re-analysis of their data that found no effect on gender diversity (see here)
• Courty and Cisyk find that concussions are more likely to occur in NFL games that are higher stakes (open access)
• Hamermesh and Kosnik look at the factors associated with decreases in scholarly productivity at older ages
• Di Maio and Sciabolazza use data from 2013-2018 to look at how variations in the individual-level intensity of conflict exposure affect labour market outcomes for Palestinians living in the Gaza Strip
• Lange and Sommerfeld find that crime rates in Germany were not affected by refugee arrivals (ungated earlier version here)
• Cole and McCullough describe beer pricing across the California beer market (both wholesale and retail), and provide a clean dataset for others to use (open access)
• Truc et al. show that economics was and remains the least interdisciplinary of the social sciences, despite a turn toward interdisciplinarity in the 1990s (ungated earlier version here)
• Castro-Pires, Chade, and Swinkels disentangle adverse selection and moral hazard (ungated earlier version, interestingly not including Casto-Pires as a co-author, here) - seems heavily theoretical based on a quick read

Book review: Chokepoint Capitalism

A couple of weeks ago, I reviewed Jonathan Taplin's 2017 book Move Fast and Break Things, noting that, although the impotence of antitrust law in the face of big tech firms is an important problem, that book lacked a central thesis and as a result, didn't provide a path forward. That is not the case for Chokepoint Capitalism, by Rebecca Giblin and Cory Doctorow, which I just finished reading.

Giblin and Doctorow take aim at the same targets at Taplin, but they offer a compelling underlying story grounded in business economics: that large platform firms create 'chokepoints' between the creators of cultural content and consumers, and exploit those chokepoints to the detriment of creators. The argument here is not that consumers are made worse off. In fact, the platform firms generally benefit by creating extra value for consumers (as I noted in this recent post), and some (e.g. Google or Facebook) provide their services to consumers for no monetary cost. These firms often take advantage of network effects - the value that each user receives increases as the number of users increases - but even without network effects, the platform firms find ways of locking in their customers. Then, having all (or a large majority) of the consumers on their service allows these firms to extract profits from the creators of the content, who need to go through the firm to get to the consumers. It is this monopsony behaviour [*] that is the true problem in these markets, and which goes mostly ignored by antitrust law.

How do these firms create chokepoints? Giblin and Doctorow spend the first part of the book outlining the many and various ways that chokepoints are created, giving detailed examples of the rise of chokepoints across many industries. They later summarise that:

...businesses fortify themselves against competition by aggregating copyrights on an industrial scale and by taking advantage of network effects, licensing mazes, regulatory capture, horizontal and vertical integration, and self-preferencing. All this keeps competitors out and lets middlemen muscle their way in between audiences and culture producers to capture a greater and greater share of the money that flows from one to the other.

The book is very well written, and easy to read. A lot of the anecdotes and stories are sobering, but there is also occasion to make the reader smile, such as this (which references the ongoing legal battle between Apple and Epic Games):

We didn't believe East German bureaucrats who insisted that the Berlin Wall's purpose wasn't to keep the people locked in, but rather to stop outsiders from breaking into the workers' paradise of the German Democratic Republic. We shouldn't believe Apple when it insists that preventing interoperability is just a way of enforcing its customers' preferences. Apple can easily prove that its customers don't want to escape its walled garden: just let Epic install a gate and see if anyone goes through.

I'm sure that Apple would not be pleased to be equated with East Germany! What really sets this book apart though is the focus on solutions in the last part of the book. Rather than considering a re-working of antitrust legislation as the only path forward, Giblin and Doctorow present a number of potential solutions. As they summarise:

Creative workers and producers deserve a better deal - one that delivers them a dignified and fair share of the wealth generated by their work. We've shown some of the key actions that can get them there, like enshrining transparency and interoperability rights, simplifying licensing, facilitating collective action and cooperative ownership, putting time limits on copyright contracts, and mandating minimum wages for creative work.

Ultimately though, Giblin and Doctorow note that it will take collective action to drive systemic change:

Individual solutions aren't going to get workers a fair go any more than recycling is going to fix climate change. They might move the dial, but they won't achieve the fundamental change we need to save the world. If we're going to successfully countervail the enormous power of today's robber barons, it will be by collectively combining to do so.

To be most effective, this collective action will need to occur at a global, rather than national or local, level. Again, it is systemic change that is required, and an international collective approach is necessary to avoid large firms playing countries off against each other.

One aspect of the book that I particularly liked is that Giblin and Doctorow apply a more critical lens to their proposed solutions than many other authors would, noting for instance the practical limitations of remuneration rights. Rather than undermining their argument, this critical view adds substantial weight to their arguments.

That isn't to say that I agree with everything in the book. I'm not convinced by the argument on job guarantees for cultural workers or that cooperatives are a scalable solution, for example. And at one point, Giblin and Doctorow write that:

The cause of chokepoint capitalism is oligarchy, the concentration of wealth and power into too few hands.

However, I suspect that many of us would agree that oligarchy is an effect of the chokepoints, not just a cause. The rise of Google or Facebook alone should convince us of that point - no oligarchy was required up front to ensure the success of those firms.

Nevertheless, this is an excellent book on the topic. It is well worth a read for anyone who is interested in how the cultural markets (in particular) got to the point that they are at now, and how society might resolve these issues in the future.

*****

[*] A monopsony is a firm that is the only buyer in a market. This distinguishes it from a monopoly, which is the only seller in a particular market. The term monopsony is often also applied when there are few buyers (even though, technically, that would be an oligopsony).

The dissolution of English monasteries, the rise of the gentry, and the Industrial Revolution

There are many theories that purport to explain why England experienced the Industrial Revolution first (rather than France, say, or China). For example, the CORE Economics textbook, which I use in my ECONS101 class, focuses on changes in the relative price of coal and labour (alongside the availability of cheap resources arising from colonial conquest and slavery). It's an interesting theory, and has support from data, but probably isn't the whole story.

Another theory posits that the Dissolution of the English Monasteries in the 1530s freed up land from feudal tenure and created a rising class of gentry, who had stronger incentives for investment, innovation, and commercialisation. This theory is attributed to the historian Richard Tawney, and is the theory that is tested in this 2021 article by Leander Heldring (Northwestern University),  James Robinson (University of Chicago), and Sebastian Vollmer (University of Göttingen), published in the Quarterly Journal of Economics (ungated earlier version here).

Specifically, Heldring et al. use a variety of data at the Parish level before and after the Dissolution, and compare parishes that had monastery-owned land ('monastic parishes') with those that didn't ('non-monastic parishes'). They find a number of important results that are consistent with Tawney's theory, including:

...we first use data on the presence of markets in 1600 and the survival of perpetual copyhold into the nineteenth century. We find that former monastic parishes are more likely to have a recurring market, and are less likely to be unencumbered by feudal copyhold tenure, consistent with our interpretation of the shock...

We use a unique census from 1700 that records the number of gentry in each town and village in England and Wales to measure the presence of the gentry... We find, consistent with Tawney, that gentry are more likely to be present on formerly monastic lands. We also find that monastic lands experienced more rapid conversion and thus subsequently had fewer Catholics...

Using census data, we show that monastic parishes employ a smaller share of the working-age male population in agriculture in 1831 and a commensurately larger share in commercialized sectors, like trade and handicraft. Moreover, using data on all textile mills in England in 1838, we find that monastic parishes are more industrialized than non-monastic parishes...

