On Free Trade Ideology

According to conventional “wisdom” free trade leads to prosperity. Usually the idea is based on a version of David Ricardo’s (1817) theory of “comparative advantage” which is taught in most high-school economics classes. There is, however, a fundamental problem with that theory, as was shown by Frank Graham in 1923, and unfortunately that problem tends to be ignored. In the following, I will briefly summarize Ricardo’s theory and Graham’s correction thereof, and discuss why the latter is ignored and the effects and implications of that neglect.

Ricardo’s model

Assume a world with two countries, “country A” and “country B”, and two commodities, corn and cars. Both country A and country B produce corn and cars, and both have a labor force of 20 “units”. (These “units” might be 10,000 workers or 1 million work hours, or whatever. This doesn’t matter.) In both countries, half the labor force produces corn and the other half produces cars. All of the production is consumed at home; there is no trade between A and B. There is one important difference between countries A and B, however, and that is that:
– in country A, 10 units of labor produce 40 units of corn, and 10 units of labor produce 40 units of cars; while
– in country B, 10 units of labor produce 45 units of corn, and 10 units of labor produce 35 units of cars.
(As in the case of labor, it doesn’t matter what exactly these “units” stand for. For convenience, I’ll omit the word in the following and will just write “cars” and “corn” rather than “units of cars” and “units of corn”.)

Country B is slightly better at producing corn, while country A is slightly better at producing cars – that is, country B has a comparative advantage in corn and country A has a comparative advantage in cars. David Ricardo showed that in such a situation both countries will benefit from specialization and trade. (As mentioned, there is something wrong with his argument, but we’ll turn to that below.)

If in country A 14 units of labor produce cars and 6 produce corn, and country B does the opposite, then the total production of cars and corn increases, and if the two countries trade their surpluses than both will get richer. Here are the numbers:
In country A: 6 units of labor produce 24 corn, and 14 units of labor produce 56 cars. (Productivity is identical to the previous example: 4 units of production per unit of labor in case of either commodity.)
In country B: 14 units of labor produce 63 corn, and 6 units of labor produce 21 cars. (Productivity, again, is unchanged from the previous example, but differs between commodities: productivity of corn is 4.5 units per unit of labor; that of cars 3.5.)

The total (world) production of corn is 87 units (up from 85 in the previous example); that of cars is 77 units (up from 75). The ratio of corn to car production is approximately 1.1, and if the two commodities are traded, one would expect them to be traded at approximately that ratio. If the two countries trade 15 cars for 17 corn (according to the approximately 1.1 ratio) then:
– country A ends up with 41 corn and 41 cars (up from 40 and 40); and
– country B ends up with 46 corn and 36 cars (up from 45 and 35);
and consequently, both countries benefit from specialization and trade indeed. Free trade makes everyone richer. Or does it?

Graham’s correction

It should, perhaps, be noted here that the model thus far assumes specialization followed by free trade, but that this order is by no means necessary. Rather, free trade may lead to specialization because if country A and B open up their borders without prior specialization, then the more expensive cars and cheaper corn from country B will effectively force that country to specialize in corn, while the reverse applies to country A.

This is not what’s wrong with Ricardo’s model, however. Rather, the problem is that the standard model assumes that productivity is constant, while Ricardo was actually well aware that it is not. Elsewhere Ricardo argued that farmers always use the most fertile land first and that if they are for some reason forced to expand and make use of less fertile land, then productivity declines. This is nowadays usually called “diminishing returns to scale”. The opposite effect, called “increasing returns to scale” also occurs: it requires less labor per car to produce 1000 cars than to produce only one car. This is the reason why most commodities are mass-produced.

Agriculture, fishing, mining, and several other (mostly primary) industries are characterized by diminishing returns. The best land, best fishing waters, best ore layers, and so forth are always used first, and expansion to lesser land, lesser waters, and lesser ore will reduce the overall productivity. Conversely, reducing production will generally increase productivity, because the least productive land, waters, and ores will be taken out of production first.

Manufacturing industry, on the other hand, is characterized by increasing returns. Mass producing things is almost always cheaper in manufacturing. The extreme case is (probably) software: the first copy of a new program costs a lot to produce, but producing additional copies costs almost nothing. (In case of national economies rather than single factories increasing returns are a bit more complicated and are mostly driven by feedback loops between industry, supply chains, education, and spill-overs.)

