You should freeze your eggs. Contrary to popular myth, egg freezing works very well and if you freeze your eggs in your twenties or early thirties, you have a very good chance of having a child.
European leaders are looking to copy Australia's example and cut migration from boat-bound refugees but they are in danger of learning the wrong lessons. Offshore detention was the most widely publicized aspect of their refugee policy but it didn't work. Turnbacks were much cheaper and more effective.
Ben, Aria and Pieter discuss different articles in the new issue of Works in Progress. They discuss how Britain lost its position as the world leader in nuclear power, why ASML is so successful, how envy killed the first bus, and how cool neo-traditional temples are.
Buy your copy here: https://worksinprogress.co/print/
You can also listen to the episode on Spotify, Apple Podcasts and YouTube.
Transcript
Ben Southwood [00:00:00]: The mainstream media is lying to you. ‘Sobering study shows challenges of egg freezing"‘. That’s the New York Times. ‘The failed promise of egg freezing’. That’s what Fox calls it. ‘Success rates for frozen eggs vary widely but rarely go above 30%’. That’s what the Financial Times says. Glamour Magazine says ‘the odds are stacked heavily against you’. This is all extremely misleading; it’s true that women in their forties who freeze their eggs, which is most women who freeze their eggs, have relatively low success rates, but that’s because they have relatively low fertility. If women freeze their eggs in their twenties, then when they unfreeze those eggs and use them, they have the fertility of their twenties. Eggs age, but uteruses don’t as much.
The oldest woman to ever give birth was 72, but she did it with donor eggs. Where did I learn these facts? Issue 23 of Works in Progress. So I’ve brought my colleagues, Aria Schrecker, Peter Garicano, I’m Ben Southwood, to discuss the latest issue of Works in Progress, issue 23.
Pieter Garicano [00:01:00]: Ben, I don’t even think that’s the most interesting fact in that whole article.
The much more interesting story the article tells you is that you’ve been lied to about fecundity curves. People often think that your probability of giving birth or your fertility, declines very gradually between ages 18 and 35 and then very steeply as women enter middle age. This is the story you hear in school, this is the kind of graph you see, this is the kind of general wisdom that people will accept. The problem is this is totally not true. It’s based on the fact that in the past we couldn’t observe actual fertility, which is like your probability that you’ll give birth if you’re having unprotected sex. We could just observe how many people were having children.
So we have these big data sets of German peasants. We have the datasets of Iranian farmers. It turns out the fertility curve correctly measured, including just non-pregnant women, is totally linear. It declines just as much between 18 and 25 as it does between 25 and 30. And importantly, this means that actually it’s quite a bit higher as you get older than what is commonly assumed when you look at the concave line.
Ben Southwood [00:02:00]: The key thing is, this isn’t looking at someone’s fertility throughout their life, in which case it would catch all of the children they had had previously to the study or after the study because they were pregnant or they’d recently given birth. Instead it’s just looking at if they had a child this year. And one of the main reasons you wouldn’t have a child this year is because you had a child last year, or you’re going to have a child next year.
Pieter Garicano [00:02:24]: If you actually plot monthly birth rates in non-pregnant women, it’s an extremely steep, basically linear curve between 18 and 40. It doesn’t go slowly and fast. It’s every single year that you wait, it becomes harder and harder to become pregnant.
Ben Southwood [00:02:39]: But there is no steep decline after age 35. It’s flat. There’s no particular year where fertility drops off a cliff. It’s just a steady decline throughout a woman’s life.
Aria Schrecker [00:02:49]: So fertility aging is much more like regular aging than it’s like going through puberty.
Pieter Garicano [00:02:56]: Exactly. And of course the key point is that what is aging is not the ovary or the uterus, it’s the eggs themselves.
Women start their lives with 2 to 6 million primordial follicles, of which roughly 400 will be menstruated. But the vast majority of them will die off throughout a woman’s life at a very linear rate. Which means that when you’re doing, for example, egg harvesting to freeze them, it becomes increasingly hard.
Now Aria, you may know something about this.
Aria Schrecker [00:03:21]: I may have some relevant information. I’m currently about to lay about 44 eggs on Saturday.
I’m going through the process of egg freezing currently. Top line review is, it’s not as bad as people say. People say it’s very painful, that you get very bad mood swings, that it’s very rough on your body, that it probably won’t work.
There are lots of things that you generally hear when people are weighing up whether or not to do it. They say it’s expensive, and it is kind of expensive, but not for what you’re buying, which is the insurance of getting to have children in the next 10 years or so, being able to have children maybe when I’m 40, which I might not be able to otherwise.
Ben Southwood [00:04:16]: Ruxandra and Luzia, the authors say that the world capital of fertility for that purpose is Spain. Are you going to Spain?
Aria Schrecker [00:04:39]: I’m not going to Spain. I’m going to a lovely place in London.
Pieter Garicano [00:04:42]: I don’t think they say the world capital is Spain. they say Spain is very cheap and also generally a nice place to be. And so if you’re going to be there anyways, you might as well take a day off.
Ben Southwood [00:04:53]: That’s the starting point for a flood of international egg tourists turning up and getting their eggs frozen.
Aria Schrecker [00:05:04]: I’m not against it. It’s much more convenient to do it in London. The price differential doesn’t seem to be that extreme. My husband would also need to come to the place that we extract them so the costs compound because we’re not just doing egg freezing.
Ben Southwood [00:05:24]: What is a typical price to pay?
Aria Schrecker [00:05:29]: A typical price would be about £5,000 in Britain and then £200 per year for keeping them frozen.
