April 6, 2024: Kuznets Curves
Good evening. I am back to topics that were more typical of my pre-Lent writing, and today we’ll look at Kuznets curves, a foundational idea in environmental economics. An environmental Kuznets curve (EKC) is the hypothesis that some environmental impact, be it air pollution, water pollution, or whatever, tends to rise with a country’s wealth to a point, but when wealth rises beyond that point, impacts level off and decline. This will be mostly about EKC, but we’ll look at a few other Kuznets curve hypotheses. For readers who work with this material, this post may be mostly review, and for everyone else, I hope it is a helpful introduction to an important topic.
The name is that of Simon Kuznets, an economist who developed the theory in 1955 that became known as the Kuznets curve by specifically looking at wealth inequality. Kuznets hypothesized that countries in an early phase of industrialization tend to face rising levels of inequality, but as industrialization proceeds, inequality tends to lessen again.
Kuznets’ original thesis hasn’t stood up very well. Especially since the 21st century, rising concerns about inequality would seem to invalidate the hypothesis. In 1997, Nielsen and Alderson argued what most of us know well: whatever empirical basis the inequality Kuznets curve had in the mid-20th century, the trend has reversed by the late century. Here’s a paper from 2012 which asserts the opposite of Kuznets’ thesis: inequality tends to decline in low-income countries and rise in high-income countries. Acemoglu and Robinson, two highly prolific and well-regarded contemporary economists, on the question of whether development leads to less inequality gave the rather unsatisfying answer of “sometimes” in 2002. In defense of the argument, here’s some more recent empirical support in Spain. Here’s some evidence for late 20th century countries.
This paper by Martínez-Navarro, Amate-Fortes, and Guarnido-Rueda is interesting. It finds that the Kuznets inequality hypothesis is not supported when we use the usual GDP metric to measure a country’s development (logarithm of GDP per capita to be precise), but it is when we use other metrics of development, such as the human development index or the share of agricultural in GDP. Interesting, but I wonder if there is an element of assuming the conclusion here.
As mentioned above, an environmental Kuznets curve (EKC) is the notion that some environmental impact tends to rise, peak, and decline with rising income. This is a well-studied topic in environmental economics, going back to Susanna Hecht in 1985, and we’ll take a fast, whirlwind tour through some results, though it is simply not possible to be comprehensive here.
Before we get into research, let’s prime our intuition. Most citizens of wealthy countries who are my age or above will remember how much worse certain environmental problems used to be. Los Angeles used to have a notorious smog problem; now it is much better. I am old enough to remember when acid raid, rather than angry protestors, was the biggest threat to public statues. After great fear in the 1980s, the ozone layer is on track for recovery. Leaded gasoline for cars used to be the norm; now it is banned everywhere in the world. In the 1970s, catalytic converters became standard, greatly reducing air pollution from cars. Ten years ago, there were fearful articles about the horrid air pollution in Chinese cities; now, sulfur dioxide, PM2.5, and nitrous oxide are greatly lowered from 2013 levels. It hardly needs to be pointed out progress on environmental challenges is far short from where we want it to be, but it is important to consider what we have already achieved.
Now for some nitty gritty. Sulfur (sorry Brits) dioxide reduction is one of the most dramatic environmental success stories of the 20th century, but the presence of an EKC for SO2 is debated. This paper regards the evidence as being weak at best. This paper focuses on China and finds evidence. This paper does not find an EKC with nitrogen dioxide in China. This paper looks at particulate matter in China and finds a maybe, sorta EKC. For air pollution in general, this paper and this paper find an EKC.
This paper finds an EKC with lead. This paper does not find an EKC with various forms of water pollution. This paper finds an EKC with ozone-depleting substances. This paper finds an EKC with construction waste management. This paper finds an EKC with nitrogen and phosphorus fertilizer usage. This article describes an EKC with nitrogen runoff.
On deforestation, this paper does not find evidence for an EKC, while this paper does, emphasizing that better property rights and environmental policies can reduce deforestation at a given income. In Romania (and probably many others places), the opposite of an EKC is found with urban land use. This paper finds an EKC with cropland use.
Now we come to greenhouse gas emissions. Several wealthy countries have reduced per-capita greenhouse gas emissions. From 1990 to 2021, the United Kingdom reduced per-capita GHG emissions by 50% despite an increase of GDP per capita of >40% over the same period. A person might retort that they have simply offshored emissions rather than eliminated them. When we consider consumption-based emissions, which is an attempt to measure a country’s emissions, plus the emissions embodied in what they import, minus emissions embodied in what they export, then the UK reduced per-capita emissions by ~35% from 1990 to 2021, so this decoupling is very real and not just a matter of offshoring. Similar patterns were documented for France, the United States, Finland, Sweden, and Germany.
