So, while procrastinating I read an article about the supposed ‘neomercantilism’ that china is using and other south east asian countries such as South Korea used to industrialise. This sent me into a massive spiral of researching the ‘mercantilism debate’ which eventually culminated in me deciding to read Friedrich List’s National System of Political Economy.

In the current dialogue, mercantilism is mostly used as a slur, or at best a strawman philosophy. People make disparaging references to countries trying to horde bullion and assume that that’s the whole philosphy. Pundits throw about “mercantilist” policies as a generic boo-light: “I disagree with this policy and it looks vaguely protectionist, therefore it must be mercantilist and therefore discredited” seems to be the logic. As a natural contrarian, whenver I see a school of thought treated in such a manner, I am inclined to investigate, so I decided to get it from the horse’s mouth: that apostle of mercantilism, Friedrich List.

I’m only about a quarter of the way through the book, but I had so many notes I decided to write up a quick blog post about it already.

In theory, economics claims to be a purely descriptive field, unmarred by normative consideratoins. It’s function is purely to answer economic questions. I ask: “what will the economy do if I (or the government, or whoever) pursue policy X?” And economics is to have no position on whether X, or the result of X, is good or bad.

In practice, though, this isn’t how it works. Economics, or at least economists,  do perceive economics through a normative sheen. Growth is good. Efficiency is good. Poverty is bad. Inequality is (often) bad. And so on. As normative commitments go, these are pretty unobjectionable, but it’s still good to be aware that we hold them.

These normative commitments form a kind of implicit ideology which is shared by most economists. It is implicit because it is so rarely attacked, or defended, or really thought about at all. Pretty much everyone who studies economics  internalises its tenets, usually without questioning them at all. They seem self evident.

That is why it is so interesting to read a work like List’s, because he operates from under a completely different implicit ideology, and it is this diference that makes so much of the ‘debate’ seem like the two sides are just talking past each other.

A classic example is the ‘criticism’ of mercantilism that it unfairly privileges producers over consumers. Yes it does. That’s the whole point.

The implicit ideology behind mainstream economics is global utilitarianism. The purpose of the economy is to fulfill human wants (generally operationalised as consumption of consumer goods). Economic policies should be made to maximise global welfare, not specifically that of a single country or other unit. People’s utility functions should be treated equally. We should not privilege producers over consumers, or people from nation A over people from nation B. The main goal of economists should be to maximise consumption. People work so as to consume.

Imports are good, because they are what we consume. Exports are bad because they something we are forced to do to pay for imports. A trade deficit, then, means a great bargain. We are effectively getting all this free stuff we don’t have to pay for, at least not right away.

List operates instead  with a different implicit ideology which I’ll call “national producerism”. The world is split into nations. The purpose of economics is to make nations strong, where strong is effectively defined as manufacturing capability. Utility functions should definitely not be treated equally. Economic policymakers should only care about their own nation. The welbeing of random people in other nations is irrelevant. Actually, as national power is mostly relative, policies that hurt other nations (or at least their manufacturing) are probably good, even if they have no beneficial impact at home.

Consumption is just an unfortunate implulse a nation’s people have that must be satiated to some degree, but otherwise subordinated to goals of national greatness.

Exports are good, becaues they demonstrate surplus manufacturing capability. They also displace the market share of other countries’ industries in their own markets. If a nation’s industries are to prosper, they must have markets beyond their home nation, so they must export. Imports are bad because it is through them that foreign manufactures compete in home markets.

It is worth noting that despite the common idea that mercantilists think all imports are bad  i.e. they advocate autarky, - this is not what List says. He sees importation of manufactured or finished goods as bad, and of raw materials as good. His ideal country would perhaps be imperial Britain, importing raw materials from its empire all over the world, and exporting finished goods back to those countries, massively undercutting and stifling foreign industry in the process.

Imperial britain also conquered a huge empire, started the industrial revolution, and spent a century as the top economic power despite having a relatively small population and not especially blessed with natural resources. Maybe there’s something in mercantilism after all.

This reminds me of a funny experience I had when reading A Splendid Exchange by William Bernstein. The overarching thesis of the book was basically that free trade was good, protectionism was evil, and that today’s globalism is wonderful in every way.  Yet there were also some… inconsistencies. The most glaring one was the Calico Acts where Britan banned the import of dyed Indian textiles. He spends a lot of time condemning the move saying it hurt English consumers and decimated the Indian textile industry, and that it was a completely self-harming and intellectually indefensible move. But a few pages later he was talking about Britain’s thriving textiles industry which spearheaded the industrial revolution. Huh.

