I would not say that economists prefer Heckscher-Ohlin to Ricardian. In the old days (70s) people work mainly on Heckscher-Ohlin because it specifies exactly where comparative advantage comes from (endowments). At that time people thought that the productivity differences of the Ricardian model were way too vague and not specific enough. Tides have turned, however. Eaton and Kortum (2002) put the Ricardian model on the center of everybody's screen.

All trade theories deal with relevant aspects. They should not be seen as competing but rather as complementary. Therefore Seev Hirsch (1967) merged product cycle theory with Heckscher-Ohlin theory. Jones 1970) mergerd HO and Ricardian theory. Helpman and Krugman (1985) merged monopolistic competition and HO theory. Trefler (1995) combined differences in technologis, endowment and preferences. If you choose between any too or more you ignore relevant arguments of those you choose against.

Thomas raises a very important conceptual point about economic theories, which is that they attempt to give us an insight into a real world process. At issue here is the source of gains from trade. If you ask many non-economists, you might find that they do not see gains from trade but instead see one side as winning and the other as losing. This is, of course, nonsensical from an economist's perspective unless the trade was coerced or carried out under false pretenses or under conditions of uncertainty (selling a counterfeit good or a 'lemon'). We can start with a two-good Edgeworth Box diagram and random endowments and identical preferences and get gains from trade. This doesn't tell us that it is a good theory, but it provides a useful insight. Ricardo wanted to show that there were gains from trade even if Country A had an absolute advantage in producing both good X and good Y over Country B. Comparative advantage is a source of gains from trade. Another useful insight but not a complete theory of trading. H-O theory considers relative abundance of capital and labor. Adam Smith's division of labor as a source of wealth, which requires trade to realize such gains. None of these theories, arguably, come fully close to explaining real world trade flows but each of them has an insight. These insights are essential in standing up to bad policy ideas because policymakers rarely consider things that are unseen (hats off to Bastiat on that one) and these insights help us argue through the unseen consequences of policies.

I believe that the reason for this preference is that Ricardo's statement of comparative advantage has been generally misunderstood. As recently as 2002 – 185 years after Ricardo's formulation – Ruffin came up with the correct interpretation of the famous four numbers in chapter 7 of the Principles. Ruffin (2002) convincingly argues that Ricardo’s numbers are not unit labor coefficients, but rather the quantities of labor needed to produce some unspecified amounts of wine and cloth traded by England and Portugal. Based on this interpretation of the numbers, I have argued that most of the criticism towards the Ricardian model of economic textbooks – including the constant labor cost assumption – does not apply to Ricardo's original formulation of comparative advantage. See attachment.

Preprint Comparative Advantage and the Labor Theory of Value

Chong Xiang talks about Eaton and Kortum (2002). Although I strongly support Ricardian theory, I have to say that Eaton and Kortum's model contains a crucial defect in it. I have discussed it on another question page:

Why did Eaton and Kortum model perform so badly?

https://www.researchgate.net/post/Why_did_Eaton_and_Kortum_model_perform_so_badly?_tpcectx=profile_questions

 As Derek Anderson points it, Eaton and Kortum model is more directed for empirical study and admits too ad hoc a hypothesis on its treatment of inputs.

 However, Ricardian (or Ricardo-Sraffa) trade theory has now a very general theory and it is welcome that somebody will try to use the new theory in empirical works. (See the two of my papers below.)

 Hak Choi misunderstands that Ricardo's theory is based on labor theory of value. It is true that Ricardo talked much about pure labor input case. He also talked the case where the ratios of direct labor and indirect labor (material inputs and tools and machines) are equal between industries. However, he knew very well that in general this assumption does not hold. His true theory is the cost-of-production theory of value. In this cost, Ricardo included the (normal) profit. See Ricardo (1951) The Principles of Political Economy and Taxation (Sraffa edition), the original note, p.47. Library of Economics and Liberty, Chap. 1. Note 7.

My theory on Ricardo-Sraffa trade economy does not depend on labor theory of value in any sense and still is based on the tradition of the classical value theory.

See also

The Revival of Classical Theory of Values

https://www.researchgate.net/publication/269393496_The_Revival_of_Classical_Theory_of_Values 

and 

On Ricardo's Two Rectification Problems

https://www.researchgate.net/publication/266737921_On_Ricardo%27s_Two_Rectification_Problems 

As for Heckscher-Ohlin model, it has too many conceptual and empirical defects.

Conceptually, HO model assumes that 

(1) every country has the same technology, and 

(2) factors of production are assumed to be given.

 Assumption (1) implies the factor price equalization theorem holds, which is not only false but positively harmful, because this conception disfigures systematically the trade and industrial policies. 

