The cost and demand differences in any form do matter in international theory, H-O model is one of the explanations.

Dr. Arjun Yallappa Pangannavar

Are you saying that any explanation in international trade theory goes as far as it argues cost and demand differences?

You have posted a question:

• Is there need for redefining 'Economics' under the present context?

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

In your explanation of your question, you put that

• We are worried about sustainable development and inclusive growth that poses various challenges before the world.

Do you think that HO-models contribute to gain necessary knowledge that helps to promote sustainable development and inclusive growth? Suppose it gives you such knowledge, your knowledge will be baseless if HO models are fictitious and irrelevant. 

The basic principle of the trade theory is 'cost-differences' on supply side and 'demand conditions' on demand side. Cost is being measured in terms of labour-units, real cost, money-cost but different theories give different reasons for cost-differences such as division of labour, factor endowment, technology, etc; meanwhile domestic demand conditions,elasticity of reciprocal demand also play vital role; the main aim of the trade is to get gains, directly and indirectly; In an open economy, international trade, like other factors, also play role in economic development, by giving due importance to protection and development of environment and natural resources of the country which is called sustainable development. the free movement of factors and commodities/services between regions and countries helps to attain price-equality.. technology improves efficiency and reduces costs that generates gains.

Dear Yoshinori, thanks for the comment.

Dr. Arjun Yallappa Pangannavar, 

I thank you for your comments. I have again a question to you.

You raised many factors which may cause differences of costs and those which may influence the demand of a country. They may be factors or conditions that you keep in mind when you think of international trade.  That is all right.

However, how are they related to Heckscher-Ohlin model? Heckscher and Ohlin wrote in prose. They may be their understanding of the facts of international trade. Your account of HO-model may be closer to their original theory. HO model or more precisely HOS model is quite different from the original theory. It has its own formula. This is the meaning of a model.

For example, the standard HOS model of two-country, two-commodity, two-factor economy is composed of  two identical production functions, which represents technologies of each country, two identical utility functions of each country, and endowments of two factors for each country.  Production functions are supposed to be homogeneous of degree 1, continuously differentiable, convex. In the standard model case,  two nations can differ only by their amount of two kinds of endowments. 

Excepting NS-HOS model, the Heckscher-Ohlin model (at least in its standard model) does not assume that two countries have different technologies. It does not assume differences of taste of two countries represented by their utility. The only  different conditions are state of endowments, say for example capital and labor which are thought to be uniform quantities.

In the standard Heckscher-Ohlin model case, it is the difference of proportions of two endowments in each country. If country A is more capital rich (or capital abundant) with regards to country B, A will export good that is more capital intensive than others.  Then,  country B is more labor rich and  exports more labor intensive good and import capital intensive good from A (Heckscher-Ohlin theorem). These are almost all you can argue in the standard HOS model. Of course, you can derive some other theorems such as Factor Price Equalization, Stolper Samuelson, Rybczynski. However, these theorems do not change the situation setting.

Then,  does this model tells you something useful for you? No story about sustainable growth. No story about environment. Contrary to your understanding, factors (capital and labor) are not permitted to move across countries. Technologies may improve, but in such a way that two countries have the same production functions. It simply means technologies are always the same between countries and there is no competition in technology  development. Do you think this image that HOS model give corresponds to what you have in mind?

Dear Yoshinori, thanks for the comment and refreshing of knowledge. The H-O theory, particularly Ohljn, advocated that international trade is a special case of inter-regional or inter-local trade; difference in money, policy, political units etc are the factors differ domestic trade from international trade but the basis of trade, domestic or foreign, is same i.e. cost differences in form of commodity-price differences. It is true/valid even in the present globalised economy, the national economy became part of global economy. Is it not the validity of  H-O theory?

Dear Arjun,

please note that my original question is about HO-model. Not about Ohlin's theory. Do you still support HO-model? If so, by what reasons?

I never read Heckscher's book in trade theory. It seems we have no English translation. Ohlin's book was first published in English in 1933 and the revised version was published in 1967. in 1930's there were tripartite controversies between Ohlin, Jacob Viner and Gottfried Haberler. Chipman in his famous survey in Econometrica 1966-67 classifed Viner and Haberler as neo-classical and Ohlin as modern theorist. However, if we read now his book smells as old as other two. We read many facts and explanations, but feel few theoretical arguments.    

I have Japanese translation of the 1967 edition and sometime consult it, but it is hard to rad it through. It is a complex book and it is hard to tell which part is really related to HO-model. For example, in Chapter 2, Ohlin argues tendency to factor price equalization, but this is quite different from factor price equalization theorem. The theorem first established by Samuelson claim that factor prices of two countries are equal provided some conditions are satisfied. Samuelson himself thought it unrealistic.

If you support HO-model, you must support all theorems derived from the model. if you support HO-model, how do you reconcile the model with various empirical studies? There are many such results, but few of them confirm the prediction of the model. How do you explain all these facts?     

Thank you Yoshinori for sharing your views on H-O Theory/Model. I would like to share some my views/understanding of  H-O Theory on international Trade:

Heckscher – Ohlin Theory:     Heckscher–Ohlin theory of international trade is also known as ‘General Equilibrium Analysis’ or ‘Factor Endowment Theory’ or ‘Modern Theory’ of international trade.

    Heckscher at first time propounded the factor endowment theory of international trade in 1919. But his student, Bertil Ohlin, a Swedish economist, made this theory more popular by building it as modern theory in 1933 in his famous book ‘Inter-Regional and International Trade’. Thence, the theory is popularly known as ‘Heckscher–Ohlin Theory’ or H-O theorem.

 (A) Objectives of H-O Theory: The H-O theory has the following objectives:

 (i)  To criticize the classical theory as it is based on unrealistic assumptions and a clumsy dangerous tool of analysis.

(ii)  To explain that differences in relative commodity prices are the causes for or basis of both inter–regional and international trade.

(iii)  To explain that differences in relative factor endowment or scarcity are the causes for differences in cost and relative commodity prices.

(iv)  To explain that the change in foreign exchange rate translates relative price differences into absolute price differences. It enables each country to take specialization in production of the commodity which suits it.

(v)  To explain that the reciprocal demand also plays important role in determination of commodity price and exchange rate.