Ben Southwood [00:05:37]: So I remember being quite interested in the piece about how the eggs winnow down from the starting point. Is that something you’ve experienced?
Aria Schrecker [00:05:49]: Before you even start any hormones, they just want to make sure it’s actually going to work if they do it. They’ll give you an ultrasound and they’ll look at your ovaries to make sure they look healthy and they can roughly get a ballpark number of how many available follicles you have that plausibly would develop into mature eggs.
They literally just do a count; you can see it on the screen and they you can see the little black holes that the ultrasound doesn’t go through. I don’t really know what the typical numbers of something like this were. I had high twenties in both ovaries, and then once you start the injections, most of those do actually start to mature.
And so when I say I’m going to lay about 44 eggs, that’s what I’m referring to. There are 44 mature follicles. After that point, you should expect a winnowing at each potential stage, once they take the eggs out of the follicles, only about four fifths of them will end up being mature eggs.
And after that point, only a certain number of those are actually going to be successfully fertilized. After that, only a certain number of the successful fertilizations are going to actually develop and grow. And after that, only a small proportion of those are going to end up not having any chromosomal abnormalities.
And then when it comes to implantation, you lose a lot in implantation as well. You should expect from 44 follicles to result in something like five or six children.
Ben Southwood [00:07:14]: The message of this piece is that egg freezing and embryo freezing are both much more effective than many people have been led to believe.
Aria Schrecker [00:07:49]: When you’re talking to the doctors and nurses, many of them express that they wish all of their patients came to them at a younger age.
They wish they were working with people in their twenties. They’re saying part of the reason the symptoms aren’t that bad is because you’re quite young, we don’t have to pump you with as many hormones. This is actually just much easier.
Obviously to some extent, everyone who works in that industry are trying to sell their goods, but you can tell that they’re actually very enthusiastic and in favor of more people doing this while they’re young.
Pieter Garicano [00:08:19]: They say in the piece, the average age someone currently does this in the United States is 37 or 38. It’s extremely late to be going through a procedure like this one.
Aria Schrecker [00:08:28]: It’s very sad that when tech companies and private insurance schemes offer egg freezing, people often view it quite cynically, as the company trying to basically buy their workers extra time and trying to hope that they’re encouraging them to delay making family formation decisions.
It’s very rare that that would be persuasive to people or that companies are thinking in that sort of homo economicus kind of way. There are lots of people who have this perk available to them as a side thing from their job that they never think about that it would actually be essentially free for people in 20% of young employees.
Also, I’ve decided not to sell my eggs. This also has an interesting reason, because you can’t sell eggs in Britain. That’s fine. I can travel abroad and I can take my eggs with me. But the main market for egg selling is America.
And because I was born in 1996 in Britain, we had mad cow disease and you cannot donate any blood or organs or anything in the United States if you lived in Britain for more than six months, from a period that ends at the end of 1996. So technically, I’m eligible to sell my eggs in America because I was born in August.
I did not live in Britain for six months. I wasn’t around for six months in 1996, but I get screened out by all of the donor websites pretty quickly. And so I would have to come up with a bespoke brokerage agreement. And that seems like a lot of work. But if you tick like a bunch of boxes, a whole bunch of them saying you a normal weight, are you, did you go to a good university?
Especially if you’re South Asian, because Indian people love having children and they hate donating eggs. You can actually make a lot of money from selling your eggs. So if you are worried about being able to afford it, you could earn between 30-50 grand per cycle of eggs.
You can come up with a bespoke arrangement, so if you want to freeze them, then you can also plausibly come up with a freeze and share agreement with a couple that would like a couple of your eggs.
Ben Southwood [00:10:44]: So basically, egg freezing is really awesome and it’s great that we have this technology.
Pieter Garicano [00:12:58]: My hope with this article is that it corrects some pretty big misconceptions. In this case, you’re being told it doesn’t really matter if you’re 25 or 35. It really does matter, and the fertility curve has a different shape, and just having that information in your mind will allow people to make more informed decisions.
But correcting big misconceptions is a broader theme in this issue because the article immediately after the egg freezing article is about how Australia stopped the boats by Amelia Wood. Now, you may have heard of things like the Rwanda deal in Britain or Netherlands has the Uganda deal. Italy has the Albania deal. The idea behind all these deals is that if you have Illegal immigrants, or asylum seekers which come to you by boat or by land, and they arrive in your country to process their asylum requests. You can process their requests offshore and in the case they are rejected, they’re no longer in your country. Many countries around Europe and indeed around the world are really keen to try out this model of offshore processing. And why are they all doing this at the same time? Well, because there’s one country in the world, which had an asylum problem, and they solved it by doing offshore processing; Australia. And in this article by Amelia Wood, Amelia pretty definitively proves that in fact, offshore processing is not what Australia did, and is not necessary at all if you would like to reduce the amount of legal immigrants arriving in your country.
Ben Southwood [00:14:32]: So explain how she does this?
Pieter Garicano [00:15:29]: So Australia’s had very low rates of boat migration since the seventies.
Initially it was people coming from Vietnam, and then towards the eighties it really started picking up and you start getting people coming from the Middle East as well. They often fly to Indonesia, and in Indonesia they’d pay a smuggler to bring their boat to Australia. And in the late nineties the Australian populists had enough and a liberal, which is in Australia conservative Prime Minister, decides to put a halt to it.
And so he does two things. One thing is that he started offshore processing. He takes people who come by boat and brings them to Christmas Island and to Nauru, which are territories or islands near Australia, and processes them there. But the other thing they do is turnbacks. A turnback is that the the Australian Coast Guard encounters your ship, or the Navy encounters your ship and then very gently either tows you back or invites you on board and brings you back to the beach.