We find that the presence of formerly monastic properties in a parish is positively and significantly correlated with patenting, enclosure, investment, and agricultural yield.

To summarise, monastic parishes were more likely to have markets, were less likely to have feudal tenure of land, had more gentry (and fewer Catholics), were more commercialised, had more textile mills, and had more patenting and agricultural investment, than non-monastic parishes. Heldring et al. note that:

Our results suggest that the end of monastic restrictions on the marketability of one-third of the land in England and relative incidence of customary tenure, itself directly linked to feudalism, were important for fundamental economic change. The lagged abolition of feudal land tenure in France and Germany may be behind why England pulled ahead on the world stage in the eighteenth century.

This is a really interesting and thorough paper, using a variety of the best available data. It doesn't provide the whole story for why England experienced the Industrial Revolution before other countries, but it does provide good evidence of some of the precursors that likely contributed to it.

[HT: The Dangerous Economist, last year] 

Dealing with the Doritos crunch externality

This story on FoodDive last month made me laugh:

While PepsiCo’s Doritos chips are popular with gamers, the loud crunch they make has long been a source of frustration.

The beverage and snacking giant estimated that 85% percent of U.S. gamers have consumed Doritos in the past three months. But at the same time, nearly a third of individuals reported that other people’s crunching distracts them from playing well and impacts their performance.

To “help gamers keep the crunch to themselves,” Doritos is debuting what it calls “Doritos Silent.” Gamers download Doritos Crunch Cancellation software and when the technology is turned on, the software detects the crunching sounds and silences it while keeping the gamer’s voice intact.

So, PepsiCo has released software that will cancel the noise from crunching Doritos that disturb other gamers. This is very overdue! [*] There is a serious side to this story though - PepsiCo has helped to reduce a negative externality problem.

An externality is the uncompensated impact of the actions of one person on the wellbeing of a third party. Externalities can be negative (they make the third party worse off) or positive (they make the third party better off). We call them externalities because they lie outside the decision that created them - that is, some of the costs or benefits are external to the person whose action creates them. In this case, the gamer eating Doritos imposes a cost on other gamers, who have to suffer the loud crunching noise over their headsets - it is a negative externality. Relative to the socially optimal level of Doritos eating by gamers, gamers will eat too many Doritos, since some of the costs of their Dorito-eating are passed onto other gamers.

Now, the Coase Theorem (named for the late Nobel prize-winner Ronald Coase) suggests that, if private parties can bargain without cost over the allocation of resources, they can solve the problem of externalities on their own (that is, without government intervention). However, notice the key phrase here is 'without cost'. When you have many gamers interacting with each other, some (many) of whom consume Doritos, then it will be quite a complex task to get all parties to agree to a solution to this externality problem. When there are many parties to an agreement, the transaction costs (specifically, coordination costs) of arrive at the agreement will be high. So, they would hardly be bargaining without cost, we therefore can't rely on the Coase Theorem.

If private parties can't solve the externality problem themselves, that usually means we must rely on a public policy solution. That is, the government would need to step in. But not in this case. Enter PepsiCo, and their Doritos-noise-cancelling software. If gamers can't hear the Dorito crunches of other gamers, then the negative externality is eliminated.

Now the question is, will they release an update to eliminate Cheetos noise, or the noise of slurping a Pepsi?

[HT: Marginal Revolution]

*****

[*] It turns out that Discord did something similar a few years ago.

This week in research #2

I enjoyed sharing research links last week, so I'll do it again (and it will be a regular feature from here on). Here's what caught my eye in research this week:

• Sintos conducts a meta-analysis on whether inflation worsens income inequality (answer: inflation has a small effect)
• Miller, Segal, and Spencer look at the effects of the COVID-19 pandemic on domestic violence in Los Angeles (open access)
• On a similar note, Bassier et al. look at the impacts of COVID-19 on poverty in South Africa (open access)
• Kim demonstrates a negative impact of the COVID-19 pandemic on sales of Chinese beer in South Korea (ungated earlier version here)
• Meehan et al. use PISA data for New Zealand to look at the labour market impacts of basic reading and mathematics skills of young people (open access)
• Unsurprisingly, Chopra et al. find that economics studies with statistically insignificant results are perceived by both PhD students and journal editors to be less publishable, of lower quality, less important, and less precisely estimated, than studies with statistically significant results (ungated earlier version here) - but would their study have been published if the results had been statistically insignificant?

Supermarkets' 'sawtooth pricing' of alcohol

NBR has been running a series of articles on supermarket pricing this week, based on new daily product-level data from Ordian (called PricePulse). In the latest article (paywalled) in the series this morning, Maria Slade covered the pricing of alcohol products, noting the presence of 'sawtooth pricing':

One of the clearest trends it uncovers is sharp and almost weekly movements in alcohol prices...

For example, the average price of a dozen 330ml Heineken lager bottles ranged by 39% between mid-September and this week.

The price see-sawed every seven days, from a low of $19.89 to a maximum of $32.45...

This pattern of alcohol pricing was the same in each individual supermarket NBR checked.

I was interviewed by Slade on Monday, and some of my comments on the reasons for sawtooth pricing made it into the article:

As NBR has conducted its Price Check series this week based on the PricePulse data, commentators have pointed out that huge variations in products with no seasonality attached to them are part of a strategy of obfuscation by the supermarkets.

“I think what we’re seeing is that it’s incredibly difficult for consumers to make rational decisions when it comes to groceries,” Massey University Business School professor of marketing Bodo Lang says.

Creating confusion is not the only factor driving sawtooth alcohol pricing, Waikato University professor of economics Michael Cameron says.

There are two key groups of alcohol shoppers – people who are motivated by price, and those who buy regularly; both are captured by seesawing pricing.

Regulars won’t change their habits and take advantage of a special, so retailers make a good margin on them regardless, he says.

“Then, at the same time, when the price is low, you take the consumers who are price conscious.

“So, it doesn’t make sense to price low all the time for those price-conscious ones, especially when you haven’t got much competition,” Cameron says.

I want to take this opportunity to explain my comments in a bit more detail. Obfuscation doesn't make much sense as an explanation for sawtooth pricing. There isn't much for supermarkets to gain from making consumers confused about pricing, because the consumers can see the price on the shelf and make a decision based on that. As I note in my ECONS101 class, making consumers confused about the price makes them less likely to buy it, not more likely to buy it. Obfuscation is much more effective (and profitable) if the 'true' price can be hidden from consumers until they actually have to pay. That's the strategy that is employed in 'drip pricing' for example (see here). It's also why you should always get a quote from a contractor, or risk a sharp surprise when presented with the invoice.

A better explanation for what the supermarkets are doing is a form of price discrimination. That's when a seller sells the same product to different consumers for different prices. This isn't a textbook example of price discrimination, so it requires a bit more explanation.

Consider a market with two supermarkets (A and B), and two types of consumers [*]. The first type of consumers are regular shoppers. They shop on the same day each week, at the same supermarket, and usually buy the same products. The second type of consumers are price conscious. They shop around, looking for the best deals each week.

