Evidence that tradeable pollution permits can work

This week, my ECONS102 class covered externalities. As part of the topic, we spent a bit of time considering the economics of pollution control, where the government has three main options: (1) regulation (command-and-control); (2) Pigovian taxes; or (3) tradeable pollution permits. I've posted before about the latter two options (see here and here).

Tradeable pollution permits often attract attention from my students, as it seems surprising that granting polluters 'the right to pollute' might be an effective way of reducing pollution. However, economic theory suggests that this can be both an effective and a cost-effective way of reducing pollution. And the research reported in this 2025 article by Michael Greenstone (University of Chicago) and co-authors, published in the Quarterly Journal of Economics (ungated earlier version here), provides some compelling evidence that tradeable pollution permits are effective.

Greenstone et al. collaborated with the Gujarat Pollution Control Board (GPCB) in India to design and experimentally evaluate a particulate-matter emissions market, the first market of its type anywhere. They describe the experiment as follows:

GPCB launched the market for industrial plants in and around Surat, Gujarat, a rapidly growing city of 7 million people, in 2019. Under the command-and-control status quo, plants are mandated to install abatement equipment and are sporadically inspected in person by government regulators and auditors to check that they meet limits on the concentration of pollution emissions... For the present experiment, GPCB mandated a sample of 318 large, coal-burning plants to install continuous emissions monitoring systems (CEMS) to measure the total mass of particulate matter (PM) emitted, compared with the measurement under the status quo of pollution concentrations during spot visits... The emissions market experiment then randomly assigned 162 out of 318 plants to the market and 156 control plants stayed under the command- and-control regime.

The tradeable pollution permits market worked much as we describe in class:

GPCB set a cap on the total mass of particulates that could be collectively emitted by all treatment plants over a compliance period. They allocated permits to treatment plants, with permits summing to 80% of the cap distributed for free, in proportion to plant emissions potential, and 20% sold off in weekly auctions. Thereafter, treatment plants could trade permits with each other. At the conclusion of each compliance period, any treatment plant that did not hold enough permits to cover their emissions was subject to fines based on the size of the shortfall.

As I note in my ECONS102 class, the advantage of allowing pollution permits to be traded is that plants with relatively low abatement costs (low costs of reducing pollution) have an incentive to sell their surplus permits and abate pollution instead, while plants with relatively high abatement costs have an incentive to buy permits rather than undertake costly abatement. Transferability is one of the important features of efficient property rights, and having permits that are tradeable helps to ensure that.

Greenstone et al. look at compliance (did they have enough permits to cover their emissions), and then compare treatment and control plants in order to estimate the effect of the permit scheme on particulate emissions and variable abatement costs. They find that:

Treatment plants complied—held enough permits to cover their emissions—in 99% of plant-periods. By contrast, the compliance rate with concentration standards at baseline was 66%...

Second, the treatment reduced particulate emissions by 20%–30%, relative to control-plant emissions in the command- and-control regime...

Our third main finding is that the market reduced variable abatement costs by 11% at a constant level of emissions.

Taken altogether, those results provide strong field experimental evidence that the permits scheme worked as intended, that treatment plants were compliant, and that emissions decreased while also decreasing abatement costs. Greenstone et al. then conduct a cost-benefit analysis of an expansion of the market. Based on a range of assumptions on the mortality effects of particulate pollution, they estimate that the benefits of the market are at least 25 times larger than the costs. Needless to say, that suggests that tradeable pollution permits have been strongly worthwhile in this setting, and that they would be worth exploring in other settings as well. And in a postscript in the conclusion to the paper, Greenstone et al. note that the GPCB has launched another particulate market in the largest city in the state, Ahmedabad. So clearly, the GPCB was sufficiently convinced that they decided to extend the approach elsewhere.

[HT: Marginal Revolution, back in 2024]

Read more:

• The inefficiency of New Zealand's emissions trading scheme
• Pigovian taxes vs. tradeable pollution permits when clean technology becomes available
• The Greens vs. Labour on carbon emissions, taxes and permits

Taking window shopping to the next level online

When a consumer consumes a good or service, they receive utility (satisfaction, or happiness) from that consumption. That is part of the standard theory underlying consumer behaviour in neoclassical economics. Behavioural economists like the Nobel Prize winner Richard Thaler, in contrast, may distinguish between two types of utility that a consumer receives when they buy a good or service. First, there is acquisition utility, which is the net benefit from acquiring the good or service relative to what the consumer gives up to get it. This is essentially the neoclassical utility from the good or service, after allowing for the cost of purchasing it. Second, there is transaction utility, which is the utility received from how good or bad the deal itself feels to the consumer, compared to what they expected to pay for the good or service.

Transaction utility arises when consumers feel like they are 'getting a good deal'. The better the deal, the greater the transaction utility. One problem is that transaction utility is largely temporary. While acquisition utility essentially lasts as long as the good or service that is purchased, transaction utility may last only as long as the transaction. For that reason, transaction utility may explain the phenomenon of 'buyer's remorse'. The purchase seems like a great idea at the time of purchase, but later, once the transaction utility has dissipated, the purchase doesn't seem like such a great idea at all.

Ordinarily, acquisition utility and transaction utility come as a bundle. When you buy a good or service, you get both. That is, until now. An article in Rest of World reported last month:

This week, I placed orders for a $44,860 Patek Philippe hand-engraved watch, a $12,500 Hermès handbag, a $9,800 Tiffany diamond ring, and a $7,350 Cartier Love bracelet in yellow gold.

I don’t need any of them. I certainly can’t afford them. Thankfully, they’ll never arrive.