Nevertheless, an EKC isn’t as easy to find as we might hope. It isn’t observed for some regions in Russia. It is observed, albeit in a limited manner for CO2 and two other greenhouse gases in EU countries. This paper sort of finds evidence for an EKC with GHG emissions, but it finds that emissions tend to decrease and converge at a non-zero level with increasing wealth, rather than decrease to zero. This paper argues against such an EKC.
All right, that’s enough for now. Obviously I can’t mention everything, and I haven’t tried to be fully representative of the literature. What I want to convey is that, while EKC has been a staple of environmental economics for almost 40 years, the topic is still controversial, basic results are disputed, and whether the EKC model is valid depends a lot on context (geographical, which environmental impact is considered, etc.).
Rather than get too deep into the weeds, I want to highlight some theoretical issues with the EKC. From the abstract of John Luke Gallup, whose paper arguing against a Kuznets curve for inequality I cited above,
There has never been good evidence for a pattern of rising inequality in low-income countries and falling inequality in higher income countries. The only evidence that appears to support the Kuznets hypothesis is the cross-sectional pattern of inequality levels across countries, although the Kuznets hypothesis is an assertion about the path of inequality within countries.
This is a form of the “reading history sideways” fallacy, which goes something like this. Countries tend to get wealthier over time. Therefore, poor countries today are what rich countries used to look like, and rich countries today are what poor countries will look like in the future.
The biggest problem I have with the EKC model is that it obscures the causal mechanism behind environmental improvement. In New Things Under the Sun, Matt Clancy has a good review of this topic. The best explanation, which comes from several of the papers he discusses, particularly this one by Acemoglu and others, is by considering how an economy evolves when there is a market force directing which technologies are developed and deployed. As GDP per capita goes up, the marginal value of more wealth decreases, while the marginal value of environmental improvement increases, and so the economic case for investment in cleaner production increases. This matches the intuition that environmental quality is a luxury belief that is held more keenly in a wealthier society.
Many EKC critics make a strawman criticism that the EKC asserts that wealth solves environmental problems automatically. It clearly does not, and I imagine that very few people think that it does. For whatever limitations the EKC model has, I would contend that it fits the data in almost all cases far better than the I=PAT model, which I discussed last year. The I=PAT model decomposes an environmental impact I as the product of population (P), affluence (A), and technology (T), implicitly treating these three terms as independent, when that is clearly wrong.
For what all this means from a practical view, the EKC is one of the most potent empirical tools for those of us who believe that an attempt to solve environmental problems by curbing wealth or population is a wrong-headed approach. The I=PAT approach is favored by those who do advocate such an approach. I have no hesitation in asserting that the former group have by far the better half of the argument.
One more topic related to Kuznets curves. Along with environmental concerns, various forms of existential risk—artificial intelligence run amok, nuclear war, a bioengineered pandemic, and whatever else your dark imagination can conjure—are out there. In September 2020, Leopold Aschenbrenner published a paper, Existential risk and growth, with the Oxford’s Global Priorities Institute, in which he argues that a Kuznets curve may exist with existential risk.
In Aschenbrenner’s model, the present is a unique “time of perils”. If societal wealth continues to grow, then just as there is a economic case for investing more in clean production, there is an economic case for investment more in safety. It is thus wrong-headed to try to halt economic growth and technological development because new technologies may be dangerous. In the long run, safety is maximized by plowing ahead with growth and investing in safety later.
I find Aschenbrenner’s paper to be a bit too theoretical to place great confidence in the results. That problem is unavoidable when dealing with existential risk, as a society with sufficient empirical experience with existential risk will not be producing any econ papers. But it is thought-provoking. And again, the “shut it all down” approach to risk cannot claim any credible evidence for its own.
Stray
Coming out of my Lenten media fast, I played Stray last week. There are several versions; the one I played is for PlayStation 5. I have some very positive things to say about it.
Stray’s unique feature is that the player assumes control of a cat. The game opens with the protagonist and his family scampering across the ruins of a long-abandoned city. While the cats are jumping across a pipe, the protagonist slips and falls into the city. You then have to navigate the city, find out what happened, and find a way out. To prevent spoilers, I’ll leave the story at that.
Stray has one of the best told stories I’ve seen in a long time. It’s fairly short, and it can present a tense action sequence in a nightmare fuel setting, followed a few minutes later by jumping on to a table and knocking pieces from a mahjong game all over the floor. There is so much fun cat stuff to do, such as knocking flower pots off ledges and “playing” a piano by walking on the keys. As a cat lover, I did not come close to tiring of the game’s cat humor, but it has a very serious story to tell as well, and the humor does not detract from that. If you don’t cry during the final scene, then you have a heart of stone.
