Monday, 31 August 2020

What income and wealth tax rates do Americans want?

In an interesting new article published in the Journal of Public Economics (appears to be open access, but just in case there is an ungated earlier version here), Raymond Fisman (Boston University), Keith Gladstone (Princeton), Ilyana Kuziemko (Boston University), and Suresh Naidu (Columbia University), investigated how much income and wealth tax Americans prefer.

Specifically, they collected data from several surveys using Amazon Mechanical Turk over the period from 2014 to 2019, plus the Understanding American Survey (UAS) in 2019. In each survey, respondents were presented with scenarios of hypothetical people with different levels of income and net wealth (assets minus debt), and asked how much tax the hypothetical person should pay. Using that data, Fisman et al. are able to extract the income and wealth tax rates preferred by the sample, on average. You can try the survey out for yourself here

They found that:

...their chosen tax bills imply a linear tax rate on income of approximately 13–15%, in line with past work...

When we restrict the relationship of the tax bill and wealth to be linear, the implied average tax rate on wealth is about 1.2% in our baseline estimate.

They also find that the source of wealth matters for their research participants' preferences on taxing wealth:

Preferred taxes on wealth from savings are 0.8%, versus over 3% on wealth from inheritance.

The results are reasonably similar across all of the various survey samples (the mTurk sample is of course less representative than the UAS sample, so it was useful that they checked robustness across different samples), including being fairly stable across samples taken four years apart. The results seem sensible enough, and imply that Americans are not averse to wealth taxes. They also imply a preference for trading off somewhat lower income taxes in exchange for a tax on net wealth.

That last point is the one issue I have with this research. Taxing net wealth seems to me to be a little fraught. Should someone with a $2 million dollar house and a 95% mortgage, and a person with no debt but $100,000 in a retirement savings account, pay the same amount of tax on wealth? That's a normative question that is not easy to answer. However, it seems to me that Fisman et al. have shown a sensible way to better understand taxpayers' preferences in terms of these normative questions. It would also be interesting to know whether, if presented with fiscal estimates based on the implied tax rates they chose, their preferred taxes would change. These are useful questions for future research, especially on how people interpret taxes on net wealth differently or similarly to taxes on gross assets (not accounting for debt).

Thursday, 27 August 2020

Book review: Mastering 'Metrics

Last year, I reviewed Mostly Harmless Econometrics, by Joshua Angrist and Jorn-Steffen Pischke, where I noted that:

...this is a book that is not for the fainthearted undergraduate economics student...

Fortunately, Angrist and Pischke have a more accessible book that I just finished reading, Mastering 'Metrics. This book is everything for undergraduates that Mostly Harmless Econometrics is for graduate students, and more. I found it incredibly accessible and readable, which sets it well apart from pretty much every econometrics textbook on the market. Angrist and Pischke use real-world econometric examples from the research literature to illustrate each topic in a very applied way. The underlying theory is clearly articulated in the text, as well as explored in more (mathematical) detail in the appendix to each chapter.

The topic coverage is a broad sweep across the main tools in the modern econometrics toolkit: randomised trials, regression models, instrumental variables, regression discontinuity designs, and differences-in-differences, which Angrist and Pischke label the 'Furious Five' of econometric research, with a hat tip to Kung Fu Panda. Like their other book, the key goal the authors have in mind is for budding econometricians to be able to extract causal estimates of relationships from a variety of data.

I think this may be the most enjoyable econometrics book I have ever read. Many of you are probably thinking that statement doesn't set a very high bar, because most econometrics books are drier than a particular arid corner of the Sahara, and almost impenetrable to anyone without high-level mathematics skills. However, there is an astronomical gap between this book and its nearest rivals. If I was teaching undergraduate econometrics, I would not hesitate to use this as the required textbook.

And even if you're not an economics student, but you want to understand the quantitative methods that underlie a large proportion of economics research (including many of the papers that I discuss on this blog), this book would be a great place to start. Highly recommended!

Monday, 24 August 2020

The rise of speakeasy gyms during the coronavirus lockdown

In my ECONS102 class, several of the examples I use to illustrate the supply and demand model involve trading of illegal goods and services (like illicit drugs). It is important to realise that prohibition (making certain goods and services illegal to trade) doesn't eliminate the market, it just pushes the market underground. So, I found this recent NPR Planet Money story interesting:

My friend Evelyn is an immigration lawyer, and she recently had a meeting at a foreign consulate in downtown San Francisco. (Her work makes it hard for her to talk to reporters, so we're not using her last name). As she walked toward the building's metal detectors, the security guards told her she couldn't bring her backpack in, so she had to leave. She worried this would make her late, so she frantically began searching for a safe place to stash it. She walked down the street, and her eyes caught a gym storefront with one of those garage-style, roll-down metal doors. It was slightly open...

"Oh, we're not open," said one of the trainers.

What Evelyn uncovered can only be described as a speakeasy gym. You know, illegal, hush hush, like the underground bars during the Prohibition era. These underground gyms appear to be popping up everywhere, from LA to New Jersey.

One fitness freak in Ann Arbor, Michigan, turned to Reddit to get their fix. "Anybody want a home gym partner or know of a speakeasy gym?" they asked — assuring readers in a follow-up post, "not a cop." "That is exactly what a cop would say," responded someone in the thread.

Welcome to the COVID-19 Prohibition era, when gym rats have gone underground.

Governments can legislate all they want, but prohibiting stuff with eager buyers and sellers is super hard, says Jeffrey Miron, an economist at Harvard University who has spent three decades studying prohibitions. Miron, who these days is legally working out in his basement, says there's a simple lesson that emerges from his studies: "Prohibitions don't eliminate things. They drive them underground." And that comes with a whole host of unintended consequences...

