Comparative advantage and when to blow up your island

post by dynomight · 2020-09-12T06:20:36.622Z · LW · GW · 39 comments

This is a link post for https://dynomight.net/2020/09/11/comparative-advantage-and-when-to-blow-up-your-island/

Contents

39 comments

Economists say free trade is good because of "comparative advantage". But what is comparative advantage? Why is it good?

39 comments

Comments sorted by top scores.

comment by johnswentworth · 2020-09-12T17:56:00.004Z · LW(p) · GW(p)

One important thing I'd include: while adding more people (i.e. more than two) creates the possibility of individuals becoming worse off, it also very quickly removes most of the incentives for strategic negotiation behavior (i.e. hiding information, faking skill, threatening to blow up islands, etc). Even with just a dozen people/islands, multiple people have to form a cartel in order to achieve high price for a particular good, and it only takes one defector to break a cartel.

Replies from: Viliam, fraidykluofficer.com
comment by Viliam · 2020-09-23T13:25:23.689Z · LW(p) · GW(p)

In general, yes, but there can be other factors that reduce the possibility to interact with many possible partners.

Geographical local monopolies -- there are thousands of islands in the ocean, but most of them are too far from your home. You could replace your nearest trade partner with someone further away, at an extra cost; and if your nearest trade partner pushes you too far, you will do it. But within that interval, the negotiation is important.

Upfront transaction costs -- even if the trade partners are equivalent, but it is costly to start interacting with another one (you have to do a complicated background check, you need to adapt to their specifics), this again creates an extra cost of switching, and an interval within which it is about negotiation.

Both can apply at the same time.

There is also a gray line between "cartel" and "people doing the same thing, acting selfishly, but updating on their competitors' past actions". To make it simple, imagine that that a fair price for a ton of bananas is $100. (Fair price = what would be the market balance if anyone could trade with anyone, in a world with zero transaction costs.) But there is a $8 cost for trading with someone who is not your geographically nearest trade partner. In this situation, the banana buyers can individually precommit to buy at e.g. $95, because they know that you will prefer to sell them for $95 rather than sell someone else for $100, pay $8 for transit, and only keep $92.

Now imagine the banana buyers have a website, where they publicly share their experience. (This is perfectly legal, right?) And there is this highly upvoted article called: "Don't buy bananas for $100, you can get them for $95 using game theory". It becomes common knowledge that the banana sellers suck at negotiation (they don't have an analogical website), and that most banana buyers only pay $95. -- Armed with this knowledge, you can now precommit to only pay $90 for a ton of bananas next year, because now it is known that the best price your neighbor can get from anyone else is $95.

How many iterations can happen, depends on the exact shape of diminishing returns. For example, even if I was willing to pay $100 for my first ton of bananas, but using my power of precommitment I already got them from my neighbor for $85, I am probably not willing to pay $100 for the second ton of bananas. Suppose the second ton of bananas is only worth $90 to me. But to obtain it, from someone who is not my neighbor, I would have to pay $85 + $8, which is more. So I will not defect against the new equilibrium. -- Here I act almost like a cartel member (my first ton of bananas is worth $100 to me, and at the end I only buy one ton, and yet I precommitted to not pay more than $85), but I am still only following my selfish incentives, and at no point I am sacrificing a potential extra profit in favor of keeping the balance.

I feel like I am reinventing here the Marxist class conflict, in a more general form, with emphasis on sharing negotiation tactics. The essence is that one side shares their negotiation tricks, which work individually even if no one else is using them (this is what makes it not a cartel), but quickly become a new standard if shared; and the new standard -- and the common knowledge thereof -- becomes a more powerful leverage (this is what makes it cartel-like in effect) in the following iteration of negotiation. The power to say: "Yes, you noticed that I am using this dirty trick against you, but we both know that all my competitors use exactly the same trick, so you cannot punish me by switching to another. And it is perfectly legal, because we coordinated this publicly. Your side as a whole sucks at negotiation, my side successfully turned it into a global leverage, and you as an individual face an uphill battle here."

comment by fraidykluofficer.com · 2020-10-16T15:49:21.339Z · LW(p) · GW(p)

What we see in reality is that companies tend to consolidate into bigger and bigger conglomerates.

It's easy to see why.

As small companies compete, you naturally get market leaders. As these companies get larger they become more efficient at producing goods and services. They invest in mass production techniques in order to produce goods more cheaply than their competitors. They buy raw materials at cheaper prices because they buy in bulk. They expand specialization amongst their workforce. The bigger they get, the easier it is to make money.

When two market leaders merge they achieve massive economies of scale. This forces others to merge in order to compete, leading to ever greater concentration. Monopolies often buy their rivals.

It's not a matter of being "evil", it's just a natural outcome of capitalism. If they don't, a competitors will.

Replies from: johnswentworth
comment by johnswentworth · 2020-10-16T16:10:57.018Z · LW(p) · GW(p)

What we see in reality is that companies tend to consolidate into bigger and bigger conglomerates.