(B) Statement of H-O Theory: Heckscher’s view:

Heckscher gave the statement of the theory as the cost–differences are the causes for trade and the differences in factor–endowment or scarcity in two countries result into differences in factor–intensities and thus differences in cost of production of same commodity in different countries. For Ex: Australia and England are the two countries. Australia is rich in land–recourses (land endowment) England is rich in capital (capital endowment) so land intensive commodity is suitable to Australia because land is cheap; capital intensive commodity is suitable to England because capital is cheap. Land-intensive commodity, says wheat, is cheap in Australia and dearer in England. On the other hand, capital intensive commodity, say machine, is cheap in England and dearer in Australia. In other words, the cost of production of wheat is low in Australia and more in England; the cost of production of machine is low in England and more in Australia. Thus trade takes place between Australia and England. Australia exports wheat to England and imports machine from it. Similarly England exports machine to Australia and imports wheat from it. Thus specialization in production of commodities, wheat in Australia and machine in England, takes place and both countries will be benefited from trade. 

 ·         Bertil Ohlin’s view:

    Bertil Ohlin too agree with Heckscher’s views that (i) differences in relative factor–endowment or scarcity result into differences relative factor–prices (ii) difference in relative factor–prices result into difference in relative factor–intensities which cause differences in cost of production of  same commodity in different regions or nations and (iii) cost difference is the basis of trade.

    Ohlin had gone one step ahead than Heckscher (his teacher). Ohlin said that cost–difference is not only the cause for trade but differences in relative commodity prices is also cause for regional trade or international trade. In market, demand and supply conditions determine the commodity prices.

    If, (i) demand and supply conditions are alike, (ii) factor–endowment or scarcity is alike and (iii) relative prices of commodities are same in different regions or nations, trade will not take place.

    Hence, Ohlin states the theory as “Trade occurs because there are differences in relative commodity prices caused by differences in relative factor prices as a result of differences in the factor–endowments between countries.”

 (C) Illustration of H-O Theory: The factor-endowment theory explains the factor–endowment or scarcity in terms of two criteria:

(a) Physical Criterion: The difference in factor–endowment or scarcity in different countries is cause for difference in relative factor intensities, difference in relative factor prices and thus cause for cost–difference of same commodities in different countries, which is the basis of specialization and international trade. The difference in factor–endowment or scarcity is explained in physical terms in the following example:

    Australia (A) and England (E) are the two countries. Australia is rich in land and England is rich in capital. In other words, Australia is poor in capital and England is poor in land. Endowment shows richness and scarcity shows poverty.

   The relative land and capital endowment or scarcity of Australia and England, in Physical turns, is expressed as:

  (C/L)E >(C/L)A

  (C/L)E is capital intensity in England

 (C/L)A is capital intensity in Australia

  It can be verified by assigning numerical values as stated in the following table-1

 Table-1 : Factor-Endowment or Scarcity

Two–Countries             Two – Factors  (in Units)             Factor-intensities 

                                         Land (L)       Capital (C)                     C/L           L/C

Australia                        100                    50                        50/150        100/50

England                         50                    150                        150/50         50/150

 (i)Land intensities in Australia and England are:

 (L/C)E < (L/C)A i.e.          (50/150)E < (100/50)A

     Thus Australia is land rich or endowed country and England is land poor or scarce country.

 (ii)Capital intensities in Australia and England are:

(C/L)E > (C/L)A i.e.          (150/50)E >(50/100)A

    Thus Australia is capital poor or scarce country and England is capital rich or endowed country.

   Therefore land–intensive commodity, say wheat, becomes cheap or low cost commodity in Australia and dearer or high cost commodity in England. Similarly the capital intensive commodity, say machine, becomes cheap or low–cost commodity in England and dearer or high cost commodity in Australia. It is because of the land is cheaper in Australia and capital is cheaper in England.

   If the factor endowment is same in both countries, then, factor intensities will be same in both countries:

i.e. (C/L)E = (C/L)A

    Hence, there will be no cost-differences, thus, there will be no trade between two countries.

 (b)Price Criterion: The factor–endowment or scarcity can also be expressed in terms of prices of factors (P):

 (PC/PL)E< (PC/PL)A

           Here, PC = Price of capital, PL = Price of land

          (PC/PL)E is capital intensity in terms of price in England.

          (PC/PL)A is capital intensity in terms of price in Australia.

          It can be verified by giving numerical value in terms of Australia Dollar (A$) to capital price (PC) and land price (PL) for ex: PC = 10, PL = 5 in Australia and PC = 5, PL = 10 in England. Then the factor–intensities in terms of prices are:

 (i)Land price intensities in England and Australia are:

          (PL/PC)E >(PL/PC)A

i.e.     (10/5)E > (5/10)A           Land is cheap in Australia 

 (ii)Capital price intensities in England and Australia are :

          (PC/PL)E

Dear Arjun and Dear All Readers,

I thank Arjun for a very long explanation and illustration of Heckscher-Ohlin's theory. It is repectable that he knows Heckscher and Ohlin's doctrine from their original texts. I have been imagining that there were but  a very few people who dare to read those texts. I  have been imagining that almost all trade theorists who support HO-theory think by means of a HO-model. It is amazing that Arjun thinks through Heckscher and Ohlin's original thoughts. This is always a necessary attitude and that is why we need history of economic thought. 

However, as I have repeatedly emphasized, HO-theory and HO model are quite different things. It seems Arjun never understands or does not want to understand this.

It is Paul Samuelson who has formulated the model. He may have got inspiration from  the HO-theory, but the model has its independent existence since 1940's.  It must be better called Heckscher-Ohlin-Samuelson model, as some people do so.

Perhaps the standard understanding of  HO-model can be illustrated by the Wikipedia's article "Heckscher-Ohlin model". Please, Arjun, read this page once and consider if it is the same theory that you call Heckscher-Ohlin theory.

Let us discuss Heckscher-Ohlin theory's validity or usefulness after we have a good review and understanding of the validity and usefulness (or invalidity and even toxic effects for economic thinking) of HO-model. I will promise.    

amazing...it is very interesting to discuss..

"Paul Samuelson who has formulated the model. He may have got inspiration from the HO-theory, but the model has its independent existence since 1940's. It must be better called Heckscher-Ohlin-Samuelson model, as some people do so"

Dear Al Afif and our dear readers,

Paul Samuelson was a genuine hero of my student days. No one was  surprised if  Samuelson argued any of economics problems. It seemed he was omnipotent. Samuelson's contribution to international trade theory was an "ordinary" things.

Here is a list of Samuelson's papers in the domain of international trade theory.