This a textbook example of a classic unintended consequence of prohibition, Miron says. When markets get pushed underground, quality control tends to go down. In the case of drugs, this means potentially finding rat poison in your weed. When it comes to gyms in the COVID-19 era, it means potentially creating fitness environments that are even more likely to spread the virus than if they were legal and regulated. "When you drive something underground, your ability to regulate it goes away," Miron says...

Higher prices are another classic unintended consequence of prohibition. With less competition and higher risks in black markets, entrepreneurs can charge extra. The money-making opportunities of black markets lead to other classic side effects of prohibition: violence and corruption. "Disputes tend to be resolved violently because the participants in an underground market can't sue each other in state or federal courts," Miron says. Mobs and gangs function as quasi-governments that use violence to keep order and enforce property rights. But it's hard to imagine illegal gym operators turning to Tommy Guns and gang warfare to resolve their business disputes.

Putting aside the issue of gang warfare between rival illegal gym operators, let's consider the effects of prohibition on the market for gym services. First, let's assume that there are stiff penalties for gym operators who open during a lockdown, but no penalties for gym members who attend the illegal gym. This situation is illustrated in the market diagram below. Without the lockdown, the market operates in equilibrium with a price of P0, and there are Q0 gym memberships. Penalties for gym owners who operate during the lockdown increases the costs of gym operation, shifting the supply curve up from S0 to S1. This increases the equilibrium price of gym services to P1 (the higher prices noted in the quote above), and the number of operative gym memberships decreases to Q1.

Now consider an alternative, where there are penalties for gym owners (as shown above), but also penalties on gym members who flout physical distancing rules by attending the gym. In this case, not only is there a decrease in supply (from S0 to S1), but there is also a decrease in demand (from D0 to D2), because gym members face the risk of being penalised if they are caught. Assuming that the penalties on gym members are smaller than the penalties on gym owners, then the shift in demand would be much smaller than the shift in supply (as shown below). The equilibrium price of gym services increases to P2, and the number of operative gym memberships decreases to Q2. [*]

It would be interesting to see if this point from the article happens:

The longer gym shutdowns last during the COVID-19 prohibition era, the more likely people will evade them. And keep in mind it's summer. Come this fall and winter, millions of workout fiends in cold climates could have fewer legal options to exercise. Speakeasy gyms could have an even greater demand.

That would raise the price of speakeasy gym services even further. Prohibition doesn't eliminate markets - it just pushes them underground.

[HT: Marginal Revolution]


[*] If the decrease in demand was larger than the decrease in supply, then the net effect on the equilibrium price would be a decrease. However, either way, we can be sure that the number of operative gym memberships will decrease.

Saturday, 22 August 2020

How Costco gets firms to compete with themselves

 In my ECONS101 class last week, we covered pricing and non-price strategy for firms. This is a topic that you wouldn't see in a 'typical' first-year economics principles class, because it covers some very applied business economics and managerial economics - pretty important stuff for business students to understand, but not seen by most principles lecturers as important enough to squeeze out all the excitement of teaching the principles of cost curves (!).

Anyway, one of the strategies that I cover in this topic is situations where a firm finds it profitable to compete with itself. The examples I use are firms like Unilever, which makes several different brands of laundry powder, effectively making its own products compete with each other. However, by crowding the market with its own brands, Unilever makes it more difficult for other firms to compete. There is limited space on supermarket shelves, so if you can fill up those shelves with your own products, then it increases your profits. The supermarkets play along, because it is lower cost for them to deal with one supplier that can supply a range of products (and therefore make it look to the customer like a wide variety), rather than dealing with many suppliers.

So, I was really interested to read this article by Adam Keesling last month, which illustrates something fairly similar:

If you’re anything like the nearly 100 million people worldwide who have a Costco membership, you probably love Costco’s Kirkland Signature. You can get two dozen cage-free eggs for $6.50, or a 1.75-liter bottle of French vodka for $19.99. 

But despite these products’ exceptional prices, their quality doesn’t suffer at all. In fact, the exact opposite is true. Many of their products pass purity tests with flying colors. 

Kirkland also has a passionate and loyal fan base — not something you typically find with a private label brand. One guy even got a Kirkland Signature tattoo on his left arm and held his 27th birthday party at the Costco food court. 

Kirkland’s success defies our intuition and experience. Shouldn’t lower prices lead to lower quality products? How can they offer rock-bottom prices but still have some of the best products around? 

The answer is this: they get the best manufacturers in the world — who already have products on Costco shelves — to make Kirkland products. Yeah, you read that right. While customers might not know it, Kirkland products are often made by the same manufacturers who make the branded products that sit next to them on the shelves. 

Now, we don't have Costco in New Zealand. However, the explanation of the economics underlying how Costco gets its suppliers to compete with themselves is eye-opening:

Shortly after learning that brands would actually manufacture and package Kirkland products for Costco, I wondered how it worked. Why would a brand — say KIND Bars or Jiffy Peanut Butter — create a product for Costco and use the Kirkland brand? And make it 1% better than their own product? 

In a normal Costco purchasing example, a company might sell their product for $0.95 and Costco might retail it for $1.00. This would result in a 5% margin (in reality the margin is a bit lower, but let’s use these numbers for estimates). 

This $0.95 would be the brand's revenue. If we use some industry averages for margins — marketing budgets average 24% for CPG brands, and gross margins hover around 40-50% — then we can estimate their cost profile.