Another interesting thing to note is that basically every industrialised country, during the beginning of their industrialisation adhered to a substantially protectionist and mercantilist policy. The industrial revolution began in Britain not during the 19th century’s apogee of free trade liberalism, but under the auspices of the Navigation Acts. In the US, the economy first really took off under the American System, which remained in place till the 1970s. ( An interesting note is that, probably coincidentally, the 1970s were also the time when most of our current malaises began such as the technological slowdown, the productivity slowdown, and the growth of real wages in the middle class (-/get the graphs)). Germany and France both industrialised under heavily protectionist regimes. If we go into the 20th century, then the Soviets achieved impressive results under a heavily autarchic and “non-traditional” economic model. As all the Asian tigers such as first Japan post-Meiji, then Korea, and now China industrialised they did so from within a heavy blanket of protectionist measures, only slowly shedding them as they caught up to become more developed economies. The only exceptions are the city states such as Singapore and Hong Kong, which is not surprising as they function much more as mercantile hubs than industrial heartland and so benefit more from free trade. Even so, I know that Singapore, at least, followed some very careful industrial policy, as chronicled by Lee Kuan Yew in his books.

History cannot be run twice, and so does not allow experiments, but to me this seems a bit suspicious. Generally, it seems that free trade is really the result, not the cause of economic prosperity. It is only once a country has already become successful that free trade becomes appealing.

Hilariously, this is also List’s position. He is not a blanket protectionist. Rather he sees protection as a necessary stage in economic development, that of industrialisation. More generally, List sees development as having four stages which for simplicity I will call undeveloped, agrarian, industrialising and industrial.

An undeveloped country in this stage is one in which there is very little formal economy at all. The vast majority of the economic activity is subsistence farming, which is almost completely unaffected by any events outside the local village. Cities are small to non-existent. The political system tends to be a rule by local chieftains. The country has very little involvement in the global economy. It’s trade is negligible. Most of its importsf are luxuries for the tiny elite.

An agrarian economy is one that is beginning to develop. The primary occuption is still farming, but it is no longer subsistence farming but large scale and highly organised farming. Typically of cash crops. The country is dominated by large estates which produce crops primarily for export. In return the slightly larger elite can import a lot more luxuries, producing a serious aristocracy. Trade is beginning to take off. Cities are larger and can support significant amounts of artisans as well as the beginnings of intellectuals. This is bascially a ‘developed’ preindustrial civilisation.

This level of development is probably a relatively stable equilibrium without industrialisation.  The civilised parts of the ancient world, as well as Europe through much of the medieval period and renaissance were agrarian civilisations. As were China and India throughout their premodern histories.

List believes the best trade policy for developing from an undeveloped to an agrarian economy is free trade. An undeveloped country has a lot of ‘spare’ human capital being used inefficiently in subsistence farming. And they are already trained for farming and are good at it. Undeveloped countries will therefore have a comparative advantage in producing agricultural goods for export. So under free trade investment will flow into the country which will tend to open up more farmland, and promote the consolidation of subsistence units into large estates. So the country will slowly develop into an agrarian economy.

But once a country is agrarian, how can it become industrial? List argues that protection is needed for this step. The agrarian country’s comparative advantage is in large scale farming, not in industrial goods. If any industries do try to develop unaided, they will be destroyed by the competition from already industrialised countries whose comparative advantage is already in producing industrial goods.

This is the infant industry argument, and on the face of it it seems reasonable if you are optimising for national manufacturing capability and not for global economic efficiency. It takes time  for a nation to build up the huamn capital, infrastructure, and networks required to be able to sustain competitive industry without any protections. Without this time the industries will fail. If we view industrial sectors as more important for long term prosperity as agricultural, than this is effectively just a long term investment. We sacrifice some short term pain for long term gain.

Why should we make such a sacrifice? The key point, which does not seem to be appreciated by economists is that different sectors have different externalities. Large ones. Externalities big enough to shape the whole culture, institutions, and hence development prospects of the country.

But first let’s discuss the “knockdown counterargument” deployed against Mercantilism. Adam Smith and Ricardo’s concept of comparative advantage. The argument is simple. England is better at making cloth than Portugal. Portugal is better at making wine than England. If they trade, both will benefit over if they remained autarkic. Moreover, this is true even if England is absolutely superior in both wine or cloth, becuase there is an inherent tradeoff between the two. Furthermore, to maximise the benefit, England should specialise in cloth and produce more cloth (until, perhaps, diminishing returns make this no longer efficient) and Portugal should specialise in wine. (-/check to see that this is the actual argument and I’m not missing anything dumb). This is all well and good, and in this model free trade is absolutely the superior policy. The problem comes with what the model leaves out. It completely ignores second order effect.

If a country’s comparative advantage at time t = 0 is to produce wine, then they will tend to produce more wine. And investors will tend to invest in more vinyards,  a greater proportion of the population will be employed and trained in viniculture and so on. This will naturally lead to the comparative advantage increasing. Small differences in initial conditions can quickly lead to complete, and irreversible specialisation.