What should Supermarket A do when pricing a product? It could set a high price all the time. The regular shoppers would buy the product, but the price conscious shoppers wouldn't. It would be moderately profitable to adopt this strategy. However, there are two problems with this. First, Supermarket B could undercut the price, capturing market share, and reducing the profits from the high price for Supermarket A. Second, if Supermarket B also sets a high price for the product, then the lack of competition in the market may come under scrutiny from competition authorities. Instead, Supermarket A could set a low price all the time. Both types of shoppers would buy the product from Supermarket A. However, that may not be very profitable (unless the supermarket can use the product as a loss leader, inducing customers to buy more of the other products that it sells), especially if Supermarket B matches the low price (since Supermarket A wouldn't gain customers at the expense of Supermarket B in that case).

A third option is to 'sawtooth' the price, alternating between a high price and a low price each week. This could actually work best because of the presence of the two different types of consumers. In the high-price week, the regular shoppers buy from Supermarket A, but the price conscious do not. In the low-price week, both types of shoppers buy from Supermarket A. This is more profitable for Supermarket A than always having a low price, because the high-price week is more profitable than the low-price week. It may (or may not) be more profitable than always having a high price, but it avoids being under-cut by the other supermarket, and avoids the unwelcome scrutiny of the competition authorities.

In this case, sawtooth pricing is a form of temporal price discrimination, since the two types of consumers end up paying different prices for the product (on average). The regular shoppers pay a relatively high price on average (the average of the high price and the low price), while the price conscious shoppers pay a low price on average (since they only buy in the low-price week). When price discriminating, a firm wants the more price conscious consumers to pay a lower price, and sawtooth pricing achieves that.

It is worth noting that there are two reasons that a sawtooth pricing strategy works in this case. First, it works because the regular shoppers are unwilling (or unable) to adjust the timing of when they buy their groceries. If these regular shoppers recognised the pattern in the prices, they might be able to hold off on buying for a week, then buy a double amount of the product in the low-price week.

Second, it only works because of a lack of competition. With two supermarkets, it is easy to maintain sawtooth pricing, since each week one supermarket can price high and the other price low, then alternating between weeks. The more additional supermarket competitors you add into the market, the more likely it is that one of them will adopt a low-price strategy, undercutting the others in every week. However, the extent to which that is possible depends on how successful the new low-price competitor can be in attracting the regular shoppers away from their usual supermarket.

These last two points provide an obvious solution to sawtooth pricing, if we believe that it's a problem. First, making consumers aware of the pricing patterns will allow them to exploit it to their advantage. I'd expect sawtooth pricing to break down fairly quickly if consumers adjust their behaviour enough. The PricePulse tool is supposed to be rolled out for consumers in due course, so that might have some effect. Second, having a more competitive market makes this pricing strategy less likely to be effective. The Grocery Commissioner has their work cut out for them in achieving that, though, even if it is on their wish list.

To me, the only thing that is left unexplained (to some extent) is why it is alcohol products that seem to be most subject to this sawtooth pricing. I haven't seen the raw data, but I wouldn't be surprised to have seen this strategy employed for many products. As yet, I don't have a good answer to this question.

 *****

[*] This explanation also works if you have many different types of consumers (what economists call heterogeneous demand), but it is simplest to explain for two types.

Does studying economics make you selfish?

It's a common trope that studying economics makes people more selfish (see for example this New York Times article (paywalled), or this NPR article). However, is it true? As I've noted before, studying economics appears to have no effect on moral reasoning. However, most papers on similar topics do a pretty poor job of actually teasing out the effect of studying economics (see here or here, for example). Often, this is because the studies don't do a good job of controlling for who selects into studying economics in the first place. So, any effects might be attributable to differences in students before they study any economics.

So, I was interested to read this new article by Daniele Girardi (King's College London), Sai Madhurika Mamunuru (Whitman College), Simon Halliday (University of Bristol), and Samuel Bowles (Santa Fe Institute), published in the Southern Economic Journal (open access). Girardi et al. look at the effect of studying intermediate microeconomics on measures of self-interest (derived experimentally using a dictator game), reciprocity (derived experimentally using a trust game), perceptions of self-interest and reciprocity (measured by survey responses to how other participants would respond in the dictator and trust games), and some measures of policy preferences.

The research participants were students at the University of Massachusetts Amherst:

Students from four different intermediate microeconomics courses and from one course outside of the social sciences comprise our sample. A course in “Nutrition and Metabolism” serves as a control non-economics course. The economics courses vary: two courses (which we call Conventional I and Conventional II) are fairly standard intermediate microeconomics courses using Pindyck and Rubinfeld (2012) and Perloff (2011); a third (Post Walrasian) course uses Bowles and Halliday (2022) and focuses on strategic interactions and contractual incompleteness alongside standard topics of optimization (crucially it contains behavioral experiments and models of social preferences); finally, the fourth course (Conventional plus social preferences), is an online course using Frank (2008). The four intermediate microeconomics courses all had the same enrollment prerequisites and identical description in the online enrollment system.

Students were surveyed at the beginning and the end of their course, which allowed Girardi et al. to employ a difference-in-differences approach. Essentially, this involves looking at the difference between students enrolled in the economics courses and students enrolled in nutrition course, before and after their course. If the change in behaviour or preferences for economics students was different than the change in behaviour for nutrition students, this could plausibly be attributed to the effect of studying economics.

Girardi et al. find that:

...a one-semester intermediate microeconomics course has little to no effect on experimental measures of social preferences or on expectations about other people's social preferences. Our estimates of the effect on measures of altruism and reciprocity are close to zero and do not differ across the differing content of the courses. We also find little evidence of an effect on the students' policy preferences or political orientations. The one exception concerns immigration: studying intermediate microeconomics (whatever the course content) seems to make students less opposed to highly restrictive immigration policies.

Even the last result about opposition to immigration policies only appears in one analysis, is based on responses to a single question, and is not robust to alternative specifications of the policy preference variables. So, studying economics appears to have no effect on selfishness or policy preferences. However, Girardi et al. note that:

The results could depend on the fact that the main effect of studying economics occurs at the introductory level, or that a single semester is too brief an exposure to produce a detectable effect.

That's possible (although Girardi et al. are quick to point out that that interpretation would be at odds with one other study that found a larger effect at intermediate than at introductory level). However, I would suggest that there is another good reason to doubt these results. The Economics Department at the University of Massachusetts Amherst is well known within the discipline as being a heterodox department. It is not mainstream, and it seems to me that it is unlikely that it attracts a majority of mainstream economics students. The types of students who self-select into studying economics at Amherst may well be students who would not be swayed into selfishness by discussions of rational utility maximisation. Alternatively, even courses that are 'mainstream' by name may approach the subject very differently at Amherst than would occur at a more traditional economics department. This study tries to address this issue by including 'conventional' intermediate economics courses, that are taught using commonly used textbooks. However, the way that lecturers approach the teaching of the subject matters too, and that is very likely to be different at Amherst than at other universities.

So, unfortunately this study may tell us little about how the teaching of economics at a traditional economics department affects student selfishness. We likely need a broader study, that uses a similar approach, but either has a sample at a more traditional economics department or, even better, looks across several economics departments at different universities.

Read more:

• Does education in economics make politicians corrupt?
• The impact of business and economics education on moral competence
• This study tells us nothing about whether studying economics make students less ethical

First-year-fees-free vs. last-year-fees-free tertiary education

Shaneel Lal wrote in the New Zealand Herald yesterday:

The new Government will change Labour’s first-year fees-free tertiary education policy to fees-free for the last year...

At first, it seems the change is to incentivise students to finish their degree, but on closer inspection, it is made primarily to make up money following the National Party’s failure to convince New Zealand First to lift the foreign buyers’ ban to fund National’s tax cuts...