Keesling estimates that the manufacturer earns about 16% profit on selling their standard product at Costco. Then:

What about Kirkland products? As mentioned above, Costco aims to save customers 15-20% on the Kirkland products compared to their branded counterparts. So in our example, retail price would be $0.80 per unit for the Kirkland products. If that’s the only adjustment, the brand would be in a tricky position. Marketing expense is $0.20, non-marketing expense is $0.60 and the retail price is $0.80. That wouldn’t leave much margin for either the brand or Costco. In fact, it would leave none at all. 

But here’s the kicker: with the Kirkland products, brands don’t have to spend nearly as much money on marketing. They don’t need to pay for Facebook ads or run television campaigns. The only remaining marketing expense might be the account reps associated with the Kirkland account. Let’s say this lowered the marketing expense from $0.20 per unit to $0.05 per unit.  

In this case, Keesling estimates that the manufacturer earns about 14% profit on Kirkland products. So, while it's not the same level of profit as for their standard product, if selling the Kirkland product is what it takes to get your standard product on Costco shelves, then many manufacturers would be willing to do so. After all, Costco is bringing in revenue of nearly US$150 billion per year, so the manufacturers know they are going to sell a lot of products if they are on Costco shelves.

And that is how Costco gets firms to compete with themselves.

[HT: Marginal Revolution]

Tuesday, 18 August 2020

Should we be worrying about cardboard recycling theft?

I've written a number of posts about rational crime, including onion theft, honey theft, and avocado theft. Here's some of what I wrote in the first of those posts:

Rational decision-makers (including criminals) weigh up the costs and benefits of an action, and will take the action which offers the greatest net benefits. That doesn't mean that every decision-maker is a calculating machine, but at least we can usually say that if the costs or benefits of an action change, then people may make different decisions. In other words, economists recognise that people (including criminals) respond to incentives.

So, when the price of onions increases, we might expect to see more onion thefts. Why? The benefits of onion theft have increased, while the costs (in terms of the risk of punishment) probably haven't much changed. We can describe two mechanisms for why this would increase onion thefts. First, career vegetable burglars (or maybe just the generally criminally-inclined) recognise that there are larger profits to be had by stealing onions for resale. So, they steal more onions (or maybe they start stealing onions). Second, ordinary people now face higher costs of purchasing onions. So, perhaps stealing onions becomes a lower cost alternative for them, so they steal rather than purchase. Either way, increases in onion theft.

In each of those cases (onions, honey, and avocados), an increase in the value of the good raised the benefits of theft, and incentivised more theft. The latest example of this phenomenon follows that same pattern, as described in this article from BBC News:

 Thieves are making a fortune from stealing used cardboard that's been left out to be recycled, and selling it on. This means that legitimate recycling firms, and the city and other local authorities who take a cut from their sales, are missing out on tens of millions...

While figures are not available for how much recycled cardboard is stolen globally, experts say it is very much a worldwide problem. And there are vast amounts of money to be made.

The annual value of the legitimate trade in recycled cardboard and other papers is expected to climb to $5.4bn (£4.1bn) by 2024, up from $4.3bn in 2017. This increase is not surprising when you consider the continuing rise in online shopping, and the fact that most consumer goods are delivered to you in cardboard boxes that are made from recycled fibre (said to be 93% recycled for boxes in Europe)...

Like any commodity, the price of recycled cardboard ebbs and flows according to global demand. Simon Ellin, chief executive of UK trade body The Recycling Association, says the current price is between £70 and £80 per tonne.

"But at the start of coronavirus it spiked to £130. Getting hold of cardboard was a bit 'name your price' at the time, because with everyone stuck at home there was a big rise in online sales," he says.

The higher price of cardboard creates an incentive for the theft of cardboard. However, the reaction of the 'victims' of this theft is different from the other examples I noted above:

But does the problem exist in the UK? Mark Hall from recycling firm says he "wouldn't be surprised".

"The problem is that recycled cardboard is so untraceable," he says. "It is not as if it has a tracker on it.

"And theft is hardly ever reported by companies, because why would they? If magic pixies have nicked their waste cardboard then it means they don't have to pay a firm like mine to come and pick it up. So they are going to keep schtum [quiet]."

This is very much the opinion of the shopkeeper we spoke to in Madrid.

The thieves benefit from the stolen cardboard, since they can on-sell it at a profit. However, it seems that the companies whose cardboard recycling is stolen also benefit (because they don't have to pay for the recycling collection). In that case, does this even count as a crime?

Read more:

Sunday, 16 August 2020

Now cheesemakers want an anti-dumping tariff on cheese

Last month, I posted about New Zealand farmers arguing for an anti-dumping ban on imported frozen potato fries. The farmers were not successful in their arguments. However, that hasn't stopped the next group from trying, this time cheesemakers are looking for new tariffs on cheese, as reported by Rural News:

Simon Berry, managing director of Whitestone Cheese and spokesperson for New Zealand Specialist Cheesemakers Association on EU tariffs and trade, says up to 25% of retail cheeses are imported – mostly subsidised European cheeses.

With imported cheeses often selling for around half the price of local ones New Zealand producers are struggling.

Berry says Kiwi cheese producers can’t compete with cheap European product flooding into the market and wants an anti-dumping duty to be placed on some imported speciality cheeses.

This week, my ECONS102 class is covering international trade, so I thought it might be timely to discuss why import tariffs (such as the 'anti-dumping duty' the cheesemakers are arguing for) are bad in terms of economic welfare. Consider the market below. The demand curve shows the demand from domestic consumers, while the supply curve shows the supply by domestic suppliers. If there was no international trade, the market would operate in equilibrium with a price of PD, and QD cheese would be traded. The consumer surplus (the difference between what the consumers are willing to pay, shown by the demand curve, and the amount they actually pay, PD) is the area AEPD. The producer surplus (the difference between the price that producers receive, PD, and their costs, shown by the supply curve) is the area PDED. Total economic welfare is therefore the area AED (made up of consumer surplus and producer surplus).

