Book review: How Economics Explains the World

I have really enjoyed reading books by Andrew Leigh in the past, including The Economics of Just About Everything (which I reviewed here), and Randomistas (which I reviewed here). So I was really looking forward to reading his 2024 book How Economics Explains the World, especially after having the pleasure of introducing Leigh as keynote speaker for the ANZRSAI Conference in Canberra last December (as I noted here).

In this book, Leigh tries to condense the entire history of humanity into 194 pages, and teach us a little bit about how economics explains some of that history along the way. As he explains:

This small book tells a big story. It is the story of capitalism - of how our market system developed. It is the story of the discipline of economics, and some of the key figures who formed it. And it is the story of how economic forces have shaped world history.

To try and do all those things in a short book is no easy feat. And unsurprisingly, the book does not go into depth on any of the historical events or historical figures that it covers. That shallowness is both the book's greatest strength, and its greatest weakness. It makes the book very easy to read, and that is also helped by the beautiful style of Leigh's writing. It also allows Leigh to cover a very large swathe of economic history and the history of economic thought. However, the book only lightly skims the surface on any particular topic. So, while the book gives the reader a taste of some economic history, and a taste of some of the history of economic thought, it is really just a small taste and will definitely leave many readers wanting more.

Interestingly, I ran into an issue with this book that I haven't seen before. My copy of the hardcover had a misprint, where pages 57 through 88 were printed twice, and then the book carried on from page 121 after that. So, the book also left me wanting more, because I missed out on 32 pages of interesting content! I'm sure these sorts of misprints must happen occasionally, and so it may be surprising (given the number of books I read) that this is the first time I have seen it. None of this is a complaint against the author - I'm sure he had nothing to do with this error!

As with all Andrew Leigh books I have had the pleasure of reading, I really enjoyed this one. I don't think that readers will learn a whole lot of economics from it, because it lacks the depth to provide that. However, it does link economics to some important historical events, and briefly summarises some of the key changes in economic thought over time. That may prompt some readers to dig a little deeper on those topics, and for that reason the book is valuable, and I am happy to recommend it.

Order matters in the Great British Bakeoff and Japanese comedy competitions

If you are being judged as part of a competition, is it better to be judged first, or last, or somewhere in the middle? In theory, it shouldn't matter. If judges were impartial and purely rational decision-makers, then your ranking shouldn't depend on where you are in the order of judging. However, judges are human, and humans are typically not purely rational. So, does the order of judging matter?

That is the question addressed in this recent article by Maira Reimão, Rachel Sabbadini, and Eric Rego (all Villanova University), published in the journal Kyklos (sorry, I don't see an ungated version online). Their main analysis uses hand-collected data from 14 seasons of the Great British Bakeoff (which really means that they watched every episode of the show and recorded the details they needed for their dataset). They limit themselves to the 'technical challenge' portion of the show. Importantly:

In the technical challenge of each episode, all contestants are required to bake the same dish from a pared-down recipe provided by the judges. When the time allotted is over, the contestants place their dish on their randomly assigned spot on the judging table... The judges blind taste each bake, moving in order from their right to left on the table. The footage is also shown in this order, as judges move from dish to dish, and we are confident from the verbal cues that this is the order in which they try the dishes...

The random allocation of the order that dishes are tasted (and rated) by the judges is important, because it allows Reimão et al. to test whether the order in which a dish is judged matters for its ranking. However, not content with looking at the Great British Bakeoff (GBBO), they also:

...complement this analysis with data from international versions of the GBBO, including The Great Canadian Baking Show and The Great Kiwi Bake Off. Data collection for this portion was limited by accessibility, but we have information from thirteen seasons across six additional English-speaking countries: Australia (4 seasons), Canada (4) Kenya (1), New Zealand (2), South Africa (1), and the United States (1).

That's a lot of baking shows to watch! The judges rank every dish from worst to best, which means that the analysis can take advantage of the full ranking (rather than just relying on whatever was selected as the best). This makes the analysis a bit more efficient. Overall, using the GBBO data they find that:

... dishes tasted first are statistically significantly more likely to be ranked higher by GBBO judges than those tasted later... This primacy effect is quite large—in an episode with 10 dishes, the magnitude represents an advantage of one spot in the ranking. More broadly, dishes tasted first are 14–15 percentage points more likely to be ranked in the top half of all dishes in the technical challenge than those tasted later...

Reimão et al. then look at whether it is just the first dish that receives a boost, or whether order effects are apparent all through the order. They find that there is:

...no significant relationship between order and ranking beyond the first dish, and dishes tasted in the first third are no less likely to be ranked in the top half than dishes tasted later.

In other words, it is the first dish that is rated higher than all other dishes, rather than an effect that occurs all through the order. Reimão et al. then go on to show similar effects in the international editions of the show, which suggests that this effect might be generalisable to other settings.

Some further evidence is provided in this recent article by Real Arai (Ryukoku University) and Ryosuke Okazawa (Osaka Metropolitan University), published in the Journal of Economic Behavior and Organization (open access). They look at the effects of order in the judging of a Japanese comedy television show. As they explain:

We analyze data from ‘‘Bakusho On-Air Battle’’ and ‘‘Onbato-Plus’’, competition-style comedy shows held by NHK (Nihon Housou Kyokai, Japan Broadcasting Corporation) once a week from April 1999 to March 2014.

Similar to GBBO, there is some randomisation involved:

In every contest, ten comedy groups, possibly including solo performers, give their performances in turn. Except for the time limit of approximately five to six minutes, there are no restrictions on the performance style and number of group members.

The performance orders are randomly determined using a lottery before each contest begins. Comedians participating in the contest retrieve a ball from a box containing ten balls with numbers on them indicating the order of performance, and perform in the order of their selection...

The contests adopt a step-by-step procedure as its judging system. One hundred amateur program viewers serve as judges and evaluate each performance. Immediately after each performance, the judges are required to choose whether it is worth broadcasting. If the judge approves of the performance, he or she casts a favorable vote. If not, he or she does not vote. Given that there are no quotas, judges can approve as many groups as they want, although the program airs only the top-five performances... Voting takes place after each performance, but the results are announced only after all performances have been completed.

Arai and Okazawa look at how the order of performance affects the average vote share and the probability of being selected in the top five performances, using data from 4774 contests. They find that:

...vote share increases by approximately 9.3% for the first position and 3.1% for the last position compared with the second through ninth middle positions. Both the estimates are statistically significant at the 1% level...

...the probability of a comedian group earning a slot in the top five increases by approximately 20% if assigned to the first position. The first-position effect on the probability of being in the top five is statistically significant and robust, even if we control for the additional covariates. This also shows that being assigned to the last position increases the probability of earning a slot in the top five by approximately 5%...

Arai and Okazawa then look to explain why the first position earns a better ranking:

A promising explanation for the first-position advantage may be the calibration effect. The judges tend to give default evaluations of early performances to preserve the freedom of the judging process. This may give an advantage to the first competitors in a contest with numerous successful applicants. The implication is that an advantageous position in the sequential evaluation of a contest will depend on how competitive it is.

Perhaps. When judges are unsure how the rest of the competitors will rate, they may tend to rate the first competitor close to the middle of the range (a default). They then rate subsequent competitors relative to the first one they rated. This might also apply when judges are only asked to provide their rating (or ranking) after considering all of the competitors. I think there would need to be more research in order to unpick these effects further.