This is just the standard argument that trade promotes the division of labour, but with countries instead of firms. And this is great for the consumer. Portugal produces more wine, which makes it cheaper for English consumers. Portugal can also afford to buy more cloth from England for Portuguese consumers. English factories and Portuguese vinyards flourish while consumers in both nations get cheaper, better goods. Everyone is happy.

The trouble with this argument is that it assumes that it is always equally ‘good’ for a country to be specialised in wine or cloth. Or, more generally, in sector A over sector B. This could be the case, but it is at least non-obvious. I mean, what country would want their comparative advantage to be in agriculture, today? The fate of the countries that have remained predominantly agricultural today is dire.

Industrialisation is not just a mechanism by which a country can produce more units of cloth. It is a massive revolution to their society. It has significant positive externalities. It is necessary for greater literacy, universal education, more democratic or at least less extractive governments, greater urbanisation and so on. These all lead to a positive feedback loop in which countries which started industrialising began to have a better workforce, trained for factories so new industies can be developed more easily, which had booming middle classes and better less corrupt governments, all of which allowed them to gain a comparative advantage faster in the new industries. (Generally, having a comparative advantage in new industries is much better than in old, as the new typically have much more potential to grow). Countries which specialised in industrialisable goods such as cloth  - England in the toy example - ended up reaping a wealth of positive externalities and boomed far beyond their competitors who specialised in non-industrialisable goods - Portugal - who remain rural, backwards, and poor.

So if not for the global consumer, but for the leaders of Portugal, perhaps List might be on to someting. If you need any convincing of the inequality of the two paths, then you only need to compare the historical trajectories of England and Portugal since Adam Smith first wrote his parable.

This can just be thought of as an extension of the fact that there are winners and losers of free trade. And economists will tell you that though this is true, trade is Kaldor-Hicks efficient, so that the losers can be compensated for their loss by the winners, so everything is fine. What this analysis ignores is whether this compensation ever happens. In reality, it rarely does because the winners are typically the ones with more wealth, and hence more power. In fact, the opposite often occurs. Wealthier countries tend to be stronger, and so exact tribute from their weaker neighbours, rather than the reverse.

This is true even within countries. Everyone loves to kick a dog when it’s down. Industrial workers lose their jobs due to offshoring caused by free trade agreements: “Haha dumb racists, learn to code. We’ve got cheap iPhones now.”

It thus seems reasonable that, at least for the losers, protection and mercantilist policies can sometimes be better for them. This is all that List wants. He is not concerned with the health of the (global) economy in general. His economics is agent-relative. To him economics is a field which agents such as countries can query to ask, what is the best policy for me to follow, and mercantilism is generally optimised for that. Modern economics is typically more cosmopolitan and utilitarian. It tries to be agent-neutral. It asks: what policies are best for everyone as operationalised by what policies maximise total utility.  The mercantilism “debate” then seems to be mostly about whether mercantilist policies is not preferable just for individual agents, but for humanity as a whole - i.e. whether mercantilist policies are viable under the implicit ideology of modern economics.

It would be impressive of mercantilism if this were the case, as it is definitely not optimised for this. Even so, I will try to sketch an argument as to how this could be possible.

We can think of free trade in terms of mathematical optimisation. Free trade is fundamentally a hill-climbing algorithm. There is no “direction” to it, and it can only transition to states which are directly profitable from its present state. As such it may get stuck in local maxima, unable to reach better states because that would require running agaisnt a profitability gradient for a significant length of time. It is almost certain that these local maxima exist: the space of possible economies is almost certainly extremely nonlinear and nonconvex. The question, then, is whether protection allows one to break out of the gradient flow to eventually find a better maxima. This question is effectively synonymous to whether we have knowledge of the underlying economci state-space. If we do, then we should be able to use protection to nudge the optimiesr towards better maxima. If not, then we’re better off just defaulting to free-trade as most random nudges will be deleterious.

Do we have such knowledge? I would argue that we do, at least for some cases. For instance, we know fairly well how to industrialise an economy. We have case-studies from Britain, Germany, France, Russia, America, Japan, and then recently from the Asian Tigers (who pretty explicitly tried to copy the earlier models).  An undeveloped country trying to follow these models will probably better off than one simply trusting to a laizzez-faire approach. And because industrialisation and development bring such positive externalities, then the gains of a more rapid mercantilist industrialisation phase might outweigh the negatives globally and thus result in a utility gain overall.

Of course, however, in most cases we don’t have such knowledge. For instance, we know much less about the optimal economic policies for modern post-industrial economies as there has generally been very little experimentation and time to analyse their performance. Here it may be better over all to be more cautious and stick to the decent but un-optimised free trade policy.