It's easy to see why.

I recently finished reading a book on theory of the firm, a branch of economics which studies where the boundaries of companies end up and why they end up there. For instance, why do companies in-house some functions and outsource others? What's the equilibrium number of companies in an industry? Why don't companies end up larger or smaller than observed?

Based on that background knowledge... your empirical claim here is true only in a relatively narrow subset of industries, and the "easy to see why" part is just wrong.

There are many industries where companies do not tend to consolidate, and even in industries where companies do consolidate there's usually an equilibrium with new companies constantly entering the industry. Larger companies are not always more efficient at producing goods and services - size introduces inefficiencies of its own [? · GW], and industries vary in the extent to production costs decrease at scale (sometimes production is cheaper at smaller scale!).

On the "why" side of things, there's no inherent reason why large-scale production has to take place within a single company - many of the benefits of scale can be achieved via multiple independent companies pooling resources (as is commonly the case in e.g. the finance industry). This was the first and primary insight which really kickstarted the theory of the firm - why do some industries see a few giant companies, rather than many smaller companies with resource-pooling agreements? Presumably there are trade-offs between the overhead of contracts and the overhead of company management, and those trade-offs are exactly what the theory of the firm studies.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-19T17:20:45.261Z · LW(p) · GW(p)

There are many industries where companies do not tend to consolidate

 

Can you give me a few examples? I'll list a few important industries:

  • Automotive
  • Construction
  • Electronics
  • Financials
  • Healthcare
  • Insurance
  • Internet
  • Oil and gas
  • Pharmaceutical
  • Retail
  • Telecommunications

 

In each one of these, you'll find a bunch of big players.

For example:

  • Automotive
    • Volkswagen (Germany)
    • Toyota (Japan)
    • Daimler (Germany)
    • Ford (US)
    • Honda (Japan)
    • General Motors (US)
  • Internet
    • Google
    • Facebook
  • Retail
    • Walmart
    • Amazon
    • Costco

 

Even if you consider new entrants (such as Tesla) we are still talking about a few companies that dominate the industry.

Can you list any important sector where we don't see consolidation?

Replies from: johnswentworth
comment by johnswentworth · 2020-10-19T18:11:42.457Z · LW(p) · GW(p)

Restaurants, car dealerships, spas and hair salons, construction, plumbers and electricians, doctors and lawyers. Every industry dominated by small businesses.

Even in some of the sectors you list - like automotive manufacturing - we haven't seen much net consolidation. We haven't seen a lot of new entrants, but's it's not like the number of car manufacturers is rapidly decreasing either. It's at an equilibrium, and that equilibrium has a lot more than just one company - which is not something you'd see if economic forces generally favored consolidation.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-20T14:46:14.918Z · LW(p) · GW(p)

Restaurants, car dealerships, spas and hair salons, construction, plumbers and electricians, doctors and lawyers.

 

What you are saying is that services can be provided by small companies.

Fair enough.

But we still can see consolidation there.

Starbucks uses a tactic known as ‘clustering’. They’ll build several cafes right in the same area to obliterate competition. This costs a lot of money, but they can afford it... They even use a strategy called ‘predatory real estate’. They pay more than market rate rents to keep competitors out of a location. 

 

 

Even in some of the sectors you list - like automotive manufacturing - we haven't seen much net consolidation. We haven't seen a lot of new entrants, but's it's not like the number of car manufacturers is rapidly decreasing either. It's at an equilibrium, and that equilibrium has a lot more than just one company - which is not something you'd see if economic forces generally favored consolidation.

The American automotive market consolidated from more than 250 manufacturers in 1909 to fewer than 50 by 1930. How many companies are there now?

One more example:

At the end of 1985 there were 18,000 banks in the United States. By 2007, this had been reduced to just 8,534, and since then has dropped further. Today, the ten largest U.S. financial institutions own more than 50% of total financial assets.

Replies from: johnswentworth
comment by johnswentworth · 2020-10-20T18:17:16.004Z · LW(p) · GW(p)

Of course there have been particular cases where an industry consolidated during a particular period. You made a much stronger claim: that industries in general tend toward consolidation. Pointing to two or three examples where industries consolidated does not provide much evidence for such a claim. On the other hand, pointing to examples where industries did not consolidate provides significant evidence against such a claim.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-21T11:32:58.060Z · LW(p) · GW(p)

Of course there have been particular cases where an industry consolidated during a particular period.

 

That period being... any moment in time.

 

You made a much stronger claim: that industries in general tend toward consolidation. Pointing to two or three examples where industries consolidated does not provide much evidence for such a claim.

 

I didn't point to "two or three examples", but eleven business sectors dominated by huge conglomerates.

 

On the other hand, pointing to examples where industries did not consolidate provides significant evidence against such a claim.

 

You responded with a single sector: services.

Even then, I gave a counter-example showing how big companies can drive small companies away.

And you see that happening all the time: Starbucks, McDonalds, etc.