However, I wonder if the order in which the fees-free year applies makes a difference. The act of having to apply for a student loan and having a debt at 17 would put off some high school leavers from getting tertiary education. University (or other tertiary education) is a daunting place to start, so adding a bureaucratic and complicated process such as applying for a student loan and having a debt of thousands could be frightening for young people.

The Government could address the concern by simplifying application processes and providing students comprehensive support to get applications in. Students would also benefit from how the loan repayment works. However, in my opinion, student support is generally appalling, particularly at universities...

The millions the Government saves from the policy change aren’t being put back into students. It will likely end up in the back pockets of uber-wealthy New Zealanders. If the Government doesn’t ensure students are not falling out of loopholes, it will rob many of their education.

Lal makes some good points. In particular, the quality of student support could definitely improve (especially financial support). However, I think that overall, the move from first-year free to last-year free is a good one, and I'm not convinced that it will reduce student numbers appreciably.

First, let's examine one key part of Lal's argument: that "the act of having to apply for a student loan and having a debt at 17 would put off some high school leavers from getting tertiary education". If applying for a loan is enough to deter 17-year-old first-year students, wouldn't it also be enough to deter 18-year-old second-year students? They have to apply for it sometime. It will be just as daunting at age 18 as it is at age 17.

However, putting that aside, the question of whether it is better to have the first-year free, or the last-year free, is an open question. As Lal notes, in terms of the overall cost to the student of their education, it should make no difference if the first year or the last year attracts no fees (except for the fact that fees do go up each year, making the last year more expensive in nominal terms [*]). However, since according to Lal only 58 percent of students complete the final year of education (and I have no reason to doubt that estimate), the cost to the government is greatly reduced by the change.

Will the number of domestic tertiary students decrease? I see little evidence of a large boost in the number of tertiary students when the first-year-fees-free policy was introduced. Here's what the data on total domestic tertiary equivalent full-time students (EFTS) looks like (source here, in the second Excel file on that page):

The first-year-fees-free policy was introduced in 2018. The trend in domestic student numbers continued its downward slide that started in 2010, through to 2020. The increase at the end is a post-COVID bump in 2021, but notice that the downward trend has resumed after that. Of course, this data isn't for first-year enrolments, which could be trending differently. However, I teach first-year university courses, and it wasn't apparent to me that there was any increase in enrolments in 2018. Of course, we don't know what the counter-factual is. Perhaps domestic tertiary enrolments would have been even lower without the policy.

Now, if we believe for a moment that the quantity of enrolments did increase as a result of the policy, there is good reason to believe that the quality of enrolments would have decreased at the same time. If the best prepared students already enrol in tertiary education, then by definition any additional enrolments must come from among students who are less well prepared for further study. We'd get more tertiary students, but the average (and marginal) student would be lower quality. This is difficult to assess, but I haven't noticed a substantial drop in the quality of first-year students. However, perhaps the margin on which additional enrolments were generated wasn't additional enrolments at university, but additional enrolments at wānanga and polytechs, and the quality of those enrolments decreased? The trend for polytech (Te Pūkenga) and wānanga enrolments was similar to the overall trend (same source as above):

Some of my colleagues were concerned that there might be an increase in the number of 'ghosts' enrolling in first-year tertiary study. 'Ghosts' are students who enrol, and then do literally nothing - they don't attend any classes, don't complete any assessments, and possibly don't do anything at all except completing their enrolment. Since the monetary cost of enrolment is eliminated by first-year-fees-free, there was an expectation that we would see more of those students as a result. I don't know what the numbers are generally, but the proportion of 'ghosts' in my first-year papers is basically identical this year to what it was in 2019. Again though, the margin for additional enrolments may have been at wānanga and polytechs, so perhaps they saw more 'ghost' enrolments?

Overall, there's little evidence of an increase in the quantity of domestic students as a result of the first-year-fees-free policy, so it seems unlikely that there would be a decrease in the quantity of domestic students when the government changes the policy to last-year-fees-free instead. Overall, we'll likely get the same number of graduates, for a lower total cost to the government.

*****

[*] And the last year might be more expensive in real terms as well, if university fees go up faster than the rate of inflation. However, allowable fee increases for 2024 and 2023 were both below the inflation rate.

Read more:

• First-year-fees-free papers are not free if you fail

Book review: Move Fast and Break Things

Robert Bork, a professor at Yale Law School, almost singlehandedly changed the way that antitrust law was applied in the US. His interpretation of antitrust law was heavily influenced by Chicago school libertarian thought, and gave primacy to consumer welfare, rather than competition. This change has directly led to the consolidation of many industries, as well as the rise of big tech firms, where consumer welfare is high (as alluded to in yesterday's post) even though competition may be minimal. The impotence of antitrust law has allowed the libertarian founders of big tech firms such as Facebook, Google, and Amazon, to monopolise culture and undermine democracy. That is essentially the message underlying Jonathan Taplin's 2017 book Move Fast and Break Things, which I just finished reading.

Taplin has led an interesting life, including stints as a tour manager for The Band, a film producer, and most recently as an academic. He brings his considerable experience of the entertainment scene to the book, with many interesting stories and examples throughout. However, the book lacks a central thesis, and is little more than an extended polemic against the big tech firms, and like all polemics it lacks balance and offers little in the way of insight.

And because of the lack of a central thesis, the book often loses its way. There are whole sections on the activities of hackers, ISIS beheading videos, Silk Road, GamerGate, and Chinese surveillance, that left me wondering how they fit into the overall picture. Answer: other than as part of a whole "look how bad big tech is" argument, they don't fit. Whole parts of the book simply come across as Taplin shouting at clouds.

Of course, there are serious problems that Taplin identifies. The best polemics encourage the reader to imagine a better world. Move Fast and Break Things doesn't do that. Given that antitrust is the issue, the best that Taplin can offer is the banal recommendation to revise the way that antitrust laws are applied. He also suggests that music production could move to a co-op structure. The problem with co-ops is that they work well when goods are homogeneous, but not when they are heterogeneous. Can you imagine Taylor Swift and Jay-Z signing up to the music production co-op and sharing their worldwide earnings with struggling artists?

Taplin also insufficiently explains references to some basic economics. At one point, he suggests that Google and Facebook create negative externalities in the same way that oil extraction does. I struggle to see how one person giving their data to Google or Facebook makes other people worse off, which is what a negative externality implies. I'm sure that you could make a case for it, but Taplin doesn't do that. Similarly, he uses Akerlof's 'market for lemons' (which economists would know as adverse selection), but doesn't show how asymmetric information about YouTube videos leads the market for videos to fail. In fact, that market is especially successful. A better metaphor drawn from economics would have been Gresham's law - the bad quality videos drive out the good quality ones.

Big tech firms are responsible for some negative things, alongside some good things. However, simply bringing together a collection of stories about their negative impacts doesn't help us to understand how the negative impacts may be prevented or mitigated. If you are interested in these questions, I wouldn't recommend this book, and instead offer Doctorow and Giblin's Chokepoint Capitalism (which I am partway through reading) as a much better alternative.

How much is your data worth to Google?

Our data is valuable to Google. Google's business model is based on using what Google knows about us to target advertisements. The more data that Google has about us, and the more accurate that data, the more precisely they can target advertising, and the more an advertiser will be willing to pay for that advertising.