Instead of Amazon, I spent the past week browsing a new breed of websites known as “dopamine sites,” a trend that emerged in South Korea. These websites — like Dopamine Shop and FoodNeverComes — recreate the entire ritual of online shopping: You search for products, compare reviews, add items to your cart, enter a shipping address, place an order, and even track your delivery. 

Then … nothing happens. No money changes hands. No package arrives...

Purchasing and shopping are not necessarily the same act. The former is about acquiring and the latter is more of a ritual. The pleasure from dopamine websites comes from the ritual.

In some ways, this is just a high-tech version of window shopping. People have always been able to get some enjoyment from browsing goods that they have no intention of buying. But these websites go a step further by recreating the transaction itself, right through to placing the order and tracking its delivery, while removing the actual receipt of the good or service that was 'purchased'.

The article argues that the pleasure comes from ritual. However, at least some of the pleasure may be transaction utility. There is no acquisition utility here. The consumer knows from the outset that they will never receive the watch or handbag. But the 'consumer' still receives some value from the 'transaction'. These websites may come about as close as possible to offering transaction utility on its own. And since consumers are never charged for their purchase, this is a low-cost way for consumers to receive that transaction utility. All it costs is the opportunity cost of their time spent browsing the website.

It is interesting to consider what the business model of Dopamine Shop or FoodNeverComes might be. The Rest of World article is silent on this point. One possibility is advertising or affiliate links to real products and retailers. After all, retailers might value the attention of people who are shopping without buying, in the hope that some of that activity eventually spills over into 'real' purchasing behaviour.

Whatever the business model, these websites provide an interesting example that demonstrates just how much utility consumers may get from the process of buying, even if they don't actually buy anything.

[HT: Marginal Revolution]

Opportunity cost makes the news

Opportunity cost is one of the most underappreciated concepts in economics, and yet it is fundamental to good decision-making. Whenever we choose to use our resources for one thing, we give up what we could have done with them instead. The opportunity cost of something is the cost of foregoing the opportunity of using the resources for something else. More specifically, the opportunity cost is measured as the value of the next best alternative that is foregone.

Despite its importance, it is surprisingly rare to see opportunity cost mentioned in the media, even in business or economics stories. So, it was a delight to see this recent article in the New Zealand Herald:

Financial adviser Niran Iswar says almost every rental property he’s ever owned has lost money – and he thinks more people are coming around to the idea that it’s not always a surefire way to make money.

Iswar, who is head of accounting, wealth and advisory at Float, said once he counted the rates, insurance, maintenance and the opportunity cost of money tied up in rental properties, every rental he had held had gone backwards, except one that worked because it was bought at the right time “which is luck dressed up as skill”.

When making a decision about how to invest their savings, an investor has many alternatives to choose from. Each alternative comes with an opportunity cost - the return they could have earned from the best of the other alternative investments. The economic cost of an alternative includes all of the explicit costs, as well as the opportunity cost. In the case of rental properties, Iswar notes the rates, insurance, and maintenance, as well as the opportunity cost of the savings tied up in the investment.

Iswar is essentially saying that, once you take the opportunity cost into account, the cost of investing in rental properties (including the opportunity cost) exceeds the benefits. That is what he means when he says that the rental properties "have gone backwards". The savings would have been better off invested in some other alternative. As an example, an investor might be attracted to a rental property investment that offers an annual net return of six percent, but fail to consider that an alternative investment of comparable risk offers eight percent. The opportunity cost of the rental property investment is eight percent return foregone from the other investment. So although the rental property earns a positive net return of six percent, relative to the next-best alternative it actually generates an economic loss of two percent. By investing in the rental property, the investor would give up an eight-percent return in order to earn six percent.

Now, there is one important caution to note. In the context of financial investments, a straight comparison of returns ignores the role of risk. Different investments come with different risks, and different investors will have different appetites for risk. So, where investments differ substantially in risk, it is their expected returns adjusted for risk that should be compared. Only where the alternatives have broadly similar risks is a more straightforward comparison of returns appropriate.

Finally, while opportunity cost is not often mentioned explicitly, I imagine that many investors are implicitly taking it into account. Anyone who weighs up alternative uses of their savings and chooses the alternative that offers the best risk-adjusted return is already thinking in terms of opportunity cost. But making the opportunity cost explicit is useful, because it reminds us that simply earning a positive net return doesn't necessarily mean that an investment is a good one. We need to consider what else could have been done with the savings instead. That is why it was refreshing to see opportunity cost brought to the fore in the New Zealand Herald story.

Supermarket structural separation stupidity

This week, the National Party proposed that, if re-elected, it would pursue structural separation of the supermarkets (as reported in the NBR, One News, and the New Zealand Herald). This was interesting timing. My ECONS102 class has just covered firms with market power and monopolies, and one of the regulatory options available to government is structural separation. If this news had broken a week earlier, this would have made a great example for an assessment question. Instead, I'll have to console myself with a blog post.

In short, the proposed structural separation is nonsensical. Foodstuffs is currently comprised of two cooperatives - Foodstuffs North Island, and Foodstuffs South Island. The two cooperatives run three brands of supermarkets: Pak'n'Save, New World, and Four Square. The National Party is proposing to separate the business into two, with Pak'n'Save as one business and New World and Four Square as another.

There are several elements of the proposal that are plain dumb. The first problem with the proposal is that it would require a substantial reorganisation of the two existing geographically-based cooperatives into two nationwide businesses. This proposed reorganisation is somewhat ironic, given that the Commerce Commission declined a Foodstuffs merger into a nationwide business in 2024. Both of the two new supermarket chains would then have to create their own warehousing, logistics, and other backend systems, separate from the other new chain. And the two pre-existing warehousing and logistics (and other systems), which are geographically separated at present, would need to be split up in order to do this. In its own house, the government is busy trying to reduce back-office functions to increase efficiency. And yet, it seems quite happy to generate inefficiency in a private sector business?

