# Evidence on the Heckscher-Ohlin Model: Does It Work in Reality?

**Channel:** Deep Learning Economics
**Source:** https://www.youtube.com/watch?v=dhSBOmKfwj0
**Transcript page:** https://www.withtranscript.ai/video/dhSBOmKfwj0

## Chapters

- 0:01 — Leontief Paradox and Initial Evidence
- 3:59 — Explaining the Leontief Paradox
- 8:22 — Cross-Country Skill Intensity Evidence
- 11:23 — China's Trade Pattern Over Time
- 13:47 — Heckscher-Ohlin Model Conclusions

## Transcript

**[0:01]:** Here, let's check out the empirical evidence for the actual only model. And uh the first uh piece of empirical evidence is so-called the Leontief paradox. Okay? So, here um we are looking at the factor content of the US exports and imports for the year 1962. The first line we see is uh amount of capital, so here means machinery equipments, tools used to produce $1 million worth of uh US imports and exports. Okay? And the second line shows the amount of labor used to produce $1 million worth of uh US import and exports. Okay? The third line we are uh trying to find the capital-labor ratio. In other words, dollars of uh capital per worker. So, here we use the first line divided by the second line. That's how we get this ratio here. Okay? Now, um if we assume that the US is better capital endowed, okay?

**[1:25]:** Then um labor. According to the actual model, we should export more capital-intensive goods, right? And import uh more labor-intensive goods. Because that capital-intensive goods uh are, you know, where our comparative advantage is, right? However, here by comparing these two numbers, we find that the capital-labor ratio is actually lower among the US exports than our imports. In other words, based upon these two numbers, uh the evidence shows that we are actually export labor-intensive uh goods and import capital-intensive goods. Okay? So, this would be against the actual model, right? Now, um over the years, economists come up with um several possible explanations okay, for this paradox, okay, for the evidence which is not support uh uh supportive of the actual model. And uh the first thing we look at is the same technology across countries. That assumption we made um when we discussed the actual model, okay?

**[2:59]:** We said that the technology is assumed to be the same across countries. And um in reality, it is uh it could be the case that uh US has much more technologies or more sophisticated technologies than other countries. So, what we export is um instead of the capital-intensive products, we're actually exports uh technology-intensive ones. Okay? And of course, you know, in in this um evidence, the capital-labor ratio does not really capture uh the level of technologies okay, in different countries. So, that could be a missing factor, okay? Uh which is quite important, but not consider not being considered in this case, okay?

**[3:59]:** The second impossible explanation is unusual data used. So, here uh it is possible that, you know, in 1962, that year uh something unique happened, okay? And that uh specifically led to this um um numbers, okay? These numbers. Uh the second I'm sorry, another uh possible uh case would be, you know, here we just focus upon the United States, okay? So, the US might be uh unique in some way, um like different from other countries, okay? However, um I think recently uh economists look at other countries, okay? And find um similar things, okay? In other words, uh in those countries uh which um should be export more like capital-intensive goods, but they actually um the uh the actual number or data shows that they they actually imported more uh capital-intensive goods and exported more labor-intensive goods, okay? In other words, the US is not ice isolated case, okay?

**[5:28]:** Um the third uh which could be the most um convincing explanation is uh here we fail to distinguish between the skilled and unskilled labor. So, we just say, you know, labor per million dollars, right? Or capital-labor ratio, but we did not um you know, break down and look at the skilled versus unskilled or no skilled, right? And uh we get this um idea by looking at the last two rows in the table. So, here um it shows the average years of education per worker, and we do find that among the US exports, this number is larger than our imports. In other words, you know, we do find that um among our exports, there are more um skilled labor-intensive ones than our imports, right? the last line is uh proportion of engineers and scientists in workforce, okay? And then we find this number is higher among our exports than our imports, okay?