The consolidation is a process, it didn't finish yet.

Replies from: johnswentworth
comment by johnswentworth · 2020-10-21T16:39:07.574Z · LW(p) · GW(p)

Services are not "one sector", they're generally considered a category of sectors, and they comprise a majority of all first-world economic activity. If something is true of the service sectors, then it's true of the majority of the economy.

Your eleven business sectors are not all dominated by huge conglomerates. Automotive production is, but automotive sales aren't. Construction isn't. Electronics isn't. Finance is mixed bag, depending on the sub-sector. Healthcare isn't. Insurance has big conglomerates on the backend, but lots of small independent salespeople on the frontend. Internet infrastructure is dominated by large companies, but the volume of small e-companies operating on that infrastructure is massive. Oil and gas is more concentrated than most of these sectors, but I'd still guess that you've never looked at the numbers enough to notice the long tail of medium-to-small oil producers (I have seen some of those numbers). Pharma is also relatively concentrated, and there the pressure toward aggregation is real. Retail has a long tail. Telecoms is concentrated, though it's become less concentrated in recent decades.

Note that, in each of these sectors there do exist large conglomerates. That does not mean that the industry is dominated by conglomerates. The big conglomerates are highly visible and salient; the small companies are not.

So out of these, you're right on maybe half of them if I'm generous. Overall, it sounds like you're making a really strong claim without ever having looked at the data, and the data does not back the claim.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-22T12:19:20.815Z · LW(p) · GW(p)

 Electronics isn't.

 

Isn't it? Think of a subcategory.

I could go on and on.

 

Retail has a long tail.

Retail is a prime example of consolidation -- think of Walmart,  

Sales by the 20 largest food retailers totaled $515.3 billion in 2016, accounting for 66.6 percent of U.S. grocery store sales, up from 42.2 percent in 1996. Amazon acquired Whole Foods in the summer of 2017. (Source)

Fast forward to 2020, "Amazon and Walmart are like two elephants wrestling, and all the other retailers in the U.S. are the grass" (Source).

 

 

Telecoms is concentrated, though it's become less concentrated in recent decades.

Telecom is another classic example of consolidation.

A long antitrust case completed in 1984 led to the old AT&T being broken into seven regional Bell operating companies and the much smaller new AT&T.

Less than 30 years later, in 2009 the Department of Justice started to look into whether AT&T and Verizon were abusing the market power they had amassed in recent years.

Now... you don't have to take my word for it. Google for "consolidation in _____ industry" and you are going to see the phenomenon being repeated everywhere.

Or read "America’s monopoly problem, explained by your internet bill".

Replies from: johnswentworth
comment by johnswentworth · 2020-10-22T17:40:19.579Z · LW(p) · GW(p)

You still seem to be missing the key point: if you want to claim that industries tend toward concentration in general, citing particular concentrated industries isn't going to cut it. You need to argue that industries which don't concentrate don't exist (or are at least rare), not merely that industries which do concentrate do exist. That requires a more comprehensive view. You cannot get there just by picking particular industries and arguing that they're concentrated.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-23T10:28:04.683Z · LW(p) · GW(p)

You still seem to be missing the key point: if you want to claim that industries tend toward concentration in general, citing particular concentrated industries isn't going to cut it.

 

Well... I gave you some examples in trillion-dollar industries and asked for counter-examples.

Many of your counter-examples -- car dealerships, spas, hair salons, etc -- are niche markets.

Few of them were multi-billion industries, and they provide more examples of consolidation.

Just google for "consolidation in restaurant industry" and you will find articles like "6 reasons for restaurants' massive consolidation wave", "Deals and consolidation dominate restaurant industry".

How about the small restaurants? According to "Food Delivery Consolidation: Good For Now, But Not For Long", platforms such as Uber Eats and GrubHub are taking 10% to 40% of gross transactions:

"With the rise in demand for their service, combined with the dire situation COVID-19 has created for most of the population, platforms such as Uber Eats and GrubHub are facing increased scrutiny over restaurant fees that can range from 10% to 40% of gross transactions, according to restaurant owners. (...)

In the future, retailers and customers can expect to pay even more in delivery fees, as the few delivery conglomerates monopolize the market and limit retailers’ ability to sustain a profitable business."

comment by Viliam · 2020-09-12T14:48:02.586Z · LW(p) · GW(p)

I believe this is a very important topic, so thank you for addressing it! But making a focus on comparative (as opposed to absolute) advantage at the beginning seems to me like distraction from the main problem, which occurs regardless of whether the advantage is absolute or not.

It is a separate (and very important) fact that mutually beneficial trade can occur in situations where one party doesn't have an absolute advantage at anything. (At least, if we ignore the transaction costs.) Without economical education, people are often unaware of this, but if you take an interest in economics, this will be one of the first things they will teach you.