Now consider international trade. Say that the world price of cheese is PW, and is lower than the domestic equilibrium price PD. Let's also assume that the world market can supply as much cheese as New Zealand consumers want at the price PW. If the domestic market was opened to international trade, domestic consumers would find it cheaper to buy cheese from the international market and would do so. Domestic consumers would increase their purchases of cheese to Qd1. Domestic producers would have to meet the world price of PW, since no consumer would be willing to pay more than PW for cheese (since they can buy cheese on the world market for PW). Domestic production of cheese would fall to Qs1. The difference between Qs1 and Qd1 is the quantity of cheese imported. In terms of economic welfare, consumer surplus increases to the area AFPW - consumers are made a lot better off. Producer surplus decreases to PWGD - producers are made a lot worse off, so it is no surprise that they would argue against open international trade of cheese. However, total economic welfare is now the area AFGD. The gains to consumers more than make up for the losses to producers, and society is better off in terms of total welfare, by the area EFG - this represents the value of the gains from trade of cheese.

Now consider what happens if the government puts an anti-dumping duty (a tariff) on cheese imports. This is illustrated in the diagram below. Now, if consumers purchase from the world market for the price PW, they have to also pay the tariff. So, effectively the price for consumers increases to PW+T. They will reduce their total purchases of cheese to Qd2. Domestic producers now don't have to compete with the price PW, they have to compete with the price PW+T. So, the domestic production of cheese increases to Qs2. The quantity of imports of cheese falls to the difference between Qs2 and Qd2. In terms of economic welfare, consumer surplus decreases to ABK - consumers are made worse off by the tariff, compared with free international trade. Producer surplus increases to KCD - producers are made better off by the tariff (which is why they would argue in favour of it). Taxpayers receive some benefit from the tariff, in terms of government revenue. This is the quantity of imports (Qd2-Qs2), multiplied by the per-unit value of the tariff (T), which is the area CBJH. That tariff revenue is part of economic welfare, because the government can use it to pay for schools, roads, hospitals, and so on. Total economic welfare now is made up of consumer surplus, producer surplus, and government revenue - it is the area ABCD+CBJH [*]. Compared with free international trade, total welfare is smaller by the area BFJ+CHG - this is the deadweight loss of the tariff.

Import tariffs decrease total welfare. They make domestic producer surplus bigger, but make consumer surplus smaller, and the loss in consumer surplus is larger than the gain in producer surplus. As a whole, society is made worse off.

Having said that, there are arguments that could be made in favour of this particular anti-dumping duty. One argument that we discuss in our ECONS102 class is unfair competition - where international firms have lower costs due to lighter regulation than domestic firms, international firms are placed at an unfair advantage in competing with domestic firms. The cheesemakers here are arguing slightly differently:

Berry says there has been a gradual growth in imported EU cheese over the past three years.

With strong marketing and distribution networks, these high-volume subsidised EU cheeses are undercutting NZ cheesemakers.

If EU cheesemakers have large subsidies, then that means that they can sell at lower prices and still make a profit. That puts them at an advantage relative to New Zealand cheesemakers, who are unsubsidised. The question for a policy maker though, is whether acting to reduce this unfair competition provides benefits that are sufficient to offset the loss of economic welfare from the anti-dumping duty. How much cost should the New Zealand consumer bear in order to support the New Zealand cheesemaking industry?

Read more:

Thursday, 13 August 2020

Book review: A Farewell to Alms

As I noted in yesterday's post, I recently finished reading Gregory Clark's book A Farewell to Alms. In the book, Clark covers thousands of years of economic history, and tries to provide some answer to three interconnected problems of economic history:

Why did the Malthusian Trap persist for so long? Why did the initial escape from the trap in the Industrial Revolution occur on one tiny island, England, in 1800? Why was there the consequent Great Divergence?

I wrote about the answer to the first question in my post yesterday. Clark pins the answer to the second question down to changing societal norms:

So the market nature of settled agrarian societies stimulated intellectual life in two ways. It created a demand for better symbolic systems to handle commerce and production. And it created a supply of people who were adept at using these systems for economic ends. While living standards were not changing, the culture, and perhaps even the genes, of the people subject to these conditions were changing under the selective pressures they exerted. All Malthusian societies, as Darwin recognized, are inherently shaped by survival of the fittest. They reward certain behaviors with reproductive success, and these behaviors become the norm of the society.

So the Industrial Revolution:

...was the product of the gradual progress of settled agrarian societies toward a more rational, economically oriented mindset...

On why the Industrial Revolution happened in England, rather than Japan, China, or somewhere else, he notes that:

China and Japan, with their longer history of settled stable agrarian systems, were independently headed on a trajectory similar to that of northwestern Europe during the period 1600-1800. They were not static societies. However, this process occurred more slowly than in England. Two important factors may help explain this. Population growth was faster in both China and Japan than in England in the period 1300-1750. And the demographic system in both these societies gave less reproductive advantage to the wealthy than in England. Thus we may speculate that England's advantage lay in the rapid cultural, and potentially also genetic, diffusion of the values of the economically successful throughout society in the years 1200-1800.

These points about the source of the Industrial Revolution and why it occurred in England rather than anywhere else challenged my preconceptions, as well as challenging how that material is taught in my ECONS101 class (although I'm not entirely culpable here - I am mostly following the required textbook for the ECONS101 paper). It's good to read something that challenges our priors, and Clark supports his arguments with a wealth and breadth of data and illustrations that I found quite convincing.

