.png)
Economic research subjects often cause confusion: for example, what is the benefit of studying now the events of a distant past? Scholars are doing this research not for the sake of finding solutions that worked well dozens or hundreds of years ago – those are really useless today, but for the sake of answering another, less obvious question. Konstantin Egorov, a professor of the University of Antwerp and a New Economic School alumnus, tells what that question is.
Economic research very often seems to be out of touch with real life. Instead of practical and pressing issues, for some reason they delve into the details of Finnish reparations to the Soviet Union after World War II, the continental system of Napoleon I, or the transition of shipbuilding from wooden to steel vessels. It is clear that such events are unlikely to happen again, and therefore, it would seem, they should not be of any interest to anyone except for the most dedicated specialists. Why, then, do most economists insist that such research help them answer relevant questions much better than even the most modern experiments?
In short, the key issue is that in economics, even very reliable results often cannot be replicated in real life. For example, after Donald Trump’s first presidential term, many great papers have been published about the effects of his tariffs on the economy. Now the situation seems to be repeating itself: the same president is imposing the same tariffs. In this case, can we expect the same results? Unfortunately, not. This time, Trump has apparently managed to abuse foreign consumers so much that even fairly friendly Canadians began to boycott American goods, regardless of how much more expensive they had become. In other words, the same measures towards the same countries after only a few years have much stronger effects: the situation has changed so much that the previous outcomes cannot be expected.
Even the most reliable results, i.e. the ones based on experiments, are not immune to this issue. In one of my favorite examples, weaker students greatly benefited from being next to stronger ones when they were placed in the same class within a research project. But when the university administration began to purposefully pursue such a policy, the strong students, apparently, recognized their goal and stopped communicating with the weak ones, negating the positive effects of such a proximity. It turns out that even the best experiment that one can have in an economic study cannot be reproduced in real life simply because it is not possible to accurately reproduce all the same conditions.
This fundamental difficulty of conducting truly refined experiments in economics is not only disappointing, but is also the reason why many representatives of STEM disciplines do not (and quite rightly) consider economics to be a science in the same sense as, for example, physics. This is partly why, instead of close-to-reality and practical research and experiments, economists have to deal with subjects which they think can give them a more reliable and reproducible answer. And to do this, they often need to ask much narrower questions.
One of the almost eternal and very important questions in economics is the question of the effectiveness of the so-called industrial policy. Many believe that it helped the wealthiest countries of the 19th century to get to that level of prosperity. Episodes of the most impressive growth of the 20th century, including Japan and South Korea, are often attributed to it too. This topic is still hot today.
Although scholars do study all proven industrial policy measures, these measures are still fraught with the same reproducibility problems as studies of Trump's tariffs. Each solution is so complex and depends on a such large number of specific features that even if it is successful, one cannot even dream of accurately reproducing all the conditions. Instead, economists have to put up with testing each of them individually.
The first and simplest question is whether there are any industries where temporary government intervention can pay off after a few decades? It is clear that this is possible. Moreover, it is not hard to find nice examples of such industries, i.e. industries with the necessary characteristics. But the key and least obvious issue here is quantitative, namely: are those characteristics sufficient to pay off the initial costs? This is the very narrow question that economists had to work with instead of providing a ready-made solution for a successful policy.
Interestingly, it is often more difficult to answer a more modest and narrow question than an ambitious one. To test just one of the necessary conditions of an industrial policy, you must knowingly fulfill all the others. It turns out that economists need to find cases when an industrial policy is so perfect that it is very difficult to doubt any of its components, except for one. And this is precisely for the sake of finding such improbable cases that economists dig into the farthest corners of dusty archives, instead of collecting data on more real-life policies that actually interest them.
One of these incredible and perfect cases happened in Finland at the end of the Second World War. The USSR demanded that Helsinki pay reparations, but historical circumstances have developed in such a way that these payments 1) were limited to eight years and 2) were made not with money, but with a set of certain industrial goods. In order to guarantee their production, the Finnish government began to help industries included in the Soviet list.