• Samuelson, P.A. (1938) Welfare Economics and International Trade. American Economic Review 28(2): 261-266.
• Stolper, W.F. and P.A. Samuelson (1941) Protection and Real Wages. Review of Economic Studies 9(1): 58-73.
• Samuelson, P.A. (1947) Foundations of Economic Analysis. Cambridge, Mass.: Harvard University Press, 1947.
• Samuelson, P.A. (1948) International Trade and the Equalisation of Factor Prices, Economic Journal 58 (June 1948), 163-184
• Samuelson, P.A. (1949) International Factor Price Equalisation Once Again. Economic Journal 59 (June 1949), 181-197
• Samuelson, P.A. (1951-52) A Comment on Factor Price Equalisation. Review of Economic Studies 19:121-122.
• Samuelson, P.A. (1953) Prices of Factors and Goods in General Equilibrium. Review of Economic Studies 21: 1-20.
• Samuelson, P.A. (1958) Frank Knight's Theorem in Linear Programming," Zeitschrift fur Nationalokonomie 18 (Heft 3): 310-317.

 Samueslon (1947) concerns only on the duality theorem between production function and cost function. We can know that Samuelson's interest in international trade started very early in his career.

 The best introduction to the history of HO-model and Samuelson's contributions is Chipman's Survery Part III. Modern Theory: 

• Chipman, J.S. (1966) A Survey of the Theory of International Trade: Part 3, The Modern Theory. Econometrica 34(1):  18-76.

You will see that almost whole story is concerned with HO model, Samuelson's contribution and later development.

As for the standard formulation of 2×2×2 HO model, I do not know the historical details. If anyone among our readers know this, please inform us.

thanks alot Mr..i'd love ur paper about new theory about intenational values. it's very usefull and i got some reference to support my research...doumo arigatou gozhaimasu...

Adrian Wood has produced one book and numerous articles on HO theory and its application to explaining North-South trade;the thing though is that he explicitly does not have capital as a factor endowments:

http://www3.qeh.ox.ac.uk/pdf/qehwp/qehwps157r1.pdf

Thank you, Jonathan.

As for North-South HO-model, we are discussing it in Dr. Tahsen Alqatawni's question:

What are the pragmatic problems with Heckscher-Ohlin model?

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

It is mainly based on based Hellier's survey report that I have cited in my quesion. You can read it at

https://www.researchgate.net/publication/234058657_The_North-South_HOS_Model_Inequality_and_Globalization

Please have a look on my posts with titles:

• The history of North-South HOS models shows a typical science philosophy of majority of economists (post #21)
• How Hellier estimates the traditional North-South HOS model. (post #22)
• How economist added modifications after modifications: (post #23)
• Final overall assessment (post #24)

I also posted some other arguments on HO-models in general. See my posts which start with

• There are many confusions concerning international trade theory. Many people understand that HO theory is a natural generalization of Ricardian theory. But this is a grave error and misunderstanding. (post #5)
• There are some sharp differences between Ricardian trade theory and HOS theory. (post #8)
• The Heckscher-Ohlin theory focuses as the raison d'être of international trade the difference of factor proportions between countries. In the simplest form, it assumes labor and capital as two major factors of production. (post #11)
• Apart from its poor predictive power, HO theory is conceptually defective. Followings are some naive questions. Can somebody explain? Why does nobody raise these simple questions? (post #12)
• I have read Joel Hellier's paper that is uploaded in RG. (post #13)
• I quickly read Kremer and Maskin's paper. So the following is a rough impression. (post #19)
• A globalization paradox? (post #20)

Chapter The North-South HOS Model, Inequality and Globalization

I have read Adrian Wood's paper

• Heckscher-Ohlin in theory and reality

that Jonathan Perraton ·recommended two-post above. It is well known that Adrian Wood worked fecund and fruitfully in the field of wage studies. I have once bought one of his books (not Globalization and Wages )but had no time to read it through.

Adrian Wood in this paper has a quite ambivalent stance. In short, he claims that "HO-theory" (sic, it will be better to call it "HO-model") works empirically but have trouble in theory. He recommends two change of interpretations:

 (1) Remove the capital form the category of endowments.

(2) Think ratios and not absolute values.

I have no idea how to think about his proposals.  At least, I have to reflect on this for more time. Does anyone including Jonathan have a good idea?   

Lefteris Tsoulfidis has send me an interesting paper:

Paraskevopoulou, Tsaliki, and Tsoulfidis (2016) Revisiting Leontief's paradox. International Review of Applied Economics 2016.

The second section of this paper is a concise survey on empirical studies related to Leontief paradox.  I here reproduce Table 1 they dressed in a little different form.

HOS and HOV model tested

Leontief                      53, 56           USA                              R

Stolper & Roskamp 61                  East Germany           C

Wahl                           61                  Canada                       R

Bharadwadi             62                    India                            R

Baldwin                    71,79               USA                             R

Hirsch                       71, 76             28 countries               C

Leamer                     80                   USA                              C

Tatamoto & Ichimura 59               Japan                            R

Stern & Maskus       81                   USA                               58 R;  72 C

Tsao                           80                  USA, Taiwan                R

Bowen, et al.             87                  USA  & 27 counting     R

Duchin                       90                    USA                               R

Trefler                         93, 95            33counties                   R

Davis, et al.               97                     Japanese regions      C

Hakura                      99                     Various countries       C

Davis & Weinstein  01                     10 OECD countries    R

                                                             & the rest of the world 

Wolf                             04                   USA                                 R

Source: Paraskevopoulou, Tsiliki, and Tsoulfidis (2016) p.5 Table 1. Summery of Studies.

Note:  R: reject HOS or HOV theory, C: confirm or does not reject. Each column indicates repectively Author of the study,   Year of the data studied,  Country or countries studied, and the result of the study.

I have changed "Yes" and "No" to Leintief's paradox to Reject (R) or Confirm (C) one of hypothesis derived from the HO model. To speak accurately  "No" to Leontief's paradox is not a Confirmation. It simple means that a test did not rejected the hypothesis. As I am here only concerned with the validity of HO-theory and how many Rejects were detected, I changed all NO to a confirmation of the theory.

Now when we discuss the validity of a theorem, a single study that is firmly established  (no miscalculation, no manipulation of data, sufficient data, and others)  is in principle sufficient to reject a hypothesis. If we have such a result, we have to think that the theory has something wrong.

If we see the table above, we count 12 R and 6 C (Stern & Maskus's study is counted as two tests).  Confirmed (or more exactly expressed "not rejected") tests counts one third of all studies. This is a clear evidence that HO model (HOS and HOV included) is definitely rejected.

