International Trade and globalization

Transcription

1 International Trade and globalization Chapter 3: Globalization, employment et wages: The factor endowment model of trade Thierry Mayer September 18, 2011

2 Part I The factor endowment model of trade

3 Introduction Ricardo vs Heckscher-Ohlin The hypotheses of the Ricardian model are so restrictive that they make it impossible to analyse key issues in trade: 1. Internal redistributive effects of trade can be very large. One needs several factors of production. 2. Countries are almost never specialized in the production of one good. The Heckscher-Ohlin model introduces the possibility for production factors to be substituted.

4 Introduction Ricardo vs Heckscher-Ohlin The hypotheses of the Ricardian model are so restrictive that they make it impossible to analyse key issues in trade: 1. Internal redistributive effects of trade can be very large. One needs several factors of production. 2. Countries are almost never specialized in the production of one good. The Heckscher-Ohlin model introduces the possibility for production factors to be substituted.

5 Introduction Ricardo vs Heckscher-Ohlin The hypotheses of the Ricardian model are so restrictive that they make it impossible to analyse key issues in trade: 1. Internal redistributive effects of trade can be very large. One needs several factors of production. 2. Countries are almost never specialized in the production of one good. The Heckscher-Ohlin model introduces the possibility for production factors to be substituted.

6 Introduction Ricardo vs Heckscher-Ohlin The hypotheses of the Ricardian model are so restrictive that they make it impossible to analyse key issues in trade: 1. Internal redistributive effects of trade can be very large. One needs several factors of production. 2. Countries are almost never specialized in the production of one good. The Heckscher-Ohlin model introduces the possibility for production factors to be substituted.

7 Introduction What is the added value of HOS? The model is helpful to analyse the link relating (internal) inequalities to trade. Key message of the HOS model: International trade generates a net gain for all counties (as in Ricardo). but within a county, there are winners and losers. The original HOS model: 2 production factors (capital and labor). Can be reinterpreted as skilled and unskilled labor. Helpful in the political economy of trade liberalization. Why do people form coalitions against free trade?

8 Introduction What is the added value of HOS? The model is helpful to analyse the link relating (internal) inequalities to trade. Key message of the HOS model: International trade generates a net gain for all counties (as in Ricardo). but within a county, there are winners and losers. The original HOS model: 2 production factors (capital and labor). Can be reinterpreted as skilled and unskilled labor. Helpful in the political economy of trade liberalization. Why do people form coalitions against free trade?

9 Introduction What is the added value of HOS? The model is helpful to analyse the link relating (internal) inequalities to trade. Key message of the HOS model: International trade generates a net gain for all counties (as in Ricardo). but within a county, there are winners and losers. The original HOS model: 2 production factors (capital and labor). Can be reinterpreted as skilled and unskilled labor. Helpful in the political economy of trade liberalization. Why do people form coalitions against free trade?

10 Introduction HOS helps understand labor market issues in an open economy Is trade with Southern countries to blame for the increase of inequalities in Northern countries? HOS makes it possible to answer this question in an analytical framework that can quantify the role played by trade compared to the other usual suspects : biased technical progress less powerful unions less redistributive policies immigration

11 Introduction HOS helps understand labor market issues in an open economy Is trade with Southern countries to blame for the increase of inequalities in Northern countries? HOS makes it possible to answer this question in an analytical framework that can quantify the role played by trade compared to the other usual suspects : biased technical progress less powerful unions less redistributive policies immigration

12 Assumptions Setting 1. 2 countries, H and F goods, X and Y production factors, K and L. 4. identical production technologies across countries. 5. different capital intensities between the two goods. 6. different relative factor endowments between countries. 7. constant returns to scale. 8. no distorsion. 9. identical and homogenous preferences. 10. homogenous factors perfectly mobile between sectors and perfectly immobile between countries.

13 Assumptions Setting 1. 2 countries, H and F goods, X and Y production factors, K and L. 4. identical production technologies across countries. 5. different capital intensities between the two goods. 6. different relative factor endowments between countries. 7. constant returns to scale. 8. no distorsion. 9. identical and homogenous preferences. 10. homogenous factors perfectly mobile between sectors and perfectly immobile between countries.