Aria Schrecker [00:16:25]: Where do they tow you back to?
Pieter Garicano [00:16:26]: Indonesia.
Ben Southwood [00:16:27]: How far is Indonesia from Australia?
Pieter Garicano [00:16:29]: Extremely close, like 200 miles.
Pieter Garicano [00:16:32]: Christmas Island in particular. So they were heading for a offshore Australian territory at Christmas Island, which is off the coast of Java. So it’s really not far at all. It’s also because the Australian navy and Coast Guard were basically camping outside the Australian EEZ.
They just wait for people to come. And so it wasn’t as if they were dragging them a hundred miles. They would drag them 20, 10 miles back.
So these both things at once, that offshore processing was very high profile, very controversial, and then Turnbacks were actually done in secret because this whole process was classified in part because Indonesians, as you might expect, were not very happy about this. And they complained. That’s the end of the first wave from Australia. This doesn’t give us any evidence which of the two things actually works, but Australia’s actually interesting and the media shows this because they had a liberal, a left wing prime minister who came afterwards, who ended both programs.
And as expected between 2012, to 2013, the whole thing happened again; a huge wave of people came from Indonesia, 30,000 people tried to move to Australia. And there they actually staggered the two approaches. So first they tried offshore processing, and then a few months later they started Operation Sovereign Borders, where they did a very large scale term expiration.
And if you look at the time of arrivals, almost no one is deterred immediately after the introduction of offshore processing. It’s only once turnbacks are happening that people stop trying to come. And the most important and the most surprising bit of evidence that she gives in this whole article is that since 2014 Australia hasn’t done any offshore processing. They secretly wound it down, and since 2014, they’ve only done turnbacks and nonetheless they’ve had zero arrivals overseas.
Aria Schrecker [00:18:19]: Why doesn’t offshore processing work?
Pieter Garicano [00:18:20]: It might work, but it’s extremely expensive. Even Australia had to spend billions and billions of dollars on their programs in Nauru.
Ben Southwood [00:18:29]: And it gets full.
Pieter Garicano [00:18:30]: It’s so expensive that you only have a capacity of 1000 to 2000 people. And so almost everyone that was taken off offshore was actually, eventually released into Australia because when too many people came at once, they decided to clear the camps by bringing people to Australia to post them there.
And so offshore processing is so hard that it is incredible. And so it doesn’t work as a deterrent.
Aria Schrecker [00:18:53]: So it’s just a gateway drug to onshore processing.
Pieter Garicano [00:18:56]: You could call it that. The people who are being processed see it that way. They expect that de facto they will be posted onshore.
Aria Schrecker [00:19:05]: Also depending on the situation you’re fleeing, maybe just being offshore processed is better than what you’ve left behind. So you don’t mind languishing in those camps.
Pieter Garicano [00:19:15]: The camps seemed pretty bad. Part of the story here is that a mix of circumstances, the fact the camps were so bad that like the NGOs made a very big deal of it and the press made a big deal of it, and the fact that Indonesia hated Turnbacks.
And so the Australian government kept it secret, which meant that almost everyone who studied the Australian migration story only heard about the offshore processing.
Aria Schrecker [00:19:57]: So do we think anywhere else is going to do Turnbacks? Could Britain do turnbacks?
Pieter Garicano [00:20:06]: So the problem is that Turnbacks currently are seen as illegal on the ECHR because it’s collective punishment. You’re always meant to treat everyone as an individual.
Ben Southwood [00:20:23]: Firstly, that’s inconsistent with the Turkey deal -So after Libya fell, you get the migration crisis coming to Europe from North Africa and the Middle East.
And in 2016, if I remember correctly, they signed a deal with Turkey, which said we’re going to turnback every migrant and you’re going to take them. You can turnback every migrant over the border, you’re going to take them, but we’ll take someone else from you. So they’re not treating them as individuals.
They’re saying that migrants are an undifferentiated mass and we are accepting people, different people from the people who walked over the border, we’re turning about the ones who come over the border and we’re giving asylum to the same number of people, but just without an incentive to come over the border.
And that seemed to have worked in that case. But is that not inconsistent with ECHR?
Pieter Garicano [00:21:17]: Officially, the Greeks say they do no pushbacks.
Aria Schrecker [00:22:08]: Seems like it would be in a lot of European countries’ interests because obviously, everyone’s kind of slightly defending their borders against each other, probably to focus a lot more Navy effort over there.
Pieter Garicano [00:22:20]: Yeah, and also it seems much more- It’s both cheaper than doing a Rwanda deal or an Albania deal. And also it seems much more humane.
Aria Schrecker [00:22:28]: what’s very appealing about this framework is it appeals to my sense of what I would call decision theory about how you should devise rules for your country.
So a lot of people like to think of policy, immigration policy and refugee policy is a very classic version of this, where they see the problem that they have on hand at the moment they have it. So they see that they’ve got these people, a lot of them are very sympathetic people and you decide, okay, what do we do?
How many of these people do we want to take? Whereas you should actually be thinking, what rules do we want to set earlier so that we have the best way of making these decisions in the future? And turnbacks are appealing because obviously if you’re picking your refugees that you want to take from the whole world of refugees, you’re not going to simply choose the people who happen to end up on your shores.
There are loads of much more appealing things you can do. Like the Hong Kong refugee scheme is a perfect example of this. But your press and your general political establishment cannot be relied upon to allow you to make those decisions when you’re actually pressed with the problem. So it is nice that you can kind of take that option away from them and give yourself the ability to make those rules by doing the turnbacks instead of the processing where you are already stuck with those people.