What we should think is why HO model  still collect a support in textbooks and is commonly accepted as plausible explanation of the international trade.    

Duhem-Quine thesis

What I have written above is based on a naive, simplism on scientific development. As Duhem-Quine thesis tells us, a successful rejection implies there is  something wrong in the total structure of a theory. So a defense of  HO model is possible in any possible ways. Perhaps those young economists who have learned about negative results on Leontief paradox and the predictive power of the HOV models may have heard some pretexts why we get such negative result and what kind of attenuation circumstance is to be considered. Still many professors and students still prefer to conserve a theory by a fear of loosing a theory at all.  In all cases, adding minor changes to a theory is possible.

It was Imre Lakatos who proposed to distinguish hard core and protective belt  of a theory. In the case of HOS model, it seems that the theory itself is so simple and it is difficult to add or change a protective belt.  Even though, many pseudo-ideas in the methodology of science  surround us and people want to continues to believe finding one reason or anther that tells them to ignore the empirical results.

I admit that it is always difficult to judge the validity of a science. It is a synthetic judgement. It always requires some jump in thinking. No body knows if the HOS theory can survive or not by adding some supplementary conditions or propositions.

As HOS model is a partial theory for a particular case of General Equilibrium Theory (GET), it would be difficult for those people who still believe in GET.  This is one of reasons that many people want to continues to believe in HOS and HOV models in contradiction to a manifest fact summarized in Table 1 in Paraskevopoulou, Tsaliki, and Tsoulfidis (2016). 

Dear all,

Today, I read all the valuable comments, particularly like always, wonderful explanations of Prof. Shiozawa. Only a question that is raised in my mind is that in the book by Giancarlo Gandolfo (2013) entitled "International trade theory and policy" on page 87, the author argues "The Heckschur-Ohlin-Vanek Generalization". I'd be thankful if you could explain whether we can use this approach as a substitution of H-O.

Dear Ehsan,

HOV is just the topic I want to explain in the next post. Please wait for a moment, because it takes a bit of time.  

If I tell my present judgment, HOV is already refuted by many of empirical works. I have not noted it, but researches made by  

Bowen, et al.             87                   USA  & 27 counting     R

Trefler                         93, 95            33counties                    R

Davis, et al.               97                    Japanese regions      C

Davis & Weinstein  01                    10 OECD countries     R

and listed in the table were based on HOV model.

The unique "successful" case is  Davis, Weinstein, Braford and Shinkai (1997). They have applied HOV model and checked if factor content equations are satisfied.

In my understanding, this is not an empirical test of HOV theory as international trade theory (or an extension of HOS theory). if HOV model is to work as international trade theory it must comprise a theory how to evaluate each factor of all countries an equivalent quantities.

Let me explain by the case of labor force. Labor  of an hour of countries A, B, ... , and is evaluated as wA, wB, ... , wC. Vanek thought that international trade is an exchange of factors, when each of them is evaluated appropriately. In international economy, the same factor of a different country may have different values. In the case of a country labor and capital have approximately the same value. Then HOV equations means they are simple transfer of physical contents of factors. This is always possible by a suitable IO table analysis. I doubt if  Davis and others have done this.

In three other cases, because there are big differences of wages, their empirical tests were all negative, if they do not modify their estimation by introducing new variables like worker's productivity. It is almost equivalent to see the same identity relations in a different expression. People who read those papers are requested to see through such a mathematical magic. Most of authors in this researches are not ware of this magic.   

Does anyone posses a PDF of Vanek's paper? I have promised to discuss Vanek's paper Factor Proportions Theory: The N-factor Case. Kyklos 21(4):749-756. I once taken a copy, but have lost it.  

A lot of literature has been generated/generating on H-O Theory and its empirical model versions, in favour and against, but its main theme that the differences in commodity price is the basis of  inter-local or inter-regional and inter national trade; even the multi-countries-multi-commodities-multi-factors models supported the main claim of H-O theory.

Dear Arjun,

You claim that H-O Theory's main theme is summarized in the following proposition. 

• the differences in commodity price is the basis of inter-local or inter-regional and inter national trade;

However, on this level, Ricardian theory (at least in the new theory of international values) has the same property as H-O theory. For the new theory of international values, please see 

New Theory of International Values: A General Introduction.

https://www.researchgate.net/publication/304717720_New_Theory_of_International_Values_A_General_Introduction

The difference between Ricardian theory and H-O theory must be argued upon the more precise theoretical structure. 

Working Paper The New Theory of International Values: An Overview

In this post, I do not examine HO-model or HOV-model, in their logic and their empirical researches. Instead, I would like to argue briefly the difference between Ricardian trade theory and  HO- or HOV-models.

 It is true that many textbooks teach that HO-model of theory is a generalization or extension of Ricardian models. What makes the situation more complicated is that there are different Ricardian trade theory strands which we should distinguish.

Let me cite the case of Chipman's survey (1965-66) A Survey of the Theory of International Trade. Econometrica 33(3), 33(4), 34(1). This survey is divided into three parts:

• Part 1: The Classical Theory
• Part 2: The Neocalsscial Theory
• Part 3: The Modern Theory

Part 1 covers Torrens, Ricardo and Mill, Part 2 Marshall, Edgeworth, Haberler, Viner, Lerner, early-Leontief, and Meade. Part 3 starts from Heckscher and Ohlin theory, but Chips thinks that the modern approach did not really come into its own until with the work of Lerner and Samuelson.

However, Ricardo's theory of international trade is very different from that John Stuart Mill. Marshall, Edgeworth, Haberler and Viner are in their essence following John S. Mill. Then, the naming such as "classical theory" is not appropriate at all. Trade theory from John S. Mill to Haberler and Viner should be called neocalsscial theory of international trade. Calling those theories classical theory is a misnomer just like that of Keynes who called Marshall and Pigoux’s theory classical theory.

For this point, you can read Jorge Morales Meoqui's working paper (2016).

Ricardo’s numerical example versus Ricardian trade model: A comparison of two distinct notions of comparative advantage

https://www.researchgate.net/publication/48375755_Ricardo%27s_numerical_example_versus_Ricardian_trade_model_A_comparison_of_two_distinct_notions_of_comparative_advantage

Although Meoqui does not mention, there is a strand which is more close to Ricardo. It is Frank D. Graham (1890-1949). He does not appear in Wikipedia, English. In Douglas A. Irwin (1996) Against the Tide, Graham is introduced as a person who defended protection on the basis of increasing returns to scale theory (Chap. 9). However, he is a one of few economists who examined the trade and international value on the tradition of Ricardo. His book (1948) Theory of International Values is a monumental work just before the arrival of activity analysis. Lionel McKenzie and Ronald Jones partly succeeded Graham's work but were soon absorbed into general equilibrium framework.