It was a lucky coincidence that the basic conditions for a successful industrial policy were there. The main threat of any such policy is that a protected industry, having grown strong thanks to government support, is lobbying for permanent protection rather than temporary one. Then you will get a product that is not truly competitive in world markets, but is in demand only when the government helps out. This will mean the failure of industrial policy. And in this particular case, the scholar was «lucky» that the Soviet Union wanted from the Finnish government not the protection of certain industries, but goods of proper quality. In other words, it created incentives not to put the excess profits from government support into owners’ pockets, but to invest them in production modernization. Moreover, these incentives disappeared after eight years, regardless of the decisions of the Finnish government.
The reparations terms excluded another factor that is a common cause of a failure of industrial policy – the subjective choice of industries. Governments often support industries that are «close» to them instead of the most promising ones (it doesn't even matter what are the reasons – corruption, national security, or other). Japan and South Korea seemed to choose the most promising industries, striving to make a leap in development. But even in their case, it is now very difficult for us to figure out to what extent their growth is due to industrial policy, and not to other efforts aimed at the same goal. In the case of Finland, little depended on the country itself: the USSR not only chose which goods it needed, but also, apparently, was not particularly interested in the capacity of the Finnish economy at that time. As a result, industries with no special ties to government or growth prospects received support.
Anyway, after the payment of reparations and the end of state support, many of the sectors of the Finnish economy chosen by the USSR continued to develop faster than others and faster than the same industries in neighboring Norway. It turned out that in the post-war Finnish economy, quite randomly there were many industries that could pay off a successful industrial policy.
What makes this case more reproducible than the case of Trump's tariffs or putting weak students together with strong ones? One may think that it should be exactly the opposite, because Trump was re-elected for a second term and repeated his tariff policy, while the Finnish economy, even if it has to pay reparations again, has already changed beyond recognition.
Indeed, the surface-level and fairly clear finding about the effectiveness of the solution in each of the mentioned studies is not so important. Economists value them primarily for the less obvious result of testing just one part of the solution. In other words, the typical economist, like the average layman, didn't care about Finnish reparations. But this research gave a convincing answer to the narrow question mentioned above: it showed that there can be quite a lot of industries that can pay off a successful industrial policy. And this very finding turned out to be reproducible in a wide variety of conditions.
In particular, the well-known research on the impact of Napoleon's continental system on the development of French economic sectors is both similar to and different from the research of the Finnish case. They are both dedicated to an episode of distant history that very few people are interested in. At first glance, this makes them completely different from each other, or from any situation that we may face in real life. But for economists, this study essentially replicates the one about Finland. It is just that instead of temporary reparations, French industries faced a temporary blockade, which in fact protected them from British competition. And instead of industries that were randomly selected by the Soviet Union, the blockade protected the north of the country more. The smugglers still carried British goods to France, but they couldn’t use the shortest northern route. Instead they had to take a roundabout route through the south. In both regions, British goods rose in price due to the increased delivery risks, but in the north the price rise was higher due to logistical costs: the route to get there became the longest.
This research shows that the textile industry, which was equally well represented in both the north and south of France before the blockade, after its start has gone through greater modernization and development in the more protected north. Moreover, this gap persisted until the end of the 19th century. The main advantage of this research is that it showed the existence of industries suitable for industrial policy.
But it cannot be used as a standard solution for the development of the textile industry in general. This was confirmed, in particular, by a research on the blockade of the Chinese textile industry during the First World War. The case was generally similar, but this time there was no modernization. Apparently, because the industry has advanced a lot over the century and it was not possible to modernize it without imported equipment that was unavailable during the blockade.
For the same reasons, the research on the shipbuilding industry in the late 19th and early 20th centuries is valuable. It describes a relevant industry in completely different conditions. In the era of wooden ships, the United States dominated the shipbuilding industry, but the transition from wood to steel gave a huge advantage to England where steel prices were much lower. This advantage disappeared at the end of the 19th century with the discovery of new iron ore deposits in North America. But even this temporary support was sufficient, and the British industry retained its advantage.
The reproducibility of such cases somehow resembles a formula of the origin of life in the universe. Some say that on Earth, life appeared thanks to a combination of extremely