Ben Southwood [00:23:46]: So the public is a principle and it has its agent, the government. Government is not completely reliable on delivering what the public wants at all times. One thing the public wants is we want these refugees but we don’t want those refugees. Right? There are many in the public who are very sympathetic towards Hong Kong and Ukraine war refugees, and then there are other cases of people they refuse. If you see a protest, people say they’re all male. If they were refugees from a war, they would not all be male.
And that might not be true actually, because of who can get there, etc. But when people say stuff like that, they’re worried that people are taking advantage. What you are suggesting, which I like, is that you can only get the trust that allows you to have the other refugees if you bind yourself in the case of some refugees.
My understanding is that Australia has quite a generous and well respected asylum system relative to most European countries, which are, where most European countries are voting far right parties or most European countries are seeing a lot of distrust in mainstream parties.
One of the causes is distrust in the immigration system. But it, see, my understanding is that that’s not true in Australia.
Pieter Garicano [00:25:05]: this is the most convincing point in favor of this policy. Is that Australia now has more refugees per capita than any country in Scandinavia.
In fact, it takes three times as many people right now than it did at the very height of the boat wave. So it turns out by being very predictable and consistent by building trust amongst the public, you actually get a much, much more generous overall policy than you do in this kind of uncontrolled principle agent world.
And in Australia, in fact of all the countries pulled by Ipsos in OECD the Australians were by far the most positive on asylum. In fact, 60% of the Australian public think refugees make a positive economic contribution to the country, which is really incredible compared with Western European countries where people are quite skeptical.
Aria Schrecker [00:25:55]: This is a white pill. I would say.
Pieter Garicano [00:25:57]: It’s a huge white pill. You can do a thing which is both cheaper, more humane and allows you to have a much more generous immigration policy.
Aria Schrecker [00:26:04]: The thing that really is appealing about it is that I have always kind of believed that Anglo countries aren’t actually racist.
And then you see all the new news from Britain and America and you think, oh no, what if they actually are? And it’s very satisfying to say, I think that Australia can continue to have actually quite a high proportion of people who aren’t white, quite a high proportion of refugees. Just the sense that they’ve got their borders under control is enough to make them happy with that circumstance again.
Ben Southwood [00:26:33]: I do think that ideological entrepreneurs, including far right racist ideological entrepreneurs, have an easy time of it when most mainstream parties won’t tackle issues that the public has - like that a significant subsection of the public has a strong opinion on. I don’t think that’s an excuse for people doing nasty stuff or having nasty opinions, but that it’s probably a predictor.
We were just talking about the principal agent problem between the public and its representatives, and did you know that in 1965 Britain had almost more reactors - nuclear reactors than all the rest of the world put together?
Pieter Garicano [00:27:18]: Because you talk about it every single day.
Ben Southwood [00:27:20]: Because I talk about it every single day and it was not true. So I originally thought they had more reactors than all the rest of the world put together, because a source I saw that seemed plausible and then another source backed it up, said that in fact, Britain had 19, the rest of the world had 21 put together, so it wasn’t far off, but it wasn’t quite true when I said that originally.
Aria Schrecker [00:27:42]: Can we slightly change the bounds of the question so that two of them don’t count?
Ben Southwood [00:27:53]: You can, so here’s one way of really changing the bounds.
So I have said many times that Britain had the first civil nuclear reactor. Britain did in fact have the first full scale civil reactor delivering power to a grid. Now if you take away all of those adjectives, it’s not definitely true.
Both the Soviet Union and the US can claim a certain first reactor that produced power. It’s just a question of where you set full scale and supplying energy to the grid. But because it was all in such a short time period and during 1952 this, everyone was getting there at the same time.
One of our articles is by Alex Chalmers, who sadly no longer works at Works in Progress. But while he was here, he did wonderful things on Nuclear energy. And in one year his incredible advocacy was clearly the main reason why John Fingleton’s review of nuclear regulation was so influential.
And the government said that they were going to repeal things, but he tells a very interesting story of British nuclear power. Britain, relatively early during the 1950s, quickly built 20 reactors in 10 years, we planned and delivered the first one in three years.
We commit to it, come up with the design for it, completely construct it, and it’s plugged in within three or three and a half years, maybe four years. And then that continues happening through the program. One of the big memes of the moment on how you should deliver nuclear reactors is that you should do it in fleets. The countries that do it best today, like Korea, and some of the countries that have done it best in the past, like France, have taken the design and then copied and pasted it around the country. In France, they rule that if you find a change that will be good, they will implement it, but during the next wave, as to not delay things or change things immediately.
Aria Schrecker [00:30:12]: That’s also our rule for editing the podcast.
Ben Southwood [00:30:14]: It’s also our rule for editing the podcast. And it is a great rule. Britain didn’t do this. Britain basically had totally flexible designs. The Magnox reactor, every Magnox reactor is totally different.
Britain builds out a bunch of reactors quickly; they do it with a really flexible design, they do it very cheaply. By today’s standards, these are small modular reactors, although obviously they’re neither small nor modular.
These are 80 megawatts or up to 200 megawatts, the last wave of Magnox reactors. But then after 1960s, despite building out the Magnox reactors pretty cheaply, and despite doing it quickly, and despite being a nuclear leader, after the 1960s everything goes wrong.
And there’s a common story of why this is, which is that when we built the Magnox reactor, it was a special kind of reactor design that no one else decided to do at scale; using the Advanced Gas-Cooled reactor, known as the AGR Advanced Gas Reactor. In the world today, about 75% of reactors are something called a pressurized water reactor. Which is the simplest kind of a reactor you can imagine.
You just have a core chain reaction going off, heat, some water that heats some other water. It’s like this, it’s the simplest possible reactor.