Meoqui (2016) rightly points that many commonly accepted interpretation concerning Ricardo's trade theory, but does not mention when and where this peculiar, strange story started. Astonishingly, it was John Stuart Mill who disfigured whole story. Marshall and Edgeworth only followed Mill. On this point, please read my working paper

Mill's Reversion and an Origin of Neoclassical Revolution

https://www.researchgate.net/publication/305810200_Mill%27s_Reversion_and_an_Origin_of_Neoclassical_Revolution

In view of this long misinterpretation of Ricardo and his trade theory, it is quite natural that HO- and HOV-models are interpret as an extension of Ricardo's theory.

Preprint Ricardo’s Numerical Example Versus Ricardian Trade Model: A ...

Working Paper An Origin of the Neoclassical Revolution: Mill's "Reversion"...

Please see the paper by Naoki Yoshihara and Kazuhiro Kurose (2014) 

The Heckscher-Ohlin-Samuelson Model and the Cambridge Capital Controversies and comments there:

https://www.researchgate.net/publication/319501242_The_Heckscher-Ohlin-Samuelson_Model_and_the_Cambridge_Capital_Controversies

Article The Heckscher-Ohlin-Samuelson Model and the Cambridge Capita...

I am happy to know that this question that dates more than a year is still read continuously. Last week ending December 31, 2017 received 13 people who came to read this question page.

Sarbajit Chaudhuri and Sugata Marjit published a new paper: International trade and quality of labour

Article International trade and quality of labour ☆ E-print

In this paper (p.583, the second page of the paper), Chaudhuri and Marjit put it:

• [T]here is no denial of the fact that rising wage gap is a phenomenon in many of the developing countries as well. Because the relative wages have moved against unskilled labour in the developing economies in which unskilled labour is the abundant factor is contrary to the predictions of the standard Heckscher-Ohlin trade model with Stolper-Samuelson theorem at its core calls for theoretical explanations.

The authors point that there are many contributions (Feenstra (2004) and Zhu and Trefler (2005), Anwar and Sun (2015), Barua and Pant (2014), Chaudhuri and Yabuuchi (2007, 2008), Marjit and Acharyya (2003), Pi and Chen (2016), Pi and Zhou (2014)) have tried to explain this contradictory facts to Heckscher-Ohlin trade model by modifying and introducing some new elements. However, we can interpret this from a totally different perspective. It signifies that it is time to abandon Heckscher-Ohlin trade model. The efforts of Feenstra and others show that they are busy in defending Heckscher-Ohlin model by modifying its protective belt. This is exactly what Imre Laktos https://plato.stanford.edu/entries/lakatos/

named as efforts to modify the protective belt in order to save the hard core of the theory (see section 3.4 of the cited paper).

Heckscher-Ohlin model faced in the past three major refutations: (1) Leontief paradox (1960's), (2) Fails of Heckscher-Ohlin-Vanek model, a variant of Heckscher-Ohlin model (1990's) and (3) Fail of Stolper-Samuelson theorem (2000's). We may add the fourth paradox which is the evident fact that contradicts Factor Price Equalization Theorem, i.e. the existence and persistence of big wage gaps between countries.

It seems for me that it is now time to abandon Heckscher-Ohlin and factor proportion theory in general.

Sir, in this paper, the issue of wage inequality is not important. We have developed an incentive contract to get rid of assymetric information problem that has been linked to international trade. This mechanism also helps to distinguish between two types of labour endogenously. These are the main contributions of the paper. I would like to have your comments on the contributions of this paper.

Sarbajit Chaudhuri

You asked me a comment on your paper with Sugata Marjit

Article International trade and quality of labour ☆ E-print

The "answer" I have posted above yesterday (one day ago) is not a comment on your paper. I have only cited a part of your paper concerning Heckscher-Ohlin model that may give us a hint for considering the question that I have posed.

You must be confusing two different replies. I have sent you a "reply" on your request personally through RG e-mail system. The above "answer" is different one. Probably you have not read my "reply". If you have, you must have thought two "replies" were the same.

You may not be interested on the validity of Heckscher-Ohlin trade model, but this is an important issue for international trade theory. In your paper, you have put it:

• Our work is mainly related to two different literatures: (i) the literature on trade liberalization and wage inequality; and, (ii) the literature on asymmetric information and wage differentials.

Even if your are interested in the second part (i.e. asymmetric information and wage differentials), readers have a freedom to argue the first part (i.e. trade liberalization and wage inequality) and for this point Heckscher-Ohlin trade model is still the issue. This is the every question I have posed in this question page.

Yixiao Zhou and Harry Bloch recently published an interesting paper:

Zhou and Bloch (2018) Wage Convergence and Trade. World Economy

https://www.researchgate.net/publication/329494786_Wage_Convergence_and_Trade

It is a survey of wage differences across countries. This is a precious research, because many papers written on the theme of wage and international trade argue the reasons of widening wage gaps inside of a country. I admit that it is an important question

in the policy discussion. However, we should not forget that more important theme lies with regards to wage rates. It is the question of international disparity of wage rates between countries.

This question has been put aside partly because Hecksher-Ohlin model has no theory with which we can analyze wage gaps between countries. If we consider that the prime target of development economics is to raise average income level of people of a nation, and the latter is determined by (real) wage rates, this is the most neglected question in international trade theory.

Heckscher-Ohlin model contains as one of four major theorems the Factor Price Equalization Theorem (FPE theorem). Simply stated the theorem stipulates that factor prices of all countries are equal when the world economy is in a FPE sector.

Labor is one of production factors and wage rate is a price of that factor. Then FPE claims that wage rates of all countries are equal provided that the sufficient conditions of the theorem are satisfied.

A simple observation is that wage discrepancies in the world are enormous. Zhou and Bloch (2018) have shown that wage rates differ to the rate 1 to 20 between low-wage and high-wage countries. This is rather obvious but precious data, because comparison of wage rates are so rare.