What Frank Graham showed is that these effects need to be taken into account when modeling the effects of free trade.

If country A specializes in cars, then its productivity in cars will increase because car production has increasing returns. And its productivity in corn production will also increase because corn production (as an agricultural sector) has diminishing returns, and therefore, reducing production and concentrating on the most fertile land increases the average productivity. Hence, county A will not produce 24 corn (with 6 labor) and 56 cars (with 14 labor) as suggested above, but (for example) 27 corn and 61 cars.

If country B specializes in corn, then it productivity in corn will decrease because corn production has diminishing returns, and its productivity in cars will also decrease because car production has increasing returns. Hence, country B will not produce 63 corn (with 14 labor) and 21 cars (with 6 labor) as above, but (for example), 58 corn and 18 cars.
(A technical note: increasing and diminishing returns are here modeled as L¹⁺ᴿ×F in which L is units of labor, R is –0.25 for corn and 0.25 for cars, and F is the value that results in the formula returning the same numbers as in the first example in case of a 10/10 division of labor.)

The ratio of total (world) production of corn and cars in this scenario (85/79) is still close to 1.1, but a bit lower than before, and the two countries could trade 15 cars for 16 corn. The result is that:
– country A ends up with 43 corn and 46 cars (up from 40 and 40) or a total of 89 units of consumed commodities (up from 80); and
– country B ends up with 42 corn and 33 cars (down from 45 and 35) or a total of 75 units of consumed commodities (down from 80).
In other words, while country A benefits from specialization and trade, country B does not. In the contrary, country B gets poorer rather than richer.

Playing around with the numbers doesn’t really change this conclusion: if one country specializes in commodities with increasing returns, and another country specializes in commodities with diminishing returns, then the first country gets richer and the second gets poorer.

There are, of course, two other possible combinations. If both countries specialize in commodities with increasing returns, then both countries will benefit from specialization and trade. (For example, if corn would have the same increasing returns as cars, both countries would increase their total consumption to 84 or 85 units.) If both countries specialize in commodities with diminishing returns, then – unless specialization is extreme – positive and negative effects more or less cancel each other out and the effects on the two economies are negligible. (But if the two countries have 3/17 divisions of labor, rather than 6/14 as in the above examples, then the negative effects are stronger, and both countries get poorer.)

The conclusion from all of this is obvious: a country should not open to free trade before it has the capability of specializing in an industry (or in multiple industries, preferably) with increasing returns. (Because specializing in diminishing returns industries will make you poorer rather than richer.) This is hardly a new insight, however. In the contrary, it was well known to Mercantilist economists and related schools of thought (such as National System economics and the German Historical school) from the 16th century onward. Ricardo’s argument was exactly intended to prove the Mercantilists (etc.) wrong. But they weren’t wrong, … or at least not in this respect.

Mercantilism and colonialism

That the Mercantilist emphasis on building up a competitive industry (before opening up to trade) works is also shown by economic history. Under the influence of variants of Mercantilist economics, Great Britain closed its borders to trade and built up its own manufacturing industry. Only after British manufacturing had grown to become sufficiently large and competitive did trade policy change and did the government promote (mostly) free trade.

Significantly, Adam Smith and David Ricardo – two of the most influential enemies of Mercantilism and defenders of free trade – were British economists living and working in this transitional period from Mercantilist protectionism to free trade ideology. That they became so influential is largely because Great Britain’s economy was now sufficiently developed to profit from free trade. In other words, Smith’s and Ricardo’s ideas were useful to the economic elite to “scientifically” defend their material interests.

The United States, under the influence of Friedrich List’s and Alexander Hamilton’s National System economics which was closely related to Mercantilism, followed a similar path. Protectionism reigned until a sufficiently competitive manufacturing industry had established itself, after which free trade ideology was adopted.

In Ricardo’s example, the two countries were England and Portugal (rather than “A” and “B”), and trade between those two in the commodities of his example was almost free (i.e. there were low tariffs). The results were disastrous for Portugal, because it killed off Portugal’s small manufacturing industry, while it gave an economic boost to England. In other words, Ricardo’s example – in actual, historical practice – followed the third scenario (i.e. that based on Graham’s variant) rather than Ricardo’s, and free trade made Portugal poorer.

