The explanation of leontief's mystery
Ken et al. (Ken Neng,1965; Keesing, 1966) thinks that one of the important reasons of Leontief's paradox is that the capital defined by Leontief only contains material capital (such as machines, equipment, factories, etc.). ), and completely ignored the human capital. Human capital is the stock of skills and production knowledge embodied in people. The income or reward of human capital investment lies in improving a person's skills and profitability, and in improving the efficiency of economic decision-making in market economy and non-market economy. This means that American labor force contains more human capital than foreign labor force. Adding this part of human capital to physical capital will make the capital intensity of American exports higher than that of imported substitutes.
Ken pointed out in his research 1965 that many paradoxes about the h-o model, including Leontief's paradox, are due to the traditional trade theory adopting the concept of special capital. (Note: Because the trade theory analyzes a long-term phenomenon, the capital in the trade theory is the capital in the sense of "waiting", not a collection of a series of physical assets. Ken believes that it is inappropriate for traditional trade theory to regard capital in the sense of "waiting" as an element that directly enters the production function, which is the root of many paradoxes and contradictions. Based on the research of Vanic, Schultz and Becker, Ken broadened the concept of capital in traditional trade theory and established a model to explain many phenomena that h-o model can't explain.
In order to understand the concept of capital in Ken's trade theory, it is necessary to briefly introduce the research of Vanic, Schultz and Becker. When studying the relationship between American foreign trade and natural resources, Vanic found that American trade is actually protecting American scarce natural resources rather than capital (see Table 2), because the substitutes imported by the United States are actually resource-intensive.
Table 2 (Note: Tables 2- 7 are quoted from P. Kenen. Nature, Capital and Trade, Journal of Political Economy, 10 1965, pp. 437-460. ) 1947 the demand for capital, labor and resource goods for American export products and imported substitutes.
Input, output and input
Capital (thousands of dollars, 1947 price) 2,085 2,244
Labor force (annual labor force) 179 164
Resource commodities (thousand yuan, 1947 price) 340 630
Schultz and Becker found that there is a close relationship between capital and labor. They found that the United States invests a lot in training workers every year, which has exceeded the growth of physical capital investment (Table 3).
Table 3 Estimates of various capital stocks in the United States range from 65438 to 0957.
Type of capital:10 billion USD (1956 price) 1939- 1957 growth rate (%)
Physical capital 1, 270 2.05438+0
Labor force education capital 535 4.09
Vocational training for male workers 347 5.36
Training capital
Ken absorbed the ideas of Vanic and Schultz and applied them to the capital theory, which broadened the concept of capital in the traditional trade theory. Its basic idea is as follows:
1. Ken believes that the nature of capital (referring to land and labor) is the most primitive production factor in the production process. Only by investing in land and labor can land and labor produce the flow of factors serving production, otherwise, land and labor will have no effect.
2. There is a stable functional relationship among land service flow, labor service flow and investment.
3. Land labor flow and labor flow are production factors directly used in production, and capital does not directly participate in the production process.
In order to explain the mystery of Leontief, Ken used his capital theory to analyze the international trade of the United States from 65438 to 0947. In order to estimate the content of human capital in trade, it is assumed that the difference between workers' skills and human capital is entirely caused by the difference in investment in labor, and the wage difference fully reflects the difference between workers' human capital and skills. Under the above assumptions, Ken estimated the capital needed to transform a completely unskilled worker into various types of skilled workers (Table 4). Then, using Leontief's statistical data, that is, the percentage of all kinds of workers in million-dollar export products and imported substitutes (Table 5), combined with the calculation results in Table 4, Ken estimates the human capital contained in American import and export products (Table 6).
Finally, Ken explained the mystery of Leontief with the estimation of human capital demand of import and export products and Leontief's statistical data. Ken found that Leontief's paradox could not be completely eliminated with the yield of 12.7% (conservative estimation); However, using 9% yield can completely eliminate Leontief paradox (see Table 7).
Table 4 (Note: These estimates underestimate the human capital of all kinds of workers, because they ignore the expenses needed to produce a manual worker. ) 1959 wage income of American workers and human capital of various workers (unit: USD)
The average wage of workers * exceeds the human capital of manual workers+
The average wage rate of return is 9%
12.7%
Professional technicians 94146010 47336 66790
Civilian secretarial staff 5 935 2 53219 937 28131
Skilled workers 5 982 2 579 20 365 438+065 438+0 28 638
Production worker 49131510189416782.