14 Endowments and intensities Factor endowments H is capital-rich: ( ) ( ) K K > L h L f Factor Intensities. Y s production is capital-intensive: ( ) ( ) K K > L L Contrary to the technological model, technical coefficients are variable. Y X

15 Illustration of factor endowments / intensities in 1984 Factor endowments Country K L K/L (000USD) India South Korea USA Switzerland Factor intensities (USA) Industry K L K/L (000USD) Oil refining Transport Shoes Clothing

16 Production side of the model Production function, described by isoquants: X = F (K,L) 1. Decreasing returns to each factor. 2. Constant returns to scale. The behavior of the producer in based on iso-cost curve: C 0 = wl + rk, On any isoquant, K = C 0 /r (w/r)l K L = w r dx = X X X dl + dk = 0 dk L K dl = L X Condition for an efficient production: w r = X L X K K

17 Production side of the model Production function, described by isoquants: X = F (K,L) 1. Decreasing returns to each factor. 2. Constant returns to scale. The behavior of the producer in based on iso-cost curve: C 0 = wl + rk, On any isoquant, K = C 0 /r (w/r)l K L = w r dx = X X X dl + dk = 0 dk L K dl = L X Condition for an efficient production: w r = X L X K K

18 Production side of the model Production function, described by isoquants: X = F (K,L) 1. Decreasing returns to each factor. 2. Constant returns to scale. The behavior of the producer in based on iso-cost curve: C 0 = wl + rk, On any isoquant, K = C 0 /r (w/r)l K L = w r dx = X X X dl + dk = 0 dk L K dl = L X Condition for an efficient production: w r = X L X K K

19 Production side of the model Production function, described by isoquants: X = F (K,L) 1. Decreasing returns to each factor. 2. Constant returns to scale. The behavior of the producer in based on iso-cost curve: C 0 = wl + rk, On any isoquant, K = C 0 /r (w/r)l K L = w r dx = X X X dl + dk = 0 dk L K dl = L X Condition for an efficient production: w r = X L X K K

20 Production side of the model Production function, described by isoquants: X = F (K,L) 1. Decreasing returns to each factor. 2. Constant returns to scale. The behavior of the producer in based on iso-cost curve: C 0 = wl + rk, On any isoquant, K = C 0 /r (w/r)l K L = w r dx = X X X dl + dk = 0 dk L K dl = L X Condition for an efficient production: w r = X L X K K

21 Two goods / Two factors / One country X = F x (L x,k x ) Y = F y (L y,k y ) K = K x + K y L = Lx + L y k y = K y /L y > k x = K x /L x Y is capital-intensive and X is labor-intensive.

22 Isoquants K K K/L Y 0 k y K X 1 =20 k x X 0 =10 L 0 L 0 X 0 L

23 The Production-possibility frontier (PPF) Y KX X 3 LY 0Y X 2 X 0 A X 1 B C C B D Y 0 0X Y 2 Y 1 KY LX 0 X Y 3

24 Equilibrium in autarky Y A p a 0 X

25 Equilibrium under free trade Y C 2 Q 1 A C 1 p 1 * Q 2 p a p 2 * 0 X

26 Consequences of the differences in terms of factor endowment K Y Y h k h Y h Y f K h Y f E h k f E f X f X h 0 L h L 0 X h X f X

27 Autarkic equilibrium in the HO model Y Y h A h Y f A f p h I f I h p f 0 X h X f X

28 The theorem Y Y h Q h I * h Y f B h C h C f I * f B f Q f p * 0 X h X f X

29 The theorem In autarky, p h > p f. Opening-up makes p h fall and p f increase, until: 1. Relative prices are equalized 2. This relative price is compatible with goods market equilibrium on the world market. Openness result in an equalization of relative prices of goods. The two countries continue to produce both goods after opening. Both countries gain to trade. The Heckscher-Ohlin theorem: A country exports the good using intensively the factor it has in relative abundance.