When Great Britain took control of India, it destroyed Indian manufacturing industry and forced India to focus on much more primitive diminishing return industries. The products thereof where then transported to Great Britain, where they were the inputs for England’s burgeoning manufacturing industry, which sold some of its products back to India – products that the Indians used to make themselves. While England benefited greatly from this arrangement, India – being forced to give up what little manufacturing industry it had – involuntarily steered a course for poverty and underdevelopment. Other colonial powers followed the British lead and forced their colonies to focus on diminishing return industries such as agriculture and mining, and the IMF and World Bank continue this policy until this day. The motivation has changed, however. While colonial powers denied their colonies a manufacturing industry for entirely selfish, Mercantilist reasons, the IMF and World Bank enforce free trade and government austerity on the developing world because they believe in the mainstream economic dogma that free trade makes everyone richer.

The effects are the same, however, the former colonies – now “developing countries” – are not allowed to (actually) develop. They are effectively forced to be like country B in the third example above (or even to be extreme variants thereof, committing its full labor force to the equivalents of corn), while the rich, industrialized countries are more like country A. And as in that example, the rich countries benefit from free trade, while the poor countries don’t. (If colonialism is defined by denying the colonies a manufacturing industry and independent economic policy, then the world-order is now more colonial than ever.)

The hypocrisy is stunning. The West doesn’t allow the “developing world” to use the exact same policies that they used themselves to get rich.

According to Ricardo’s pseudo-scientific theory, everyone benefits from free trade. This idea became a dogma of mainstream economists and their political allies (centrists, neo-liberals, as well as many other liberals and conservatives). But the dogma is false, and this has been shown in actual practice over and over again.

Look, in addition to the examples above, at the countries where IMF and World Bank took over after the fall of “communism”, for example. De-industrialization lead to economic decline and poverty. And in the EU, similar things happened, albeit on a slightly less devastating scale. Dictated by Germany, neo-liberalism, and mainstream economics free trade ideology reigned (and continues to reign) and countries were (and are) forced (either directly or by “the market”) to specialize in their comparative advantages. For Germany, with its strong and advanced manufacturing industry, this worked out great. For Southern Europe, … not so much. Spain, Portugal, Italy, and Greece de-industrialized, and their governments – having their hands tied behind their back by EU-enforced austerity and free trade rules (including restrictions on subsidies any other kinds of interference with the economy) – couldn’t do anything about it.

Ideology and democracy

The term “ideology” is used with (at least) two different meanings. It can refer to a more or less coherent set of beliefs about how society should be organized and what the scope and aims of government should be. And it can refer to the set of values and beliefs adhered to and defended by a social class (in the broadest sense of “class”) and that serves the interests of that class. Thus “ruling class ideology” (or “dominant ideology”) is the set of values and beliefs that serves the interest of the ruling class. Usually, when the term “ideology” is used in this second sense, it refers to the values and beliefs (serving the interests) of the socio-economic and/or political elite (or “ruling class”) particularly. To de-mask ideology (in this sense), ask yourself the question “Whose interests are being served?” by some particular value or belief. If the answer is that that value or belief (and/or your adoption of that value or belief) serves the interests of those who are already in preferential positions (that is, the rich and/or powerful), then you’re dealing with ideology.

Free trade ideology is ideology in this sense. In the third example above, country A benefits from free trade, while country B gets poorer. This is because country A focuses on the production of cars and country B focuses on the production of corn. In other words, the industrialized country benefits from free trade, while the agricultural country doesn’t. It works the same way in the real world. If all the countries involved in free trade would be industrialized countries, then – assuming that the model is right – they would all benefit from free trade, but countries that have no competitive manufacturing industries and that are forced to focus on sectors with diminishing returns do not. In other words, free trade makes rich countries even richer, and poor countries even poorer. So, to ask the question mentioned above: “Whose interests are being served?” Well, obviously, the interests of rich countries (and particularly the rich and powerful in those countries). So indeed, free trade ideology is ideology indeed.

As is the case with most ideologies, not everyone adheres to free trade ideology, however.

While the insight that a manufacturing industry with increasing returns is essential for economic prosperity seems to be forgotten by most 20th (and 21st) century economists, it is – much to the chagrin of mainstream economists – not completely forgotten by politicians and by the “general public”. Many governments that officially adhere to free trade ideology simultaneously use various policies to protect infant industries and other sectors of the national industry that politicians and governments consider vital to national interests (except, of course, when they are not allowed to by international agreements, the World Bank, and/or the IMF).