We currently give our data to Google for free. It is a condition of using free services like Google Search, GMail, Google Maps, and so on. However, imagine for a moment that, instead of giving our data to Google for nothing, Google had to pay us for our data. How much do you think Google would be willing to pay each year for your data? Would Google be willing to pay $1000 per year? $10,000? $50,000? Google is incredibly profitable. Surely our data must be incredibly valuable.

In a post last month on the Asymmetric Information substack, Dave Heatley provides an indication of the answer to how valuable our data are to Google:

Last week, a Google witness in the US v. Google anti-trust trial accidentally blurted out that Apple gets a 36% cut of Google’s ad revenue from being the default search engine in Apple’s Safari browser... This was the final piece of data I needed to scrape together an estimate of the average search revenue that Google receives from each Apple user... As Apple users are, on average, richer than non-Apple users, we can treat this as an upper bound on the value to Google of an individual search user...

How many Apple users are there?

As of February 2023, Apple had more than 2 billion active devices. While iPhones predominate, this total also includes iPads, Macs and Apple Watches. This corresponds to 1.4bn individual users...

What does Google pay to advertise to them?

Google reportedly paid US$18 billion a year to be Apple’s default search engine in 2021. More recent estimates put this at around US$20bn.

That means Google is paying Apple just US$10 per year per device, or US$14 per individual Apple user, to be able to show search-related advertising.

The big reveal: just how valuable is your data to Google?

And how much revenue does Google make from that “investment”? In the ongoing US v. Google trial, University of Chicago professor Kevin Murphy, appearing for Google, revealed that Google pays Apple 36% of the revenue it earns through search advertising.

Using that figure, Google’s average annual revenue is US$28 per Apple device, or US$39 per Apple user. After paying fees to Apple, net revenue is at most 64% of that — that is, US$18 per device or US$25 per user. (The actual figure will be smaller, as it costs Google to operate their platform for both searchers and advertisers.)

So, there you have it. Your personal data is worth no more than US$25 (NZ$43) per year to Google.

If Google was required to purchase access to our data, Google would be willing to pay no more than US$25 per year to do so. If you wanted to hold out for a better deal, Google would simply say, "no thanks", and you'd lose access to their services.

So, if the value of our data to Google is just $US25 per year, why are Google's revenue and profits so high? The answer is scale. If Google has four billion regular users, each providing Google with data worth US$25 per year, that is up to US$100 billion of revenue per year (and in fact, Google's advertising business generates even more revenue than that). Even though each user's data is not worth much, the sheer number of users means high revenue and profits for Google.

Now think about the other side of our bargain with Google. How much value does Google actually give us in exchange for our data? I'd suggest that it is much more than US$25 per year. Think about all the Google services that you use. If you had to pay real money for them (and for any alternative to them), how much would you be willing to pay? We don't have an answer to that question, but we do know that people would be willing to pay between US$600 and US$2000 for access to Facebook (see here and here), which they also receive for nothing. I'd suggest that Google provides much more value than Facebook does, and so we'd be willing to pay more for Google services than for Facebook.

Taken altogether, we give our data to Google, and that data is worth US$25 per year to Google. Google provides us with services that are worth far more than US$25 per year, probably in the high hundreds of dollars or more per year. If we marketised both sides of this exchange, we could charge Google US$25 per year for our data, but Google could charge hundreds of dollars or more per year for their services.

We may gripe about Google (and Meta, and other big tech firms), and there are definitely negative aspects to their activities (more on that in my book review tomorrow). However, when you actually look at it, the big tech firms generally do appear to be giving us a good deal.

This week in research

In today's post, I thought that I would try something new. I read a lot of research, but there is far more published research than I have time to read. Part of the reason for this blog is for me to keep track of interesting research that I've read and might like to find later. However, if I don't read the research, I won't blog about it and then can't find it easily.

So, I'm going to start blogging each week (hopefully) a quick take on some of the research that caught my eye that week. Some of the research I will read and blog about in detail at a later time. Others will simply fall to the bottom of my virtual 'to-be-read' pile and eventually be deleted unread.

I must provide a quick disclaimer first: These posts are just research that caught my eye this week, drawn mostly from the emailed tables of contents I am subscribed to. Regular readers of this blog will know that my interest is reasonably eclectic, but mostly limited to papers in economics journals. Anyway, I hope that you will find it interesting.

In this week's list:

• Sodini et al. use a neat natural experiment to identify the causal impacts of home ownership in Sweden (ungated earlier version here)
• Liu and Netzer use survey response times to overcome the problems of ordinal data on subjective wellbeing (or happiness), which I've blogged about before here and here (ungated earlier version here)
• Mo, Wu, and Yuan look at the effect of air pollution on performance in eSports
• On a similar note, Gao, Zhang, and Nan look at the effect of air pollution on migration in China
• Bansak, Dziadula, and Zavodny look at the value of a 'green card' in the marriage market for Chinese migrants to the US (ungated earlier version here)
• Veenhoven and Kegel use 75 years of happiness data to answer the question: "Is life really getting worse?"
• Girardi et al. use data from intermediate microeconomics students to look at whether studying economics makes you selfish (open access)
• Tallgauer and Schank present a conceptual framework that they claim will redefine undergraduate economics education for the Anthropocene

You're fooling yourself if you think you can land that plane

Almost everyone has thought about it at least once. You're on a plane, minding your own business when suddenly and unexpectedly, an announcement comes over your entertainment system that the pilots have been incapacitated and they are urgently looking for someone to land the plane. Would you put your hand up for this heroic task? Surely, with the aid of modern instruments and the guidance of air traffic control, you could do it. How hard could it be?

Very hard, it turns out, as Guido Carim Junior and co-authors outlined in this recent article in The Conversation:

We’ve all heard stories of passengers who saved the day when the pilot became unresponsive. For instance, last year Darren Harrison managed to land a twin-engine aircraft in Florida – after the pilot passed out – with the guidance of an air traffic controller who also happened to be a flight instructor.

However, such incidents tend to take place in small, simple aircraft. Flying a much bigger and heavier commercial jet is a completely different game...

Both takeoff and landing are far too quick, technical and concentration-intensive for an untrained person to pull off. They also require a range of skills that are only gained through extensive training, such as understanding the information presented on different gauges, and being able to coordinate one’s hands and feet in a certain way.

If you think you can land a plane, you're not alone. As the authors note:

Survey results published in January indicate about one-third of adult Americans think they could safely land a passenger aircraft with air traffic control’s guidance. Among male respondents, the confidence level rose to nearly 50%.

What this demonstrates is the positivity bias, or the Dunning-Kruger effect (both related to what some psychologists call self-enhancement), where people overestimate their ability. This is also why 12 percent of men think that they could score a point off Serena Williams (see here). One interesting point is that men appear to be more susceptible to positivity bias than women (at least, based on these two examples), which probably reflects over-confidence (which men may be more likely to exhibit - see here, for example).

Positively bias is another example of how real-world decision-makers are not purely rational, but quasi-rational. A purely rational decision-maker would never be tricked into thinking that they could fly (or land) a plane without any prior training. In contrast, a quasi-rational decision-maker times that they are much better at activities than they really are. It's not all bad though. Without positivity bias, we wouldn't be able to enjoy some of the funniest (or cringiest) moments on reality television:

Sometimes even economists get mixed up about the sunk cost fallacy

In an interesting article in The Conversation this week, Aaron Nicholas (Deakin University) wrote:

Have you ever encountered a subpar hotel breakfast while on holiday? You don’t really like the food choices on offer, but since you already paid for the meal as part of your booking, you force yourself to eat something anyway rather than go down the road to a cafe.