However, Zhou and Bloch (2018) are making a big error, because they are trying to justify the HO model. They have found that catching up rate of low wage countries to high wage countries are 4 % per year. Based on this finding, Zhou and Bloch think that FPE theorem was justified. The fact that they have think like this reveals how bad function HO model plays. If they know a different theory, for example, the new theory of international values, they have not reasoned like this. Four per cent catching up process for a country of 1/20 wage rate of an advanced countries requires more than 58 years until it arrives at 1/2 wage rate. Is it plausible to say that this is an adjustment process? I do not believe so.

Please see my comment on Zhou and Bloch (2018).

https://www.researchgate.net/publication/329494786_Wage_Convergence_and_Trade/comments

As a young phd student, actually I know no professor now still teaches HO theory in the phd series courses on International Trade. But this theory can explain some interesting facts and especially your question on EK model - "Why EK model performs so badly with data". Indeed, the correlation between the ratio of the different production factor stock and the technology is obvious. North countries are rich in capital, so they mainly develop the technology that enhance the capital efficiency, for example, in US, the financial engineering is far more developed than any other country in the world. South countries are abundant in unskilled labor, so they introduce the technology that enhance the unskilled labor productivity. Then, in this story, you can link the Ricardian and HO. The theory can be built on the Acemoglu's paper on skill-biased technology advance. Because it depends on how much the development of technology can be rewarded. So in this sense, I don't think the HO theory is totally irrelevant to trade but it is just hidden behind the Ricardian theory. Also, this explains why EK model is not perfect. Because the technology of the firms are correlated through the HO model.

Dear Xiaojie Liu

thank you for your frank argument. The fact that no professors teach HO theory in the phd courses does not matter. If most of problems that can be explored by a theory, it will cease to be taught. HO theory is still important because it is still binding economists' way of thinking. As you argued, you jump to think proportions between capital, labor and perhaps technology as factors of production. This habit of thinking is formed by HO theory. But is it reasonable?

My contention is factor proportion theories (HO theory is only an example) are all wrong.

(1) First, it has no good predictive powers. See controversies in 1990's vis-à-vis HOV models.

(2) Theoretical framework of factor proportion theories contains many serious theory troubles. To cite only two of them, (2a) measurement problem in capital theory and (2b) aggregated production functions. For the latter, see

Felipe and McCombie 2013 The Aggregate Production Function and the Measurement of Technical Change: 'Not Even Wrong' , Edward Elgar.

Aggregate production function is in a great probability misidentifying accounting identity as production functions as Herbert A. Simon pointed it in is Scandinavian Journal of Economics when he got a Nobel Prize of Economic Science.

(3) They give erroneous footlights on less important questions instead of more important issues. For example, technology is understood in a very superficial way and misleads people to wrong questions (especially the people of less developed countries). Factor proportion theories have strong tendencies to consider technology as something which can be bought by money.

You should also remind that Ricardian theory has been renovated in these ten years or so. First, what is understood as Ricardian theory in schools is not what Ricardo thought but an invention of John Stuart Mill.

See

Gilbert Faccarello (2017) A Calm Investigation into Mr Ricardo's Principles of International Trade, and

Yoshinori Shiozawa (2017) An Origin of Neoclassical Economics: Mill's 'Reversion' and its Consequences.

Second, a new general theory (the new theory of international values) in the Ricardian tradition is now build. It is much superior than four generations of neoclassical trade theories (textbook Ricardian, HO theory, New trade theory and New new trade theory). For example, the new theory of international values is a theory which permits input trade in a general form (Ricardo-Sraffa trade economy). This is the theoretical basis for the analyses of global value chains (GVCs), because GVCs are network of productions and international trade. Please note that above four generations preclude input trade by assumption or is practically a theory of small open economy as it is the case of New New Trade Theory. The new theory of international values is free from such theory questions of capital measurement and aggregated production functions, simply because it does not draw on such concepts and constructions.

Third. Another essential difference between the new theory of international values and HO theory is the theory of wages. HO theory has no good theory except factor price equalization theorem. The theorem claims that wages of all countries are the same in the factor proportion equalization cone. It is evident that this kind of theory cannot be used to analyse rich and poor counties problems, because this is practically the question that a rich country's workers gains higher wage rates than those in poor countries. The new theory of international values is a theory by which we can determine wage rate of each trading nations. Thus, workers of a nation gain five or twenty times as much as those of a poor nations. This big difference of wage rates between countries does not come from the difference of factor proportions. Wages depends on the technology (in other words, the set of production techniques that a nation [or firms of a nation] possesses) .

The first task that a poor nation must tackle is to change their production techniques to more productive ones. It does not mean to buy a new technology from abroad. By a change of production sites, productivity of production techniques can be changed tremendously. In this changing phase, foreign capitals' firms may give a good stimulation, but what is important is the efforts of workers (including directors and engineers) on the production site. It is also necessary to note that social infrastructures like roads and ports influence the total productivity of a nation, because transportation is an important part of production techniques.

Sarbajit Chaudhuri recommended Xiaojie Liu's post. It is a great pleasure that he is interested in our question. Please join us and give your opinions about international trade theory and its relations to development economics.

HO was created to explain why two countries trade, not to answer any quantitative question about the trade. The state of HO theory is more like the Ricardian before EK model which brought it into a revival.

Dear Xiaojie Liu

The HO model was created to fill an alleged gap left by the so-called Ricardian trade model, which is a misinterpretation of what Ricardo wrote in the Principles. Here is my interpretation of it, in case you are interested:

Preprint Overcoming Absolute and Comparative Advantage: A Reappraisal...

Xiaojie Liu

HO theory has many theoretical and empirical problems. To explain why two (or more) countries trade does not necessitate HO theory. Why do you want to defend such a theory?

Ricardo's example and the new theory of international value can well explain why two or more countries trade. The new theory has a much better quality, because it can explain trade situations with unemployment. Thus, it can explain not only the gains from trade, but also losses from trade. (See my paper 2017 The new theory of international values: An overview. Section 4.)

HO theory is a special explanation which was created to explain why Sweden engage in exports in woods and steal in the turn from the 19th to 20th centuries. It had a good forest and iron ore which contains low rate of sulfur and silicon. The same situation can be explained by the fact that Sweden has a good production techniques of wood making and steal making aided by abundance of natural resources.

Heckscher and Ohlin could explain Swedish export of early 20th century. But, how can you explain by means of HO theory that China exports HUAWEI smart phones? Is it because China is rich in low skilled labor? By HO theory, you cannot explain the complicated international trade. Global value chains are typical phenomena that traditional trade theories could not explain. The need for a new theory is evident.

Dear Yoshinori Shiozawa

Why do you think that classical theory of international trade, when properly understood, cannot explain global value chains?