Manual workers ++ 3403
* The average wage is the weighted average of the wages of various professional workers, and the income of farmers and freelancers represents the average wage.
+Human capital is the value obtained by dividing the value in the corresponding second column by the rate of return, which is not adjusted according to the asset life.
++Farmers and freelancers are classified as manual workers, which is different from demography and Leontief classification.
Table 5 1947 Export products and import substitutes/KLOC-Proportion of various types of workers in the labor demand of $0/million
Worker type export import
Professional technicians 13.75 12.24
Civilian secretarial staff 22.07 17.00
Technician15.1511.79
Production workers 30.05 28.38
Manual workers 18.98 30.59
Table 6 1947 human capital demand of American exported products and imported substitutes per million dollars.
Revenue (%) thousand dollars
Export import
12.7 17.56 14.95
9.0 24.78 2 1. 10
Table 7 Material capital and total capital demand per 10,000 US dollars of American export products and imported substitutes (unit: 1000 US dollars/year labor force, price 1947)
Conversion value of project human capital
Hourly income consumer price index
Export import export import
Physical capital11.6213.6611.6213.66.
Physical capital and human capital are in 21.38 21.97 25.08 25.12.
12.7% yield
When the yield is 9.0%, it is 25.39 25.38 30. 1 29.83.
Note: Human capital is first converted into the average hourly wage of manufacturing industry through bls index, and then converted into the value of 1947 through consumer price index.
Another factor related to human capital is the impact of scientific research on American exports. The "knowledge" capital brought by scientific research progress has improved the output level of the same amount of material and human resources. Even the roughest statistics show that most American exports are scientific research and technology-intensive.
Kisin's article published in 1996 pointed out that for a considerable number of manufacturing industries, international trade is based on the differences in skill endowments among countries. He believes that Leontief's discovery in 1947 is not so much a paradox as that the United States has a comparative advantage in technology and exports high-skill intensive products. In order to test his hypothesis, Keehing made an empirical analysis of the import and export of 46 industries in 14 countries 1962 industries by using the American technical table. The results show that the technology intensity of American export products is much higher than that of other countries. In order to further confirm that the United States has a comparative advantage in technology, Kising has made a single correlation analysis and rank correlation analysis on the comparative advantage of American industry, the percentage of American industrial export value in the total export of 14 countries, and the various labor demands of the United States for this industry. The results of correlation analysis further confirm that the United States has a comparative advantage in high-tech labor, and its trade model conforms to the factor endowment theory.
Third, the theoretical basis of Leontief's mystery.
Rimmer et al (Rimmer,1980; Casas, 1985:aw, 1983) thinks that the root of Leontief's mystery lies in the wrong theoretical basis of Leontief's inference. In the article 1980, Limer strictly proved that in a multi-commodity world, comparing the capital intensity between production and consumption or comparing the capital intensity between net export and production and consumption is the correct way to determine whether a country's capital is abundant relative to its labor force; Leontief's method, that is, by comparing the relative capital intensity of export goods and import goods, is only applicable to the simple two-commodity world. The following is a detailed introduction of Limer 1980' s explanation of Leontief's mystery.
The theoretical starting point of Limer's 1980 is Heckschel-Olin-Vanich equation, namely:
a t[，i]=e[，i]-e[，w] a[，i] i= 1，2…i ( 1)
Where A represents the input-output matrix of n×n, t[, i] represents the net export vector of goods of country I, e[, i] represents the factor endowment vector of country I, e[, w] represents the factor endowment vector of the world, E [,W] = ∑ E [,I], and A [,I] represents the proportion of consumption of country I in the total consumption of the world.
From equation (1), we can get two equations about factor endowment and factor demand of net export:
k[，t]=k[，i]-a[，i] k[，w] (2a)
l[，t]=l[，i]-a[，i] l[，w] (2b)
Among them, (k[, t], l[, t]) respectively represent the flow of capital and services reflected in net exports, (k[, i], l[, i]) represent the endowment of capital and services in China, and (k[, w], l[, w]) represent the endowment of the world.
In order to understand Limer's analysis, let's first introduce Williams' concept of element abundance. If the ratio of a country's capital stock to the world's total capital stock exceeds the corresponding labor ratio, that is, k[, i]/k[, w]≥l[, i]/l[, w], then the country's capital is abundant relative to labor.