Aria Schrecker [00:32:36]: I saw a good meme on the internet that was like, we’ve invented a new form of power. You invented a new form of power? Or is it just steam again? Just steam again.
Ben Southwood [00:32:43]: Yeah, exactly. It’s the same old thing. The magnox is a bit more complicated than that. The reason we went for it is that it generates plutonium, weapons grade plutonium as a side product. Britain wanted lots of plutonium at the time, now we don’t want it.
The common story is that when a new design went ahead for the Advanced Gas-Cooled reactor, the difference was that they didn’t need a pressure vessel; their view was they didn’t need a vessel at all. It was completely safe automatically, there was no way it could ever go wrong, it had built in safety features. And to be fair, they ran a bunch of them for a long time and none, never had a disaster. So maybe it was true, but there was no containment vessel.
The AGC reactors were going to run at a much higher heat than if it was with water, as you can’t heat water past a certain point, but you could run gas at a higher temperature such as 600 degrees.
So that was all going to happen and then it went wrong. Alex takes a different view. He says that a lot of things went wrong at the same time, and it’s related to the fact that they chose a bad reactor design. But if you look later on, once the Advanced Gas-Cooled reactors were privatized and handed over to professional managers, they actually ran them pretty well and they were pretty good designs.
They had some flaws. They have to be retired more early than the standard reactor designs, but it wasn’t an utter disaster. The problems were the public started getting antsy about all government projects during the 1960s and 1970s in the UK. And this is not just nuclear. Everything starts going wrong.
Everything started after a certain point, takes ages to finish, goes through constant fights with everyone who’s nearby it in a way that just wouldn’t have happened before. When they were building out the first nuclear program the Minister for Power might do an inquiry that finds that the National Parks have said that this is going to damage the national park. Then the decision is that the benefit of cheap electricity outweighs that damage, and one week later they can start construction.
It seems this problem began during the 1970s. So Alex’s view is that the technocrats who ran everything were widely accepted when they made the right decisions, but once the public as the principal lost trust in its agent, in the state to make these decisions on its behalf, everything went to pot and perhaps the technocratic model was backfiring.
Aria Schrecker [00:36:38]: And you would say that in the 1970s we had a period of high inflation. And this is why the public lost trust.
Pieter Garicano [00:36:57]: So I am skeptical of the inflation story. You know, ‘68 happens before the oil shock. And also, for example the country worst hit by the oil shock in Europe was the Netherlands, because Netherlands was most pro-Israel.
And so the Netherlands embargoed for the longest. In fact, Netherlands embargoed for so long, the French volunteered to join the Arab embargo if in exchange, they got oil from the Arabs. So the Dutch had the worst inflation shock, yet didn’t have a big upswing in social disorder. But let’s disregard inflation.
We dedicated a whole episode to this question. I’m trying to understand what you mean by technocrats, or people losing trust in technocrats. Is it that the technocrats who were in charge of nuclear mismanaged the programs such that people were then against nuclear? Is it that different technocrats who are responsible for planning screwed it for everyone else because they made all products hard to deliver?
Or is this a third thing, which is kind of this general sense of how elites want ugly things, they believe social change is not good, that modernity is going the wrong direction. Therefore I’m generally distrustful whether they are technocrats in government or business elites or art elites.
Like which of these three stories, or maybe a fourth story is actually the Ben Southwood view for why people lost trust.
Ben Southwood [00:38:33]: Here are my suspicions.
Suspicion one is that the AGR did cause the problem. Not because it was really bad, because it turns out that the AGR was still five times cheaper than the supposedly good design that we’re implementing now in Hinkley point C and Sizewell C, which is basically a pressurized water reactor.
The AGR was supposedly worse than them, but it was much cheaper in real terms than those, and they still ran them for 40 years with no accidents. Once they worked out how to use them, it was completely fine. So it wasn’t that bad of a design. Now, the thing that it hurts the program through, was by being perceived as a really bad design and by having really bad hiccups for the first 10 years. It did have some clear flaws, but the main reason they hurt the nuclear program is not because it was so unworkable, but because they actually did make it workable and they could have kept building them.
And if we continued the program even at that cost, we would have saved tens of billions of pounds in the Ukraine crisis, and on the subsidies that Britain paid for energy during that time alone. I do think that the fact that it messed up initially punctured faith at the same time as a lot of other things punctured faith.
Pieter Garicano [00:39:57]: What’s the pattern there, that the technocrats were very consumed by glamorous technological super projects? Is it just a kind of cultural fad that happened to pass over?
Ben Southwood [00:40:09]: Think of the general experience you’ve had, which is that if the technical people are in charge, they always want to do the technically perfect idea.
They don’t want any other considerations. They’re not very good at trade offs if given complete control, they’re extremely useful people to have around, maybe the most important, but you don’t want them to be in charge of everything. If it was a bridge, they’ll build the coolest possible bridge that could take a hundred earthquakes at the same time, not the cheapest bridge that does the job sufficiently.
On another note, the laws that came in weren’t designed explicitly to hamper nuclear power. To make this more general, there are lots of things that got destroyed by 1970s reforms where the thing being targeted wasn’t really what ended up getting destroyed.
So the US Clean Water Act type; they brought in an NEPA review system that means you can’t reclaim land. And we just disagree with them about whether claiming land was good and maybe we’re right, maybe we’re wrong, but we just disagree.
But there are some other things similar to nuclear power here. A lot of the review systems that were brought in could have just been written except energy projects with the national security element.
And there are contingent factors that led to that happening and then unwinding that is difficult over time, because you build up stakeholders who like the system as it currently is. Now why did people lose faith in their system as a whole? There are a wide range of things that were happening at the same time.