Assumption (2) is common for all General Equilibrium Theory, but capital goods (tools and machines) are accumulated according to the necessity of production. Overlooking this fact leads to a wrong policy for developing countries. Importation of capital goods is very important for the further economic development. HO theory gives no analysis on this crucial question. 

Empirically, it is notorious that HO model and more general HOV model badly fit to empirical data. The most well known result is Leontief paradox. Leamer's and Trefler’s works made clear that unadjusted HOV model performs no better than a simple toss. Trefler argued that the model fits well to the factor prices if he takes in consideration the differences of productivity. However, this is but a proclamation of the HO models' defeat vis-à-vis Ricardian theory.

Conference Paper On Ricardo's Two Rectification Problems

Conference Paper The Revival of Classical Theory of Values

Very interesting points Yoshinori. Thank you for sharing with us the conference papers.

Dear Konstantinos, 

I have recently finished the revision of a working paper which claims that Ricardo's original numerical example does not rely on any of the unrealistic assumptions usually associated with the textbook trade model of comparative advantage. In addition to this, it also highlights some other important differences between the two. Despite its misleading association with Ricardo's name, the current version of the textbook trade model is in fact the result of significant misinterpretations of the numerical example in the Principles.

Best, 

Jorge

Thomas Zeisemer and Jon Ryding seems to be relativists in assessing different theories. They emphasize that theories are complementary rather than competing.

I admit that some couples of theories are often complementary, but I believe that the HO vs. Ricardo case does not belong to this category.

In my opinion, technology is generally more important than resources  except for countries which are very rich in oil beds, some other mineral mines and others. There are countries that can gain enough foreign currency by the export of its crude oil or iron core and can cover all the expenditure of importing almost all consumers goods such as passengers-cars and home appliances. These are rather rare cases. Many other countries have to find some industries whose products are competitive and exportable. Here come the questions of technology.

What happens when a theory gives an insight that technology is more relevant in economic development than resource endowments?

When we want to examine the questions of technology in relation to economic development in the context of international competition, Ricardian trade theory is much superior than HO theory (or HOV [Hechscher-OHlin-Vanek] theory). The Ricardian theory supposes that countries have their own technologies expressed by the production function or production frontier. In the HO case, the "competition" is made between countries. However, this is a very bad representation of the reality and gives an erroneous insight to consider questions of the technology.

Who is the possessor of a particular knowledge (technique) of running a certain production process in a good way? It is firms and not countries. Competition occurs between firms and not between countries. The difference of countries are important, because the wage rates are sometimes very different (10 to 20 times) by countries.

Then, the theory of international trade which can contribute to the consideration of technology problems should explain at least two problems: (1) the differences of wages and how they are changed by the technological development, and (2) how each firm is struggling in order to compete foreign firms.

HO or HOV theories explain neither (1) nor (2). In comparison to this, Ricardian trade theory, now extended to includes many country, many-commodity case with choice of techniques and input trade, is well adapted to explain (1) and (2).

As for point (1), I have cited two of my papers in the post of two weeks ago. As for the (2), please see the paper of mine co-authored with Professor Fujimoto:

Inter and Intra Company Competition in the Age of Global Competition: A Micro and Macro Interpretation of Ricardian Trade Theory.

Fujomoto is famous in his empirical study in world-wide automobile industry and others. The main theme came from Fujimoto as he is always observing how Japanese firms are facing international competition (from China in particular).

As Konstantinos has written a paper

Ultra-organizational Co-opetition Dynamics: Designing micro-foundations of organizational performance,

I hope he will be interested in our paper.

He may also find two applications at the end of another paper of mine (and a presentation data in a conference): The Economics of the Great Unbundling.

https://www.researchgate.net/publication/233943529_Inter_and_Intra_Company_Competition_in_the_Age_of_Global_Competition_A_Micro_and_Macro_Interpretation_of_Ricardian_Trade_Theory?ev=prf_pub

https://www.researchgate.net/publication/272943681_The_Economics_of_the_Great_Unbundling

Article Inter and Intra Company Competition in the Age of Global Com...

Conference Paper The Economics of the Great Unbundling

• I f factor endowments are sufficiently much different between countries in HO theory, there will be perfect specialization and no factor price equalization. The country with the lower relative labour endowment will have higher wages. The zero profit situation indicates the struggle. Both models are just first principles. The competition and innovation issues are much clearer in the combination with endogenous growth. See Grossman/Helpman book 1991 or the articles. 

Dear Thomas,

of course, the factor price equalization theorem does not hold out of factor price equalization cone, but is it a good theory to analyze wage differentials between countries?