**[6:44]:** Which is consistent with again the actual model and the assumption that uh in the US, we have more relatively more um R&D capacities, okay? Or uh resources. Okay? And um Now, um we already check out this chart, right? Uh during our virtual meeting, we said that, you know, uh on the far right, and uh we put the GDP uh share here as a benchmark to compare. And we do find that, you know, um the US have has um better endowed in physical capital, R&D scientists, and skilled labor, right? And if you pay more attention to these first three uh types of resources, uh which the US is better endowed, you do find that the physical capital, relatively speaking, um is not as much endowed as the R&D scientists and skilled labor, right? So, that's why um some economists would argue that, you know, we shouldn't be surprised with the results of the Leontief paradox, okay?

**[8:08]:** Because the physical capital is not really uh the the the strongest or the most endowed uh resources we have, okay?

**[8:22]:** All right. Now, let's take a look at um the second piece of empirical evidence, okay? Which shows the skill intensity and the US ex- US imports from the following countries, okay? Now, uh two things here we need to explain before we discuss the um you know, the the figure, okay? The first one is we divide um the US imports into uh five groups according to the skill intensity, okay? So, um these five groups are least skilled, second least skilled, sec- second most skilled, and uh most skilled. I'm sorry, four four groups, right? Four groups um based upon the skill intensity, okay? So, um on the graph, you know, the green bars shows the least skilled, okay? And the purple one is the most skilled, okay? And um also based upon our in you know, these six trading partners, we we divided them into two groups, okay?

**[9:35]:** The first group which uh are demonstrated on the left-hand side of the figure are Bangladesh, Cambodia, and Haiti. And on the right-hand side, you can find France, Germany, and UK, okay? So, once again, based upon the um the common sense, we would believe that uh, on the left-hand side, these three developing economies are better endowed in labor, more specifically unskilled labor, right? And on the right-hand side, these uh, three developed European economies are better endowed in uh, skilled labor. Okay? And so, here we look at, you know, the uh, height of these bars shows uh, the share of that group of products we imported in the total uh, imports from that country. Okay? You find that here, Bangladesh, Cambodia, and Haiti, um, you know, the our imports from these countries are actually dominated by the least skilled um, sector or or products. Okay? And on the right-hand side, uh, from France, Germany, and UK, uh, we imported predominantly the most skilled uh, products.

**[11:09]:** Okay? So, this piece of the um, empirical evidence is actually in favor of the actual model. Okay? It's pretty consistent with what the model predicts.

**[11:23]:** All right. Now, after checking out the cross-country um, evidence, let's focused upon one country, China, here, and look at the over time um, evolution. Okay? Of its comparative advantage. Now, here, once again, we divided um, all the US uh, imported uh, goods from China into four groups on the basis of their uh, skill intensity. Okay? Um, it's the same as the previous one. Um, but the difference is we just look at China in different years from uh, 1983 uh, to 2012. Okay? Now, uh, according to the actual model, um, because of the economic takeoff um, in China, so it should have, you know, higher and higher income. And then, they should spend more resources um, in their public education. So, their labor force over time should be better and better educated. They should get uh, more and more skilled. And so, uh, based upon the HO theorem, then we would expect that uh, the US import of the low-skilled or least skilled uh, products from China would gradually decline.

**[12:57]:** Which again is uh, exactly what this chart shows. Okay? And uh, the US imports of relatively high-skilled or even most skilled uh, products from China should gradually uh, increase. Again, roughly speaking, that is true. And we do find an outlier here uh, from '88 to '92 because of the political um, uh, factor or influence in that uh, period gets a lower one, but again, overall, the trend of uh, US imports of these uh, most skilled products from China are actually going up. Okay? So, this is also uh, in favor of the actual model. Okay?

**[13:47]:** Now, here, let's quickly draw the conclusion of this chapter. We discussed the HO model, which is uh, two products, cloth and food, two factors, uh, labor and capital, and two countries, home and foreign. So, that's what we call 2 by 2 by 2 model. Okay? The key difference between the HO model and the previous models is here, every factor is mobile. Okay? And the the key conclusion is number one, the pattern of trade, um, which is uh, specified in the actual theory. Okay? Basically, we're saying that every economy exports its abundant resources. Okay? And then, uh, the influence upon the income distribution uh, is the Stolper-Samuelson theorem. What we're saying is, you know, the owners of the abundant factors would benefit from trade, and the owners of the scarce uh, factor would uh, hurt by trade. And we check out the empirical evidence and find it's mixed.