30 The theorem In autarky, p h > p f. Opening-up makes p h fall and p f increase, until: 1. Relative prices are equalized 2. This relative price is compatible with goods market equilibrium on the world market. Openness result in an equalization of relative prices of goods. The two countries continue to produce both goods after opening. Both countries gain to trade. The Heckscher-Ohlin theorem: A country exports the good using intensively the factor it has in relative abundance.

31 The theorem In autarky, p h > p f. Opening-up makes p h fall and p f increase, until: 1. Relative prices are equalized 2. This relative price is compatible with goods market equilibrium on the world market. Openness result in an equalization of relative prices of goods. The two countries continue to produce both goods after opening. Both countries gain to trade. The Heckscher-Ohlin theorem: A country exports the good using intensively the factor it has in relative abundance.

32 Relationship relating the price of goods to the price of factors I - The Stolper-Samuelson theorem. Trade equalizes the prices of goods, is it the same with the prices of factors? After the opening-up to trade and the rise of the relative price of the labour-intensive good, the relative price of labour also increases. With p X /p Y X and Y. The X sector demands more of both factors and the Y sector demands less of both, but not in the same proportions.

33 Relationship relating the price of goods to the price of factors I - The Stolper-Samuelson theorem. Trade equalizes the prices of goods, is it the same with the prices of factors? After the opening-up to trade and the rise of the relative price of the labour-intensive good, the relative price of labour also increases. With p X /p Y X and Y. The X sector demands more of both factors and the Y sector demands less of both, but not in the same proportions.

34 Stolper-Samuelson - I Y KX X b LY 0Y (w/r) b A X a k xb (w/r) a B B p a C A Y b 0 X 0X KY LX p* Y a

35 Stolper-Samuelson - II Y k' y w / r K Y k y k' x (w / r)* k x X X 0 p* P X / P Y 0 L

36 Stolper-Samuelson - III w / r k x (w / r) h k y (w / r)* (w / r) f P X / P Y p a p* 0 h p a f k * x k * y k

37 The Stolper-Samuelson theorem Mathematical developments show that the opening-up to trade leads to: 1. An increase in the relative price of the labor-intensive good generating an increase in the relative price of labor. 2. A magnification effect: ŵ ˆr > ˆp X ˆp Y > 0. The increase in the relative price of labor is greater than the increase in the relative price of the labor intensive good. 3. There is an increase in the real income of labor and a fall in the real income of capital

38 The Stolper-Samuelson theorem Mathematical developments show that the opening-up to trade leads to: 1. An increase in the relative price of the labor-intensive good generating an increase in the relative price of labor. 2. A magnification effect: ŵ ˆr > ˆp X ˆp Y > 0. The increase in the relative price of labor is greater than the increase in the relative price of the labor intensive good. 3. There is an increase in the real income of labor and a fall in the real income of capital

39 The Stolper-Samuelson theorem Mathematical developments show that the opening-up to trade leads to: 1. An increase in the relative price of the labor-intensive good generating an increase in the relative price of labor. 2. A magnification effect: ŵ ˆr > ˆp X ˆp Y > 0. The increase in the relative price of labor is greater than the increase in the relative price of the labor intensive good. 3. There is an increase in the real income of labor and a fall in the real income of capital

40 The Stolper-Samuelson theorem Mathematical developments show that the opening-up to trade leads to: 1. An increase in the relative price of the labor-intensive good generating an increase in the relative price of labor. 2. A magnification effect: ŵ ˆr > ˆp X ˆp Y > 0. The increase in the relative price of labor is greater than the increase in the relative price of the labor intensive good. 3. There is an increase in the real income of labor and a fall in the real income of capital

41 Factor Price Equalization: HOS theorem The Heckscher-Ohlin-Samuelson theorem: The free trade of goods equalizes relative incomes of factors between the two countries through the equalization of relative goods prices. This equalization takes place as long as both countries continue to produce both goods.

42 The limits of factor price equalization 1. Transport costs are large in a lot of industries. 2. Trade barriers such as tariffs and quotas persist. 3. Distorsions of all kinds are pervasive. 4. Technologies are not really identical. The important aspect of the HOS theorem is not its version in levels, but much more its version in tendency.