On the Great Divergence, Clark concludes that there are three reasons why developed countries diverged from less developed countries in terms of income per person:

The first is that in the preindustrial world, because of the Malthusian Trap, differences in labor effectiveness had no consequences for the average level of output per person across societies... Since the Industrial Revolution income per person has longer been constrained by Malthusian mechanisms. So existing differences in capabilities between societies could now express themselves through income per person rather than population densities...

The second is that modern medicine has substantially reduced the subsistence wage in such areas as tropical Africa, allowing populations to continue growing at incomes which are substantially below the average of the preindustrial world...

The third reason, more tentatively, is that the new production techniques introduced since the Industrial Revolution have raised the wage premium for high-quality labor...

On that last point, Clark includes one of the best explanations I have read of recent Nobel Prize winner Michael Kremer's 'o-ring theory of development'. 

Having read a number of economic history books, the challenge in writing this sort of book is to avoid either trying to create a 'grand narrative' that encompasses everything (perhaps that could be referred to as the 'Jared Diamond approach'), or getting bogged down in minutiae, especially when you have thousands of years of disparate data that you are trying to corral into a coherent story. Clark does a wonderful job of navigating in between those two extremes. And he does it with the occasional flash of humour - consider this, from one of the footnotes on trial by combat:

It is not clear, however, whether armed combat is any worse a way of settling disputes than hiring high-priced attorneys to wield the niceties of legal theory in courtroom battles.

I truly enjoyed this book, and I wish I had not taken a decade or more to get around to reading it. Highly recommended for anyone who wants to understand long run economic development, and in particular the Industrial Revolution, from an economic perspective.

Read more:

Wednesday, 12 August 2020

Gregory Clark on the Malthusian Trap

I've really enjoyed reading Gregory Clark's book A Farewell to Alms. It's been on my must-read-soon list for a long time, but I finally got to it over the last couple of weeks. I'll post a proper review of the book tomorrow, but in the meantime I wanted to focus on Clark's exposition of the Malthusian Trap - an explanation of why income per capita barely changed between ancient times (or even earlier) and the start of the Industrial Revolution. I've been a critic of Malthusian ideas in the past (e.g. see this post), so reading about this model was a good learning experience for me.

The basic model is laid out in the diagrams below (Clark credits this diagrammatic representation to a chapter by Lee and Schofield in this 1981 book). The x-axis on both diagrams is income per person. The top diagram shows the relationships between birth rate (BR) and income per person, and death rate (DR) and income per person. Notice that the birth rate increases with income per person. Higher incomes lead to more births on average. Notice also that the death rate decreases with income per person. Higher incomes lead to fewer deaths. That all seems rather intuitive. When the birth rate is equal to the death rate, the population is in a steady state (neither growing nor shrinking). That happens at an income per person of y*, and a population of N.

The bottom diagram shows the relationship between population and income per person, and this is where things get interesting. If income per person was to increase above y*, then we end up at an income per person of say y0 (with population N0). At that level of income per person, the population starts to increase over time (because in the top diagram, the birth rate is now higher than the death rate). And as population increases, it pulls income per capita down on the bottom diagram, until eventually the economy ends up back at the steady state income per person of y*.

The same thing happens in reverse if income per capita were to fall below y*. In that case, the death rate would exceed the birth rate, and the population would decrease over time. And as the population decreases, it pushes income per capita up until the economy ends up back at the steady state income per person of y*.

You might wonder why income per person and population are negatively related. Clark explains that it relates to diminishing marginal productivity of land:

In the preindustrial era land was the key production factor that was inherently fixed in supply. This limited supply implied that average output per worker would fall as the labor supply increased in any society, as long as the technology of that society remained unchanged. Consequently average material income per person fell with population growth.

In other words, because each additional person employed on the land would be less productive than other workers already working the land (diminishing marginal productivity), average productivity per worker would fall as more workers worked the land, and so average incomes would decrease as well.

And so, because the economy would always end up eventually back at the steady state income per capita of y*, income per capita remained fairly constant for thousands (or even tens of thousands) of years. Clark notes that:

English wages in 1800 on average were about the same as those for ancient Babylon and Assyria, despite the great technological gains of the intervening thousands of years.

However, the model is not just interesting for that fact. You can also use it to show some fairly perverse effects that the Malthusian Trap had on society. Consider what would happen if the general level of health in the population improved (such as through improved sanitation). As shown in the diagram below, that would imply a decrease in the death rate at every level of income - the death rate shifts down from DR0 to DR1. Now at the previous steady state income per person y0* (with population N0), the birth rate is higher than the death rate. The population will start to increase, and as a result income per person falls, eventually reaching the new steady state income per person of y1* and a higher population of N1. The improved sanitation perversely leads to lower income per person - it makes people worse off! As Clark notes:

This Malthusian world thus exhibits a counterintuitive logic. Anything that raised the death rate schedule - war, disorder, disease, poor sanitary practices, or abandoning breast feeding - increased material living standards. Anything that reduced the death rate schedule - advances in medical technology, better personal hygiene, improved public sanitation, public provision for harvest failures, peace and order - reduced material living standards.

On the other hand, consider the impact of an improvement in production technology that raises income per person at any level of population. This shifts the curve in the bottom diagram to the right, as shown in the diagram below. Society would temporarily be at a higher income per person y0 (with population N0), but since at that level of income per person, the birth rate exceeds the death rate, population would increase, with the economy eventually ending up back at the steady state income per person y*, but with a higher population of N1. As Clark notes:

In the preindustrial world, sporadic technological advance produced people, not wealth.