Economists and social scientists argue that such behaviour can happen due to the “sunk cost fallacy” – an inability to ignore costs that have already been spent and can’t be recovered. In the hotel breakfast example, the sunk cost is the price you paid for the hotel package: at the time of deciding where to eat breakfast, such costs are unrecoverable and should therefore be ignored.

The problem is, the example of the subpar hotel breakfast doesn't necessarily illustrate the sunk cost fallacy at all. At least, just because some people choose the subpar hotel breakfast, it doesn't mean that those people have fallen victim to the sunk cost fallacy.

To see why, let's first consider what a purely rational decision-maker might do. A purely rational decision-maker considers only the costs and benefits of each of the alternatives available to them. As Nicholas notes, sunk costs are unrecoverable and therefore ignored. In the case of the subpar hotel breakfast, the benefits of the hotel breakfast are low, but the costs are effectively zero (since it has already been paid for). The benefits are greater than the costs. However, going down the road to a cafe has greater benefits (better food), but also comes with greater costs (the time and effort to get to the cafe, plus the monetary cost of the breakfast). It's not certain that the net benefit (benefits minus costs) would be greater for the cafe breakfast than for the hotel breakfast, even for a purely rational decision-maker. So, just because someone chooses the subpar hotel breakfast, it doesn't mean that they have fallen victim to the sunk cost fallacy.

Now consider a quasi-rational decision-maker. Quasi-rational decision-makers are loss averse (they value losses greater than monetarily-equivalent gains), and engage in mental accounting. Mental accounting suggests that we keep 'mental accounts' associated with different activities. Quasi-rational decision-makers put all of the costs and benefits associated with the activity into that mental account, and when they stop that activity, they close the mental account associated with it. And since they are loss averse, they are reluctant to close an account where the costs are greater than the benefits. In the case of the hotel breakfast, the mental account for breakfast has the cost of the breakfast in it (even though it is a sunk cost), so a quasi-rational decision-maker is more likely to stay for the subpar hotel breakfast than a purely rational decision-maker, because the quasi-rational decision-maker wants benefits (however modest) to offset the cost of the breakfast before they close the breakfast mental account. It is mental accounting (and loss aversion) that makes quasi-rational decision-makers susceptible to the sunk cost fallacy.

Taken altogether, this suggests that quasi-rational decision-makers are more likely to stay for the subpar hotel breakfast. It does not mean that staying for the subpar hotel breakfast means that a decision-maker is quasi-rational (and falling victim to the sunk cost fallacy), since a purely rational decision-maker could decide on the subpar hotel breakfast as their better option, even ignoring the sunk cost.

The other examples that Nicholas uses are better. The best examples of the sunk cost fallacy involve decision-makers continuing an activity they have started, even though the remaining costs will outweigh the remaining benefits. The sunk cost fallacy (arising from mental accounting and loss aversion) keeps us in unpromising projects for too long, as well as unhappy relationships, and bad jobs. 

Most real-world decision-makers are susceptible to the sunk cost fallacy. That's why it's sometimes more notable when we see decision-makers not falling victim to it (see here and here, for example). However, when economists explain sunk costs and the sunk cost fallacy, we need to make sure that we are using examples that unambiguously illustrate the problem.

Shaneel Lal on enforcing lecture attendance

When I saw the headline for this recent opinion piece in the New Zealand Herald by Shaneel Lal, I expected it to make me quite angry: "Forcing students to attend lectures in person creates barriers". In-class attendance has been a hot topic of conversation among academics the last few years, as COVID decimated in-person teaching and created a new norm of students not attending. However, the new norm doesn't appear to include students substituting their in-class attendance for other effective forms of learning, and so the impact on grades has been quite negative (especially for students in the bottom half of the ability distribution). Or, the impact on grades would have been negative, if universities hadn't engaged in a systematic policy of grade inflation relative to pre-COVID norms. At least, that has been my experience, and the experience of many others that I have discussed this with.

So, I expected Lal to engage in a general beat-up on whining academics who offer little in the way of value-added in their classes and yet still expect attendance. However, Lal is much more even-handed in their article:

Last year, Victoria University of Wellington announced all second-year law students would have to attend lectures in person. Students who could not attend lectures could apply for hardship grants to access lecture recordings...

Not all students come from wealthy families. Many cannot afford to be at university without taking on additional work to cover rent, bills, food, fees, technology and the many other costs that come with being alive and at university. Students are justified to prioritise work over lectures when there are clashes.

Some may argue students facing such hardship should seek hardship grants to access lectures. VUWSA president Jessica Ye labelled the hardship application process “bureaucratic”. It creates unnecessary and additional barriers for students who are already stretched thin and disproportionately disadvantages marginalised students.

For me, the primary consideration would not be whether a student suffers hardship - it would be whether the student is keeping on top of course work. If I were a lecturer, I would have no qualms about a student not attending a lecture so long as they made time to catch up.

It would be obstructive of me to withhold a recording, punishing a student who is ready and willing to catch up.

There may also be students who are not facing any hardship and attend lectures in person but require the lecture recordings anyway because they couldn’t note all the important parts during the live lecture. This is particularly the case for content-heavy lectures where lecturers race through the material or lectures in which lecturers do not use or share their slides.

Some students learn best by pausing the lecture recording every five minutes and noting everything a lecturer says...

I am not saying there aren’t valid concerns about students not attending lectures in person. I understand the genuine concern about students not keeping up with lecture content. Watching lectures live is often the best way to keep on track. The snowball effect of not watching lectures for some students is debilitating. I also accept there are some activities a student may only benefit from if they participate in the live lecture.

Like with many things, we have to do a balancing exercise. The premise for requiring students to attend lectures is to ensure students are on top of lectures, but withholding lecture recordings creates additional barriers and is counterintuitive.

I can find very little to argue with here. Withholding lecture recordings from students does create an unnecessary barrier to student learning. As I noted in this post in 2020 (ironically, just days before we went into the first COVID lockdown), over ten-plus years of recording lectures, I have seen little evidence that offering lecture recordings affects student attendance. There are much more important factors affecting attendance, such as students having to work in order to live because the loans and allowances system is not generous enough (as I've noted here and here), and lecturers not making their classes sufficiently engaging or offering enough value-add to incentivise students to attend.

However, one aspect that Lal fails to note is the negative spillover effects of a student not attending one class on attendance in their other classes. Based on conversations that I have had with students, some students do not attend lectures when they only have one in-person class on a particular day. I can appreciate that position. When weighing up the costs and benefits of travelling to university, the costs (parking or public transport costs, opportunity costs of time, and inconvenience) are more likely to exceed the benefits (learning) if the student has only one class on a particular day, than when they have many. With many classes shifted online now, the cost-benefit calculation is more weighted against in-person attendance. This is even the case for classes where attendance is expected (or encouraged). Similarly, when a lecturer simply reads from lecture slides, uses the default textbook slides, or otherwise offers very little value-added to students, the benefits of in-person attendance are low and this may spill over onto other classes where the lecturers are more engaging.