43 The limits of factor price equalization 1. Transport costs are large in a lot of industries. 2. Trade barriers such as tariffs and quotas persist. 3. Distorsions of all kinds are pervasive. 4. Technologies are not really identical. The important aspect of the HOS theorem is not its version in levels, but much more its version in tendency.

44 24 FPE in level fails CHAPTER 1. BUSINESS ACROSS BORDERS Largest 52 Economies in 2000 (sorted by real income) Income per capita, 1000s of PPP$ United States Canada Japan Germany United Kingdom France Italy Spain Korea Mexico Brazil Russia China Equal Incomes Indonesia India Cumulative Population (mil.)

45 Part II HOS: empirical validations

46 HOS: empirical validations 1st generation: Leontief (1953) paradox US Exports should be capital-intensive and imports labor-intensive. Input-output tables of the US economy US substitutes for imports, rather than imports. Computes K m/l m K x /L x, Ratio obtained = 1,3. Table: Capital and labor required for the US production of 1 mn USD exports and import substitutes in Exports Import Capital ($ mn) Labor (person/year) K/L ($/person) 13,700 18,200

47 HOS: empirical validations 1st generation: Leontief (1953) paradox US Exports should be capital-intensive and imports labor-intensive. Input-output tables of the US economy US substitutes for imports, rather than imports. Computes K m/l m K x /L x, Ratio obtained = 1,3. Table: Capital and labor required for the US production of 1 mn USD exports and import substitutes in Exports Import Capital ($ mn) Labor (person/year) K/L ($/person) 13,700 18,200

48 HOS: empirical validations 1st generation: Leontief (1953) paradox US Exports should be capital-intensive and imports labor-intensive. Input-output tables of the US economy US substitutes for imports, rather than imports. Computes K m/l m K x /L x, Ratio obtained = 1,3. Table: Capital and labor required for the US production of 1 mn USD exports and import substitutes in Exports Import Capital ($ mn) Labor (person/year) K/L ($/person) 13,700 18,200

49 HOS: empirical validations 2st generation: first critics 1. Technology is different. One American worker is worth 3 foreign workers. What about capital? 2. Labor is heterogenous. 3. Impossible to consider imports without domestic substitutes. 4. Other factors might be important (land...). 5. US tarifs hit in particular labor intensive goods. 6. Where are the factor endowments in Leontief s method?

50 HOS: empirical validations 2st generation: first critics 1. Technology is different. One American worker is worth 3 foreign workers. What about capital? 2. Labor is heterogenous. 3. Impossible to consider imports without domestic substitutes. 4. Other factors might be important (land...). 5. US tarifs hit in particular labor intensive goods. 6. Where are the factor endowments in Leontief s method?

51 HOS: empirical validations 2st generation: first critics 1. Technology is different. One American worker is worth 3 foreign workers. What about capital? 2. Labor is heterogenous. 3. Impossible to consider imports without domestic substitutes. 4. Other factors might be important (land...). 5. US tarifs hit in particular labor intensive goods. 6. Where are the factor endowments in Leontief s method?

52 HOS : empirical validations 3rd generation : Heckscher-Ohlin-Vanek (1968) predictions A country will export the services of a factor when its share in world endowment of this factor exceeds its share of world GDP (demand in fact). Net export of a factor s services = net exports of goods factor content of goods (technical coefficients matrix). Prediction for each factor/country combination: 1. Net exports of a factor s services = share of factor endowment in the world total - share of the world GDP. 2. Signs should at least be the same, correlation should be one (harder).

53 HOS : empirical validations 3rd generation : Heckscher-Ohlin-Vanek (1968) predictions A country will export the services of a factor when its share in world endowment of this factor exceeds its share of world GDP (demand in fact). Net export of a factor s services = net exports of goods factor content of goods (technical coefficients matrix). Prediction for each factor/country combination: 1. Net exports of a factor s services = share of factor endowment in the world total - share of the world GDP. 2. Signs should at least be the same, correlation should be one (harder).