So people will forget it now, but there was a lot of worry about immigration in the 1970s in the UK and there were in lots of other countries as well. Maybe that’s a UK specific factor, but it definitely seems like it was an important new case specific factor. It is clear that high inflation periods tend to be periods with more social dynamic, people voting for weirder parties. And so that was an important factor.
And then I do think that like the elites of the post-war era had a lot of different opinions from the public they represented and they did because of their especially strong powers that they had at that time, especially in Anglo countries where they had overwhelming power to expropriate to take decisions on behalf of society. There were lots of administrative systems that would’ve stopped the German government taking decisions quite so quickly on nuclear power during the 1940s, fifties and sixties.
Whereas the British government could decide within a day whether it was starting a nuclear power program. So many of these decisions were taken in ways that the public decided it didn’t like over time, it gave them a lot of benefit of the doubt. And then when things came to a head, they came to a head altogether.
Pieter Garicano [00:43:23]: So if I had to combine the reasons, it’s a mix of contingent nuclear specific mistakes plus elites who were generally very out of touch with the population. Plus a very, very responsive political system, plus a very responsive state, plus high inflation, all of which combines to ensure that British nuclear loses its way.
Aria Schrecker [00:43:45]: So these stories about bad British technocrats should bring us onto a piece about good technocrats, good American technocrats, good Dutch technocrats. We should talk about ASML, which is another piece by Neil Hacker in the issue 23 of Works in Progress.
It’s plausibly the most important company in the world, depending on what you think about AI.
But lots of people are very familiar with TSMC and Nvidia as being the two big chip companies that you should buy stock in, that you should be watching, but both of those companies are totally reliant on this other company, further down in their supply chain called ASML.
What ASML does is they make what is called an extreme ultraviolet lithography machine. One of the features of chips is that they cut patterns in very, very small to transmit information across. They used to cut them with physical tools, but once you’re getting down to the 10 nanometer, 3 nanometer sizes, physical tools start to leave enough residue that the lines aren’t perfect.
So instead they are cut with pure energy, light, and ASML make the only machines in the world that are capable of doing this. Other companies tried to make extreme ultraviolet lithography machines, but they’ve all failed up to this point. It’s an incredibly complicated machine.
We call it the world’s most complex machine because of how many parts there are inside of it to make this very precise technology. ASML would not be here without the American government, probably without the Belgian government and possibly without the Dutch one as well.
Pieter Garicano [00:45:20]: I don’t have a big Ben Southwood theory for ASML, but I do think there’s a few interesting lessons that I learned from this piece.
For example, the American government, so ASML has EUV technology, Extreme Ultraviolet technology, because the American government gave it EUV technology in the 1990s.
There were big cuts to the Department of Energy. And so they closed this EUV program and instead they asked the private industry to come and take care of it. Intel provided most of the funding, and they were looking for a manufacturing partner to actually make the machines. At the time there was a little worry about Japanese companies, for example Canon and Nikon, which were the two primary chip manufacturers. They were excluded from the competition and instead the sole contractor were ASML and a Californian company which soon went bankrupt and was bought by ASML, which left ASML as the only participant in this consortium.
Aria Schrecker [00:46:23]: What’s particularly interesting about that is it seemed like their initial instincts were to make it only American companies, and the whole project would’ve likely failed if they’d stuck with their gut instinct there.
Pieter Garicano [00:46:34]: Well, but in some ways the product also failed because when the Americans gave the technology to ASML, they required ASML to make 55% of the components in America, it was one of these classic JVs. The technocrats wanted to have spillovers in America and make sure America still controlled the technology. But it turns out that once you hand over all the IP and all the knowhow to a company like ASML, it’s really hard to enforce the contract obligation they had to create 55% of their components in America.
Aria Schrecker [00:47:06]: And ASML now basically have the opposite stance, which is they guarantee that maybe a third of their supply chain is EU specifically because they don’t want to be disrupted.
Pieter Garicano [00:47:14]: Correct. ASML has huge suppliers in Europe, they buy a lot from Germany and Belgium in particular. They buy most of the components from Eurasia. And so the Americans — it’s kind of a time inconsistency problem here — where they invited a foreign firm, set a set of requirements, and by the time the transfer happened there was no way to enforce the rest of the contract.
Aria Schrecker [00:47:34]: They did get their primary goal, to have a significant part of the chip industry outside of East Asia, because at that point that was a big geopolitical problem for them.
The people who started the program probably have a strong preference for this to be happening in Europe than in East Asia. And they probably are happy with the results.
Pieter Garicano [00:47:52]: I don’t think this outcome we got is bad at all. The national security fears they had were very overblown, but it is notable that even in a case where you’re making a deal with a company based in a very friendly country, there is no way of actually enforcing these contracts oftentimes.
Aria Schrecker [00:48:06]: Europe’s greatest industrial strategy from America.
Pieter Garicano [00:48:51]: I also think, by the way, ASML entirely fits the model I described; when we say that the labor markets are the main problem for European innovation, because the kind of innovation ASML has done has been extremely incremental.
They were given the big jump, the big bet part they were given, and the rest of it was the kind of thing Europeans have always been very good at. And so as Neil describes in his article; ASML has some of the longest tenures of any company of its size; their CEO joined 30 or 40 years ago when he was 29.
Many of their senior engineers are people who joined during its founding in the eighties. And so in many ways, ASML is an example of incremental innovation, the kind of innovation we actually predict that Europe would be good at.