The lack of value-added and the impact of online classes have, in my experience, become much more important factors in the post-COVID teaching environment. That has meant that I have had to double-down on how I incentivise student attendance. Aside from generally trying to make my lectures engaging, I use extra credit for this purpose (as I noted here and here). I also try to make explicit the trade-off between attendance and grades early in the trimester (we have years of data showing the difference in pass rates and grades between students who attend and students who don't attend). This year, I increased the amount of extra credit on offer, and was even more enthusiastic than usual about pointing out the impact of non-attendance on grades.

Attendance in my classes has held up well, in spite of the challenges. I know that others are adopting similar approaches, and as more do, the negative spillovers start to reduce. If we want students to attend our classes, like-minded lecturers need to work together towards that goal. In fact, as Lal's article notes, we don't need to withhold lecture recordings and create barriers for students. In fact, we should be making it easier for students to engage with our classes, not more difficult.

Vice-chancellor narcissism and university performance

Over the last two decades (or more), universities have increasingly come to be managed like businesses. In that case, the role of the university vice-chancellor (or president) has come to resemble that of a business CEO. As a consequence, the types of skills a successful vice-chancellor must possess have changed. And, the types of academics attracted to becoming a vice-chancellor have also changed. If I claimed that current vice-chancellors are, on average, more narcissistic than vice-chancellors from ten years ago, I think many academics would agree.

In fact, that is one of the findings of this new article by Shee-Yee Khoo (Bangor University), Pietro Perotti (University of Bath), Thanos Verousis (University of Essex), and Richard Watermeyer (University of Bristol), published in the journal Research Policy (sorry, I don't see an ungated version online). Khoo et al. aren't interested in the level of narcissism of vice-chancellors per se, but whether vice-chancellor narcissism is related to university performance. They use data from British universities from 2009/10 to 2019/20 to investigate this question. Their sample includes 133 universities, and 261 vice-chancellors.

Interestingly, they:

...measure narcissism based on the size of the signature of the VC...

First, we obtain the signature of each VC from the university annual report, or the university strategic plan or letter when the signature is unavailable in the annual report. Second, we draw a rectangle around each VC's signature, where the signature touches its furthermost endpoint, ignoring any dot at the end of the signature or/and underline below the signature... Third, we measure the area covered by the signature by multiplying the length and width (in centimetres) of the rectangle. Fourth, we divide the area by the number of letters in the VC's name to control for the length of the VC's name.

Apparently, this is a widely used measure of narcissism <quickly checking the size of my signature>, and Khoo et al. note that it has advantages over survey-based measures because it is "unobtrusive". It is also less subject to manipulation than survey-based measures, although I was worried that the size of the signature on an electronic document (like an annual report) would create a lot of measurement error. However, in the section on robustness checks, Khoo et al. report that:

...we compare the handwritten and electronic signature sizes of the same VC, for the sample where we have both types. The size of the signature remains the same irrespective of the signature type.

Ok then. For their measures of university performance:

Firstly, the Research Excellence Framework (REF) and its predecessor until 2014, the Research Assessment Exercise (RAE), is a system for assessing the research quality of UK universities and other HE institutions... We use the overall quality of research based on the REF (formerly known as the RAE) as our research quality indicator...

Secondly, we employ the National Student Survey (NSS) which assesses teaching quality in UK universities... In particular, we employ the Student Satisfaction Score, which is the average score from across the organisation and management, learning resources, learning community and student voice sections of the NSS...

Thirdly, we use the overall university ranking, based on the Guardian newspaper.

Based on the way that they describe their results though, it seems like they use the ranking of each university, rather than the scores themselves for research quality and teaching quality. For the analysis, they perform the analysis separately for 'old universities' (those that were created before 1992) and 'new universities' (those created after 1992). The main difference between those groups of universities is that the old universities tend to be more research-focused, while the new universities tend to be more teaching-focused. The analysis for each type of university compares the university ranking two years before and two years after a change in vice-chancellor:

In particular, we rely on the universities that face a VC transition, i.e., a change from a low narcissist VC to a high narcissist VC, within the sample period. We define a high narcissist VC as one in the top quartile of the distribution. In our analysis, our treatment group consists of universities that appointed a high narcissist VC during the sample period (i.e., low-to-high transition universities). The baseline group consists of universities that appointed a low narcissist VC during the sample period (i.e., low-to-low transition universities).

Collapsing vice-chancellor narcissism to a dichotomous variable (equal to one only when a university transitions from a low-narcissism vice-chancellor to a high-narcissism vice-chancellor) deals with some of the measurement error issues I noted above. However, it does open up their analysis to some criticism, since there are many alternative arbitrary cut-offs that they could have used (and they don't report the robustness of their results to this particular choice). With that limitation in mind, Khoo et al. find that:

...a change from a low narcissist VC to a high narcissist VC is associated with a deterioration in research performance for both New and Old universities. VC Change×Narcissism Change is negative and significant, confirming that VC narcissism has a negative effect on research performance. Controlling for university as well as VC characteristics, and year and university fixed effects, a change from a low narcissist VC to a high narcissist VC is associated with a drop of approximately 16 places (VC Change×Narcissism Change = –16.07) in research performance for the sample of New universities and nine places for the sample of Old universities (VC Change×Narcissism Change = –9.57). This finding also demonstrates that New universities are more susceptible to VC transitions.

Then for teaching quality:

The coefficient of VC Change×Narcissism Change is negative and significant, indicating a drop of approximately 12 places for the group of New universities and 19 places for the group of Old universities (VC Change×Narcissism Change = –12.06 for New universities and –19.52 for Old universities).

And for overall ranking:

The transition from a low to a high narcissist VC is associated with a drop of approximately 27 places in the Guardian ranking for the group of New universities but has no effect on the Guardian ranking of Old universities.

So, narcissistic vice-chancellors lower the research and teaching performance of universities, but have more negative impact on new universities (except for teaching quality, where the impact is greater for old universities). Why do vice-chancellors negatively impact performance? It turns out that the answer may be different for old universities and new universities. Khoo et al. go on to show several additional results, including that:

...for Old universities, financial risk substantially increases with VC narcissism. Specifically, the appointment of a highly narcissistic VC deteriorates the financial sustainability of Old universities by approximately five to six [Financial Security Index] points...

Hence, the appointment of a highly narcissistic VC is associated with higher financial risk (i.e., lower financial sustainability) and lower effectiveness of the use of the resources. These results are consistent with excessive risk-taking behaviour. For Old universities, the findings suggest that highly narcissistic VCs take unnecessary risk, which might lead to a decrease in university performance...

The results for Capital Expenditures are insignificant. However, when using Expenses to Revenue as the dependent variable, the coefficient on VC Change×Narcissism Change is positive and significant at the 5 % level for the group of New universities. This evidence, although based on only one of the two measures, is consistent with highly narcissistic VCs engaging in empire-building strategies in New universities, which might be detrimental to the performance of the organisation.

I don't find the ratio of expenses to revenue very convincing as a variable, but Khoo et al. use it to suggest that narcissistic vice-chancellors in new universities are engaging in empire building. In contrast, Khoo et al argue that narcissistic vice-chancellors in old universities are engaging in excessive risk-taking behaviour. Both of these behaviours have been noted of narcissistic business CEOs as well.

So, what should universities do in order to mitigate the negative impacts of vice-chancellor narcissism? Khoo et al. recommend that:

...university councils and relevant committees should take into account and, if possible, measure the narcissism of the candidates for the role of VC... Given, however, that narcissists tend to appeal to recruiters, we also recommend that VC selection committees should undertake rigorous training that will allow them to control for this implicit bias in favour of narcissistic applicants.