54 HOS : empirical validations 3rd generation : Heckscher-Ohlin-Vanek (1968) predictions A country will export the services of a factor when its share in world endowment of this factor exceeds its share of world GDP (demand in fact). Net export of a factor s services = net exports of goods factor content of goods (technical coefficients matrix). Prediction for each factor/country combination: 1. Net exports of a factor s services = share of factor endowment in the world total - share of the world GDP. 2. Signs should at least be the same, correlation should be one (harder).

55 Factor endowments in 2000

56 HOS: empirical validations Trefler ( ): The simple model is too simple. Trefler takes 9 factors, 33 countries in Sign test works for 50% of cases. Correlation is only In reality: net factor exports are almost zero: this is the mystery of the missing trade compared to differences in factor endowments.

57 HOS: empirical validations Trefler ( ): The simple model is too simple. Trefler takes 9 factors, 33 countries in Sign test works for 50% of cases. Correlation is only In reality: net factor exports are almost zero: this is the mystery of the missing trade compared to differences in factor endowments.

58 HOS: empirical validations Trefler ( ): The simple model is too simple. Trefler takes 9 factors, 33 countries in Sign test works for 50% of cases. Correlation is only In reality: net factor exports are almost zero: this is the mystery of the missing trade compared to differences in factor endowments.

59 HOS: empirical validations Trefler ( ): The simple model is too simple. Trefler takes 9 factors, 33 countries in Sign test works for 50% of cases. Correlation is only In reality: net factor exports are almost zero: this is the mystery of the missing trade compared to differences in factor endowments.

60 HOS: empirical validations Solutions: 1. Introduce technological differences à la Leontief. 2. Ricardo clearly improves HO: The correlation increases from 0.28 to Sign test increases from 50% to 72%. HOS/HOV in its initial form is too strict. Introducing Ricardian elements leads to better results. Corrections can bring much better fit to the data.

61 HOS: empirical validations Solutions: 1. Introduce technological differences à la Leontief. 2. Ricardo clearly improves HO: The correlation increases from 0.28 to Sign test increases from 50% to 72%. HOS/HOV in its initial form is too strict. Introducing Ricardian elements leads to better results. Corrections can bring much better fit to the data.

62 HOS: empirical validations Solutions: 1. Introduce technological differences à la Leontief. 2. Ricardo clearly improves HO: The correlation increases from 0.28 to Sign test increases from 50% to 72%. HOS/HOV in its initial form is too strict. Introducing Ricardian elements leads to better results. Corrections can bring much better fit to the data.

63 Effective factor endowments in 2000

64 Part III Trade and employment/wages

65 Trade and employment/wages HOS is a convenient tool to analyse the effect of opening to trade on factors income (and inequalities). Ricardo and HOS: trade specialization there are losing sectors. Entails restructuring costs: professional training, country planning... HOS: Stolper-Samuelson theorem Scarce factors lose from trade (fall of their real income) necessity to compensate (redistributive policies). Some people clearly loose from the globalization. However, the question of the size of the impact of trade openness on labor markets remains an open one.

66 Trade and employment/wages HOS is a convenient tool to analyse the effect of opening to trade on factors income (and inequalities). Ricardo and HOS: trade specialization there are losing sectors. Entails restructuring costs: professional training, country planning... HOS: Stolper-Samuelson theorem Scarce factors lose from trade (fall of their real income) necessity to compensate (redistributive policies). Some people clearly loose from the globalization. However, the question of the size of the impact of trade openness on labor markets remains an open one.

67 Trade and employment/wages HOS is a convenient tool to analyse the effect of opening to trade on factors income (and inequalities). Ricardo and HOS: trade specialization there are losing sectors. Entails restructuring costs: professional training, country planning... HOS: Stolper-Samuelson theorem Scarce factors lose from trade (fall of their real income) necessity to compensate (redistributive policies). Some people clearly loose from the globalization. However, the question of the size of the impact of trade openness on labor markets remains an open one.