But another fact is that if we have a situation (or possibility) of mutually beneficial trade -- regardless of how specifically that happened -- and the traded resources are (sufficiently) continuous, then there is actually a whole interval of possible mutually beneficial trades (ZOPA), and it matters a lot at which specific point the trade happens. It matters so much that it actually makes sense to put the entire trade at risk in order to achieve a more profitable point (which is still profitable for your trade partner!). For some reason, this fact is taught much less frequently (at least it seems so to me).

If you understand the concept of relative advantage, but don't understand the concept of ZOPA (and other people know this), what happens is that you get many trade offers that all give you epsilon extra benefit, and at the end of the day you are confused: "If I am participating in so many mutually beneficial trades, why am I not so rich?" Yeah, the technical answer is that your profit margin is low; but the important question is: why?

Even worse, low profit margin may seem like a natural consequence of trading at an efficient market (which again is a concept you do learn in Econ 101). So we get a "valley of bad economical literacy" where you believe that your behavior is optimal, and you even have seemingly solid scientific arguments to prove it, but in fact you are leaving a lot of money on the table. That's because you are an "educated stupid" and whenever you meet a market imperfection [LW · GW], you allow someone else to pick up the banknote from the street. (For the purposes of this article, the market imperfection is that there are two different islands, instead of thousand similar ones.)

I wonder what other similarly important concepts I am unaware of...

EDIT:

The way I used to describe this topic is: Imagine that if you work alone, you can make $100 a day. There is another person who can also make $100 a day working alone. But if you work together, you can make $300 a day. How would you split the money?

Now, the second scenario: If you work alone, you can make $100 a day. Another person approaches you and offers to work together, because that would be more productive; at the end of the day the person will give you $120. Would you accept the offer? You don't know how much money the other person can make alone, and how much money they will keep if you work together; they refuse to tell you. Do you actually need this information in order to make a rational decision? Obviously, $120 is more than $100, therefore...

Replies from: fuego
comment by fuego · 2020-09-15T17:58:50.426Z · LW(p) · GW(p)

I think the reason the second fact gets less attention has to do with the focus rapidly shifting to markets and prices as the mechanism by which the exchange rate is set, rather than 1-on-1 negotiation.

Not to say that the second fact is irrelevant. OPs examples of relevance of fact 1 include things like household chore splits between partners, as well as career choice -- applications where fact 2 is clearly very relevant.

But in econ 101, there is a lot of ground to cover. comparative advantage is a natural jumping off point for supply & demand Qs which are pretty much required content, whereas negotiation, game theory, bilateral decisions are usually treated as somewhat supplementary.

comment by abramdemski · 2020-09-21T19:33:58.252Z · LW(p) · GW(p)

Here's something I don't get about comparative advantage.

The implied advice, as far as I understand it, is to check which good you have a comparative advantage in producing, and offer that good to the market.

But suppose that there are a lot more goods and a lot more participants in the market.

For any one individual, given fixed prices and supply of everyone else, it sounds like we can formulate the production and trade strategy as a linear programming problem:

  • We have some maximum amount of time. That's a linear constraint.
  • We can allocate time to different tasks.
  • The output of the tasks are assumed to be linear in time.
  • The tasks produce different goods.
  • These goods all have different prices on the market.
  • We might have some basic needs, like the 10 bananas and 10 coconuts. That's a constraint.
  • We might also have desires, like not working, or we might desire some goods. That's our linear programming objective.

OK. So we can solve this as a linear program.

But... linear programs don't have some nice closed-form solution. The simplex algorithm can solve them efficiently in practice, but that's very different from an easy formula like "produce the good with the highest comparative advantage".

And that's just solving the problem for one player, assuming the other players have fixed strategies. More generally, we have to anticipate the rest of the market as well. I don't even know if that can be solved efficiently, via linear programming or some other technique.

Is "produce where you have comparative advantage" really very useful advice for more complex cases?

Wikipedia starts out describing comparative advantage as a law:

The law of comparative advantage describes how, under free trade, an agent will produce more of and consume less of a good for which they have a comparative advantage.[1]

But no precise mathematical law is ever stated, and the law is only justified with examples (specifically, two-player, two-commodity examples). Furthermore, I only ever recall seeing comparative advantage explained with examples, rather than being stated as a theorem. (Although this may be because I never got past econ 101.)

This makes it hard to know what the claimed law even is, precisely. "produce more and consume less"? In comparison to what?

One spot on Wikipedia says:

Skeptics of comparative advantage have underlined that its theoretical implications hardly hold when applied to individual commodities or pairs of commodities in a world of multiple commodities.

Although, without citation, so I don't know where to find the details of these critiques.

Replies from: Raemon, abramdemski
comment by Raemon · 2020-09-21T20:12:42.411Z · LW(p) · GW(p)

This actually has been a major question for me. 

It seems like there are two separate claims here, which is "societies tend to produce goods that are their comparative advantage" and "you, an individual, should try to do this." I'm mostly focused on the second one, and whether it applies to things like the x-risk ecosystem. People have talked as if it did apply. My guess is that insofar as there's formal math, it's much less clear and might be dominated by other considerations. 

