INTERNATIONAL TRADE AND FACTOR MOBILITY

Transcription

1 ECON 4415: International Economics Autumn 2007 Karen Helene Ulltveit-Moe Lecture 5: INTERNATIONAL TRADE AND FACTOR MOBILITY

2 1 Introduction Simple trade theory is based on the assumption on traded goods and non-traded factors. Not consistent with the real world. in the real world both movements of goods and factors are possible but entail costs, such as transport costs, non-tariff barriers costs, tariffs and other transaction costs. In a world of FPE, trade in goods substitutes for trade in factors, and there is no incentive for factors to move across borders. But the existence of trade costs implies that good trade alone will not equalize factor prices, and there is thus an incentive for factors move. What will determine which commodities goods and factors which will actually be traded in a trade equilibrium?

3 2 The model: a small open economy 2X2 HO framework Good 1: labour intensive, trade incurs trade cost t 1 Good 2: capital intensive, free trade (the numeraire) Factor n, immobile Factor k, internationally mobile at cost t k The home economy is relatively abundantly endowed with labour; so in the absence of factor movements it would have a CA in the production of good 1. Price of good 1: p f 1 =world price of good 1 The foreign wage rate is given by r 1 (p f 1,wf )=r 2 (1,w f )

4 2.1 A model without capital mobility From Figure 1 we can see: ³ k n h ³ ] < k 1 n : Specialized equilibrium in the home 1 country. Good 1 is exported, good 2 is imported (point a, Figure 1) ³ ³ n < kn h < kn C : Diversified equilibrium, 1 and since the home economy consumption is more capital intensive than the home endowment, good 1 will be exported and good 2 will be imported (point b, Figure 1) ]³ k 1 ³ k n C < ³ kn h : At prices p h 1 = p f 1 t 1 the home country will produce less of good 1 than is demanded, so there cannot be any export of good 1. Instead this will lead to p h 1 >pf 1 t 1, which shifts the rate of

5 return curve for good 1 up (ex point c, Figure 1). Since the gap between p h 1 and pf 1 is less than t 1,no trade occurs. See Figure 2 for an overview of equilibria. 2.2 A model with capital mobility The resulting equilibrium depends on trade costs related to goods, trade costs related to capital and initial endowment ratios. From Figure 3 follows that: Area I: The difference in rates of return at home and abroad is less than the cost of moving capital. No capital movement. Trade in goods if ³ ³ k n h < kn C.

6 Area II and III: In these regions t k <r h r f, capital will be imported until t k = r h r f. If the initial configuration is in II, the new equilibrium will still be one of complete specialization; the home country will import capital, export good 1, and import good 2 (since the domestic production of good 2iszero). If the initial configuration is in III, the home country will import capital and good 1, and export good 2 3 International equilibria Key reference: article by Norman and Venables in Economic Journal in 1995

7 3.1 A simple model 2X2 HO framework Good 1: labour intensive, price p i 1,i= h, f, trade incurs trade cost t 1 Good 2: capital intensive, price p i 2,i= h, f, free trade (the numeraire) Factor n, immobile, price w i Factor k, internationally mobile at cost t k, price r i Fixed coefficient technology The home economy is relatively abundantly endowed with labour; so in the absence of factor movements it would have a CA in the production of good 1. This implies that we restrict the analysis to the area under the diagonal in Figure 4.

8 Diversified equilibria in both countries Assumes that transaction costs has a negative impact on consumers utility in the exporting country (simplifying assumption) u = Y e (p 1 ) T e (p 1 )=index of cost of living T = utility that is "wasted" on international transactions T = t 1 tradeingood1+ t k trade in capital In terms of the numeraire, the cost of transaction services is e (p 1 ) T, so the unit cost of transactions will be e (p 1 ) t 1 for good 1 and e (p 1 ) t k for capital.

9 3.1.1 Trade in goods but not in capital Forwhatendowmentswilltherebenotradeingoods, given that there is no trade in capital? Along the diagonal, relative factor endowments are the same in both countries, so relative factor and product prices are also the same. Hence there will be no trade and ³ ³ c 1 c = x1 2 h x. 2 h For some endowments near the diagonal, relative prices will be sufficiently close to make the price differencessmallerthanthetransactioncosts inwhich casetherewillbenotradeandp h 1 > pr 1 e(ph 1 )t 1. The boundaries for the no-goods-trade area (see Figure 4) are accordingly given by Ã! h c1 = Ã x1! h. (1) c 2 x 2

10 p h 1 = pf 1 e(ph 1 )t 1 (2) Trade in capital but not in goods Under what circumstances for what endowments will there be trade in capital, given that there is no trade in goods? Answer: there will be capital trade if the autarky differences in rates of return exceed transaction costs for capital. The boundary of the no-capital trade endowment area must therefore be given by differences in autarky rates of returns equal to the unit of transaction cost for capital, i.e. (since the home country is a potential capital importer): r h = r f + e(p h 1 )t k (3)

11 To allow the two cases of no-trade in goods versus notrade in capital to be compared, we use that the equilibrium rate of returns is given by x r = p 1 1 w n 1 = x 2 w n 2 (4) k 1 k 1 k 2 k 2 and rearranging we may write x w = p 1 1 r k 1 = x 2 r k 2 (5) n 1 n 1 n 2 n 2 and solving the latter of the two equalities for r, weget x 2 x n p 1 r = 2 1 n 1 k 2 n k. (6) 1 2 n 1 Since coefficients are fixed, the difference in autarky rates canbewrittenas with r h r r = B(p h 1 pf 1 ) (7) B x 1/n 1 k 2 n k (8) 1 2 n 1 Using (3), (7) and (8), we have that p h 1 = pf 1 e(ph 1 )t k/b

12 Theboundariesfortheno-capitaltradearea(seeFigure 4) are accordingly given by Ã! h c1 = Ã x1! h. (9) c 2 x 2 p h 1 = pf 1 e(ph 1 )t k/b (10) The no-capital-trade boundary will look just like the nogoods-trade boundary, but will be closer to the diagonal if t k /B < t 1, and further from the diagonal for t k /B > t 1.

13 3.1.3 Trade in capital or goods? Different outcomes are possible depending on transaction cost of capital relative to transaction cost of goods. If t k /B < t 1, two possible outcomes: If countries are fairly similar (endowments within area A), there will be no trade at all. If countries are sufficiently different in terms of endowments to be outside the capital-trade boundary (area B or C), there will be trade in capital but no trade in goods. the home country will import capital, have no trade in good 1, and therefore (since there is balanced trade overall) export good 2. The factor content of consumption must lie on the capital-trade boundary, because this

14 gives the relative prices of good 1 in each country, and hence the consumption ratios. If t k /B > t 1, the capital-trade boundary is irrelevant, since the product price differences can never exceed t 1. So trade in goods will ensure that the rate of return differences are smaller than the unit transaction cost for capital. Two possible outcomes: If countries are fairly similar, there will be no trade at all. If countries are sufficiently different in terms of endowments to be outside the no goods-trade boundary, there will be trade in goods only. The home country will export good 1, and import good 2. The existence of transaction costs implies that trade patterns may turn out rather surprising.