68 Trade and employment/wages: earning inequalities USA Relative Wage Figure 1.1: Relative Wage of Nonproduction/Production Workers, U.S. Manufacturing

69 Trade and employment/wages: earning inequalities OECD Evolution des écarts de salaires Etats-Unis UK Sources : OCDE (1996)

70 Trade and employment/wages: earning inequalities OECD Evolution des écarts de salaires Japon Sources : OCDE (1996)

71 Trade and employment/wages: earning inequalities OECD Evolution des écarts de salaires France Allemagne Italie Sources : OCDE (1996)

72 Trade and employment/wages: earning inequalities OECD Evolution des écarts de salaires Etats-Unis France UK Japon Allemagne Italie Sources : OCDE (1996)

73 estimated to translate into job and wage losses that are approximately three times larger than those that would result from an equally large shock a Trade and employment/wages: earning inequalities low offshoring industry. The rise in the responsiveness of employment and wages to demand shocks has been strongest in manufacturing industries, OECD ALITY WIDEN to rnings 1 Note: The figure shows that in all countries except Ireland, Japan and Spain, the earnings of the 10% best-paid workers increased more than the earnings of the 10% least-paid workers, over the period (i.e. earnings inequality widened). 1. Full-year, full-time workers. The data shown are consistent over time, but not entirely comparable across countries owing to differences in pay reporting periods and coverage of workers. 2. Unweighted average of countries shown in the figure. Source: OECD, Earnings Distribution Database.

74 Trade and employment/wages: France (very long run) Part du premier décile de revenus en France et Cf. Thomas Piketty

75 Trade and employment/wages: unemployment OECD 15 Taux de chomage UK 10 5 USA Source : Eurostat

76 Trade and employment/wages: unemployment OECD 15 Taux de chomage 10 5 Japon Source : Eurostat

77 Trade and employment/wages: unemployment OECD 15 Taux de chomage France Italie 10 Allemagne Source : Eurostat

78 Trade and employment/wages: unemployment OECD 15 Taux de chomage UK France Italie 10 Allemagne 5 Japon USA Source : Eurostat

79 Trade and employment/wages: unemployment France Taux de chomage qualifiés / non-qualifiés Moins qualifiés Plus qualifiés

80 Trade and employment/wages: unemployment France Skill Premium (diff in unemployement rate) tertiary education/no diploma year skill premium, 11 years out skill premium, 5 10 years out skill premium, 1 4 years out source: INSEE

81 Trade and employment/wages: share of skilled workers 28 France : Part de l'emploi non-qualifié (%) % de l'emploi total 25 % de l'emploi équiv. temps plein

82 Trade and employment/wages: share of skilled workers 60 Taux de variation de l'effectif en fonction de la qualification initiale 27 (Informatique) 40 Taux de variation de l'effectif, (Viande) 53 (Transf. mat. plast.) 51 (Presse édition) 37 (Conserves) (Habillement) 10 (Sidérurgie) 19 (Prod. pharm) 33 (Aéronautique) Droite de régression y = 1.46x R 2 = Part des qualifiés dans la main-d'oeuvre en 1977 (cf. encadré I.1 pour la définition des travailleurs qualifiés) CF. Cortes et Jean

83 Trade and employment/wages Evolution of inequalities Anglo-saxon countries: growth of inequalities. Other developed countries : growth of unemployment (+ inequalities facing unemployment). Evolution of the employment structure Increase of the relative employment of skilled workers (in the overall economy and in each sector). Is it a matter of labor demand or of labor supply?

84 Trade and employment/wages Evolution of inequalities Anglo-saxon countries: growth of inequalities. Other developed countries : growth of unemployment (+ inequalities facing unemployment). Evolution of the employment structure Increase of the relative employment of skilled workers (in the overall economy and in each sector). Is it a matter of labor demand or of labor supply?

85 Trade and employment/wages Evolution of inequalities Anglo-saxon countries: growth of inequalities. Other developed countries : growth of unemployment (+ inequalities facing unemployment). Evolution of the employment structure Increase of the relative employment of skilled workers (in the overall economy and in each sector). Is it a matter of labor demand or of labor supply?