These results have important implications outside of the context of television competitions. Selection panels for awards or job interviews must consider applicants sequentially. Teachers grading assessments must consider their students' work sequentially. In both of these cases, it is worth considering whether there is a bias in favour of those who are considered first. Do the first interviewees have an advantage? Does the first student whose work is marked have an advantage? If so, then we need to consider some way of mitigating that bias. When it comes to grading, I often come to the end of an assessment, and go back and look over the first ones that I marked, in order to assure myself that I haven't been overly easy (or hard) on the first ones. Perhaps that approach needs to be adopted more widely.

Judges are not impartial and purely rational. These results tell us that judges are affected by a 'serial position bias', where the order that they consider competitors affects their ranking (and most positively for the competitor that is considered first relative to all others). With this in mind, the next time you are a contestant on the Great Kiwi Bakeoff, you want to hope that your technical challenge is judged first.

The Cristiano Ronaldo effect on the Saudi Pro League

This past week, my ECONS102 class covered labour markets. Part of that topic is a discussion of superstar and tournament effects, which are explanations for why, within a particular labour market, some workers get paid a lot while most workers get paid very little. As an example, in the labour market for actors, the top actors get paid a lot, while the 'average' actor barely earns enough to get by (or doesn't earn enough to get by, which is why so many aspiring actors work as waitstaff at restaurants).

Superstar effects arise when the worker (the superstar) earns a lot of value for their employer. Labour markets with superstar effects generally have two features:

1. Scale – the top performers can satisfy the demand of a lot of consumers (which is more likely when the output is non-rival). That generates a very high value (technically, the value of the marginal product of labour, or VMPL) for the employer, with little additional cost; and
2. Non-substitutability – the particular job (or skills, or ‘style’) performed by the top performer is unique, and cannot be easily replicated by the ‘average’ worker.

Because of the high value created by the superstar, and non-substitutability, employers compete fiercely over these top performers, and so they will receive a very high wage. Movie stars provide a good example of the superstar effect in action.

In some labour markets, workers are rewarded for their relative (rather than absolute) performance. In these markets, workers essentially compete for a ‘prize’ – maybe a raise or a promotion, and they only need to be a little bit better than the second-best person in order to ‘win’ the prize. These labour markets are said to exhibit tournament effects. In these markets, the extra pay for the top workers arises not because they create more value for the employer, but as a way of incentivising workers to work hard (in order to try and 'win' the tournament). Good examples of tournament effects in action are the pay for CEOs, sports stars, or hedge fund managers.

How can we distinguish between superstar and tournament effects? Some markets actually have both, like the market for sports stars. However, for the high wages to be a superstar effect, the worker must generate much more value for the employer than alternative workers do. And that brings me to this recent article by Dominik Schreyer (WHU – Otto Beisheim School of Management) and Carl Singleton (University of Stirling), published in the journal Contemporary Economic Policy (open access).

Schreyer and Singleton look at the impact of Cristiano Ronaldo on the Saudi Pro League, after he was surprisingly signed by the Al Nassr club a few days after the 2022 FIFA World Cup. They note the potential for superstar effects here, specifically:

The Ronaldo signing, and subsequent player moves, could attract international tourists and foreign investments to the country, help market the TV product abroad...

Schreyer and Singleton look at the impact of Cristiano Ronaldo on stadium attendance, as a measure of his positive impact on the league. Their data comes from 240 matches played by 16 different clubs during the 2022–2023 season, noting that Ronaldo made his debut for Al Nassr on January 22 and played in 16 matches over the course of the rest of the season. In their preferred regression specification, Schreyer and Singleton find that:

...the estimated average effect of Ronaldo playing at home is 20% points of capacity, and the average effect of him playing away is 15% points, significantly different from zero at the 10% level (two‐sided test).

Moreover, when they look at whether there was a general impact of Ronaldo joining the league on attendance, they find that:

...the post‐Ronaldo‐playing period of the season 2022–2023 was associated with generally higher attendance demand across all matches, by 3% points of stadium capacity, significantly different from zero at the 10% level (two‐sided test), with a further significant 17% point effect when he played at Al‐Nassr's home.

Schreyer and Singleton conclude that this is evidence in favour of a superstar effect. However, I am not entirely convinced. How much additional value is Ronaldo generating for Al Nassr and the Saudi Pro League? Mean stadium capacity in Schreyer and Singleton's sample is just 26,000. So, an increase of 15-20 percentage points is an increase of 4000-5000 people in attendance at the game. That isn't going to generate anywhere near enough additional revenue to cover Ronaldo's salary of €180 million per year. Even with jersey sales and an increase in advertising or sponsorship revenue, this will not break even for Al Nassr. On the other hand, for Saudi Arabia generally this might be a good deal. Schreyer and Singleton note that:

The Ronaldo signing... [could] legitimize KSA's other foreign sports investments, including the 2021 takeover of Newcastle United FC...

Perhaps those broader benefits are worth more than Ronaldo's salary? Saudi Arabia may value the legitimisation quite highly (hence all the accusations of sportswashing). Still, it seems to me that at least part of Ronaldo's salary is a tournament effect. If Al Nassr had signed Kylian Mbappe instead, I'm sure they would have paid a hefty salary for the privilege. Instead, Mbappe is being paid the comparatively pauper-like salary of €36 million at Real Madrid. That suggests that the next best player earns far less than Ronaldo [*], which is indicative of Ronaldo's salary being a tournament effect.

*****

[*] We could argue endlessly about who are the top and second-best players. However, choose any from this list of the top-paid footballers, and the argument still holds up.

How not to demonstrate that income inequality impacts economic growth

Many studies have estimated the relationship between income inequality and economic growth (see here and here, for example). Fewer studies have attempted to estimate a causal relationship between the two variables. Unfortunately though, that's what most of us are really interested in. Does higher inequality experience inhibit economic growth?

Theoretically, the causal relationship is not straightforward. Rising shares of income among the wealthy may hold back consumer demand, because the rich save a higher proportion of their income. That would mean that more unequal countries have lower GDP. Alternatively, governments may respond to inequality by redistribution such as progressive taxation, which reduces work and profit incentives and reduces growth. Or, high or rising inequality may reduce trust in government and undermine institutions that are critical for economic growth. On the other hand, the rich save a higher proportion of their income, and those savings are then used for investment spending, which increases economic growth. And more inequality means that those who succeed will receive very high incomes, creating incentives for entrepreneurship and innovation. So, even if inequality causes economic growth, it is unclear if inequality should cause economic growth to be higher, or lower.

This recent article by Zixiang Qi, Bicong Wang (both Beijing Wuzi University), and Yaxin Wang (Chinese Academy of Social Sciences), published in the American Journal of Economics and Sociology (sorry, I don't see an ungated version online) attempts to establish the causal relationship between inequality and economic growth in the long run, using data from 99 countries over the period from 1980 to 2018. Qi et al. find that there is an inverted-U shaped relationship between inequality and economic growth. That is, economic growth is lowest when inequality is low, and when inequality is high, but higher when inequality is middling). However, there is a major problem with the analysis.

In order to establish a causal relationship, Qi et al. rely on an instrumental variables analysis. Instrumental variables analysis involves finding some variable that is correlated with the endogenous variable (income inequality), but uncorrelated with the dependent variable (economic growth, measured as the growth rate of per capita gross national income). That means that the only effect of the instrument on the dependent variable must be through its effect on the endogenous variable.

Qi et al.'s proposed instrument is the age-dependency ratio. Here's where the problem lies. Population ageing has a direct effect on economic growth. The reason why is explained in this post. In fact, Qi et al. even acknowledge this themselves, when they write that:

...Japan's economy has been in a period of secular stagnation for several decades because of ageing population.