Yoshinori Shiozawa Your point on China is not right. When China first got into the WTO, it mainly exported textile, toys and furniture which are very low-skill intensive. So Chinese government supported these small factories by giving low-interest loans and almost neglected the protection of the labor supply side to keep the wage low to remain the price of good competitive. (This is equivalent to the development of the unskilled labor technology) After 10 years, the capital is accumulated and because of the one child policy, Chinese children get more education than before. With the factor intensity changes across time, you can see China's export moves up and up along the supply chain which can be explained by the HO theory. Therefore, what we really need is a dynamic version of the HO theory.

Xiaojie Liu

To want to explain everything by factor proportions is fruitless efforts. Heckscher-Ohlin-Vanel theory (HOV theory) is a generalized version of

HO theory. It tried to explain world trade by considering various kinds of factors of proportions. In 1990's, Daniel Treffler and Edward E. Leamer were the leading economists on this topic. They counted 10 to 12 factors including forest and pasture. Of course they distinguished 2 or 3 different classes of labor. But, how many kinds of products are there in the world. At least millions. In the Japan Standard Commodity Classification System, we see more than 10 thousand items. To want to determine specialization pattern of these large number of varieties of products by a dozen of production factors is an impossible attempt.

In reality, the results of empirical results were very bad. (Treffler was more critical than Leemer who wanted to defend the validity of HOV theory.)

In the modern economy, there are more than tens of millions (or even thousands of millions) of products. We can consider production techniques for each of those products which may differ according to firms. HOS and HOV assumes that countries have the same production function and want to explain the trade pattern and production specialization by the difference of factor proportions. It was a reckless attempt from the first.

When there was no theory which can replace HO theory, we were obliged to use it. It was the case that a bad or poor theory is better than nothing. But, the situation changed. We have now much superior theory. Why do you hesitate to change to the new theory? Of course, it is hard to understand at first, but you will soon be accustomed to use it.

Dear Jorge Morales Meoqui

It is important to note at first that I am considering to explain complex networks of productions and trades that we observe in the 21st century world economy, whereas you are mostly interested in interpreting and explaining Ricardo's four magic numbers. The aims are completely different. What you call classical theory of international trade cannot be a framework for analyzing Global Value Chains for the reasons I will explain in the following.

I agree with you that dichotomy between absolute and comparative advantage is useless and misleading. We can put off these two concepts and related arguments. I also agree with you that John Stuart Mill was mainly responsible of the present use of the dichotomy.

We can explain traders' motive in a uniform way by the rule that you have cited in your paper from Ricardo's Principles:

• The motive which determines us to import a commodity, is the discovery of its relative cheapness abroad: it is the comparison of its price abroad with its price at home.(Principles, I, 170)

As Ricardo added after this,

• In a country exports hats, and imports cloth, it does so because it can obtain more cloth by making hats, and exchanging them for cloth, than if it made the cloth itself. (Principles, I, 170)

Ricardo follows the same logic that Viner called "eighteenth-century rule" and that you are proposing to name "classical rule of specialization."

What is lacking in your papers and all other papers including Gilbert Faccarello , Roy J. Ruffin, and Andrea Maneschi is the recognition that Ricardo could not find the law that determines wage rates of different nations. In the last paper of yours (Overcoming absolute and comparative advantage), no words like wage or wage rate appear. You are totally missing this crucial point as others do.

Ricardo wrote that

• The same rule which regulates the relative value of commodities in one country, does not regulate the relative value of the commodities exchanged between two or more countries. (Principles, I, 133)

In other words, the domestic rule of value does not apply as a rule of international values. You have cited this paragraph in your paper, but do not explain why the rules are different in the case of inside of one country and the case of international trade.

Ricardo knew the reason: immobility of work forces (or population) and capital. In a country, wage rate is almost the same between industries. They are very different between countries. Thus the wage and profit rates for different countries are not necessarily the same across countries. If we can find a theory how wage rates are determined between countries we can get a unified theory of value. (The new theory of international values contains theory of domestic values.)

If we consider a very simple case where commodities are made directly by workers without any input products (This is quite unrealistic but Ricardo often assumed this situation). Suppose the profit rate is 20 percent for two countries Portugal and England. Suppose also that wine production of a quantity requires 100 hours of labor in England and 90 hours in Portugal. How can we calculate the money costs of two cases? If we do not know the wage rate of both countries, we cannot know the two costs and compare them. If we know them, the rest is a matter of elementary calculation.

Let us then assume that an hour of labor in Portugal is 10 pence (converted to English money terms) and 8 pence in England. The production cost of cloth of the specified quantity is 800 pence in England and 900 pence in Portugal. In our case, including profit, the full production cost or value of cloth is 960 pence in England and 1080 pence in Portugal. English cloth manufacturing is more competitive than Portuguese. The same calculation is possible for wine making. The wine production requires 120 hours in England and 80 hours in Portugal for the quantity that is exchanged in international trade with the same value as the cited quantity of cloth. The full cost of wine is 1052 pence in England 960 pence. Thus, Portugal is more competitive in wine production. In the international trade, cloth of 960 pence made in England is exchanged with wine of the same value made in Portugal.

Even in this simplest case, in order to know the cost of production (including profit, if we want to know the full cost in the modern terms), it is necessary that we know the wage rates in both countries (if the profit rate is uneven, we should count this difference two). Without knowing wage rates of two countries, how can you know the money costs of production (or full costs including profit) and how can you compare two values?

Ricardo could not present a unified rule or theory of value that holds both for domestic and international exchanges. This is the problem of determining terms of trade that John Stuart Mill thought Ricardo left unsolved. But, this naming is unfortunate, because it is not the terms of trade that are really important. What matters is the wage rate differences between nations. This misleading problem setting still continues in the theory of trade like Theory of Unequal Exchange and many development economics arguments.

One of the reasons of Ricardo's failure is his theory of wage that considered that real wage level was determined by the necessity to reproduce the labor power. In the time of Ricardo and Marx, real wage rates were very stable. It was reasonable to think that they are determined socially by customs of countries. However, the wage rate is a function of the sets of production techniques that nation have. The new theory of international values is first of all a theory that can determine wage rates between countries. In the case when final products are only traded (there were no such periods, because even in the time of first Industrial Revolution, England imported cotton flower which was sine qua non material for cotton industry), the wage rate can be determined by the single set of production techniques. However, when input trade is important, wage rate of a nation is not determined by the set of production techniques it possesses. Wage rates are determined in relations to sets of production techniques for all countries. When the sets of production techniques change by technological evolution, the wage rates change. Some countries increase its wage level quite rapidly whereas other countries suffer from stagnating wage levels. They are not determined by the relative proportion of capital and labor as HO theory assumes. Wage rates are functions of sets of production techniques and there is no other way to improve them without changing the sets of production techniques.