Aria Schrecker [00:49:46]: This is why lots of the chip industry is in either Asia or Europe, it seems like it fits that business model more because it seems like it is just squeezing out more and more efficiency out of something that we already have rather than coming up with something totally new and disruptive.
And it matches my stereotypes about East Asian and European economies.
Ben Southwood [00:50:05]: That makes sense. But you have to remember that the chip industry is now fabs and designers, right?
Pieter Garicano [00:50:13]: It’s mostly just files. Just software.
Ben Southwood [00:50:15]: So it’s split perfectly along the lines that you would expect if the Peter story of the world is true, they’ve actually broken the firms down the middle so that the Americans can do the radical design elements, and in fact, Britain to some extent and the Europeans can do the incremental innovation.
Pieter Garicano [00:50:33]: Another great fact that I really enjoy in this issue of Works in Progress is in our article about buses by Samuel Hughes. The bus actually is not very radical technology. You need stable roads, you need pretty sophisticated carriages, but you can do it with horse drawn carriages, which we’ve had for the last 5,000 years. So why does it takes so long for bus to come about? Well, we invent buses once in the 1660s. It turns out Pascal, famous for Pascal’s wager invents the bus, the very first bus service in history, and he has a bus that runs around Paris.
Pascal was defeated by the great enemy of all innovation, which is regulation. The workers of Paris are very angry to see in the 1660s to see the bourgeois in their fixed little route coming around through the streets of Paris, and they riot and they make the bus illegal.
And then the bus technology is lost for the next 150 years until 1824, when it is once again invented by a different Frenchman who has a bathing house, I believe in Rennes or in Nantes.
He just has a service where he brings people to his bathing house. And then he realizes that people who know that he has a bathing house service will get on his carriage and then get off later, without reaching the bathing house because they know he is going to pick them up anyways.
And he figures out this is actually something you can commercialize. And within six years, the bus technology spreads from France to London, to Manchester to Birmingham. It spreads across the ocean. Like the moment people figure out you can just run a carriage in predictable routes, predictable times, the bus technology is everywhere, unfortunately for the inventor of buses, there’s a market crash in 1829 and he kills himself.
And the tragedy is that his bus company, which he invented, is actually the main founding company of the current Parisian power transport monopoly. And so the guy who invented buses, still lives on both the fact that his buses exist and his company exist, RATP.
Aria Schrecker [00:53:02]: Actually this segues surprisingly well onto Ben’s piece. I have always thought since UberPOOL and obviously since the great death of UberPOOL in London, that we are clearly very, very close to not not needing buses exactly, but being able to have a bus like thing where you can hire it and it gets a whole bunch of people in the same vehicle and moves you around.
And it feels like once we have autonomous vehicles regularly driving on London roads, that should be a very, very easy technology to be accessible to everyone.
Pieter Garicano [00:53:32]: Will it really? Will we have autonomous vehicles driving everywhere on London roads? Or will we have total gridlock because I read a piece arguing in fact we’re doomed to gridlock unless we do one simple fix.
Aria Schrecker [00:53:47]: Unless we impose a very particular form of tax slash regulation.
Ben Southwood [00:53:51]: So I’m not the only person to have ever said this. I’m not the originator of this thought, lots of people have predicted that autonomous vehicles, without some sort of change in our institutions would lead to gridlock, because you can have an office, you can have a bed, you can have a little bar with your friends. Like all of these things are possible. They reduce the cost of sitting in traffic to basically zero.
Once that’s zero, then you are completely indifferent about sitting in traffic. Though there might be some people who still need to get where they’re going- for those people, it’s still an enormous problem. I call this the Ogallala trap, after the Ogallala aquifer, which is this gigantic aquifer under the Midwest, which is why all the way from Texas up to Montana is some of the most productive farmland in the world. In the 1940s, farmers invented these centrifugal pumps that were cheap. Everyone could get and could pump tens of thousands of gallons of water per week. And so it started off as a seemingly endless resource, the water down there in the aquifer, and now many places have cut a hundred feet off this apparently endless resource and in decades, in some places it’ll be almost dry. Obviously they need to come up with some way of charging for it, replenishing it, etc. And there are many cases, the passenger pigeon, for example, survived until the telegraph allowed people to send quick messages about where the passenger pigeon flocks were going. This drove it extinct within a few years. Once they got sonar and all the other tools that allow- and diesel boats that could stay out at sea for a long time, the Atlantic whales died quickly after that. Common pool resources will get destroyed by technology in certain cases.
My opinion is that the roads are going to be just like that. Now, that’s actually not a very original contribution, this is well known. The thing that is slightly original from me is that there’s a wide perception that charging people more to use the roads is just completely impossible. That’s mistaken.
There have been lots of successful road pricing schemes introduced around the world. People often artificially restrict their range of what paying for the roads counts as so they say only Singapore has introduced variable time of use pricing on all its major expressways, so therefore only Singapore has road pricing, everywhere else just has a kind of tolling system. But tolls are also paying for the roads on the margin, and the interesting thing is that these schemes work on two very simple principles; with a toll road, tolls are very unpopular if you put them on roads that people are already driving for free.
If you build a new road and put a toll on it, people might grumble. There are always whingers out there, but people will think this is a new thing, I can choose whether I want to use it or not. They consider the roads they already drive on to be something they can’t choose.
Don’t change the rules of the game. So if you want, you can charge on new things, but not on old things. And the second is if you want to charge for stuff, then giving people stuff in return makes it much easier to sell.
And so the classic example of this would be how in most European countries, fuel duties and road taxes pay for more than the cost of building and upkeeping roads, right? So people are very willing to pay for that, which in most countries it’s been justified as something you pay in exchange for getting the roads.