So, Qi et al. should know that their instrument is not a valid instrument, and yet they press ahead and use it. At that point, I think we can safely disregard the rest of their results. It is interesting that they found an inverted-U shaped correlation between income inequality and economic growth. But that is all it is - a correlation. We need better methods to establish a causal relationship, and this paper simply doesn't live up to its promise.

Read more:

• Latest research seems to support the Kuznets hypothesis
• Income inequality and economic growth in Australia
• Is there an S-shaped relationship between inequality and per capita GDP?

This week in research #85

Here's what caught my eye in research over the past week:

• Cox and Figueroa (with ungated earlier version here) show that, contrary to previous work, the Spanish Inquisition did depress scientific research in Spain, using data on scholarly interactions among early modern Spanish book authors and data on affiliation with secular educational institutions
• Ugalde (with ungated earlier version here) looks at how high school students in North Carolina react to standardised test performance labels regarding their advanced math and English enrolment decisions, and find that female students labelled as not proficient in math are less likely to enrol in advanced math courses than their proficient-labelled peers, with similar (but smaller) effects in English

The Big Five personality traits and earnings

Human capital is the education, training, experience, intelligence, motivation, and other human factors that workers use to be successful in their job. In effect, it is the capital that is bound up within the workers as a person, and inseparable from them. When economists measure human capital, we usually use education as a proxy. Sometimes we use job experience, or measures of intelligence or skill. Sometimes we use a combination of two or more of those measures.

It is less common for economists to use more intrinsic measures of human capital, such as personality. However, it seems likely that personality traits should affect earnings. So, I was interested to see what this 2023 meta-analysis article by Melchior Vella (University of Essex), published in the journal Bulletin of Economic Research (open access), would turn up. A meta-analysis is a way of quantitatively combining the results of many studies into a single overall estimate of the relationship. If done well, it can also account for publication bias (where studies that fail to show statistically significant effects are less likely to be published).

Vella looks at studies that estimate the relationship between earnings and the Big Five personality traits, which are:

...openness to experience (ability to be creative, curious, intellectually engaged, honest/humble, and inquisitive), conscientiousness (self-discipline, punctuality, and organized and general competence), extraversion (how talkative, friendly, energetic, and outgoing the person is), agreeableness (the tendency to be kind, charitable, warm, and generous), and neuroticism (fear, worry, paranoia, and stress)...

While the Big Five has copped a lot of criticism (for example, see here), the taxonomy is still widely used, which means that there are a lot of studies that Vella can draw upon in the meta-analysis. Altogether, after a thorough search and screening process, they include 52 studies, with 1307 estimates of the relationship between the Big Five personality traits and earnings (most studies have many estimates of the relationship, with various different combinations of control variables, different specifications, or robustness checks). Applying a random effects model (which is quite common in meta-analysis, since it assumes that the true effect sizes vary across studies, not just due to random sampling error, but also due to real differences between the studies), Vella finds that:

For openness to experience, the true effect size is 0.019, indicating that a one standard deviation increase in openness to experience corresponds to a 1.92% increase in earnings. Similarly, conscientiousness (θ = 0.016, 1.61%) and extraversion (θ = 0.003, 0.30%) are positively correlated with earnings, whereas agreeableness (θ=−0.017, −1.69%) and neuroticism (θ=−0.018, −1.78%) show negative correlations.

Vella checks for publication bias in the results, and finds that there is statistically significant publication bias in estimates for conscientiousness, agreeableness, and neuroticism. After adjusting for publication bias, the relationships become much smaller, and are only statistically significant for openness to experience (positively correlated with earnings) and agreeableness (negatively correlated with earnings). The other traits are not statistically significantly related to earnings, once publication bias is controlled for.

Vella then looks at heterogeneity across studies, identifying what factors are associated with the size of the reported estimates. This analysis identifies that studies that fail to account for demographic characteristics (age), family background (parental education and/or income), socioeconomic status, education, occupation, or cognitive ability, are likely to have biased estimates of the relationship between the Big Five personality traits and earnings. Since most studies won't control for all of those factors, that explains part of the reason why there is publication bias, since studies that have biased estimates may be more likely to show statistically significant effects, if the true effect is small or zero.

Overall, this meta-analysis finds evidence that people who have higher openness to experience (ability to be creative, curious, intellectually engaged, honest/humble, and inquisitive) earn more, and those who have higher agreeableness (the tendency to be kind, charitable, warm, and generous) earn less. Other personality traits don't appear to matter (or the effects are too small to be measured). Vella stops short of trying to explain why it is that openness to experience and agreeableness matter, which is just as well. These studies, and therefore the meta-analysis as well, only show the correlation between the personality traits and earnings. They do not show that the personality traits cause differences in earnings. Nevertheless, it probably wouldn't hurt for people to be a little more open to experience (although I'm less keen on people being more disagreeable!).

When drug prices fell, Mexican drug cartels pivoted to avocados

One of the determinants of the market supply of a good is the price of other goods that the seller could sell instead (what we refer to as 'substitutes in supply'). For example, consider a seller that could use their resources to produce two goods, A and B. Say that the price of A decreases, while the price of B stays the same. The seller will respond by using less of their resources to produce A, and therefore using more resources to produce B. The quantity of A supplied will decrease, while the supply of B will increase. [*]

The Waikato Economics Discussion Group looked at an example like this at our meeting today, based on this recent article by Itzel De Haro (University of Navarra), published in the journal World Development (open access). De Haro looks at whether, when the price of heroin in the US fell because of the introduction of fentanyl, the Mexican drug cartels pivoted to the avocado industry instead. In short, De Haro finds evidence consistent with that story, and consistent with the theory of the supply curve outlined above. Let's break down how De Haro showed this.

De Haro first notes that:

Mexico is also the main exporter of heroin to the U.S., responsible for 90% of the heroin consumed in the country... Consequently, fluctuations in heroin demand in the U.S. can significantly impact opium production, cartel competition and violence within Mexico...

However, beginning in 2013, the demand for opioids shifted due to the introduction of Fentanyl in the U.S...

The decrease in the U.S. demand for heroin directly affected the demand for raw opium paste used to produce heroin. As a result, prices of opium paid by drug cartels to opium farmers in Mexico decreased significantly, falling 50% to 80% between 2017 and 2019...

In other words, the price of heroin decreased, and heroin became less profitable for drug cartels. Poppy farming also became less profitable for farmers So, it is natural for the cartels (and poppy farmers) to look for alternative (and more profitable) uses of their resources. De Haro built a crop suitability index for poppy and avocado growing for the eight states in Mexico that are the main focus of the study. The index is based on the FAO’s Ecological Crop Requirements (Ecocrop) database, and historical precipitation and temperature data obtained from AgMerra, along with land elevation data from INEGI. The two suitability indexes look very similar geographically. From Figure 5 (panels (c) and (d)) in the paper:

Notice that the areas that are most suitable for growing avocados (the darker green areas on the map on the left) are to a large extent the areas that are most suitable for growing poppies (the darker blue areas on the map on the right). So, farmers have a choice of using the land for growing poppies or avocados. And cartels have a choice of targeting poppy farmers or avocado farmers in a particular area.

De Haro then looks at violence and cartel activity data across those eight states over the period from 2011 to 2019. Specifically, De Haro looks at the overall homicide rate, the homicide rate for agricultural workers, the homicide rate for males aged 15-40 killed by a firearm (as a proxy for inter-cartel violence), and drug cartel presence (based on various measures derived from data from the Mapping Criminal Organizations (MCO) project). The main regression specification looks at whether US fentanyl deaths in the previous year affects violence (or cartel activity) in the current year, differently in areas that are more suitable for growing avocados and areas that are more suitable for growing poppies, controlling for a variety of other variables.