Now, let's introduce Limer's theoretical framework for judging the relative abundance of factors, and its main points are the following five theorems.
Theorem 1
Only the following formula can be satisfied:
k[，i]/(k[，i]-k[，t])>l[，i]/(l[，i]-l[，t]) (3)
The capital of country I is abundant relative to the labor force.
Starting from the theorem 1, we can get the following three inferences. If a country is relatively rich in capital, that is, k [,i]/k [,w] > l [,t]/l [,w], the following three inequalities hold:
k[，i]/l[，i]>k[，c]/l[，c] (3a)
-k[，i] l[，t]>-l[，i] k[，t] (3b)
-k[，c] l[，t]>-l[，c] k[，t] (3c)
Where (k[, c]l[, c]) represents the flow of capital and services embodied in consumption, therefore, k [,c] = k [,i]-k [,t], l[, c] = l [,i]-l [,t].
From inequalities 3b and 3c, we can draw the following conclusions. If country I is a net exporter of capital service flow and labor service flow, only when the capital intensity in net export is greater than that in production or consumption, that is, k [,t]/l [,t] > k [,i]/l [,i] or k [,t]/l [,t] > k [,c]/l.
Theorem 2
If the sign of net export value of capital service flow and labor service flow is opposite, the factor of positive net export value is the factor of relative wealth of the country.
Theorem 3
If the signs of capital service flow and labor service flow reflected by a country's net export are opposite, then only when the country's capital is relatively abundant, that is, K [,I]/K [,W] > L [,I]/L [,W], the capital intensity of export goods will exceed that of imported goods.
Theorem 4
In a multi-commodity trade world, the order of factor density of export commodities and import commodities, such as k [,x]/l [,x] > k [,m]/l [,m], cannot determine the relative abundance of the country's factors.
Theorem 5
If there are only two kinds of goods, one is export goods and the other is import goods, the order of capital intensity of import and export goods is consistent with the order of factor richness of the country, that is, only when k [,I]/k [,w] ≥ l [,I]/l [,w], k [,x]/l [,x] ≥
On the basis of the above theoretical analysis, Limer re-examined Leontief's findings in 1 947 by using the statistical data of Leontief and travis (tables1,8 and 9). As can be seen from Table 9, from 65438 to 0947, the United States is a net exporter of capital service flows and labor service flows. Therefore, it can be seen from Theorem 4 that the theoretical basis of Leontief's inference is that it is wrong to judge the relative abundance of American capital and labor by comparing the relative order of capital intensity of export goods and imported goods. The correct method, as stated in theorem 1, should be to compare the relative capital intensity of net export and production, or the relative capital intensity of net export and consumption. From table 10, we can find that the capital intensity of American net exports is far greater than consumption, so we can be sure that American capital is abundant relative to labor. (Note: If the net export value of American service flow is positive, it means that the per capita national income level or per capita consumption level of the United States is lower than the world average, which is not true. Therefore, a new paradox appears in the Heckschel-Olin-Vanich equation. For the discussion of the new paradox, please refer to francois r.casas's Leontief Paradox, Continue or Solve, Political Economy, June 1985, pp. 6 10 ~ 6 15. )
Table 8 Other Trade and Factor Endowment Data of the United States 1947
Trade and factor value
Exports166.784 million USD.
Imported substitutes: USD 665,438+USD 75.7 million.
The net export value (kt) of capital service flows is $23.45 billion.
Net export value of labor flow (lt) 1 9.9 million years of labor.
The capital intensity (kt/lt) in trade is 1 1, 783 USD/year.
Table 9 (Note: Table 8 ~ Table 10 is quoted from E.E. Leamer's Identification of Leon Tiff's Paradox, Journal of Political Economy, June 1980, pp. 495 ~ 503. ) American production, net export and consumption 1947.
Net export consumption of production *
The capital is 328,565,438+09 million USD 23,450 million USD 305,069 million USD.
472.73 million years of labor, 1.99 million years of labor, 45.28 million years of labor.
Capital/labor: 6949 USD/annual labor: 65438 USD +065438 USD +0.783 USD/annual personnel: 6737 USD/annual labor.
worker
* Consumption data is calculated by the following formula: consumption = production-net export.
Fourthly, other explanations of Leontief's mystery.