If once we have a theory by which to determine wage rates of all countries, the pattern of trade is easily determined by comparing full costs of production for each country. We can know what kind of Global Value Chains emerge when we know the relative wage rates.

Classical theory of trade (if there exists such a theory) lacks a theory by which to determine (relative) wage rates of nations. Without having this theory, you cannot know which country is competitive in which product. If you have a theory of wage rates between countries, what you need is only the comparison of cost of production. However, the theory of wage rates is most difficult one and most of trade theorists assume that they are given as natural variables and do not understand that they need a theory and such a theory exists already.

I have talked as if wage rates can be determined with no reference to pattern of international specialization. In fact, the two are determined simultaneously. This is one of reasons why many attempts have failed before me. The new theory is in fact a subtle theory that you must study attentively.

Dear Yoshinori Shiozawa

I hope you don’t mind if I confront you with direct quotes from the Principles that are at odds with your claim that Ricardo could not present a unified rule or theory of value that holds both for domestic and international exchanges. The passages are also quoted in my paper.

“The real and ultimate regulator of the relative value of any two commodities, is the cost of their production, and not the respective quantities which may be produced, nor the competition amongst the purchasers” (Vol. 1, 344).

“All that I contend for is, that it is the natural price of commodities in the exporting country, which ultimately regulates the prices at which they shall be sold, if they are not the objects of monopoly, in the importing country” (Vol. 1, 375).

For Ricardo, natural price was only a different name for cost of production (Vol. 2, 46).

Ricardo further specified what he meant when he stated that “the same rule which regulates the relative value of commodities in one country, does not regulate the relative value of the commodities exchanged between two or more countries” (Vol. I, 133), namely, that “the quantity of wine which she [Portugal] shall give in exchange for the cloth of England, is not determined by the respective quantities of labour devoted to the production of each, as it would be, if both commodities were manufactured in England, or both in Portugal” (Vol. 1, 134-135).

This is the actual difference between domestic and international exchanges. I also briefly mention in the paper the explanation given by Ricardo for it: the different rates of profits between countries, which fail to equalise because capital is less mobile between countries than within a country.

If capital would freely flow towards those countries where it could be most profitably employed, Ricardo stated that “there could be no difference in the rate of profit, and no other difference in the real or labour price of commodities, than the additional quantity of labour required to convey them to the various markets where they were to be sold” (Vol. 1, 136).

Therefore, I consider that Ricardo had a unified rule for determining the relative value of commodities in national and international exchanges. He also had a theory regarding natural wages:

“The natural price of labour is that price which is necessary to enable the labourers, one with another, to subsist and to perpetuate their race, without either increase or diminution.

The power of the labourer to support himself, and the family which may be necessary to keep up the number of labourers, does not depend on the quantity of money which he may receive for wages, but on the quantity of food, necessaries, and conveniences become essential to him from habit, which that money will purchase. The natural price of labour, therefore, depends on the price of the food, necessaries, and conveniences required for the support of the labourer and his family. With a rise in the price of food and necessaries, the natural price of labour will rise; with the fall in their price, the natural price of labour will fall (Vol. 1, 93).

Dear Jorge Morales Meoqui

I have explained myself in detail. I have no other words to add.

Bernard C. Beaudreau has published in 2015 a paper:

Bernard C. Beaudreau (2015) Rehabilitating the Factor-Proportions Hypothesis. Modern Economy 06(04):436-457

https://www.researchgate.net/publication/276099819_Rehabilitating_the_Factor-Proportions_Hypothesis

that tries to rehabilitate the Factor Proportion Hypothesis (FPH). Heckscher-Ohlin theory is the protype of FPH. This implies that means Professor Beaudreau tries to rescue already wretched theory by adding new elements (factors). His main idea is to add technology-creating capability or "ability to produce visionaries and scientists" as a new factor for countries or regions. He may have thought that he saved (or rehabilitated) the FPH. However, what he could do was only to show that proportions of traditional factors (labor, land, and capital) are not important in determining international (or interregional) trade patterns. What he proved is that it is the technology that a nation (or an area) possesses, or more precisely the production techniques that firms of the country or region possess.

Introducing "visionaries and scientists" permits him to use these factors as a proxy of techniques. If there is no theory which focuses on technology, Beaudreau's research project may have been useful until a new technology-based theory gives birth. However, such a theory already exists. See my paper:

Shiozawa, Y. (2017) The new theory of international values: an overview. In Shiozawa et al. (eds.) A New Construction of Ricardian Theory of International Values. Springer

Of course, this paper was not yet published at the time when Beaudreau's paper was published. However, there was a preceding paper before this, although it was a bit difficult to understand the real range of the new theory:

Shiozawa, Y. (2007) A New Construction of Ricardian Trade Theory--A Many-country, Many-commodity Case with Intermediate Goods and Choice of Production Techniques. Evolutionary and Institutional Economics Review 3(2): 141-187.

The readers of this discussion page are requested to read Yoshimichi Mirakami's paper:

Preprint Globalization and Income Inequality in Latin America: A Revi...

This is a survey on various arguments concerning what were the determining factors that moved wage inequality in Latin American Countries since 1980's.

One of topics of the paper is how Heckscher-Ohlin model (Stolper-Samuelson theorem in particular) was relevant in explaining the movement of inequalities.

Murakami follows carefully the arguments made in defense of and against Heckscher-Ohlin model. The researchers have found clear cases that contradict original Stolper-Samuelson theorem, but many economists tried to modify the model in order to justify the apparent contradiction by adding new factors which may influence the results.

These efforts reminds me typical defensive trials of orthodox theory that Tomas Kuhn presented when anomalies of the theory were found. HO model backed by General Equilibrium theory of Arrow and Debreu type is very flexible and easy to modify the model so that the new model eliminates the anomaly.

This kind of defensive reactions can continue forever until a new good theory is build. Therefore it is necessary to make a decision whether it is time to abandon the theory or conserve the theory by modifying it. This decision making is the most difficult one, because it requires the most synthetic understanding of the state-of-the-art of the concerned field. It needs a big jump, but in my opinion it is time to jump.

Murakami's paper provides enough information in making this decision.