It still feels vaguely like "what is my comparative advantage" within a particular community aimed at a particular task should be a relevant factor. 

The very crude algorithm I think I've been doing is "look at the list of things I seem particularly good at" (that I might have absolute advantage in), and then look at the things that are in-demand (which, I might plausibly turn out to have comparative advantage at). That at least narrows the search space a bit to "things that are good hypotheses for being my comparative advantage."

Replies from: abramdemski, dynomight
comment by abramdemski · 2020-09-22T16:17:01.731Z · LW(p) · GW(p)

I once had a discussion with Scott G and Eli Tyre about this. We decided that the "real thing" was basically where you should end up in the complicated worker/job optimization problem, and there were more or less two ways to try and approximate it:

  1. Supposing everyone else has already chosen their optimal spot, what still needs doing? What can I best contribute? This is sorta easy, because you just look around at what needs doing, combine this with what you know about how capable you are at contributing, and you get an estimate of how much you'd contribute in each place. Then you go to the place with the highest number. [modulo gut feelings, intrinsic motivation, etc]
  2. Supposing you choose first, how could everyone else move around you to create an optimal configuration? You then go do the thing which implies the best configuration. This seems much harder, but might be necessary for people who provide a lot of value (and therefore what they do has a big influence on what other people should do), particularly in small teams where a near-optimal reaction to your choice is feasible.
comment by dynomight · 2020-09-21T23:03:15.826Z · LW(p) · GW(p)

In principle, for work done for market, I guess you don't need to explicitly think about free trade. Rather, by everyone pursing their own interests ("how much money can I make doing this"?) they'll eventually end up specializing in their comparative advantage anyway. Though, with finite lifetime, you might want to think about it to short-circuit "eventually".

For stuff not done for market (like dividing up chores), I'd think there's more value in thinking about it explicitly. That's because there's no invisible hand naturally pushing people toward their comparative advantage so you're more likely to end up doing things inefficiently.

comment by abramdemski · 2020-09-21T20:05:26.037Z · LW(p) · GW(p)

OK. It seems there are results for more than 2 goods, but the results are quite weak:

Thus, if both relative prices are below the relative prices in autarky, we can rule out the possibility that both goods 1 and 2 will be imported—but we cannot rule out the possibility that one of them will be imported. In other words, once we leave the two-good case, we cannot establish detailed predictive relations saying that if the relative price of a traded good exceeds the relative price of that good in autarky, then that good will be exported by the country in question. It follows that any search for a strong theorem along the lines of our first proposition earlier is bound to fail. The most one can hope for is a correlation between the pattern of trade and differences in autarky prices. 

Dixit, Avinash; Norman, Victor (1980). Theory of International Trade: A Dual, General Equilibrium Approach. Cambridge: Cambridge University Press. p. 8

comment by Raemon · 2020-09-20T03:48:08.824Z · LW(p) · GW(p)

Curated. I've read several explanations of comparative advantage over the years, and I found this to be among the most clear and accessible ones that I've read. I also liked the juxtaposition with ZOPA. I found Villiam's followup comment [LW · GW] additionally helpful for solidifying how a couple different economics principles fit together.

comment by fraidykluofficer.com · 2020-09-16T07:16:53.269Z · LW(p) · GW(p)

The author discusses impossibly perverse behaviors such as blowing up one's island, but forgets to mention the most common anti-competitive practices such as dumping and product tying.

For example, the other guy would price coconuts and bananas well below the Zone of Possible Agreement (dumping), and you would be so glad to buy everything from them -- destroying your industry and increasing his comparative advantage.

Of course, he is not altruist, so, when you are completely dependent, he will raise his prices above the ZOPA -- and you have no option because it would cost you a lot (time and effort) to rebuild your industry and regain your comparative advantage.

If you try to produce bananas, he drops the price of bananas, making profit from coconuts; if you try to produce coconuts, he drops the price of coconuts, making profit from bananas.

When you start to recover one industry (say, coconuts) he could force you to buy coconuts if you want to buy bananas (tying).

And if you ever managed to rebuild both industries, all he had to do was to bring his prices below ZOPA -- and you would have to decide if you still believe in free and unregulated markets, or avoid the same mistake, imposing tariffs and regulation.

---

You can observe this kind of imbalance in the real world: developing countries export iron for $100 per metric ton ($0.05/pound) and buy high value-added products (phones, tablets, computers) for $500/unit.

Developed countries use the profit they make from high value-added products and services to subsidize their farmers, pushing the price down. For instance, wheat costs $188.75/ton ($0.10 per pound).

Without subsidies, the farming industry would be decimated from rich countries, and prices would go up. Of course, there is no such thing as "free market".

Markets need regulation to prevent anticompetitive behavior.

Replies from: Viliam
comment by Viliam · 2020-09-23T20:58:53.212Z · LW(p) · GW(p)

In this example it is assumed that the entire island is literally owned by one person. So, if you wish, this person may be a metaphor for a strong centralized government.