According to Bryan Caplan (2006) most citizens of democratic countries have “systematically biased beliefs about economics”; namely, most citizens favor Mercantilist (or Mercantilist-like) economic policies rather than the policies prescribed by mainstream economics. Because Caplan uncritically subscribes to mainstream economic dogma, his conclusion is that the public is ignorant and wrong (or “systematically biased”). However, given that mainstream economics is completely and disastrously wrong about the effects of free trade (as shown above, as well as in a vast literature) and Mercantilism is at least partially right, it turns out that the general public actually favors more sensible economic policies than professional economists. (Ha-Joon Chang (2010) also argued that you don’t need economists to have sensible economic policies. It turns out that he is right.)

Jason Brennan (2016) builds on Caplan’s flawed analysis to argue against democracy and for “epistocracy”, a political system in which only the most knowledgeable have the right to vote (or otherwise influence policy). (Caplan’s and Brennan’s books fit in a long tradition of mainstream economists arguing against democracy, partially because it is not enough like a market for their preferences, and partially because it is “irrational” by their standards.) Like Caplan, Brennan uncritically accepts mainstream economics (or actually, he is a mainstream economist and also wrote a couple of books in defense of neo-liberal capitalism and libertarianism, the most extreme form of free market fundamentalism), and completely ignores the fact that the people he considers most “knowledgeable” are actually completely wrong. More importantly, however, he also ignores that something very much like the system he prefers is already in effect.

As shown by Wolfgang Streeck (2016) and others, since neo-liberalism took over in the 1970s political power over many aspects of the economy has gradually been transferred from parliaments (and thus indirectly from the public) to institutes outside democratic control. “Independent” central banks were the keystone in this process, but international agreements and various other legal and institutional changes also took away more and more power to implement economic policy from parliaments. In the current situation, there are very few – if any – countries in which the parliament has any significant power left to make and/or change economic policy. Economic policy has been transferred to institutes outside democratic control, institutes controlled by mainstream economists, neo-liberals, and other believers in the pseudo-scientific dogma of free trade. In other words, when it comes to economic policy, something much like Brennan’s vision has already been realized.

Unfortunately, the results are disastrous. The continuing economic crisis that started a decade ago, poverty in the “developing” world, the economic troubles in Southern Europe, and a long list of other economic problems (in addition to environmental devastation) are not natural disasters but the result of economic mismanagement. If the public favors Mercantilist policies indeed, then they are more knowledgeable about sensible economic policy (perhaps because they are less influenced by ideology) than those who are supposed to be, and a more democratic approach to economic policy would have resulted in a very different world with less poverty and less economic (and environmental?) disaster.

The consequence of the gradual retraction of the economy from democratic control is that free trade ideology has become virtually untouchable. It has been embedded in institutions on which politicians and citizens have hardly any influence, and it is defended by layers of ideologically motivated obfuscation and pseudo-science that are continuously reinforced by mainstream economics (which is more like a religious sect than a science) and the perverted “common sense” they have managed to create.

But free trade is not a panacea that solves all problems and automatically makes everyone richer. As Frank Graham showed, if Ricardo’s simplistic model is made just a tiny bit more realistic by taking into account that productivity changes with levels of production, then it turns out that some countries (and regions) do not benefit from free trade, and even get poorer as a result of free trade. In general, a rich country with a well-developed manufacturing industry will benefit from free trade, but a poor country that does not have such an industry and that depends mostly on agriculture and mining will not. Rather, that poor country will be denied a chance to built up a manufacturing industry and will only get poorer. Free trade makes the rich richer and the poor poorer. Claiming that it benefits everyone is an ideologically motivated lie.


  • Brennan, Jason. 2016. Against Democracy. Princeton: Princeton University Press.
  • Caplan, Bryan. 2006. The Myth of the Rational Voter: Why Democracies Choose Bad Policies. Princeton: Princeton University Press.
  • Chang, Ha-Joon. 2010. 23 Things They Don’t Tell You About Capitalism. London: Penguin.
  • Graham, Frank. 1923. “Some Aspects of Protection Further Considered”, The Quarterly Journal of Economics 37.2: 199-227.
  • Ricardo, David. 1817. On the Principles of Political Economy and Taxation.
  • Streeck, Wolfgang. 2016. How Will Capitalism End?: Essays on a Failing System. London: Verso.

Leave a Reply

Your email address will not be published. Required fields are marked *