Moreover, the cost-benefit analysis that Sense Partners completed for the Ministry of Business, Innovation and Employment demonstrates how sensitive the case is to assumptions about supply-chain costs. The preferred analysis assumes supply-chain costs rise by one percent, generating an estimated net benefit of $2.9 billion over 20 years. However, if supply-chain costs rise by two percent, the estimated benefit falls to around $920 million, and if supply-chain costs rise by more than two percent, the costs outweigh the benefits. That doesn't leave much margin for the proposal to go wrong before society is worse off overall.

The second problem is that this proposal leaves the other major player in the supermarket sector, Woolworths, structurally unchanged. While the proposal would reorganise Foodstuffs into two separate nationwide supermarket groups, Woolworths would be free to continue its current operations, which also includes three brands: Woolworths, FreshChoice, and SuperValue. Admittedly, the two businesses are not structured in exactly the same way. Woolworths operates its Woolworths supermarkets directly, while FreshChoice and SuperValue are locally owned stores franchised through a Woolworths subsidiary. So, there may be good reasons why the same form of structural separation would not be appropriate for Woolworths. But the obvious question then is, if structural separation is the answer to weak competition in the supermarket sector, why is restructuring only one of the two big market players the right approach?

The third problem is that this proposal ignores another, arguably more obvious, form of structural separation that could be enacted, which could be applied symmetrically to both of the large supermarket players. This is to structurally separate wholesale and distribution from supermarket retail. Currently, both Foodstuffs cooperatives and Woolworths benefit from substantial economies of scale in purchasing, wholesale, and distribution. Those economies of scale may themselves create a barrier to entry, because a new supermarket chain that cannot obtain groceries on similarly competitive terms faces higher costs and will struggle to compete effectively.

Structurally separating wholesale and distribution from retail could potentially address that problem by allowing existing and new retailers to access the scale economies of the established distribution networks on fair and non-discriminatory terms, without having to build their own nationwide wholesale and distribution systems. This isn't a new idea (for example, see here and here). Wholesale ownership separation was explicitly considered in a government-commissioned study in 2022, although that analysis raised significant concerns about whether an independent wholesaler would remain viable after structural separation. Despite those concerns, this remains an alternative worth considering, and the Labour Party announced a policy along these lines earlier today. Its proposal would require Foodstuffs and Woolworths to run their wholesale businesses independently from their retail businesses, although it would not require the wholesale businesses to be separately owned.

As I note in my ECONS102 class, structural separation can be an effective way of regulating a monopoly. Typically, it is used with natural monopolies, where increasing competition across the entirety of a sector is not straightforward. For example, in New Zealand we structurally separated the telecommunications backbone (now run by Chorus) from the retail. We also separated electricity lines businesses from generation and retail, although generation and retail were allowed to remain integrated in the so-called 'gentailers'.

Now, supermarket retail itself isn't obviously a natural monopoly, but wholesale and distribution may contain some of the characteristics that we associate with natural monopolies, including high fixed costs and large economies of scale. Regardless, it is clear that to date government's various measures have been largely ineffective in increasing competition in the supermarket sector. So, structural separation may be more effective than what has been tried so far. However, what is being proposed by the National Party is only one possible form of structural separation. The more important question is where in the supply chain separation would do the most to reduce barriers to entry while preserving the economies of scale that keep costs down.

This week in research #144

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

• Tompsett (open access) finds that the opening of new bridges over major rivers increase per capita economic activity

Also new from the Waikato working papers series:

• Luengo et al. investigate the causal impact of AI advice on price discovery in a controlled asset-market experiment, and find that access to AI advice significantly reduces mispricing relative to the baseline, regardless of whether the AI advice was aligned with long-term fundamentals, or myopic and based on current prices

Try this! Pull incentive sizing tool

My ECONS102 class covered intellectual property rights this week. As part of that topic, we discuss how governments might incentivise the development of new intellectual property without necessarily relying on strong intellectual property rights. One of the potential alternatives is pull funding - where the government funds successful research and development efforts once they have been developed.

In class, I focus on particular examples of advance market commitments and rewards or prizes. Advance market commitments involve governments making a commitment to buying the goods and services from intellectual property, after the IP is developed. Rewards or prizes involve the government offering to pay the reward or prize to whoever successfully develops some particular intellectual property.

That discussion leaves an important question unanswered. Pull funding creates an incentive for firms (or individuals) to develop some particular intellectual property. But how effective is that incentive? Or, turning that question around, how big does pull funding have to be to incentivise firms (or individual) to work on solving a particular problem?

The Market Shaping Accelerator team (a collaboration between researchers at Dartmouth College, the University of Chicago, and the Center for Global Development has provided a tool that gives us some idea of the answer to that question. You can find the tool here (and associated blog post here). Choose whether you want to use a prize, an AMC, or an AMC with milestones. Then decide on the number of stages in the development process (as well as how long each stage is, the cost, the probability of success of each stage, and how much cost can be covered from other sources). Then provide a discount rate for firms, and the firms' target probability of success. And then hit calculate!

The tool reports the number of firms that would be incentivised by your pull funding, the overall chance that at least one firm succeeds, and the size of the prize required. The tool also provides a comparison with push funding - where the government pays for research and development up-front, regardless of whether it will be successful or not. And the tool also allows you to compare different scenarios.

There is far more to this tool than I can do justice to with a short blog post. There are a lot of models running behind the scenes in this tool, and a lot of assumptions. Fortunately, the documentation gives you all the detail.

If you are interested in push funding, and how it can provide incentives for firms to develop new ideas, I recommend that you try the tool out for yourself. Enjoy!