Destroying your production capacity is a strategic mistake, and exposes you to blackmail in the future. A smart owner (or a smart centralized government) would not let that happen. If you want to give me free bananas, okay, I will take them; but I will still keep my banana plantation ready. That way, I get free bananas today and keep my ability to produce bananas tomorrow.

(And the other side of the same coin is that a smart owner -- or centralized government -- will try to expand their future production capacities. For example, if today it is for me more profitable to grow bananas than to write computer software, I might strategically decide to write software anyway, at least part-time, because two or three years later my software-writing skills are likely to increase dramatically, while my banana-growing skills would probably remain the same. So the comparative advantage of tomorrow may reward me for writing software, but in order to get there, I need to accept some disadvantage today.)

That said, another question is whether subsidies are the best way to keep your production capacity, and what amount of subsidies is optimal. (Of course, the farmers will always say "more is better" for obvious reasons.) If we discuss real-life agriculture, I would even challenge which types of products should we subsidize: if the goal is to prevent starving, we probably do not need to protect our meat production -- if the other countries keep giving us cheap meat, let them; and if they suddenly stop doing that (in the unlikely case that all meat-subsidizing countries would coordinate to do this in the same year), we may have a year or two of mostly vegetarian diet, but no one is going to die.

In other words, although some protection of production capacity is strategically important, it doesn't necessarily follow that the farming subsidies, as we know them now, are anywhere near the optimal solution. (Specifically, I think that subsidies of meat production are completely unnecessary -- it is unlikely that all other countries would stop subsidizing meat at the same year, and in the unlikely case that would happen, we would survive anyway.)

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-09-25T12:57:01.965Z · LW(p) · GW(p)

Destroying your production capacity is a strategic mistake, and exposes you to blackmail in the future. A smart owner (or a smart centralized government) would not let that happen. If you want to give me free bananas, okay, I will take them; but I will still keep my banana plantation ready. 

 

The article clearly suggests that the inefficient farmer should stop working on coconuts and work full-time (8.5 hours a day) on bananas; so he wouldn't have time to keep the secondary plantation ready.

That's the recipe to become dependent from a stronger partner, something that the article suggests is a good strategy for both parties. But is it?

Let's not assume bad intentions from any side.

Everything goes well for a couple of years... but the efficient farmer realizes he is working only 2 hours a day, and decides to increase his production, working 4 hours per day. (Again -- let's not assume bad intentions; he just wants to expand his production)

Boom! He no longer needs to buy bananas.

What should the inefficient farmer do? Ideally he would reduce the banana production, but he can't produce coconuts overnight. He must sell bananas cheap to buy coconuts.

The bananas:coconut exchange rate jumps from 2:1 to 4:1. That means he must double his production, working 20+ hours per day, while the other farmer comfortably works 4 hours per day.

Of course, he can't do that for long and goes bankrupt before his coconut plantation is ready.

Please note that this doesn't even assume bad intentions from the efficient farmer; it's just a natural result from the free trade between unequal partners.

That's why, in the real world, countries must impose tariffs, quotas and other types of regulation.

 

And the other side of the same coin is that a smart owner -- or centralized government -- will try to expand their future production capacities

I agree with you -- and that's exactly what the efficient farmer did, in the present.

And because the inefficient farmer decided to focus solely on one industry, he couldn't keep the other.

(Something that the article suggests would be beneficial for both parties.)

 

That said, another question is whether subsidies are the best way to keep your production capacity, and what amount of subsidies is optimal. 

It's a sane strategy if you don't want to become dependent.

People can live without iPhones; but they can't live without food.

 

I would even challenge which types of products should we subsidize: if the goal is to prevent starving, we probably do not need to protect our meat production -- if the other countries keep giving us cheap meat, let them; and if they suddenly stop doing that (in the unlikely case that all meat-subsidizing countries would coordinate to do this in the same year), we may have a year or two of mostly vegetarian diet, but no one is going to die.

 

You are underestimating the time and effort to re-create an entire industry, once it is destroyed...

We are talking about decades!

 

In other words, although some protection of production capacity is strategically important, it doesn't necessarily follow that the farming subsidies, as we know them now, are anywhere near the optimal solution.

 

Do you think it would be a smart strategy, for the developed countries, to become dependent?

I don't think so.

Subsidies may not be the "optimal" solution -- but is good enough to protect their interests.

(And, of course, it is not very good for the developing countries.)

comment by Rafael Harth (sil-ver) · 2020-09-12T11:49:00.899Z · LW(p) · GW(p)

There's no simple simple math to decide what point in the ZOPA we settle on.

Actually, you could model this as a bargaining problem. The only issue that the basic theory models only one resource that is to be distributed among two people, whereas now we have two (coconuts and bananas). This means it requires both of us to be honest about our preferences for how we weight bananas vs coconuts. Given that, bargaining theory requires that we define

A feasibility set , a closed subset of that is often assumed to be convex, the elements of which are interpreted as agreements. is often assumed to be convex because, for any two feasible outcomes, a convex combination (a weighted average) of them is typically also feasible.