And drivers have generally been the people who pushed it into existence in the beginning because they wanted more money for roads, which they wanted to drive on. So if we were to use both of those things to think about auto autonomous vehicles potentially causing massive gridlock, we would put a tax on autonomous vehicles now, perhaps a surge pricing style tax, as with an Uber when it’s in higher demand. The people who already use Waymo’s in the small number of places where it exists won’t like it, could become much more expensive, but almost no one uses autonomous vehicles. But once they become omnipresent, it’ll be built into what people think of as using it.
And the second thing is that if you really want to build a strong constituency for this tax surviving and not getting ground down by inflation or changes, make it go to something that the people who pay it want. If at least 50% of this money is going towards building new roads, tunneling roads, building bridges, things like that, then if you put the money raised from this to something that drivers want, they might not oppose it as strongly.
Pieter Garicano [00:58:44]: And you have some very notable examples in your piece of this happening, especially in the United States, which people commonly associate as being very pro driving and very anti-tax. I myself live in Washington DC in Virginia, around the corner from me is a huge network of expressways, which are private highways, which you pay a huge toll for.
They’re very expensive. It costs you $30 or $40 to go from DC to the airport. And the trick is the expressways are only new roads added afterwards that are built by private companies that pay down the road with a tolling. There’s huge support for them. One of the reasons people like them is that the police don’t operate on them. So you can in principle go as fast as you want.
It’s a trust system and as far as I can tell people abuse it only slightly, and so my point being that America’s actually very supportive of tolling and paying for roads.
Ben Southwood [00:59:39]: Through Florida there are loads of cases where you add the lanes onto existing roads, so Lexus lanes they sometimes call them, but I actually don’t think it’s because the people who use them are exclusively well off people.
It’s people who need to get somewhere fast. And they have them across the US like there are enormous numbers of these additional lanes where the lanes are paid for, the road widening is paid for by charging for those lanes. And very often those lanes have varying prices based on how busy the road is.
I always enjoy driving through France. In Britain, most of our major roads are built on exactly the site of our old minor roads. Some of them are Roman roads, some of them medieval, but they’re all very old. Whereas in France they left those roads, you can still drive on them; many of them became dual carriageways, and became major roads.
But their best roads were built again by private companies, these are massive motorways near to but not the same alignments, so you always have the choice of any journey that can be done on a slower, more winding, but still good road. But you always have the choice of upgrading, and you pay about 10 euros every hundred kilometers.
Aria Schrecker [01:01:44]: I have another additional take on one of the pieces, which is I was deeply curious about the giant neo trad Hindu temple piece.
I was surprised at how colorless all of the Hindu temples are, because my family is from South India. And so if I imagine a Hindu temple, I imagine it blue and pink and green. It’s gaudy, kind of quite similar to our classical statues piece from before; I imagine these things to be painted. I was disappointed to see they’re all white. They’re very ornate. Sure, they’re like neogothic buildings, but they’re not exciting.
Pieter Garicano [01:02:37]: We all agree, temples are very pretty and it is genuinely very cool that the stone is load bearing. I never knew this. Almost every building being built today is load-bearing steel.
But the temples are almost the only buildings being built in America, in Britain, in Spain, where the actual load-bearing part of the building is the masonry.
Ben Southwood [01:02:59]: The modernistic impulse, which is similar to the gothic impulse, is to say that authenticity is extremely important, right? You should never lie with a building, and this leads them to the view that facades in themselves are lies.
And you should never have ornate, beautiful, ornate or in any way beautiful facades. You should be making a building maximally functional, and it should be truthful and show you what it actually is. In this case, the important thing is if the facade is not load-bearing, it should not look like it’s load-bearing.
So it’s obliged to have a curtain wall style aspect, or you can even see that it’s glass the whole way down, so it couldn’t possibly even hold itself up. I object to this very strongly because once you look into the history of architecture, you find that almost nothing passes their standard of authenticity and truth.
Aria Schrecker [01:04:03]: In fact, Samuel says this is maybe one of the solid trends in what people find beautiful in buildings, which is that people like to be able to see how they think it’s load bearing. That seems to be a source of comfort, people don’t like it if columns are too spindly.
Ben Southwood [01:04:18]: Unless they’re made out of steel, because then you can tell they’d be very strong anyway unless they appear to be made out of steel.
And so my view is that it’s kind of more noble if a building is not showing its structure on the outside. because that’s thumbing its nose at the modernist impulse thing, which is the wrong impulse.
Pieter Garicano [01:04:35]: I’m totally with Aria here where you need very elaborate counter signaling about the snobbish modernness, but I believe thick walls are cool.
And so I see these temples and they look very massive. They look very permanent. I don’t really care if it’s fake or not, I just think the thick wall looks cool and I’m glad they’re making them.
Ben Southwood [01:04:50]: Yeah, It’s a slightly perverse impulse of mine, but I want it to be fake, to show that fakeness is not in itself bad and that what we should care about is what the building looks like, and how we experience it and how enjoyable it is to be around, and not some idea of how true the outside is to the inside or those kinds of considerations.
Aria Schrecker [01:05:09]: Of all of our pieces this is probably one of the ones that I would most recommend be read on the physical magazine rather than online, because obviously the whole point is to look at the temples and see the descriptions and line it up, and the way that is set on the page has been a real triumph of our art team.
I’m sure it’ll look nice on the website, but this is the piece that is most enjoyable in this format.
Ben Southwood [01:05:32]: That’s all we have time for. Thank you very much for listening to the Works in Progress podcast. If you’re not a subscriber, you can subscribe by going to worksinprogress.co/print if you are a subscriber, I hope you’re enjoying issue 22. Thank you.