The Commerce Commission goes rafting

Back in June, the Commerce Commission ruled on the case of a proposed merger between three commercial rafting operators in Rotorua. As the New Zealand Herald reported (in August):

The Commerce Commission has approved the merger of Rotorua’s three commercial rafting operators, much to the pleasant surprise of those behind the bid.

The ruling, finalised in June, came after months of investigation, during which Rotorua Rafting director Sam Sutton became so resigned to defeat that he and his fellow applicants came close to ditching the proposal.

After two decision extensions amid the commission’s concerns that the merger might lessen competition, Sutton was prepared for bad news.

I briefly covered the economics of antitrust law in my ECONS102 class today. My view still remains that the Commerce Commission's criterion for determining merger applications is deeply flawed. As I noted in this 2021 post, the Commerce Commission exists to enforce the Commerce Act, and section 47 of the Act says prohibits acquiring "assets of a business or shares" if that acquisition would have the effect of "substantially lessening competition".

Now, any merger between competing firms by its very nature removes some direct competition between them. So, the Commerce Commission's decision always hinges on whether a proposed merger meets the threshold of 'substantially' lessening competition. The Commerce Act defines 'substantial' only as "real or of substance", and the case law on this makes it clear that whether competition is substantially lessened is ultimately a matter of degree. So the exercise is inherently imprecise and subjective.

The US Federal Trade Commission (FTC) has a more objective approach, which is to compute the Herfindahl-Hirschman Index (HHI) before and after the proposed merger, and to presume that any merger that increases the HHI by 100 points in an industry that is already classed as 'concentrated' or 'highly concentrated' is substantially lessening competition.[*] The FTC may then choose to contest the proposed merger (in court). That at least provides a much transparent threshold against which mergers can initially be assessed. In my view, a more economically coherent approach would be for antitrust authorities to decide on the basis of economic welfare, as I discussed in this 2021 post.

Interestingly, the Commerce Commission's decision itself illustrates the difficulty of applying the test. Only two of the three Commissioners were satisfied that the merger was unlikely to substantially lessen competition, and all three agreed that it was a "finely balanced decision". Seven months after lodging their application, though, the rafting operators finally have their answer. And now they can get on with their business, safe in the knowledge that they are not 'substantially lessening competition'.

Read more:

• The Commerce Commission, competition and total welfare

*****

[*] The Herfindahl-Hirschman Index can be calculated by squaring the market share of each firm in the market, and adding up the resulting numbers. An industry with an HHI of under 1000 is classed as 'competitive', an industry with an HHI of between 1000 and 1800 is classed as 'concentrated', and an industry with an HHI of over 1800 is classed as 'highly concentrated'. For example, an industry with just four firms, with market shares of 40 percent, 27 percent, 19 percent, and 14 percent, would have an HHI of 40^2+27^2+19^2+14^2 = 2886, and would be classed as 'highly concentrated'.

Manufacturers' response to 'supercycles' in the market for computer memory

Back in May, David Oks wrote a fascinating Substack article about the economics of dynamic random access memory (DRAM) - the memory that is used in computers, smartphones, and gaming consoles, but also importantly in servers and data centres, such as those used to train generative AI models. I'm not going to focus on the effect of generative AI in this post, but instead on this particular part of Oks's article, which caught my attention:

And that combination—capital-intensive manufacturing plus fungibility—is a punishing combination. Because memory is fungible, the industry is intensely cyclical: the entire history of the DRAM industry is a history of boom-and-bust supercycles. First, strong demand from one sector or another—like Windows PC adoption in the 1990s—drives surging prices and a wave of investment from every player; cumulative overinvestment in an undifferentiated good produces oversupply; and then oversupply leads to collapsing prices.

And because production is so expensive, those down-cycles turn out to be existential: the memory industry is marked by constant wreckage. Intel dominated the memory game in the early 1970s but left in the 1980s, opting to focus on processors. Texas Instruments and IBM, also once major players, left in the 1990s. Germany’s Qimonda collapsed in 2009; Japan’s Elpida, once the world’s third-largest DRAM manufacturer, declared bankruptcy in 2012.

In my ECONS101 class, we teach a model of dynamic supply and demand that can be used to explain the boom-and-bust 'supercycles' that Oks describes. Consider the market for memory, and assume that it is perfectly competitive - most importantly, there are no barriers to entry into the market or barriers to exit from the market.[*] The market for memory is shown in the diagram on the left below. The diagram on the right will track changes in memory manufacturers' profits over time. Initially (at Time 0) the market is at equilibrium (where demand D0 meets supply S0) with price P0, and memory manufacturers are making profits π0. Now say there is a permanent increase in demand at Time 1, to D1. This increase in demand may be because of Windows PC adoption, or some other positive demand shock. Prices increase to P1, and memory manufacturers' profits also increase (to π1). There are no barriers to entry (this is a perfectly competitive market), so the higher profits encourage new manufacturers to enter this market (or more realistically, they encourage the existing manufacturers to increase capacity). However, new manufacturing capacity takes time to bring online, and firms make their investment decisions independently. By the time all of that new capacity becomes available, supply may have overshot the level required to simply meet the higher demand. Supply increases to S2 (more producers) at Time 2. Price falls to P2, and memory manufacturers' profits also fall (to π2).

Next, at Time 2 profits are low and some memory manufacturers will choose to exit the market (no barriers to exit because this is a perfectly competitive market), or more realistically it encourages the manufacturers to reduce their capacity. This helps explain the pattern that Oks describes, with Intel, Texas Instruments, IBM, and others exiting the market. Supply will decrease to S3 (fewer producers) at Time 3. Price will increase to P3, and memory manufacturers' profits will increase to π3. So, these 'supercycles' arise as memory manufacturers enter and exit the market in response to an initial increase in demand.