You have mentioned Jones (1970). I suppose it is the paper with the title "The role of technology in the theory of international trade."

In section I, Jones examines "the general case in which production functions are allowed to differ between countries." (p.74) This is already a bit different from the standard HO model, which assumes the same technology for both countries.

In section II, Jones examines what he calls modified Heckscher-Ohlin model. He willingly admits that his modified  "Heckscher-Ohlin model represents a step backward from the earlier Ricardian tradition." (p.78) It does not seem to me that Jones examines the case of perfect specialization. What Jones examines as his "modified HO model" is the case where "the two countries differ only slightly in their technologies." (p.83) It is quite doubtful if Jones examined the case of "complete specialization". He only observed slightly modified state of the factor price equality case. By his method, he cannot examine the case where the wages differs in a large extent (workers of a country gains in the same working time ten to twenty time more than workers of the other country).    

There are another serious defect for the HO model. What is more questionable for the HO model is the assumption that factors are predetermined quantities. We can suppose that workers stay in their original country, but how about capital or more precisely input goods (or intermediate products)? They are traded as freely as final goods. Some reports that this part of trade (proportion of input trade) is increasing. As Jones admits,  "the entire concept of factor abundance is subject to question if capital is mobile between countries." (p.84)

I know for example Kar-yiu Wong's book which claims to have integrated the "trade in goods and factor mobility," but I cannot say it is very successful. It is a complicated but ad hoc theory.   

You should know that we have now a very general Ricardian theory (M-country, N-commodity case with trade of intermediates goods and choice of techniques). Why should we stick to a very restricted theory? At last account, what  Jones succeeded  is the two-country, two-goods and two-factors case. It may work as a parable, but theorems do not generally hold in higher dimensional case. What is the merit to adhere to such a restricted model?

I did not answered to the question posed by Konstantinos Biginas. I only pointed the various deficiencies of HO theory (or HOV theory). If my points are granted, the next question is why this deficient theory is preferred by economists than theories of Ricardian (Ricrado-Sraffa) trade economy.

If I am permitted, I would say frankly that HO theory is preferred by a kind of paradigmatic necessity.

(See for this concept my answer [the 5th post in p.13] with a heading

            (1) Paradigmatic necessity of theory

 to the question

• How do the economists define self-interest and rationality?   )    

Many economists have learned macroeconomics and used Solow-Swan type production functions. They cannot abandon HO theory even if they knew that the theory has many theoretical defects, because Solow-Swan type production function has become a vested interest.

https://www.researchgate.net/post/How_do_the_economists_define_self-interest_and_rationality/13 

I have criticized many aspects of HOS or HOV models. Another defect of these models is the lack of recognition that modern manufacturing has a circular structure. As Piero Sraffa has chosen as the title of his book, commodities are made by means of commodities. They do not recognize this important point and imagine that products are transformed from some local ingredients to a final products. There exists such products, of course, but they are rather exceptions.  Majority of exports of developed countries are a result of processing and assembling imported materials. HOS or HOV theories cover these facts and make developing countries concentrate exporting natural resources.

I posed a new question on the concept of "processing trade."

Why did the notion of "processing trade" not become a major concept before fragmentation and global supply chain?

https://www.researchgate.net/post/Why_did_the_notion_of_processing_trade_not_become_a_major_concept_before_fragmentation_and_global_supply_chain

The history of "processing trade" concept reveals a result of HOS or HOV modeling of international trade. My discussion in this question page is not only theoretical, but has a deep policy implication. Without a good critique of theories we risk to remain blinded by a popular theory of the time.         

I have discussed a variant of Heckscher-Ohlin model in the following question  by

Dr. Tahsen Alqatawni: 

What are the pragmatic problems with Heckscher-Ohlin model?

https://www.researchgate.net/post/What_are_the_pragmatic_problems_with_Heckscher-Ohlin_model

I discussed there in many of my posts Joel Hellier's article:

J. Hellier (2012), ‘The North-South HOS Model, Inequality and Globalization’, in: J. Hellier & N. Chusseau Eds, Growing Income Inequalities, Economic Analyses, Chapter 4, Palgrave MacMillan, 107-146.

North-South Heckscher-Ohlin Model is a variant of Heckshcer-Ohlin model adjusted to be applicable for analyzing wage inequality question and others. Heller had dressed an overall assessment of various North-South Heckscher-Ohlin Model. I have examined result obtained. My judgement is quite severe. This is my diagnosis:

• The real question starts here. Is it wise that we continue to search a new model, or is it better to change our research program? As I have repeatedly stated, state of HOS-related theories and models is very similar to the state of the astronomy just before Nicolaus Copernicus. Adding new epicycles does not resolve the problem. A new insight is required.