A disagreement, or threat, point , where and are the respective payoffs to player 1 and player 2, which they are guaranteed to receive if they cannot come to a mutual agreement.

The disagreement point equals the pair of {coconut plus banana utility for (me,you) according to our honest preferences} we get when we don't trade. The feasibility set consists of all possible pairs of of {coconut plus banana utility for (me,you) according to our honest preferences} when we trade according to any one exchange rate.[1] So point in the ZOPA corresponds to one point in the space.[2] Given this, there's a unique mathematical solution that follows from some basic properties.

If both parties accept this, this would solve a part of the problem. Not all, since we are still incentivized to lie about our preferences.


  1. Well, actually it would be the convex closure of those points plus . ↩︎

  2. I believe you could also include all of the exchange rates outside of ZOPA and it wouldn't change the solution. ↩︎

Replies from: daniel-kokotajlo
comment by Daniel Kokotajlo (daniel-kokotajlo) · 2020-09-14T09:29:36.334Z · LW(p) · GW(p)

There are other assumptions that get to other bargaining solutions, however. Also, it's unclear to me whether any of the bargaining solutions are morally acceptable in all important cases.
 

comment by ryan_b · 2020-09-21T13:45:49.962Z · LW(p) · GW(p)

What is the development of this that accounts for the different style of transactions, and for risk?

  • Before trade, X minutes yields a banana or a coconut, which allows for keeping marginal gains. In the trade scenario, this is replaced with a lump sum payment at the end of the day. It seems like we should weigh an all-or-nothing transaction differently then incrementally adding bananas.
  • Following on this, the risks are different. There's not much in the way of perverse behavior you will encounter working for yourself. Further, even outside of perverse behavior you will now absorb regular risks that affect the trade partner and their island. Sprained ankle; outbreak of rats; hurricane hits them but not you.
comment by Thomas Sepulchre · 2020-09-20T08:57:41.722Z · LW(p) · GW(p)

The ZOPA issue you raise actually disappears when the trade involves a lot of players, not only two.

Let's say we have N players. The first consequence would be the existence of a unique price. A lot of mechanisms can lead to a unique price, you could spy on your neighbors to see if they get a better deal than you do, or you could just have a price in mind which gets updated each time you get a deal or you don't - If I get a deal, that's suspicious, my price wasn't good enough, I'll update it. If I don't, I was too greedy, I'll update it - In the end, everyone will use the same price.

At this point, everyone will specialize in one good (banana or coconut) based on whether each one values banana/coconut more or less than the market does.

The ZOPA is therefore the ZOPA between the worst banana gatherer and the worst coconut gatherer. The bigger N, the smaller the ZOPA, therefore the smaller the need for perverse behaviors.

Replies from: fraidykluofficer.com
comment by fraidykluofficer.com · 2020-10-16T08:39:49.813Z · LW(p) · GW(p)

Let's say we have N players. The first consequence would be the existence of a unique price.

 

That's not true if one of the players has a monopoly.

A monopoly will extract as much as it can, delivering as little as possible.

It will be able to charge different prices from different customers. In a free and unregulated market, the monopoly can award you ("prime" customer!) or punish you -- it is the lawmaker, juror and executioner.

 

At this point, everyone will specialize in one good (banana or coconut) based on whether each one values banana/coconut more or less than the market does.

 

This scenario is not supported by reality.

What we see, in practice, is that free, unregulated market will create monopolies.

As small companies compete, you naturally get market leaders. As these companies get larger they become more efficient at producing goods and services. They invest in mass production techniques in order to produce goods more cheaply than their competitors. They buy raw materials at cheaper prices because they buy in bulk. They expand specialization amongst their workforce. The bigger they get, the easier it is to make money. Smaller companies cannot compete.

When two market leaders merge they achieve massive economies of scale. This forces others to merge in order to compete, leading to ever greater concentration. Monopolies often buy their rivals.

The only solution to this is to admit that we need regulation and laws.

There is no such thing as "free and unregulated" market. An unregulated market will not be free.

comment by ChristianKl · 2020-09-12T10:02:46.223Z · LW(p) · GW(p)

There's the other externality of becoming dependent on the bridge existing. Right now supply chains that were working well a year ago collapsed as COVID-19 lead to borders being closed. 

Replies from: luke-finney-1
comment by Luke Finney (luke-finney-1) · 2020-09-15T20:22:39.921Z · LW(p) · GW(p)

As another addition to this critical point - resilience as a theme is not even considered in these scenarios of comparative advantage. So as others have pointed out, as long as it's steady state, this works. But how long does this remain steady state in the real world? And at some point, it's really really beneficial to not 'forget' how to harvest bananas or coconuts..

comment by Mary Chernyshenko (mary-chernyshenko) · 2020-09-13T13:15:03.592Z · LW(p) · GW(p)

How interesting! And what happens when you introduce naturally fluctuating harvests? It seems that the dweller of the unlucky island has to re-negotiate to progressively worse conditions?

comment by Slider · 2020-09-16T14:01:12.137Z · LW(p) · GW(p)

Thinking about this if the diffference in gathering rates is due to gathering skill then learning the gathering technique of the more skilled gatherer would have equal impact to trade as blowing up your island.