86 Trade and employment/wages 0.5 Relative Employment Figure 1.2: Relative Employment of Nonproduction/Production Workers, U.S. Manufacturing

87 Trade and employment/wages Proportion de bacheliers dans une génération Source : INSEE première

88 Trade and employment/wages Diplôme des ans Source : INSEEE

89 Trade and employment/wages Taux de scolarité masculins (hors apprentissage) Source : INSEE

90 Trade and employment/wages Taux de scolarité féminins (hors apprentissage) Sources : Recensements et Enquêtes emploi, Insee

91 Two major explanations Biased technical progress. Trade with low-wages countries.

92 Trade and employment/wages Wood (1995)

93 Trade and employment/wages Counter-argument 1: no decrease of inequalities in the South? Counter-argument 2: Difficulties for HOS to explain trade flows. Counter-argument 3. (Krugman - in Pop internationalism). North-South trade is only a small share of OECD countries trade. not very convincing Counter-argument 4. The observed evolutions do not completely match the predictions of the HOS model. Counter-argument 5. Inequalities increase within sectors as much as across sectors

94 Trade and employment/wages Counter-argument 1: no decrease of inequalities in the South? Counter-argument 2: Difficulties for HOS to explain trade flows. Counter-argument 3. (Krugman - in Pop internationalism). North-South trade is only a small share of OECD countries trade. not very convincing Counter-argument 4. The observed evolutions do not completely match the predictions of the HOS model. Counter-argument 5. Inequalities increase within sectors as much as across sectors

95 Trade and employment/wages Counter-argument 1: no decrease of inequalities in the South? Counter-argument 2: Difficulties for HOS to explain trade flows. Counter-argument 3. (Krugman - in Pop internationalism). North-South trade is only a small share of OECD countries trade. not very convincing Counter-argument 4. The observed evolutions do not completely match the predictions of the HOS model. Counter-argument 5. Inequalities increase within sectors as much as across sectors

96 Trade and employment/wages Counter-argument 1: no decrease of inequalities in the South? Counter-argument 2: Difficulties for HOS to explain trade flows. Counter-argument 3. (Krugman - in Pop internationalism). North-South trade is only a small share of OECD countries trade. not very convincing Counter-argument 4. The observed evolutions do not completely match the predictions of the HOS model. Counter-argument 5. Inequalities increase within sectors as much as across sectors

97 Trade and employment/wages Counter-argument 1: no decrease of inequalities in the South? Counter-argument 2: Difficulties for HOS to explain trade flows. Counter-argument 3. (Krugman - in Pop internationalism). North-South trade is only a small share of OECD countries trade. not very convincing Counter-argument 4. The observed evolutions do not completely match the predictions of the HOS model. Counter-argument 5. Inequalities increase within sectors as much as across sectors

98 Trade and employment/wages It is hard to disentangle the role played by trade from the role played by biased technical progress. The direct effects of trade (Stolper-Samuelson effect) appear quite limited (research on this question finds). However, international trade can also indirectly impact labor markets: Defensive innovations. An increase in competition reduces firms markups and workers bargaining power. Those call for new models. Among those, some can even explain an increase of inequalities within sectors.

99 Trade and employment/wages It is hard to disentangle the role played by trade from the role played by biased technical progress. The direct effects of trade (Stolper-Samuelson effect) appear quite limited (research on this question finds). However, international trade can also indirectly impact labor markets: Defensive innovations. An increase in competition reduces firms markups and workers bargaining power. Those call for new models. Among those, some can even explain an increase of inequalities within sectors.

100 Trade and employment/wages It is hard to disentangle the role played by trade from the role played by biased technical progress. The direct effects of trade (Stolper-Samuelson effect) appear quite limited (research on this question finds). However, international trade can also indirectly impact labor markets: Defensive innovations. An increase in competition reduces firms markups and workers bargaining power. Those call for new models. Among those, some can even explain an increase of inequalities within sectors.

101 Trade and employment/wages It is hard to disentangle the role played by trade from the role played by biased technical progress. The direct effects of trade (Stolper-Samuelson effect) appear quite limited (research on this question finds). However, international trade can also indirectly impact labor markets: Defensive innovations. An increase in competition reduces firms markups and workers bargaining power. Those call for new models. Among those, some can even explain an increase of inequalities within sectors.