This model is very cool. I wish there was time for me to include it in the first week of my ECONS101 class, as a way of motivating the substantial change in income per person that arose after the Industrial Revolution. Indeed, that is what Clark uses the model for in his book (but more on that tomorrow).

Tuesday, 11 August 2020

The unexplained decrease in young adolescent drinking over time

The New Zealand Herald pointed me to this new article by Jude Ball (University of Otago) and co-authors, published in the International Journal of Drug Policy. The article is ambitiously titled: "What explains the decline in adolescent binge-drinking in New Zealand?" I say it is ambitiously titled because decreases in young adolescent (mid-teens) drinking have been observed in many countries (see this article for a recent review of the literature), but no studies have definitively been able to identify the factors that have led to this decline. Part of that reason is that these studies are observational, and so they can only establish causation, and there are many things that have been changing over time that might affect teenage drinking behaviour.

The Ball et al. study doesn't establish causality either, and in fact doesn't find much of anything in the way of correlations either, that might help to explain the decrease in teenage drinking. They use New Zealand data from the 2001, 2007, and 2012 waves of the National Youth Health and Wellbeing surveys, for adolescents mostly aged 13-15 years. They focus on the outcome variable 'binge drinking': consuming five or more drinks on a single occasion, within the previous four weeks. They first establish the key fact that:

Prevalence of past month binge drinking decreased substantially between 2001 and 2012 across all demographic groups by gender, age, ethnicity and school decile...

That's right. In every single subpopulation group binge drinking declined over that eleven-year span. When you also look at 2007 as well, binge drinking decreased in every group between 2001 and 2007, and decreased further in every group between 2007 and 2011. So, it's a robust trend, and it's the trend that Ball et al. then seek to explain.

They look at several variables from the surveys that might be correlated with the decline in binge drinking, including home factors (like parental monitoring, family attachment, or quality of family relationships), school factors (school attachment, and academic aspirations), leisure factors (time spent hanging out with friends, and having a part-time job), attitudinal factors (attitudes towards cigarette use, alcohol use, and cannabis use), and behavioural factors (current use of tobacco or cannabis, and whether the adolescent was sexually active).

They find that:

Trend analysis showed that, after adjusting for demographic factors, the odds of binge drinking in 2012 compared with 2001 was 0.33... Only four variables substantially attenuated the [odds ratio] when added to the model individually: ‘condones smoking’, ‘condones alcohol use’, ‘current smoking’ and ‘current cannabis use’...

...the fully adjusted model, including all of the predictors in combination, explained approximately half of the decline in binge drinking... none of the predictors, when removed from the model individually, resulted in a statistically significant shift in the OR. The factors that made the biggest (though not statistically significant) independent contribution to the trend were: condones alcohol use, current cannabis use and current tobacco use.

In other words, their main finding is that, after controlling for demographic factors (age, gender, school decile), the remaining variables didn't really explain any of the change in binge drinking over time in this age group. Ball et al. try to dress up their results in some speculation (and are even more speculative in the New Zealand Herald article), but it really is just speculation because no other conclusion is supported by their results. I did find this bit interesting though:

It is interesting to note, though, that our study found a substantial decline in parental alcohol use at home among families with secondary-school aged children, whereas surveys of the general adult population in NZ showed very little decline in prevalence of alcohol use over the same period...

That might be something worth following up on, if data allows. However, the takeaway message from this study is that, when even your correlational analysis isn't able to show anything of interest, your research question remains unanswered. To answer the question, we're going to need more (and better) data.

Monday, 10 August 2020

The projected demographic impact of COVID-19 in Hamilton

Yesterday, I posted about a new working paper on the projected demographic impact of COVID-19 in Australia. They estimated an impact on the Australian population that would see the population smaller by 1.4 million people (about 4.2 percent) by 2040. I mentioned that I had been doing some work for local councils here in the Waikato region, so I want to talk a little bit about that now (ahead of presenting on this work to Hamilton City Council next week).

As I mentioned yesterday, projecting the population using the cohort-component method involves projecting three components: (1) fertility; (2) migration; and (3) mortality. Let's talk a little bit about the impact of COVID-19 on each of those components.

Although the scariest aspect of COVID-19 is undoubtedly its impact on mortality rates, New Zealand has thankfully been spared. And provided we are sensible about managing public health at the borders, it is possible we could stay that way. The impact so far on mortality has therefore been minimal, and so my projections assumed no change in mortality (note: this was the same assumption that the Australian paper mentioned yesterday adopted, even though deaths in Australia have been rising rapidly over the last couple of weeks, in Victoria especially).

In terms of fertility, as I noted yesterday:

Despite media articles suggesting that there could be a 'COVID baby boom', that seems unlikely. In fact, fertility tends to decrease in times of economic recession.

My projections assumed no impact on fertility. Given that we are undoubtedly in a recession, that might tend to overstate future births, and therefore over-estimate the population. However, in the absence of good data on changing fertility intentions, it is difficult to pin down the impact on fertility. The Australian study assumed a small reduction in fertility, which would persist for some time. So far, unemployment rates have not spiked as much as expected in New Zealand (probably thanks to the wage subsidy), so it is possible that there will be no change in fertility. We won't know for another six months or more.

Finally, the current model that I have been using combines net international migration and net internal migration into a single net migration rate for each territorial authority. However, based on past inter-Censal population movements, we know what proportion of migration flows arise from international rather than internal migration (and I am also using that data in updating my projections model to separately account for internal and international migration - more on that in a future post). I assumed that net international migration would fall to zero for one year (from March 2020 to March 2021), before bouncing back to normal. At the time I made the projections for Hamilton City (in early May this year), that seemed like a sensible assumption, based on what we knew at the time. Compared with the Australian study, my assumptions are a mixture of their severe scenario (zero net international migration) and light scenario (bouncing back to normal after one year). Probably now, we might expect lower net international migration to persist for longer than one year, and possibly to take some time to return to normal. My projections might therefore overstate the future population as a result.