Now, memory manufacturers are not stupid. The manufacturers that remained after previous busts realised that they needed to change their approach in order to avoid these problems. Oks notes that:

And decades of collapse and consolidation left only a few players standing. In the 1990s, there were perhaps 20 meaningful producers of DRAM around the world; today there are three that account for more than 90 percent of global production. South Korea has two, SK Hynix and Samsung; and the United States has one, Micron.

And these memory makers have learned a very particular lesson from the unforgiving history of their industry: always leave demand unmet.

So, rather than rapidly increasing production in response to higher demand, the memory manufacturers have instead become much more cautious about adding capacity, deliberately allowing some demand to remain unmet. That reduces the risk that increasing supply causes a decline in prices and profits. However, the downside is, as Oks notes, that we now have a global shortage of memory as current production levels (and manufacturing capacity) are unable to keep up. That may open opportunities for new manufacturers to enter the market, even if they are less efficient (higher cost) than the incumbents. And therein lies the problem - by deliberately restricting supply tight, the incumbent manufacturers keep prices and profits high and that may eventually set off the next supercycle.

*****

[*] The market for computer memory isn't actually perfectly competitive, as there are likely to be large economies of scale in memory manufacturing. That means that a small firm entering the market would be at a large cost disadvantage to the large incumbent firms, and so small firms would be deterred from entering the market. However, there is another way of looking at this situation. Instead of considering new firms entering and exiting this market, think about the incumbent firms adding and subtracting additional manufacturing capacity. This has the same effect of increasing and decreasing supply, and leads to the same dynamic pattern in prices and profits.

Europe faces a flood of cocaine

The Financial Times reported back in July:

Cocaine production in Latin America has quadrupled in the past decade, with criminal networks in Colombia, Peru and Bolivia exploiting global trade routes to shift vast quantities of the drug. Cocaine is also increasingly shipped from Brazil into Europe via west Africa.

The UN warns that supply could soon exceed demand, increasing traffickers’ incentives to dump even more product on to Europe’s streets...

The result is that more cocaine is available in Europe today than in the 1980s — often seen as the drug’s heyday — according to the UN. Last year alone, its residue in city wastewater rose by more than a fifth, according to the EU’s drugs agency...

Lower prices and frictionless dealing on popular encrypted messaging apps have made cocaine more attainable.

My ECONS101 class covered the model of supply and demand last week, and this seems like a good example. The European market for cocaine is shown in the diagram below. The market was initially in equilibrium, where supply S0 meets demand D0. The equilibrium price of cocaine was P0, and the equilibrium quantity traded was Q0. The supply of cocaine has increased (a shift to the right, or down, of the supply curve for cocaine), to S1, due to increased production in Latin America being shipped to Europe. This lowers the equilibrium price of cocaine to P1, and increases the quantity traded to Q1.

However, the increase in cocaine supply hasn't had the same effect everywhere. The FT article also reports that:

One country where the retail price has risen rather than fallen, according to UN data, is the UK, which one former senior European police officer says may be linked either to stronger demand or to traffickers’ perception that it is riskier to smuggle cocaine into the country than elsewhere.

If cocaine traffickers believe that it is riskier to smuggle cocaine into the UK, that would decrease the supply of cocaine, and increase the price (see the diagram above, but with the change in supply reversed). That might also contribute to the increase in supply to Europe, if shipments that previously would have gone to the UK go to continental Europe instead. As for 'stronger demand', that effect is shown in the diagram below. The market was initially in equilibrium, where supply S0 meets demand D0. The equilibrium price of cocaine was P0, and the equilibrium quantity traded was Q0. If supply remained constant, and demand increased (as noted in the FT article), the demand curve shifts to the right (to D1). This increases the equilibrium price of cocaine to P1, and the quantity of cocaine traded to Q1.

The increase in demand could even lead to a price increase even with supply increasing as well. If the supply increased to S1, then the equilibrium price goes back to P0, with the quantity of cocaine traded increasing to Q1. 

Combining an increase in demand with an increase in supply is certain to lead to an increase in the equilibrium quantity. However, the change in the equilibrium price is ambiguous. Any smaller increase in supply than shown in the diagram would lead to a net increase in the price of cocaine in the UK, as would any decrease in supply. Any larger increase in supply would lead the equilibrium price to decrease.

Europe is, of course, not the only country dealing with an influx of cocaine. It has also been in the news in New Zealand recently as well. The underlying drivers are very similar, and are a recurring issue (see this post from 2017, related to the US cocaine market).

Read more:

• Cocaine makes a comeback in the U.S.

This week in research #143

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

• Magnuson and Ruggles (open access) explore the creation of the first census microdata sample at the US Census Bureau in the early 1960s (I had no idea it was so long ago!)
• Coelli and Borland (open access) find that the decline in the estimated return to a bachelor's degree in Australia over the period from 2001 to 2016 is potentially explained by strong earnings growth in low‐skill occupations, due to increases in real minimum and award wages, as well as possibly the mining boom
• Bonilla-Mejía et al. (with ungated earlier version here) find that rising violence increases intentions, plans, and preparations to emigrate from Central America, particularly to the United States
• Delavande et al. (with ungated version here) find that student attendance at a public UK university is highly dependent on timetable structure, that students compensate for marginal non-attendance at some events with increased attendance at others within the same module, and that timetable features and attendance choices do not have systematic impacts on academic achievement
• Dang (with ungated earlier version here) finds that one additional international student per thousand working-age residents of a US local labour market increases the employment-to-population ratio by 0.19 percentage points and average hourly wages by 0.48 percent
• Frank and Vickery (open access) look at how LGB students differ from their heterosexual counterparts using data from the UK from 2012 to 2020, and find that LGB men and women both have large shifts towards Humanities subjects, where they perform better, and away from law, economics, management, and STEM, and that they have lower early-career salaries than heterosexual students

The political language of economists, and the associated principal-agent problem for government

Governments often employ economists as policy analysts or consultants, to provide advice on policy. That relationship between the government and the economist is a principal-agent relationship (regardless of whether the economist is a government employee or a consultant), as my ECONS102 class covered this week. 

In a principal-agent interaction, one person or group (the agent) is given the power to decide how to use resources that ‘belong’ to someone else (the principal). In the case of economists working for government, the government is the principal and the economist is the agent. The 'resource' that the agent is using is their work time, which has been paid for by the government.

Now, the government wants the economist to use that work time to provide unbiased economic or policy advice. [*] However, the economist has their own goals and motivations. They might use that work time to scroll TikTok, and then use ChatGPT to write the advice. Or, they might use the time to provide biased advice, based on their own political preferences. Either way, this is detrimental to the interests of the government. This is the essence of the principal-agent problem (or the agency problem).

Do economists act that way? Many economists (including, at times, myself) argue that we are non-partisan, and unbiased in our policy advice, even if we hold particular political views. But is that actually the case?

That is the question addressed in this 2024 article by Zubin Jelveh (University of Maryland), Bruce Kogut, and Suresh Naidu (both Columbia University), published in The Economic Journal (ungated earlier version here). They first link over 53,000 economists from the American Economic Association member registry in 1993, 1997, 2002, and 2009 to two measures of political activity: (1) their campaign contributions (between 1979 and 2012) drawn from the Federal Election Commission's website; and (2) their signing of one or more of 35 petitions aligned with the political left or right.

Next, Jelveh et al. obtain the full text of over 62,000 academic articles and over 17,000 NBER Working Papers published by a subset of 2471 of the economists from the larger sample, between 1973 and 2011. They then rank each of the articles and papers in terms of 'ideological valence', based on the phrases that appear in it, and use that to derive several different 'ideology scores' for the writing of each economist.

Jelveh et al. then start to do some analysis of the ideology scores, finding first that:

...the fields of finance, macroeconomics and industrial organisation are more conservative, while labour is considerably more liberal than the average. Other fields, such as history and international trade, show less political valence. We further see that faculty at business schools are more conservative, as are professors affiliated with ‘freshwater’ schools, while ‘saltwater’ schools have a left-wing bent. Professors of European origin also seem to be somewhat more conservative, and there seems to be no association with Latin American origin, full professor rank or top five department ranking.

The 'saltwater schools' are predominantly those on the east coast of the US, like Harvard or MIT, while the 'freshwater schools' are predominantly those in the Midwest, such as Chicago or Minnesota. Those results, and the results by field of economics, will not surprise many people who know those schools or those fields.

The more interesting results involve the next stage of the paper, where Jelveh et al. look at several economics debates, where there is a clear left-right divide in terms of expected effects. For example, conservative economists may be more likely to believe that the minimum wage reduces employment, while liberal economists may be more likely to believe that it doesn't. Jelveh et al. take several meta-analyses on the minimum wage, and similarly political topics, and look at the relationship between the estimated elasticity in each paper in the meta-analysis, and the estimated political ideology of the authors. They find a statistically significant relationship - elasticities reported by more conservative economists tend to be consistent with more conservative policy prescriptions, while elasticities reported by more liberal economists tend to be consistent with more liberal policy prescriptions.

Economists have political leanings (as does everyone else), Jelveh et al. show that those political leanings are correlated with the results that economists report in their research. This isn't to say that economists are engaged in falsifying data to support their political beliefs. Jelveh et al. have no evidence of that. However, their results could arise if economists choose to apply methods or investigate datasets that are more likely to lead to results that are consistent with their beliefs. Or, economists may simply choose not to publish results that are inconsistent with their beliefs. Either way, this research provides some evidence that knowing the political preferences of economists may be important in interpreting the results from their research.

In my ECONS102 class, we discuss various ways that the principal can act to reduce the principal-agent problem. Those options include stricter monitoring of the agent, paying efficiency wages, or performance-based pay (or delayed payment). In this case, the government might ask the economist agent to fully document their research, or subject it to careful peer review. This would constitute stricter monitoring. However, the relationship between the government and the peer reviewer opens up an additional layer of potential principal-agent problems (what are the political preferences of the peer reviewer?). The government might pay a wage to the economist that is much higher than the equilibrium wage, hoping that would motivate them to produce higher-quality and less biased work (because if they didn't, they would lose their job and have to work elsewhere for less). Government might also use performance-based pay, or may hold back payment until after a peer review, or a replication of any analyses. However, for all of those solutions, some form of monitoring is still required in order to determine the quality of the work. There is an additional problem, though. Efficiency wages and performance-based pay work best when the agency problem involves the effort the agent puts into their work. Those solutions may be less effective when the problem arises from sincerely held beliefs about which models, methods, or evidence are most appropriate.

It seems that the principal-agent problem for governments employing economists (and, possibly, other consultants) is challenging to solve. Perhaps the best that governments can do is insist on transparency around economists' assumptions, methods, and evidence, and subject their analyses to replication and peer review. Political preferences may still matter, but good institutions can make it harder for those preferences to determine the advice that governments receive.

[HT: Marginal Revolution, back in 2024]

*****

[*] Or, maybe the government wants the economist to provide economic or policy advice that accords with the preferred policy platform of the government, rather than independent or unbiased advice. In that case, the nature of the principal-agent problem changes. Nevertheless, in both cases, what really matters is whether the economist's goals and motivations are aligned with those of the government.

Will AI make personalised pricing a reality, and is that really a bad thing?

Personalised pricing (or first-degree price discrimination) occurs when the seller sells their good or service to every consumer at a different price, as I described in this 2023 post. If executed perfectly, the seller could extract all of the consumer surplus as profits, by charging a price to every consumer that is exactly equal to the maximum the consumer is willing to pay. Fortunately for consumers, such perfect personalised pricing has remained a theoretical possibility.

But technological tools are increasingly helping firms to learn more detailed information about consumer preferences, and that allows firms to home in on consumers' maximum willingness-to-pay. The latest worry for consumers is AI, as this article in The Conversation by Patrick Dodd and Hanoku Bathula (both University of Auckland) notes:

Digital platforms can observe thousands of individual decisions. A ride-hailing platform can see which jobs a driver accepts, when they work and which incentives bring them online. A retailer can see purchases, abandoned carts and responses to discounts.

There is no strong evidence major companies already know everyone’s precise financial breaking point. But algorithmically mediated pay, personalised worker incentives, discounts and consumer offers are already real.

Notice that Dodd and Bathula also take the logic of personalised pricing for consumers, and apply it to gig-economy workers as well. Platforms such as Uber or Lyft or Doordash can increasingly use what they know about their delivery workers' preferences to determine their minimum willingness-to-accept for each delivery. The target is different (minimising how much they pay to the delivery worker, rather than maximising the price they charge the consumer), but the underlying premise of personalised pricing is the same.

Algorithms have been around for a while, though. Dodd and Bathula do not clearly lay out why they think that recent developments in AI make personalised pricing more of a reality than before. Most of what they say about 'algorithms' applies equally to statistical algorithms that have been around for years (decades, even) as to more recent developments in AI and machine learning (AI/ML). So, let me extend their argument more explicitly.

AI/ML dramatically lowers the cost of estimating individual willingness-to-pay. It can combine huge numbers of relatively weak signals about a particular consumer, learn complex patterns from the behaviour of millions of other similar consumers, experiment continually with prices and discounts, and update its estimate each time that circumstances change. That gives firms far richer models from which to estimate each particular customer's maximum willingness-to-pay (or, for their workers, to estimate their minimum willingness-to-accept). Moreover, while older statistical algorithms allowed firms to segment customers into fairly coarse categories, AI/ML allows firms to make predictions for each individual, and in real time. The better estimates from these newer models therefore allow firms to price much closer to the perfectly price-discriminating ideal. It's still not completely perfect, but it is a further improvement on what they were previously able to achieve.

Dodd and Bathula finish their article by noting the unfairness of personalised pricing. Their argument is essentially that there is asymmetry in the relationship between consumers and firms. Firms using algorithms (and now AI/ML) know increasingly more about what consumers are willing to pay, but consumers know very little about what firms are willing to accept.

However, it is worth unpacking that a bit more. Firms that don't know consumer willingness-to-pay can't raise their prices without limit, as consumers with low willingness-to-pay would stop buying from them. In practice though, personalised pricing will never be perfect. Firms may charge higher prices to consumers that they estimate have high willingness-to-pay, while offering lower prices or discounts to consumers with lower willingness-to-pay. So, relative to offering the same price to everyone, personalised pricing need not make every consumer worse off. The high-willingness-to-pay consumers are likely to be worse off, but some low-willingness-to-pay consumers may actually be better off.

Now consider which types of consumers tend to have high willingness-to-pay, and which types tend to have low willingness-to-pay. For many goods, lower-income consumers are likely, on average, to have lower willingness-to-pay, so personalised pricing could result in some of them being offered lower prices. That won't always be true though. Some lower-income consumers with few alternatives or an urgent need may have high willingness-to-pay despite having a low income. Taken together, this means that the distributional effects of personalised pricing are not necessarily straightforward. However, in some instances preventing firms from price discriminating could be making low-income consumers worse off. With that in mind, is it really fairer that firms are not allowed to offer lower prices to consumers with low willingness-to-pay?

I'm not really trying to defend price discrimination here. I'm not keen on personalised pricing for very selfish reasons - I don't want to pay more, even if I am willing to pay more! And like me, most consumers should probably not be keen on personalised pricing. But before we rail against the evils of firms price discriminating, we need to properly consider its distributional consequences. And that means thinking about which groups may be made better off by price discrimination, not just which groups are made worse off.

Read more:

• Surge pricing is coming to a supermarket near you
• What would happen if perfect personalised pricing was possible?
• Personalised pricing is now a step closer with electronic shelf labels

Can a simple email get students to study more economics?

This year, I started sending the top students in my ECONS101 class a 'student recognition letter', congratulating them on their performance and noting their overall grade and their ranking in the class. The purpose of this letter was not to try and sell these students on studying economics, but to provide them with some well-earned recognition, as well as a valuable signal for future employers that their A+ grade actually meant something (whereas in some papers, it clearly doesn't mean much).

I could have added a sentence or two encouraging those students to study more economics. They certainly have shown an aptitude for it. However, some years ago I did try to encourage the top students to study more economics, but it seemed like those efforts didn't have much impact. I stopped doing this in about 2018. And it seems I might have made a good choice, at least according to this forthcoming article in the journal Economics of Education Review (open access) by Olivia Edwards and Jonathan Meer (both Texas A&M University).

Edwards and Meer study the effect of a simple email from the professor to the top ten percent (or so) of the class in introductory economics at Texas A&M University:

...praising their performance and encouraging them to take more economics courses and considering majoring or minoring in economics if they were not already doing so.

Their analysis is based on 10,600 students who took the course between 2017 and 2023, of which 1802 received the encouragement email. Edwards and Meer apply a regression discontinuity design (RDD) approach to the analysis, essentially comparing students just above the cutoff mark for receiving the encouragement email with those just below the cutoff mark. The outcome variables Edwards and Meer look at are whether the student went on to study intermediate microeconomics, whether they majored or minored in economics, and whether they majored in an 'economics-adjacent field' (by which they mean agricultural economics or business). They also test whether the results differ by gender, under-represented minority (URM) status, and whether the student was first-in-family to study at university.