If the difference is not learnable and is due to soil and such one would think that the poor island resident would be tempted to relocate to the rich island. Such relocation could go over peacefully or it might cause tensions and border enforcing. But if it goes to enforcement it could go to violence and even threat of violence could be a form of coersion. So already in "We are in our respective islands" we are already in the realm of coersion.

comment by jmh · 2020-09-13T13:14:09.558Z · LW(p) · GW(p)

I think another point to keep in mind when thinking of comparative advantage is that it is clearly a necessary condition for trade (or at least seems obvious to me that it is) but not clearly a sufficient condition.

Even in this two good model it seems that one must accept there is an implied third good, everything else (leisure, sleep, something else?) so the analysis of the two goods is a partial equilibrium analysis. If the margins are getting changed there then the general equilibrium has now been disturbed. That will then feed back into the two good relationships.

I think one of the questions then is just what margins or constrains are being relaxed when looking to trade based on the comparative advantage.

For instance, lets assume that the two islands specialize in either coconuts or bananas and then trade at the 1:1 exchange. Production and consumption has not changed. Since they clearly had more minutes in the day to increase output on their own consuming 10 bananas and coconuts a week was optimal. If that level of consumption is not changed due to trade (generally assumed to be the case but lets ask) what happens with the time no longer needed for production?

If the additional available time is now just boredom I would think trade becomes a bit of an annoyance and trade is curtailed. If that is true on one side but on the other they other things they can do with the time a different form of friction emerges, perhaps one side claiming the other is dumping or protectionist.

comment by Simeon Pilgrim (simeon-pilgrim) · 2020-09-16T00:50:32.939Z · LW(p) · GW(p)
Concealing information. You are smart. After the bridge appears, you quality realize I might spy on you, and this would harm your negotiation position.

realizing does not have the nature of quality, I suspect it's nature of quickly however is more fitting. Trade these words or my island will blow-up.

comment by Slider · 2020-09-14T10:55:20.298Z · LW(p) · GW(p)

What irks me a little is that work is being allocated to be done by people who are bad at that job. If people who were good at the picking did all the picking collecitvely between the two of them there would be only 300 minutes of work. In the specialization arrangement there is a total of 700 minutes of work. I do wonder whether there is any way of finding an activity that takes between 300/2=150 minutes and 600 minutes to make up for the increased workload of the trading partner (say massage them).

On particular it seems weird that the bananas not being picked on the efficient island are then picked on the inefficient island for 3 times as long. It is also strange that one benefits from poor conditions elsewhere to improve their own situation. It also takes incentive away form alleviating the others poor conditions and maybe even have more trade partners with poor conditions or comparative advantages dissimilar to yourself.

Replies from: Dagon
comment by Dagon · 2020-09-14T17:18:38.774Z · LW(p) · GW(p)

I'd love to see your model - include a value of leisure, which is _NOT_ tradeable. Show your work in how you calculate this value, relative to the value of other resources (note: this request is a trap. It's likely impossible to agree on private valuation of leisure.)

I very strongly expect that there are arrangements which increase total leisure, but they do not satisfy the requirement that all transfers are voluntary and make all participants better off. They increase the leisure of those bad at their job (and worse at other jobs), but they decrease the leisure of people who are good at their job.

Replies from: Slider
comment by Slider · 2020-09-16T14:33:21.375Z · LW(p) · GW(p)

Off course a good that you are not allowed to trade doesn't have a trade value. The value of leisure is in effect already in play at comparative advantage in that the advantageous position is preferable because I get to the same end state with less time spent ie more leisure. But while I care about my own leisure I don't typically care about others.

I guess the situation could be expanded in that there is a third good (say canoes), that if we are separate I would not produce for myself and you would not produce for yourself but in contact I could produce it for your consumption. That is if am good in canoes but hate them, but you are bad at canoes but like them. It would just be a situation of aymmetries of production but more equal overall absolute production levels.

There is still the choice that one typically doesn't choose max leisure time and starving but rather some work and some food. So if you get fruit diet at 100 minutes of work or caviar diet 101 minutes or grass diet with 99 minutes it is hard to pass objective judgement on that. So choices and needs might not be completely separate.

comment by MikkW (mikkel-wilson) · 2020-09-13T15:28:29.680Z · LW(p) · GW(p)

In section IV Zopa, this text is repeated twice:

Of course, I’d prefer to pay you fewer bananas! So I’d prefer a rate to the left end of this range. Conversely, it takes you twice as long to make banana as a coconut. You’d be thrilled if I paid you 4 bananas per coconut, but you’d never accept less than 1/2 a banana for one coconut.