What do the results look like? The diagram below shows two projections for Hamilton, based on different assumptions for fertility, mortality, and net migration: (1) a medium (baseline) scenario; and (2) a low scenario, that assumes lower fertility and net migration, and higher mortality. Both scenarios start in 2013, and are calibrated to replicate the estimated Hamilton population in June 2019, before diverging. Can you see any impact of COVID-19? Barely. If you tilt your head just right, you might see that the projection gets slightly flatter from 2019 to 2020. However, the takeaway message is that, even though net international migration is projected to be zero for one year, that impact is swamped by the long-run upward trend in the total population. 

The impact is easier to see if we break down projected annual population change in Hamilton into two components: (1) natural increase (births minus deaths); and (2) net migration. This is shown in the diagram below. Here, you can clearly see that there is a sizeable impact on migration for the 2020 and 2021 years, after which things get back to normal fairly quickly. However, that impact is almost imperceptible when we look at population change in the longer term, as in the previous diagram.

You might doubt that the impact of COVID-19 would be so small. After all, it is the most salient thing to happen to population change in New Zealand since the baby boom. However, consider the diagram below, which tracks the total population of New Zealand from 1900 to 1935. In the middle of that time series is the combined impact of World War I and the Spanish Flu. Both events resulted in large increases in mortality and decreases in net international migration, relative to 'normal'. And, because mortality was concentrated in prime age adults, there was also a negative impact on fertility. And despite all of those compounding negative impacts on the population, there is only a flattening out of the population before it returned to its previous trajectory. Given that COVID-19 is a much smaller event in terms of population impact that the combined World War I-Spanish Flu, it should be no surprise that it is difficult to see any change in the projected total population.

Compared with the Australian research I discussed yesterday, my work is showing a much smaller (not zero, but small) impact on the future population of Hamilton City. I've done similar work for Waipā District, which shows an even smaller impact (because population change is drive more by net international migration in Hamilton than in Waipā). I expect to also see something similar in some forthcoming work for Waikato District. However, as I noted above, I may be overestimating future births by not adjusting fertility downwards in response to the recession, and I may be overestimating net international migration if border closures persist for some time. As in the case of the Australian research, that could be addressed by projecting alternative scenarios.

It is easy to see the world falling apart around us, and jump to doomsday scenarios for population change. However, it would take something much larger and longer-lasting than what we are observing in New Zealand right now, before we would see a substantial change in the trajectory of future population.

Read more:

Sunday, 9 August 2020

The projected demographic impact of COVID-19 in Australia

I've been meaning to post about the demographic impact of COVID-19 for some time, based on some work I've been doing for local councils in New Zealand. However, before I get to my work (perhaps in my next post), I thought I should post about this new working paper by Elin Charles-Edwards (University of Queensland), Tom Wilson (University of Melbourne), Aude Bernard and Pia Wohland (both University of Queensland), that presents some scenarios on the impact of COVID-19 on the Australian population (see also this article in The Conversation about the research).

Projecting the future population in a sensible way requires understanding that there are essentially only three things that happen to the population: (1) people are born; (2) people move from place to place; and (3) people die. So, if you can project fertility, migration (international and internal), and mortality, then you can project the population - this is the basis of the cohort-component method, which involves projecting each component of population change.

The biggest problem with estimating the demographic impact of COVID-19 is uncertainty - we don't know what the impact will be on any of the components of population change. New Zealand has been lucky - it appears the impact on mortality is minimal if anything. Australia has been similar (they have more COVID-19 deaths, but also have a larger population, and deaths in Australia are still not high compared to other countries). Despite media articles suggesting that there could be a 'COVID baby boom', that seems unlikely. In fact, fertility tends to decrease in times of economic recession. However, the biggest uncertainty comes from migration. How long international borders will remain closed is a big unknown, reduced international migration as a result of COVID-19 will have the largest impact on the future population.

Charles-Edwards et al. tackle this uncertainty by presenting three scenarios for the future population of Australia:

  1. A light scenario, which assumes a 28% decrease in net international migration for 2020/21, with a quick return to normal after one year;
  2. A medium scenario, which assumes a nearly two-thirds decrease in net international migration for 2020/21, with a gradual return to normal after four years; and 
  3. A severe scenario, which assumes zero net international migration for 2020/21, and a gradual return to normal only after eight years.
Their scenarios also include some differences in fertility and net internal (inter-State) migration (but no differences in mortality). They also run a business-as-usual No pandemic scenario. They find that:
The reference No pandemic scenario has Australia’s population reaching 27.6 million by 2025, 29.5 million by 2030 and 33.2 million by 2040... Based on the modelled scenarios, COVID-19 is expected to have a measurable and persistent impact on Australia’s population. Under the Severe scenario, Australia’s population will reach 26.6 million by 2025, 29 million by 2030 and 31.8 million by 2040; 1.4 million or four per cent fewer than the No pandemic scenario. The impact is less under the Light and Moderate scenarios, with Australia’s population reaching 0.18 million fewer and 0.50 million fewer by 2040 respectively... The impacts of COVID-19 are felt most strongly in the short term with annual population growth dropping [from] 1.40 per cent in 2020-21 in the no pandemic scenario to 1.14 per cent in the Light scenario, 0.78 per cent in the Moderate scenario and just 0.41 per cent in the Severe scenario. For historical context, Australia’s annual population growth last dropped below 0.78 per cent in 1935, and last dropped below 0.41 per cent in 1916...