There did seem to be some effect on students going on to study intermediate microeconomics. For students around the cutoff point for receiving the encouragement email, there was:

...a relatively large and statistically significant discontinuity of 8.8 percentage points (s.e. = 3.3 percentage points), an increase of about 40 percent over the baseline level below the cutoff.

However, there was little evidence of an effect on majoring or minoring in economics, with both effects being statistically insignificant. Turning to whether there were heterogeneous effects based on demographic characteristics, the effect of the email on taking intermediate microeconomics did not differ significantly by gender, but the overall positive effect appears to have been concentrated among URM students and students who were first-in-family to attend university. The encouragement email had no effect on going onto an economics major or minor for any of the demographic groups.

Edwards and Meer describe the effect of the encouragement email as "modest". I would say it was disappointing, but not surprising. If Texas A&M is anything like Waikato, then once a student has chosen a particular major (or minor), it is actually quite difficult to induce them to switch, even if they later find that some other major or minor would be a better fit for them. In part, this is due to institutional barriers, such as the difficulty in navigating the change-of-enrolment process, or the necessity to take pre-requisites that students might have missed, or that the student may have 'spent' papers on a particular major that would be lost if they switch. It could also be that students simply don't like changing their minds (a 'status quo bias').

So, while a simple information intervention may not be enough to change students’ choice of major, it does seem capable of nudging some students into taking more economics. And on the plus side, sending an email doesn't cost the professor much, so even if it only encourages some additional enrolments in intermediate microeconomics, on a cost-benefit basis for the Economics Department, it is probably a net positive. Given these results, perhaps I was too hasty in stopping the explicit encouragement of my top students to study more economics? I may have to re-draft my student recognition letters for this trimester.

[HT: Tim Harford, last year]

Read more:

• Could closing the gender gap in economics be as simple as providing students with information?
• Disappointing results on simple information interventions to close the gender gap in economics

This week in research #142

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

• Sacerdote, Staiger, and Tine (with ungated earlier version here) find that test score–optional policies harm the likelihood of admission for high-achieving applicants from disadvantaged backgrounds, meaning that the availability of test scores on an application can promote rather than hinder social mobility
• Armona et al. (with ungated earlier version here) develop and test a model of what is newsworthy to a media outlet
• Crossin et al. (open access if you set up a free account) find using data from a nationally representative sample that an estimated 71.1 percent of the New Zealand population think politicians should do more to keep people safe from alcohol harm, with majority support across the political spectrum
• Smit (open access) outlines the drawbacks to remote working that may have prevented greater internal migration from cities to the periphery in the Netherlands

Jeanna Smialek on why we should remember Maria Edgeworth

I have read several books now on the history of economic thought and, for the most part, women are conspicuously absent from those books. One obvious exception is Edith Kuiper's A Herstory of Economics (which I reviewed here), and there are a few bits in Steven Medema's The Economics Book (which I reviewed here). The latter brought both Harriet Martineau and Jane Marcet to my attention (and my two class AI tutors, Harriet in ECONS101, and Jane in ECONS102, are named after them).

A surprising omission from both books was brought to my attention by this New York Times article (ungated version here) by Jeanna Smialik: Maria Edgeworth. [*] According to Wikipedia, Maria Edgeworth was a novelist and an aunt of the much more famous 19th Century economist, Francis Ysidro Edgeworth. However, Smialek makes a strong case for the importance of Maria Edgeworth as one of the earliest voices in the developing field of economics, with her economics expressed within her fiction. However, Smialek also spends some efforts to explain why Edgeworth has been largely forgotten as an economist:

As the field professionalized, the second- and third-generation economists distanced themselves from the women and the fiction that had once helped to popularize and explain their ideas. Serious science, after all, could not possibly be for everyone. The economist Alfred Marshall referred to the women who tried to simplify economic doctrine as “parasites” in a footnote to his hugely influential textbook, “Principles of Economics,” first published in 1890 and popular throughout the 1900s...

Her literature was lost partly because of its clunkiness — its economics lessons made it harder to digest as lighter and more naturalistic stories became the style.

Her economics was lost for a different reason. Subsequent influential academics disparaged the ways that the early women of economics had presented the field. They bristled at being simplified, and they dismissed the idea that the women could have contributed something meaningful. 

Smialek also notes that Edgeworth provided editorial comments on Jane Marcet's Conversations on Political Economy, arguably the first economics textbook:

Edgeworth was as skilled an editor as she was a writer. She told Marcet that Adam Smith ought to be abridged, because he “needs it much,” whereas the utilitarian philosopher Jeremy Bentham was “absolutely impossible” to put in fewer words, because what he needed was “to be diluted.”

“The more amusing anecdote & illustration you can mix with your solid information the better,” Edgeworth wrote to Marcet. “It should be your object rather to sow seeds than to exhibit full grown plants — Your work should excite curiosity to go further.”

Clearly, there is more for us to learn about the history of economic thought from Maria Edgeworth's life and her writing. And we are in luck! Smialek has a forthcoming book about Edgeworth due to be released next month, titled The Invisible Hand of Maria Edgeworth. I'm definitely looking forward to that one, and you can expect a review of it here in due course (although, with my backlog of reading, that might not be until sometime next year!).

*****

[*] Smialek notes that Maria Edgeworth was brought to her attention by Robert Heilbroner's The Worldly Philosophers (which I reviewed here). Heilbroner's reference to Edgeworth clearly didn't make as great an impression on me, as I don't remember it!