I don't find those recommendations to be very convincing (and they aren't really supported by Khoo et al.'s results). What is supported is higher-quality governance, since in their final set of results:

The core finding that VC narcissism has a negative effect on research performance still holds after controlling for the effect of university governance; however, we note a decrease in the magnitude of the effect...

The results here are not consistent across all measures of university performance, but in all cases the measure of university governance does appear to moderate the effect of a shift from a low-narcissism vice-chancellor to a high-narcissism vice-chancellor.

So, the takeaway message from this research is the narcissistic vice-chancellors harm university performance, but having strong university governance can limit the damage. The only question remaining is, how did the vice-chancellors of these researchers' universities respond when they learned of this research?

New results on the bat-and-ball problem

A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost?

If you guessed ten cents, you would be in the majority. You would also be quite wrong. The correct answer is five cents. This 'bat-and-ball' problem is quite famous (and you may have seen it, or a question like it, before - a variant was in a pub quiz that I competed in a few weeks ago, for example). The problem is one of three questions included in the Cognitive Reflection Test, which purports to measure whether people engage in cognitive reflection, or are more prone give into 'intuitive thinking'. It also relates to what Daniel Kahneman referred to in his book Thinking Fast and Slow as System 1 and System 2 thinking. System 1 is intuitive and automatic (and gives a ready answer of ten cents to the bat-and-ball problem), while System 2 is slower and reflective (and is more likely to lead to the correct answer of five cents).

However, a new article by Andrew Meyer (Chinese University of Hong Kong) and Shane Frederick (Yale University), published in the journal Cognition (open access), may give us reason to question the theory of System 1 and System 2 thinking (or reason to question the validity of the bat-and-ball question). Frederick is the author who introduced the Cognitive Reflection Test, so the results reported in this paper should be considered especially notable.

Meyer and Frederick conducted a number of studies of the bat-and-ball problem, showing a number of increasingly disquieting results. First:

...verifying the intuitive response requires nothing more than adding $1.00 and $0.10 to ensure that they sum to $1.10 (they do) and subtracting $0.10 from $1.00 to ensure that they differ by $1.00 (they don't). Since essentially everyone can perform these verification tests, the high error rate means that they aren't being performed or that respondents are drawing the wrong conclusion despite performing them.

If respondents aren't attempting to verify their answer, encouraging them to do so may help. We tested this in five studies involving a total of 3219 participants who were randomly assigned to either a control condition or to one of four warning conditions shown below. Two studies were administered to students who used paper and pencil. The rest were web-based surveys of a broader population...

The warnings improved performance, but not by much... This suggests that they failed to engage a checking process, or that the checking process was insufficient to remedy the error...

Specifically, only 13 percent of research participants in the pure control group got the bat-and-ball problem correct. In the treatment group that received the simplest warning (which simply warned: "Be careful! Many people miss this problem"), this increased to 23 percent. There were modest increases in performance across other studies that Meyer and Frederick report (with various different wordings of the warning), ranging from -9 percentage points to +17 percentage points. They don't report a measure of statistical significance, but the magnitude of the change is not large, and warnings to check the answer don't eliminate the intuitive response. Evidently, research participants aren't great at checking their answer. Or maybe, they simply don't perform any check at all. What about being more directive that research participants should check their answer if their original answer was ten cents:

Since these warnings were ineffective, we next tried an even stronger manipulation by telling respondents that 10 cents is not the answer. We conducted eight such experiments, with a total of 7766 participants. In five studies (three online and two paper and pencil), participants were randomly assigned to either the control condition or to a Hint condition in which the words “HINT: 10 cents is not the answer” appeared next to the response blank...

In three other studies (two online and one in-lab), we used a within-participant design in which the Hint was provided after the participant's initial response. In those studies, respondents could revise their initial (unhinted) response, and we recorded both their initial and final responses...

The hint that the answer wasn't 10 cents helped substantially, but, more notably, many – and sometimes most – still failed to solve the problem...

Receiving the hint increased performance in the bat-and-ball problem by between +17 percentage points and +23 percentage points in a between-subjects comparison (comparing research participants who received the hint with those that didn't receive the hint), and between +16 and +22 percentage points in a within-subjects comparison (where research participants could change their answer after they received the hint). The latter results lead Meyer and Frederick to note that:

Though the bat and ball problem is often used to categorize people as reflective (those who say 5) or intuitive (those who say 10), these results suggest that the “intuitive” group can – and should – be further divided into the “careless” (who answer 10, but revise to 5 when told they are wrong) and the “hopeless” (who are unable or unwilling to compute the correct response, even when told that 10 is not the answer).

Why would so many research participants still maintain that the answer is ten cents, even when they are explicitly told that ten cents is not the correct answer? Meyer and Frederick suggest that:

This result has hallmarks of simultaneous contradictory belief (Sloman, 1996), because respondents who report that $1.00 and $0.10 differ by $1.00 obviously do not actually believe this. It is also akin to research on Wason's four card task showing that participants will rationalize their faulty selections, rather than change them (Beattie & Baron, 1988; Wason & Evans, 1974). It could also be considered as an Einstellung effect (Luchins, 1942), in which prior operations blind respondents to an important feature of the current task or as an illustration of confirmation bias, in which initial erroneous interpretations interfere with the processes needed to arrive at a correct interpretation (Bruner & Potter, 1964; Nickerson, 1998).

I would put a lot of this down to motivated reasoning. However, it gets even worse:

...we ran two studies on GCS in which we asked respondents to either consider the correct answer (N = 2002) or to simply enter it (N = 1001)...

Asking respondents to consider the correct answer more than doubled solution rates, but only to 31%. Asking them to simply enter the correct answer worked better, as 77% did so, though, notably, the intuitive response emerged even here.

So, when research participants are asked to consider if the answer could be five cents, more than half still get it wrong. And even when research participants were told that the answer is five cents, and directed to write down five cents as the answer, nearly a quarter of research participants still get the answer wrong. That leads Meyer and Frederick to conclude that:

...the very existence of such manipulations (and their lack of complete efficacy) undermines a conclusion many draw from dual process theories of reasoning: that judgmental errors can be avoided merely by getting respondents to slow down and think harder...

Meyer and Frederick use all of these results (and others) to suggest that people engage in an 'approximate checker' process, wherein if the intuitive result provided by System 1 is approximately correct, then the more deliberative System 2 doesn't go through a complete process of checking. They demonstrate this with some further results that show that:

As the price difference between the bat and ball decreases, participants slow down... and solution rates rise markedly – from 14% to 57%...

So, perhaps these results are not fatal for the idea of System 1 and System 2 thinking, but psychologists and behavioural scientists need to re-think the conditions under which System 2 operates, and whether it always operates optimally. The results also suggest that the bat-and-ball problem may not actually show quite what it purports to - at least, it doesn't necessarily show cognitive reflection, as even when such reflection is explicitly invoked (through asking research participants to check their answer, or telling them to consider if the answer might be five cents), many do not exhibit such reflection (or else, they reflect and still get the answer wrong. Meyer and Frederick finish by noting that:

...the remarkable durability of that error paints a more pessimistic picture of human reasoning than we were initially inclined to accept; those whose thoughts most require additional deliberation benefit little from whatever additional deliberation can be induced.

[HT: Marginal Revolution. back in September]