First, looking at violence, De Haro finds that:

...the introduction of Fentanyl in the U.S. led to increases in the homicide rates of avocado-suitable municipalities, while having the contrary effect on poppy-suitable municipalities.

Because the suitability index is just an index, the actual coefficients on the variables in the regression don't have a natural interpretation (and the interaction effects make things tricky). However, it looks like a one-standard-deviation higher avocado suitability is associated with a 7.1 percent increase in the homicide rate, while a one-standard-deviation higher poppy suitability is associated with a 6.9 percent decrease in the homicide rate. De Haro also finds an:

...increase in homicides of agricultural workers for avocado-suitable municipalities, suggesting that cartels may be using force to extract revenue from farmers, possibly to enforce extortion payments, or through violent robberies.

The effects are slightly larger here, with a one-standard-deviation higher avocado suitability associated with an 8.3 percent increase in the agricultural worker homicide rate, and a one-standard-deviation higher poppy suitability associated with a 7.4 percent decrease in the homicide rate. Finally, for inter-cartel violence, De Haro finds:

...no effect in avocado-suitable municipalities and a decrease in poppy-suitable municipalities...

This latter null effect might seem surprising. However, whether cartels are fighting over poppies or fighting over avocados, the amount of fighting may well be similar. And what De Haro finds is consistent with that. Turning to cartel presence generally, De Haro finds:

...small and no statistically significant effects on cartel presence and market concentration in municipalities suited for avocados. This suggests that, despite substantial revenue growth in the avocado industry over the past two decades, the potential profits that cartels can extract are not significant enough for them to move into these municipalities.

Another way of interpreting this last result is that the cartels don't need to move into the avocado-suitable municipalities because those municipalities are also generally those that are poppy-suitable, and therefore the cartels are already there!

So, overall, De Haro finds evidence that cartels redirected their attention from poppies to avocados, when the price of heroin (derived from poppies) fell, decreasing the profitability of poppy cultivation. The practical effect was more homicide of agricultural workers in areas more suitable for avocado growing, and less homicides of agricultural workers in areas more suitable for poppy growing.

When the price of a good falls (poppies), sellers want to sell less of it, and instead use their resources for something else (avocados) instead. It turns out that drug cartels act like rational and profit maximising sellers.

*****

[*] The difference in terminology here is important. A given supply curve holds constant all of the determinants of supply other than the price. When there is a change in price, the market moves along the supply curve (a change in quantity supplied). When there is a change in one of the other determinants of supply (other than price), the entire supply curve shifts (a change in supply). In this example, because the price of A has decreased, the market moves along the supply curve for A. The seller wants to sell less of A because the price is now lower, so there is a decrease in the quantity of A supplied. The price of B has not changed, but one of the other determinants (the price of the substitute in supply, A) has decreased. That leads the seller to want to sell more of B, so there is an increase in the supply of B. This change happens regardless of the price of B. In other words, the seller would want to sell more of B at each and every price. The entire supply curve for B will shift to the right.

Book review: Co-intelligence

Regular readers of this blog will know that I have posted reasonably frequently about generative AI and its current and future impacts on education (as I see them). It would be hard not to write about that topic, even if I didn't have a keen interest in how we can use generative AI for the good of student learning. Most of my reading on generative AI has been online (Substack, blogs, etc.), and most of my learning has been hands-on. As a change of pace though, I decided to pick up and read Ethan Mollick's 2024 book Co-Intelligence.

Mollick has been one of the key thought leaders on the use of generative AI over the last few years, and for that reason he should be taken seriously. This book collates his thoughts up to the end of 2023, and provides an excellent primer for those who have had little exposure to generative AI. Readers who are very familiar with generative AI, and especially those who follow Mollick on X and/or those who have read his research, will find little new in the book. However, for other readers, there is lots to digest.

The book first explains some of the basics of generative AI. This is not a book on the computer science or mathematic underpinnings of generative AI though. It is more of an overview of how AI models work and what to expect (and not to expect) from them. That section finishes with four very helpful principles or 'rules' for using generative AI: (1) Always invite AI to the table; (2) Be the human in the loop; (3) Treat AI like a person (but tell it what kind of person it is); and (4) Assume this is the worst AI you will ever use. Those four principles are important to keep in mind whenever we use generative AI. The book then pivots, with the rest of the book outlining a number of broad use cases for generative AI.

Although the book has a generally positive tone, it is not a rave about the value of generative AI. Mollick has realistic expectations about what generative AI can and cannot do, and happily shares examples of where generative AI has definite room to improve. The use cases are clear, and helpful, and highlight some of the limits of the technology. In this way, Mollick does his best to ensure that readers are never overawed by generative AI. For example:

And you can't figure out why an AI is generating a hallucination by asking it. It is not conscious of its own processes. So if you ask it to explain itself, the AI will appear to give you the right answer, but it will have nothing to do with the process that generated the original result. The system has no way of explaining its decisions, or even knowing what those decisions were. Instead it is (you guessed it) merely generating text that it thinks will make you happy in response to your query. LLMs are not generally optimized to say "I don't know" when they don't have enough information. Instead, they will give you an answer, expressing confidence.

It is the passages like that one that are provide the greatest value from this book for the general reader. For me, though, I most enjoyed the sections on generative AI in education ('AI as a Tutor', and 'AI as a Coach'). As a professor at Wharton, regularly using generative AI in teaching, Mollick has good insights into how generative AI will likely impact education in the future. He writes that:

...the ways in which AI will impact education in the near future are likely to be counterintuitive. They won't replace teachers but will make classrooms more necessary. They may force us to learn more facts, not fewer, in school. And they will destroy the way we teach before they improve it.

For a teacher, that is equal parts reassuring and terrifying. However, Mollick clearly sees the value in education, even in the face of generative AI. That may be self-serving, but it doesn't seem to be a blinkered view. In particular, I loved this bit about the 'paradox of knowledge acquisition':

Large Language Models seem to have accumulated and mastered a lot of collective human knowledge. This vast and tappable storehouse of knowledge is now at everyone's fingertips. So it might seem logical that teaching basic facts has become obsolete. Yet it turns out the exact opposite is true.

This is the paradox of knowledge acquisition in the age of AI; we may think we don't need to work to memorize and amass basic skills, or build up a storehouse of fundamental knowledge - after all, this is what the AI is good at. Foundational skills, always tedious to learn, seem to be obsolete. And they might be, if there was a shortcut to being an expert. But the path to expertise requires a grounding in facts...

The issue is that in order to learn to think critically, problem-solve, understand abstract concepts, reason through novel problems, and evaluate the AI's output, we need subject matter expertise... We need expert humans in the loop.

I wish that more educators (and their managers) would read this book carefully, especially those sections above. Simply outsourcing tasks to generative AI is not helping students to learn and become future experts. They need a grounding in foundational knowledge - basic concepts, models, and intuitions - before they can use that knowledge in critical thinking tasks.

Notice that the last part of the last quote links the example back to the second of Mollick's four principles. These links to the principles are sprinkled throughout the book, although at times I wish they had been made a bit more explicit. Nevertheless, Mollick writes in a style that is easy to read, and the book is suffused with interesting anecdotes and references to recent and relevant research (by Mollick, and by other researchers). Those aspects of the book will likely become out of date in due course, but not yet. Although I read the book more than a year after its release, it still seemed mostly contemporary, in a space where the technology is moving incredibly fast. The book will likely age well, and the four principles will remain relevant for some time to come. 