Another explanation for Leontief's mystery is that Leontief used a model of two elements (capital and labor), ignoring the influence of other elements such as natural resources (land, mineral deposits, forests, etc.). If a commodity is natural resource intensive, it is obviously incorrect to classify it as capital intensive or labor intensive in the two-factor model. Diab( 1956) divides American trade goods into two categories: processed goods and primary goods. In his research, he found that: (a) among the imported substitutes in the United States, primary products account for an absolute advantage, accounting for about 65%, while export commodities only account for15%; (b) In the United States, the capital-labor ratio of primary products is greater than that of finished products, while the capital-labor ratio of the two types of products in the corresponding export and import substitutes is almost the same. Therefore, Deiab believes that the high capital intensity of imported substitutes in the United States reflects the scarcity of natural resources in the United States. Vanek( 1963) provides further evidence for diab's conclusion. Vaneck found that the ratio of labor, capital and natural resources required for exporting and importing substitutes in the United States was 1.07, 0.83 and 0.54 respectively. He further believes that natural resources and capital are complementary in the production process. On the basis of the above findings, Vaneck believes that the reason for the high capital intensity of imported substitutes in the United States is that natural resources, a scarce factor of production in the United States, can only be effective in production if it is combined with a large amount of capital.
American tariff policy is an important factor in Leontief's mystery. Tariff is actually a tax on imported goods, which can reduce imports and stimulate the production of domestic imported substitutes. In the research of 1964 and 1972, Travis found that the most closely protected industries in the United States are labor-intensive industries, which affected the American trade model and reduced the labor-intensive degree of imported substitutes in the United States. Baldwin( 197 1) believes that travis' research is an important explanation of Leontief's mystery, that is, American tariffs have changed American trade patterns. However, even considering the tariff factor, the capital-labor ratio of imported substitutes in the United States has only dropped by 5%, which is not enough to explain the mystery of Leontief.
The inversion of factor density is also a representative explanation of Leontief's mystery. The inversion of factor density refers to the situation that a given commodity is a labor-intensive product in a country with abundant labor and a capital-intensive product in a country with abundant capital. Once the factor density inversion occurs, the factor equilibrium theorem no longer holds, so the Hertz-Russian theory no longer holds. (Note: For the discussion on the inversion of factor density, see Dominic Solvi Torre, International Economics, Tsinghua University Press, 1998, p. 103- 104. )
H-o theory considers problems from a static point of view. Some economists (Posner,1961; Vernon, 1966) thinks that perhaps what really affects the trade pattern is some dynamic factors, such as technological changes. Posner put forward the technology gap theory in 196 1, arguing that the United States, as the most technologically advanced country, exports a large number of high-tech products. However, when foreign producers acquire new technologies, they will eventually occupy foreign markets, even the American market, with lower labor costs. Vernon further developed Posner's theory and put forward the product life cycle theory in 1966. The main points of this theory are as follows: (1) The life cycle of general products consists of three development stages: new products, mature products and standard products; (2) At different stages of product development, the proportion of capital and labor is different. Generally speaking, in the new product stage, skilled workers are most needed; In the mature stage, marketing and capital dominate; In the standardization stage, a large number of unskilled workers and raw materials are most needed. Vernon thinks his theory is helpful to explain Leontief's mystery. Because the products exported by the United States are in the new product stage, the production process is more dependent on the input of skilled labor, while when the United States imports this product, the product has developed to the standardization stage, and the production process is more dependent on capital. (Note: The above viewpoint is quoted from Chacholiades, M. 1978. International Trade Theory and Policy, new york: McGraw-Hill Press, pp. 298-306. )
Although Leontief's mystery has been successfully explained in theory, empirical research shows that the trade patterns of countries in international trade are far from Herzog-Russia theory. (Note: In 1987, Bowen, Limer and Svekowska used Limer's theoretical framework for empirical analysis, and used more complete cross-sectional data of 27 countries 12 in 1987, and found that Herzog-Russia theory was only established in half cases. Therefore, when China is about to join the WTO, it is undoubtedly helpful for China to choose its trade model and trade policy after its entry into the WTO by reviewing various explanations of Leontief's mystery and using the empirical methods of western scholars to examine whether there is Leontief's mystery in China.
See this webpage for details: /Browse/ShowDoc.aspx? Code = 3-1&; id = 200507 1 1 10 1 1 1885256