The graphics are quite good, especially the use of lighting. I was about to rave about the PS5 version’s excellent use of raytracing, except it doesn’t use raytracing; it uses a much more established baked lighting effect to achieve similar effects. It uses Unreal Engine 4, as opposed to UE5, and the main advantage of the PS5 version is 4K resolution, 60 frames per second, and some better models such as showing more detail on the cat’s fur.
My graphical complains are minor. There are some wall clip bugs and a few noticeable physics fails. There is also an overuse of certain assets.
The gameplay is pretty simple, and this is meant to be a casual game. It is very satisfying to run and jump as a cat, jumping up on pipes and air conditioners to get to places that a person could not get to. I can better appreciate how my own cat loves to jump on the cabinets in the kitchen, even though he knows he’s not supposed to do that. The game intersperses action sequences, methodical stealth sequences, puzzles, and exploration well. My only gameplay complaint is that the game is too easy. There were only two or three points near the end where I found it to be genuinely challenging to proceed.
The music is also very good. Even if you aren’t interested in playing the game, it is good music to put on while working.
It is a common enough phenomenon that a Reddit forum is dedicated to Cats Watching Stray. My own cat was absolutely captivated. It took me around 14 hours to complete the game, spread over a week. Natsu watched almost the entire time, and with the same attention that he watches birds and squirrels outside the window. I think he enjoyed the game even more than I did.
While I have completed a playthrough, there is more challenge in finding all the items and trophies and getting a playthrough under two hours. Maybe I’ll try that some day.
Quick Hits
There were a couple of notable anniversaries during Lent that I wanted to comment on but saved until now. One of them is that on February 24, we marked two years since the full-scale war in Ukraine. I am disappointed to see that Congress is still dithering on another aid round. I also don’t understand by the Biden administration needs Congressional approval at all, since they apparently haven’t with aid for Israel. But I’ll refrain from griping about something that I can’t put the effort into understanding right now. It looks like the main thing that has gone on in the last few months has been some marginal Russian advances in the east. Also last week, the Russian Orthodox Church, whose head is an established Putin ally, described the conflict to be a “holy war”. Armchair strategists who think that a negotiated settlement can easily be attained should reflect on the difficulty that Americans have faced over the last few decades in making peace with jihadists who believe that their enemies are the enemies of God.
On March 11, we marked 20 years since the March 11, 2004 bombings of the train system in Madrid. The attack killed 193 people and was the deadliest terrorist attack in Europe since the 1988 Lockerbie bombing. It also occurred days before the Spanish general election and was meant to sway public opinion against the war in Iraq. Indeed, the bombings have been credited with swinging the election toward the socialist party, which proceeded to end Spain’s involvement in the war. Thus this is one of the few deadly terrorist attacks I can think of that explicitly accomplished its goals, and I can’t help but wonder if this “success” emboldened Al Qaeda to carry out a similar attack in London the following year.
Also on terrorism, at attack at a music venue in Russia a couple weeks ago killed 144 people. The perpetrator is ISIS-K (the Khorosan Province branch of the Islamic State, which operates mostly in Afghanistan and Pakistan). Americans will remember this group as the perpetrators of a 2021 airport attack in Kabul, which disrupted America’ badly-planned withdrawal from that conflict. They’ve carried out a few other major attacks in Pakistan and Iran, but I believe that this is their first attack in a country that isn’t or doesn’t border Afghanistan. CNN reports that ISIS-K has the ambitions, but probably not the means for now, to attack the United States and Western Europe, though it should be kept in mind that there was a foiled ISIS-K attack in Germany a few months ago. I’m sorry to say that it won’t surprise me if the U.S. is back in Afghanistan before long, and for the same reason we were there the first time.
Caspian Report has an episode about the possible Sahel Confederacy, an integration of Mali, Niger, and Burkina Faso. All three of the countries have recently had military coups, and the military juntas are more ideologically aligned with Russia’s Wagner Group than the governments they replaced. Because Niger in particular has sidelined more competent American and French counterterrorism efforts, the coups are likely to be good news for Islamist militants operating in West Africa.
To wrap up this long post on a lighter note, Donald Trump and Joe Biden have secured their party nominations, making 2024 a rematch presidential election. Past rematch elections—where the two leading candidates are the same twice in a row—include John Adams and Thomas Jefferson (1796 and 1800), John Quincy Adams and Andrew Jackson (1824 and 1828), Martin Van Buren and William Henry Harrison (1836 and 1840), Benjamin Harrison and Grover Cleveland (1888 and 1892), William McKinley and William Jennings Bryan (1896 and 1900), and Dwight Eisenhower and Adlai Stevenson (1952 and 1956).
Martin Van Buren, Millard Fillmore, Ulysses S. Grant, Grover Cleveland, Theodore Roosevelt, and Herbert Hoover are the former presidents who made serious efforts to get their old jobs back. Of these, Cleveland was the only one who succeeded.