This book was a genuine pleasure to read, and is essential reading right now. I strongly recommend it to anyone looking for a basic grounding in generative AI.

Sometimes people intuitively understand compensating differentials

Today my ECONS102 class covered compensating differentials, in a topic on the labour market. A compensating differential explains why jobs that are otherwise similar might have different equilibrium wages. For example: consider two jobs that are similar in terms of skill requirements. If one of the jobs has positive non-monetary characteristics (e.g. it is safe, clean, and engaging), then more people will be willing to do that job. This leads to a higher supply of labour for that job, which leads to lower equilibrium wages. In contrast, if the other job has negative non-monetary characteristics (e.g. it is dangerous, dirty, and boring), then fewer people will be willing to do that job. This leads to a lower supply of labour for that job, which leads to higher equilibrium wages. The difference in wages between the attractive job that lots of people want to do and the unattractive job that fewer people want to do is called a compensating differential. The compensating differential essentially compensates workers for working in jobs with negative non-monetary characteristics compared with working in jobs with positive non-monetary characteristics.

Compensating differentials don't only apply to labour markets. As one example in a rental market, see this post. Another interesting example comes from this article in The Conversation last year by Andrew Vonasch (University of Canterbury):

If you’re offered a free cookie, you might say yes. But if you’re paid to eat a free cookie, would your response be the same?

In our new research, twice as many people were willing to eat a cookie when they weren’t offered payment compared with when they were...

Research participants who were offered a free cookie plus payment thought maybe the cookies were poisoned. Or maybe someone spat on them. Or they expected they would then owe a favour to the person handing out the treats once the cookie was eaten.

Vonasch claims that their research results reveal that the research participants were making irrational economic decisions, or that "people aren’t purely economic". However, that need not be true. A rational decision-maker who intuitively understands compensating differentials might respond in exactly the same way as Vonasch's research participants.

To see why, consider the choice that the research participants were offered. Eating a cookie is an enjoyable experience for most of us most of the time. Most of us wouldn't need to be paid to eat a cookie. So, if someone offers you a cookie and offers to pay you to eat it, it is natural to wonder what negative unobservable characteristics the payment is supposed to overcome. If there wasn't anything wrong with the cookie, why would they need to pay you to eat it? That is quite a rational response to the offer.

I often apply the same approach to thinking about salary offers in jobs (for example, see here). If a job is offering a much higher salary than other similar jobs, my first thought is always: what are the negative non-monetary characteristics that the extra salary is trying to compensate for? After all, who wants to live in South Waikato? Or rural Australia?

Compensating differentials are one of my favourite concepts in economics (and that's why there are so many regular posts on this blog about them). Interestingly, it appears that research participants who are offered payment to eat cookies understand them, even if they don't realise it!

Your matcha fix is going to cost you more

The New Zealand Herald reported last month:

At a minimalist Los Angeles matcha bar, powdered Japanese tea is prepared with precision, despite a global shortage driven by the bright green drink’s social media stardom.

Of the 25 types of matcha on the menu at Kettl Tea, which opened on Hollywood Boulevard this year, all but four were out of stock, the shop’s founder Zach Mangan told AFP.

“One of the things we struggle with is telling customers that, unfortunately, we don’t have” what they want, he said.

With its deep grassy aroma, intense colour and pick-me-up effects, the popularity of matcha “has grown just exponentially over the last decade, but much more so in the last two to three years”, the 40-year-old explained.

It is now “a cultural touchpoint in the Western world” - found everywhere from ice cream flavour boards to Starbucks.

This has caused matcha’s market to nearly double over a year, Mangan said.

“No matter what we try, there’s just not more to buy.”

Thousands of kilometres away in Sayama, northwest of Tokyo, Masahiro Okutomi - the 15th generation to run his family’s tea business - is overwhelmed by demand.

“I had to put on our website that we are not accepting any more matcha orders,” he said.

Producing the powder is an intensive process: the leaves, called “tencha”, are shaded for several weeks before harvest, to concentrate the taste and nutrients.

They are then carefully deveined by hand, dried and finely ground in a machine...

“It takes years of training” to make matcha properly, Okutomi said. “It’s a long-term endeavour requiring equipment, labour, and investment.”

What has happened in the market for matcha is explained in the diagram below. The market started at equilibrium, where the supply curve S0 met the demand curve D0. The equilibrium price of matcha was P0, and the quantity of matcha traded was Q0. Demand increases to D1. If prices don't adjust and remain at the original price of P0, then the quantity of matcha supplied remains at Q0, but the quantity of matcha demanded increases to QD. There will be a shortage of matcha.

However, when there is a shortage the market will eventually adjust, and the price will increase to the new equilibrium price of P1 (where the supply curve S0 meets the new demand curve D1), while the quantity of matcha traded increases to Q1. But notice that, even if the market adjusts to a new equilibrium, the quantity of matcha increases only slightly. That's because the supply curve is very inelastic (very steep). That's because the matcha producers cannot adjust quickly to the change in price by producing more - it takes a lot of time to add productive capacity in matcha.

Finally, the price of matcha drinks is also going to increase. To see why, consider the market for matcha drinks, shown below. The market was initially in equilibrium, where demand D0 meets supply S0, with a price of P0 and a quantity of matcha drinks traded of Q0. The cost of matcha increases, so tea sellers face higher costs of production. This decreases the supply of matcha drinks to S1. This increases the equilibrium price of matcha drinks to P1, and reduces the quantity of matcha drinks traded to Q1.

Be prepared to pay more for your favourite matcha drinks.

This week in research #84

For the last two days, I've been at the International Population Conference in Brisbane. This conference only happens every four years, so it is like the Olympics of population research. Although I missed the first few days due to teaching, I still managed to see some really interesting research presented. Here are some of the highlights I found from the (last two days of the) conference:

• Isabelle Seguy (extended abstract here) presented a really interesting attempt to reconstruct parish-level population estimates for France for the 1780s, using parish-level data on births, marriages, and deaths that covers about 80 percent of all areas
• Irina Grossman (who was a keynote at last week's New Zealand Population Conference, as I noted last week; extended abstract here) presented estimates of the population living with dementia at the local level for Australia, which involved a combination of population projections with projected changes in dementia prevalence
• James Raymer (full paper here) presented on methods to derive age-sex profiles for net migration (although, funnily, he noted that he wasn't a fan of net migration as a concept, preferring to consider gross migration flows instead)
• James Carey (extended abstract here) presented on methods of measuring structural ageing, built on the 'stationary population model' (this was of particular interest to me, given my previous work in this area)
• Cassio Turra (extended abstract here) also presented on population ageing, focusing on the 'old-age demographic transition' across the world
• Jooyung Lee (extended abstract here) presented estimates of North Korea's declining fertility rate, based on surveys of North Korean refugees, who were asked about the fertility behaviour of their friends and relatives in North Korea, with the resulting estimate of 1.39 births per woman in the 2010s being far lower than estimates by the UN and other agencies

Aside from the conference, here's what caught my eye in research over the past week (a very busy week, so little opportunity to think about other research):

• Kearney and Levine have a new working paper on fertility, which documents rising childlessness at all ages in high-income countries, and finds that short-term changes in income or prices cannot explain the widespread decline, leaving them to conclude that fertility decline is being driven by a broad reordering of adult priorities with parenthood occupying a diminished role

Book review: Empire of Guns

My ECONS101 class discussed the Industrial Revolution in class last week, focusing on why it happened first in England, rather than France or China or elsewhere. By coincidence, I was just finishing up reading the book Empire of Guns by Priya Satia. A conventional view of factors underlying the Industrial Revolution in England, such as the explanation by the economic historian Bob Allen, focuses on the role of changes in wages and coal prices, and individual entrepreneurship and invention. Satia instead focuses on the role of government in driving investment in manufacturing, and in particular the manufacturing of guns. As Satia explains:

The story of Britain's transformation from a predominantly agrarian, handicraft economy to one dominated by industry and machine manufacture - the commonly accepted story of the industrial revolution - is typically anchored in images of cotton factories and steam engines invented by unfettered geniuses. The British state has little to do in this version of the story. For more than two hundred years, that image has powerfully shaped how we think about stimulating sustained economic growth - development - the world over. But it is wrong: state institutions drove Britain's industrial revolution in crucial ways.

The book moves through four sections, motivated by a story of Samuel Galton Jr. (grandfather of the statistician and eugenicist Francis Galton), and a disagreement with his fellow Quakers over his gun-making business (which was inconsistent with the Quakers' ethic of nonviolence). In the first section, Satia tells the story of the British gun trade from 1688 to 1815. The second section then discusses how guns "migrated from being an instrument of terror specifically relevant to contests over property to a weapon for new kinds of impersonal violence on the battlefield and in the streets". The third section takes the story up to the present, looking at how failures to regulate gun manufacture and trade has "distorted the theory and practice of economic development".

The book is exhaustively researched, with 67 pages of endnotes and an extensive bibliography. However, with so much material at hand, it can be difficult to keep the narrative flowing. Satia falls into this trap, with the same ground covered from multiple angles in several sections of the book. It is difficult to see how this could be avoided, but I found it distracting. The text is also quite dense and at times overly descriptive, again as can be expected given the large amount of source material. This is definitely not a lightweight and breezy treatment of its topic.

I did learn some interesting things from the book. For example, 'trade guns' were a common form of 'money', when coins and other mediums of exchange were scarce. However, guns were not a perfect form of money, as:

...guns' market value and metallic content made them a practical monetary instrument, but their perishability undermined both those sources of value, reducing them to mere commodities after all.

Given the focus on the (British) trade in guns, I was also disappointed to see only a single mention of the musket wars in New Zealand. Satia notes that it:

...consumed large quantities of British guns, killing a third to a half the Maori population.

That aspect of the story is left at that. I may be biased, but that sentence alone screams for some further exposition. I wasn't really convinced by the last section of the book either. Like the earlier sections, it was interesting, but in my view, it lacked a strong narrative thread and seemed more like a collection of related anecdotes.

Overall, this book would be a good read for a student of history, particularly those interested in a broader understanding of the context of the Industrial Revolution. But for others with a more general interest in history or economic history, there are better and more readable book-length treatments available.

Is poverty a driver of crime?

This week my ECONS101 class covered, among other things, the difference between causation and correlation. When two variables appear to move together, the relationship might be causal. We might even be able to tell a good story of why the relationship is causal. However, there may be other explanations for the relationship.

For example, take the relationship between poverty and crime. It is well established that poor people commit more crime. Is that because poverty causes people to commit more crime? Many people think so. Being poor means that people lack access to resources, and they may try to obtain those resources through crime. Alternatively, perhaps being poor leads to anger, frustration, or resentment, which leads poor people to commit more crime.

On the other hand, perhaps there is reverse causation - people who commit more crime may make themselves poorer. Being caught and punished through fines or imprisonment will reduce a person's financial resources, making them more likely to be poor. Or, perhaps there is some confounding (or a common cause), and both poverty and crime are related to some third variable. Education is a possibility, since people with more education tend to earn more (and be less at risk of poverty), and also commit less crime. The local unemployment rate might also be a confounder, since when unemployment is higher, poverty will also be higher, and unemployment is also associated with crime.

Putting all of that together, it isn't clear that there is a causal relationship between poverty and crime. We would need some careful research to try and identify whether the relationship is causal. Fortunately, we have this 2023 NBER working paper by David Cesarini (New York University) and co-authors, to provide us with some evidence. Cesarini et al. look at the impact of winning the lottery in Sweden on criminal convictions. They are fortunate in two ways. First, they have data from the register of criminal convictions on all convictions between 1975 and 2017. And second and more importantly, they are able to match the conviction data to four samples of lottery players. Their sample includes over 350,000 lottery wins by over 280,000 individuals. They also look at the effects on children (of their parents winning the lottery), where they have a sample of over 100,000 children.

The cool thing about this analysis is that Cesarini et al. can look at what happens to criminal behaviour, comparing people who are otherwise similar but win the lottery (and therefore are less poor) with those who did not win the lottery (and therefore are just as poor as before). This analysis should establish the causal impact of financial resources on crime (and therefore by extension also the causal impact of income or wealth on crime).

For the adult analysis, Cesarini et al. find:

...a positive but statistically insignificant effect of lottery wealth on criminal behavior. The point estimate of our main outcome of interest — conviction for any type of crime within seven years of the lottery event — suggests 1 million SEK (about $150,000) increases conviction risk by 0.28 percentage points (10.2%). The 95% confidence interval allows us to reject reductions in conviction risk larger than 0.16 percentage points (5.8%). We find no clear evidence of differential effects across types of offenses.

In other words, lack of financial resources does not cause crime in this sample. If it did, then the increase in financial resources arising from the lottery win would lead to less crime. Or, lack of financial resources does cause crime, the effect is very small. Turning to the effect on children, Cesarini et al. find:

...an effect of parental financial resources on child delinquency close to zero, but non-trivial effects in either direction cannot be ruled out. The 95% confidence interval for the effect of 1 million SEK ranges from a 1.36-percentage-point reduction (12.9%) to a 1.54-percentage-point (14.6%) increase in conviction risk.

Again, the central estimate of the effect of financial resources on crime is zero, albeit with less certainty in the result. The takeaway is that a lack of parental financial resources does not cause crime among children. Both results point to a lack of a causal impact of financial resources on crime. Cesarini et al. conclude that:

Our results therefore challenge the view that the relationship between crime and economic status reflects a causal effect of financial resources on adult offending.

It is likely, then, that the observed correlation between poverty and crime arises as a result of reverse causation (in my view possible, but unlikely), or confounding. That has clear implications for policy, because it suggests that focusing on reducing poverty is unlikely to have any impact on crime. Now, this study was conducted in Sweden, and these results might not hold in other contexts. However, they should make us question more strongly the prevailing view that poverty is a driver of crime.

[HT: Marginal Revolution, back in December 2023] 

This week in research #83

For the last two days, I've been at the New Zealand Population Conference in Wellington. There was some really interesting research presented, but the biggest talking point (unsurprisingly) was the recently announced changes to the Census. Here are some of the highlights I found from the conference:

• Irina Grossman presented a fast-paced keynote on small-area population projections, noting that machine learning methods do not systematically outperform simpler methods (and also noting that while many end-users say that they want measures of the uncertainty associated with projections to be reported, very few of them actually use those measures!)
• Rosemary Goodyear and Miranda Devlin presented new data on severe housing deprivation and homelessness in New Zealand which, among other things, showed that homelessness has been increasing in every Census since 2001, and that severe housing deprivation is highest among Pacific Peoples (but being 'without shelter' is highest among Māori)
• Jacques Poot reviewed various methods of modelling internal migration, using data from Australia, and similar to Grossman he concluded that more complex methods don't systematically outperform simpler methods
• Ji-Ping Lin shared details about a free dataset, the Taiwan Indigenous Peoples Open Research Data (which can be found here)
• John Bryant retraumatised the audience by talking about excess mortality during COVID-19 in New Zealand, where excess deaths were low, but interestingly mortality hasn't completely returned to the pre-COVID trend
• Andrew Sporle shared news about a forthcoming data portal for New Zealand that will include 25 years of data on 'amenable mortality' (preventable mortality)
• Marion Burkimsher summarised data on fertility change in New Zealand, noting that the fertility curve for New Zealand in 2024 most resembles the curve for England and Wales
• Several Stats NZ staff tried (with little success) to sell the audience on the administrative and survey data that will replace the five-yearly Census, although Hannes Diener clearly undersold the value of the experimental Administrative Population Census (which one of my PhD students has been working with)

Aside from the conference, here's what caught my eye in research over the past week:

• Bajaj, Jena, and Reilly (open access) use data from a now-defunct online gambling platform that created 'share prices' for soccer players, and find that the skin tone of the player matters with the darker the shade, the lower the player’s online purchase price
• Guthmann and Scheidel (with ungated earlier version here) develop a theoretical model of the economics of Greco-Roman slavery in the ancient world
• Wang, Sarker, and Hosoi (open access) find a positive and statistically significant effect of investment in analytics on NBA team performance
• Nilsson and Biyong (with ungated earlier version here) find that training hairdressers to be mental health first responders improved hairdresser-customer interactions, but had no effect on the mental health of customers, and worsened mental health outcomes for hairdressers

This couldn't backfire, could it?... Spanish slugs edition

My ECONS102 class covered unintended consequences this week. So, this story from YLE in Finland last month seemed very timely:

The population of the invasive Spanish slug (Arion vulgaris) has exploded in Finland, prompting four cities to offer six euros per litre for dead slugs.

Known as a highly destructive garden pest and even nicknamed the "killer slug" following reports of it preying on bird chicks, this species thrives in wet summers, with each individual capable of laying hundreds of eggs.

To combat infestations this summer, the cities of Lappeenranta, Turku, Kerava and Jämsä are encouraging locals to get the Crowdsorsa app, which allows residents to earn money by helping remove invasive species...

Getting a payout requires a few more steps. To earn a reward, slug killers must film a video showing the slugs being packed into one-litre containers (like milk cartons) sealed with tape and disposed of in designated bins.

The final step is uploading a video of the packing and disposal process to the Crowdsorsa app, and if everything is done correctly, the payment gets credited to the user's account.

These sort of bounty programmes have a habit of backfiring, though. The emblematic example of this is a story I wrote about back in 2015:

The government was concerned about the number of snakes running wild (er... slithering wild) in the streets of Delhi. So, they struck on a plan to rid the city of snakes. By paying a bounty for every cobra killed, the ordinary people would kill the cobras and the rampant snakes would be less of a problem. And so it proved. Except, some enterprising locals realised that it was pretty dangerous to catch and kill wild cobras, and a lot safer and more profitable to simply breed their own cobras and kill their more docile ones to claim the bounty. Naturally, the government eventually became aware of this practice, and stopped paying the bounty. The local cobra breeders, now without a reason to keep their cobras, released them. Which made the problem of wild cobras even worse.

So, how long will it be before an enterprising Finn realises that they can make money by breeding these Spanish slugs? Particularly since:

Hunting down the slugs is not always easy, and the pest can be confused with the homegrown Limax cinereoniger, or ash-black slug.

They would be much easier to hunt down if you are farming them yourself! And, when you are farming the right type of slug, there's no risk of confusing them with a local slug. It would be so much easier to claim the bounty that way, than by fossicking around hunting wild slugs. How long will it be before the Finns work this out?

[HT: Marginal Revolution]

Read more:

• This couldn't backfire, could it?... Ragweed edition
• This couldn't backfire, could it?... Possum bounty edition
• This couldn't backfire, could it?... Cane toad edition
• This couldn't backfire, could it?... Dead possums edition

Conflicting signals from attractiveness and education may reduce the chance of getting a job

There is fairly robust evidence of a beauty premium in the labour market (see the links at the end of this post for some examples). More attractive people earn more than less attractive people, ceteris paribus (holding everything else equal). One of the arguments for why this correlation may be causal (that is, attractiveness causes better labour market outcomes) is that attractiveness is a signal that conveys that the person will perform better in a particular job. [*] So, employers might employ more attractive people, thinking that they will be better workers. Another signal of good job performance is education. Employers employ more highly educated people, thinking that they will be better workers. But what happens if these two signal conflict?

That is the question that is addressed in this new article by Christopher Marquis (University of Cambridge), András Tilcsik (University of Toronto), and Ying Zhang (Singularity Academy), published in the American Journal of Sociology (ungated earlier version here). They hypothesise that:

HYPOTHESIS 1.—For higher-status positions, employers will favor applicants with both physically attractive looks and elite educational credentials over applicants who lack either or both of those valued status dimensions.

HYPOTHESIS 2.—For lower-status positions, employers will favor applicants without physically attractive looks and elite educational credentials over applicants who possess either or both of those valued status dimensions.

And for those whose education and attractiveness do not send the same signal, Marquis et al. note that:

...perhaps most interesting is the situation of applicants with discrepant status characteristics—those who are more attractive but from a less elite institution and those who possess elite educational credentials but are less attractive. In contrast to expectations from prior research, our theoretical framework implies that status-inconsistent applicants elicit ambiguous expectations and are thus less likely to be seen as a clear fit for either higher-status or lower-status positions.

Marquis et al. test their hypotheses using two studies: (1) a CV audit study in China, where they sent fictitious applications for various jobs and noted the callback rate for fictitious applicants with different characteristics (physical attractiveness, and education); and (2) a survey experiment in the US, where they asked research participants whether a fictitious applicant should be recommended for an interview for a particular job.

In the China study, they applied for jobs in the "sales" category across six cities on the job website Zhaopin.com. They applied once for each job, randomly varying the picture on the CV (since Chinese CVs typically include a headshot photo), and the education of the applicant (distinguishing between local universities ranked 10-39 on Project 985, and those ranked 40-116). Focusing on callback rates, Marquis et al. find that:

Overall, status-consistent combinations led to more callbacks than status-inconsistent combinations at a rate of approximately 1.7 to 1 (11.7% vs. 6.8%). This difference is significant statistically (P < .001) and substantively.

When looking at jobs of different status levels, Marquis et al. then find that:

...when restricting the sample to jobs with above-average salaries... the high-high combination of university status and attractiveness was more likely to lead to an interview invitation than the low-low combination (P < .10), the high-low combination (P < .05), and the low-high combination (P < .01). These differences were substantively significant as well; for example, the callback rate for the high-high combination in this subsample (24.4%) was about 1.7 times as high as for the low-low combination (14.1%)...

...for jobs with below-average salaries, the low-low combination led to more callbacks than any of the other combinations and that callback rates for the other three combinations were not significantly different from one another. The low-low combination had a substantial advantage in this subsample, generating 1.8 times as many callbacks (22.8%) as the high-high combination (12.9%).

This was consistent with their hypotheses. However, is this peculiar to the Chinese context? To test that, Marquis et al. then conducted the survey experiment in the US. They use a sample of 2020 people 'with hiring experience', recruited through the online panel Prolific. They showed each research participant a profile photo (as would be found on a professional social media site like LinkedIn), and the CV, and distinguished education based on three levels:

Using the 2022 National University Rankings of U.S. News and World Report, we selected universities from California, New York state, North Carolina, and Pennsylvania. From each state, we included (a) one institution ranked among the top 25 national universities, (b) one ranked in the 25–75 range, and (c) one not ranked among the top 75 national universities.

Marquis et al. then:

...asked participants to indicate the likelihood that they would interview the candidate for the position and the likelihood that they would give the candidate an offer, both on 1 (low) to 7 (high) Likert-type scales. We averaged these two items into an overall rating of the candidate for the job in question...

Similar to the Chinese results, Marquis et al. find that:

...for the higher-status position, the mean rating for the combination of elite university status and high physical attractiveness was significantly higher than the average of the five other combinations (5.92 vs. 5.53, P < .0003) and significantly higher than the mean of the second most favored combination (5.92 vs. 5.67, P < .04).

And for the lower-status position:

...the overall ratings for the two combinations of nonelite university status and lower physical attractiveness were significantly higher than the average of the other four combinations (6.04 vs. 5.72, P < .001; 6.01 vs. 5.72, P < .003). The ratings of the two combinations of nonelite university status and lower physical attractiveness were also higher than the mean of the next highest combination (6.04 vs. 5.77, P < .014; 6.01 vs. 5.77, P < .027).

Marquis et al. then use the US sample to look into the mechanism driving the results. Unsurprisingly:

...candidates with status-inconsistent combinations might fall between two stools: they are not perceived to be the most qualified candidates when they target higher-status jobs, and they are seen as lacking organizational commitment when they apply for lower-status jobs. Moreover, for both types of positions, these candidates generate greater evaluative uncertainty than other applicants.

When the two signals conflict, evaluators have less certainty about the applicant and rate them lower. This should be a cautionary tale for job applicants: make sure that all of your signals are conveying the same information about your quality as a job applicant, or you may find your application falling to the bottom of the pile.

*****

[*] I'm not sure that I find this argument very convincing, but Marquis et al. rely on a citation to this 1983 article (ungated version here) from a sociology journal to support it. In economics, for a signal to be effective, it must be costly, and costly in such a way that those with lower-quality attributes wouldn't want to attempt the signal. Being attractive is costly (or, at least, personal grooming is). But is it costly in such a way that people with low job performance wouldn't be able (or willing) to send the signal? I have my doubts. Anyway, I've put this argument aside for the moment, because the more interesting point in this article is what happens when two signals are conflicting.

Read more:

• The beauty premium in undergraduate study is small, and more attractive women major in economics
• New results questioning the beauty premium should be treated with caution
• The beauty premium in the LPGA
• The beauty and height premiums in the labour market
• Blonds have more fun (or rather, they get paid more)
• The beauty premium for economists
• More evidence on the blond wage premium
• The beauty premium at the intersection of race and gender
• Narrowing down on the source of the beauty premium
• The beauty premium and student grades for in-person and online education
• Is there a premium for very unattractive workers?

This week in research #82

Here's what caught my eye in research over the past week:

• Insler, Rahman, and Smith (with ungated earlier version here) use data from the United States Naval Academy, and find that peers influence students into selecting different academic paths than they would have chosen independently, and that social learning, rather than social utility, drives this outcome
• On a similar note, Murat (open access) examines the impact of host-country citizenship on immigrant students' choice of academic majors at an Italian university, and finds that obtaining citizenship reduces the likelihood of choosing math-related disciplines, this effect is more pronounced among women, and these gaps are larger among students from more gender-equal countries
• Charles, Kortt, and Harmes (open access) present a detailed analysis of what economics education is currently taught at Australian universities
• Miragaya-Casillas, Aguayo-Estremera, and Ruiz-Villaverde (open access) conduct a longitudinal study of self-and other-interested behaviour in university students following a standard microeconomics course, and find no evidence of an indoctrination effect from the study of microeconomics among business or law students

Don't expect to see a Danish Grand Prix any time soon

Big events are fun, and draw in large crowds. But by itself, that doesn't mean that big events are worth the cost. Someone, often but not always taxpayers, has to be willing to pay the cost of hosting. A rational decision-maker would only be willing to host the event if the benefits outweigh the costs. This is a point that I'll be teaching in my ECONS102 class next week, so I was interested to read this recent article by Christian Gjersing Nielsen (Danish Institute for Sports Studies), Søren Bøye Olsen (University of Southern Denmark), and Arne Feddersen (University of Copenhagen), published in the Journal of Sports Economics (sorry, I don't see an ungated version online).

Nielsen et al. focus on the case of a return of the Danish Formula One Grand Prix (which was last held in 1962, although a Danish Grand Prix for Formula Three cars was last run in 1995). They focus on this because:

In 2017, a Danish consortium of private investors presented a plan to host a Formula 1 (F1) Grand Prix... in Copenhagen in 2020, 2021, and 2022.

Ultimately, the plan fell through because it required government funding, and while the national government seemed supportive, but only if the Copenhagen City Council contributed financially. The Council ultimately withdrew its support for the event, and the idea never progressed. Nielsen et al. ask whether the Copenhagen public would actually have been willing to fund the costs of the event, which are significant:

...hosting an F1 Grand Prix in Copenhagen would cost approximately €58 million (adjusted to 2023 prices) annually, including salaries (€15.5 million) and temporary stands (€14 million)... plus an additional annual fee of between €14 million and €50 million for hosting to the rights owners, Liberty Media Corporation...

Nielsen et al. undertook a survey of Copenhagen residents, asking a hypothetical question about their willingness to pay (WTP) for a Grand Prix to be hosted in Copenhagen. Specifically:

Respondents were then asked to imagine that Liberty Media had approved Copenhagen hosting F1 in 2026, 2027, and 2028 and that the private and government funding was already in place. To make hosting conditioned on their response, they were also told that Copenhagen would only accept hosting the race if enough taxpayers would support a temporary municipal tax increase at the household level... Following this, respondents were randomly assigned to two groups. The first group was told that the tax amount that they would have to pay if Copenhagen ended up hosting F1 would depend on their household income... Respondents assigned to the second group did not receive this information and were instead asked to state their household income in the latter part of the survey...

This is an application of the contingent valuation method (which is quite a polarising method, with many debates that I have written about, most recently here). Their sample is about 2000 people, once they exclude 'protest' responses, and just 1452 in their preferred 'weak knife-edge' sample - those who were responsive to a difference in the cost of hosting the event. Based on their range of samples, Nielsen et al. find that:

...mean annual WTP (in the 3 years that Copenhagen hosts F1) estimates between €22.95 and €36.94, while the weighted models result in mean WTP estimates between €24.34 and €43.07, with €30.24... as our central estimate due to the theoretical considerations about consequentiality... Extrapolating our mean WTP estimates to the 320,825 households in Copenhagen Municipality, the aggregated annual WTP (in each of the 3 years that Copenhagen hosts F1) is between €7.36 million and €13.82 million, with €9.70 million (n=1,452, weighted) being our central estimate.

This compares unfavourably with the costs. Specifically:

...the public costs—ignoring indirect or intangible costs—would amount to between €14.4 and €21.6 million annually. Based on our central mean estimate of €9.70 million, the benefits for the households in Copenhagen make up between 44.9% and 67.4% of the public costs, which does not justify hosting F1.

The Copenhagen public are not willing to pay enough to cover the costs of hosting a Grand Prix. So, don't expect to see a Danish Grand Prix any time soon.