Those impacts are quite large (and much larger in relative terms than the impacts I have estimated for some districts in New Zealand, as I will discuss in my next post). At the State level, they find that:

In relative terms, the largest impact out to 2040, based on the severe scenario, will be in Victoria followed by Western Australia, both of which will have a population 5 per cent smaller than in the no pandemic scenario.

However, as I noted above, there is a huge amount of uncertainty here. The international migration scenarios in the Charles-Edwards et al. research are based on a survey of just six demographic experts. It's hard to see how you could do much better than this though. As they point out in the conclusion to their working paper, identifying leading indicators of demographic change (or even contemporary indicators of change) is incredibly difficult. Yet, that is what you would need in order to have demographic projections that adjust dynamically to real-time events. However, as I will demonstrate in my next post, it would be easy to overstate the impact of these events, especially when you look at them alongside the impacts of large events in the past. I think that is a real risk with this Australian research.

Wednesday, 5 August 2020

Bye-bye Rio Tinto, don't let the door hit you on the way out

Finally New Zealand appears to be about to discard the parasitic Rio Tinto from our shores, as the New Zealand Herald reported last month:
The Government appears to accept that Rio Tinto's announcement that it plans to wind-down and close the Tiwai Point aluminium smelter is final.
On Thursday Rio Tinto said it planned to close the smelter in August 2021. More than 1000 people are directly employed at the smelter with another 1600 jobs indirectly affected, the company claims.
In a statement more than two hours after Rio Tinto made the surprise decision, the Government gave no signal that it is trying to convince the mining giant to change its mind, saying there was "a degree of inevitability" about the move...
Energy Minister Megan Woods said the smelter was receiving large subsidies under the emissions trading scheme while transmission pricing plans released recently would also have lowered the smelter's transmission costs.
"This is a blow for the people of Southland and I feel for them, but we need to look to the future," Robertson said.
Woods said there was "a clear understanding" that direct subsidies were not on the table. The smelter wanted a "prudent discount" on transmission pricing. However a formal application had not been put in.
Regular readers of this blog may recall that the prospect of continuing subsidies for Rio Tinto got me pretty angry last year. My ECONS102 class covered subsidies this week, so it is worthwhile recapping why a subsidy is not good, in terms of economic welfare, and why it is even worse in the case of Rio Tinto.

To keep things simple, let's assume that the government subsidy is a direct subsidy on aluminium production (rather than a subsidy on purchased electricity, or a subsidy in the form of below-cost carbon credits under the Emissions Trading Scheme), and that it is a constant value per unit of production. Let's also assume no international trade (although I'll revisit this later in the post). We'll also assume no negative externalities of this production (which, if we included them, would simply make the subsidy even less defendable in terms of economic welfare).

The market for aluminium is shown in the diagram below. Without the subsidy, the market would operate in equilibrium, where supply meets demand, with an equilibrium price of PA, and an equilibrium quantity of aluminium traded of QA. The subsidy is paid to Rio Tinto (a seller in this market), and that acts sort of like lowering their costs of production (in fact, it is exactly like lowering their costs of production if the subsidy is in the form of lower energy costs). This is demonstrated in the diagram by the new curve S-subsidy. The lower costs mean that the sellers can sell at a lower price to consumers PG, and receive a higher effective price PF, once we factor in the value of the subsidy. The quantity of aluminium produced increases to QS, and so does the quantity of aluminium demanded.

Now let's consider what happens in terms of economic welfare. Consumer surplus is the difference between what consumers are willing to pay for the service (shown by the demand curve) and the price they actually pay. In the diagram above, without the subsidy the consumer surplus is the area FGPA. With the subsidy, the consumer surplus increases to the area FJPG. Aluminium consumers are better off as a result of the subsidy.

Producer surplus is the difference between the price that the producers receive and the producers' costs (shown by the supply curve). In the diagram above, without the subsidy the producer surplus is the area PAGH. With the subsidy, the producer surplus increases to the area PFKH. Aluminium producers (i.e. Rio Tinto) are better off as a result of the subsidy.

All sounds great so far, right? Not so fast. The taxpayer contributes into this market the value of the subsidy, which is the area PFKJPG. That area is negative welfare, because it comes with an opportunity cost - a dollar paid as a subsidy to Rio Tinto cannot be spent on health, education, public transport, or anything else that might increase society's wellbeing.

The overall effect on total welfare is that it decreases. The market operating at equilibrium has total welfare (consumer surplus plus producer surplus) equal to the area FGH, but with the subsidy, total welfare (now consumer surplus plus producer surplus, minus the subsidy) decreases to the area FGH-GKJ. There is a deadweight loss (lost total welfare) equal to the area GKJ. That is the welfare cost of this subsidy.

However, it only gets worse from there. Because the majority of aluminium is exported, the gains in consumer surplus don't actually accrue to New Zealanders. It's not clear whether we should be considering those as gains for New Zealand. Similarly, the gains in producer surplus (essentially profit) go to Rio Tinto, and almost certainly go offshore. So, it wouldn't be outlandish for us to consider the entire subsidy area as lost welfare to New Zealand. [*]

Now, of course removing the subsidy and the resulting closure of the Tiwai Point smelter will lead to some one thousand people losing their jobs. However, surely the government could spend some of the tens of millions of dollars saved from subsidising Rio Tinto on retraining those workers and/or redeploying them in other sectors. Overall, the net effect for New Zealand as a whole would be positive.


[*] In this case, then most of the consumer surplus and producer surplus in this market at equilibrium also accrues to foreigners as well.
Read more: