MEASURING COMPETITIVENESS OF TRANSITION ECONOMIES WITH RICARDO AND HECKSCHER-OHLIN

Transcription

1 MEASURING COMPETITIVENESS OF TRANSITION ECONOMIES WITH RICARDO AND HECSCHER-OHLIN Semr Daskapan Delft Unversty of Technology, Jaffalaan 5, Delft, The Netherlands, To measure compettve trade advantage of an economy trade performance ndcators (TPIs are used. Current TPIs do not measure trade performance of transton economes properly due to ther unrealstc assumptons. Frst, the current TPIs underestmate the potental of transton economes, snce they are based on ex post data,.e. before transton, and not on the current or future fgures (of to-be-exported goods. Second, current TPIs do not reflect the Rcardan model of comparatve advantage or Heckscher-Ohln s theory of factor endowments. They merely asses the export or mport shares worldwde. As such the purpose of ths paper s to propose new TPIs that resolve both defcences: the factor prces (or unrevealed comparatve advantage ndcator and the Heckscher- Ohln ndcators. eywords: Comparatve advantage, Internatonal trade, Balassa, Rcardo, Heckscher-Ohln, Transton economes. 1. INTRODUCTION Governments are frequently changng ther nternatonal competton polcy to protect ther local markets. To measure the mpact of ther polcy so-called trade performance ndcators (TPIs are used. Common measures, lke the Balassa, Mchaely and Lafay ndcators for example, use somehow a rato between the export and mport of a country to show the degree of trade specalzaton. Gven the mpact of ther results, t s of hgh mportance that those calcul can be reled on. But many authors lke Marsh and Tokarck (1994, Daskapan (2008a, and Scott (2005 are sceptcal. Especally governments of transton economes would be mslead by those unrelable TPIs. Daskapan (2008b has presented sx lacks n the current TPIs based on two leadng trade theores: the Rcardan model of comparatve advantage and Heckscher-Ohln theory of factor endowments. From al the ndcators the Balassa ndcator appeared to be best but was stll lackng (Balassa The defcences they have exposed n the current TPIs are the followng. The frst concern s symmetrcty of the TPI,.e. the ndcator turns out to produce an output, whch cannot be compared on both sdes of 1. Secondly, a more dsturbng ssue s the effect of ntra ndustry trade. As such a dstorted overpostve result wll be gven. Thrdly, another serous problem s the lack n the current TPIs to nclude not-yet-exported goods. The TPIs measure only the exported goods ex- post; based on that outcome a concluson s drawn about the country s trade specalzaton. Other specalzatons of a country, leadng to goods wth hgh trade potental, but whch are not exported (yet, are as such not exposed. Consequently, the goal of Rcardo to fnd out a country s trade comparatve advantage s mssed. An alternatve soluton to ths problem does not exst yet. Ffth, the gven TPIs do not ndcate trade specalzaton, but rather the relatve export share. Ths relatve export share can be the consequence of numerous factors other than trade specalzaton, lke poltcs and trade barrers. Sxth, measurng trade

2 performance accordng to the Heckscher-Ohln theory (Heckscher 1949, Ohln 1933 lacks an approprate method. Due to ths fact, we are not able to measure and compare dfferences n factor endowments between natons. Especally the thrd, ffth and sxth lacks are affectng the measurement of transton economes. Regardng the thrd lack, transton economes are per defnton evolvng and have not reached yet ther maturty stage. As such many potental trade sectors, wth no ex-post data on trade yet, are excluded n the current calculatons. The compettve power of a transton economy would be underestmated. As for the ffth lack, governments are mslead by (only presentng exportmport shares, snce those are n a transton economy yet low, whereas ther unmeasured potental s hgh. Regardng the sxth lack, transton economes have abundance n several (natural resources that would cause accordng to the Heckscher-Ohln theory an export mpulse. Unfortunately, a TPI based on Heckscher-Ohln theory s lackng to measure ths potental. Therefore, we expect that the current TPIs wll gve ncomplete andor dstorted results when appled to measure the trade performance of transton economes (see fgure 1. As such the purpose of ths paper s to propose two new ndcators to resolve both shortcomngs n the context of transton economes. Economst Statctcans: Use TPIs Governments: make nt. trade polcy Prvate enttes: Int. trade Fgure 1. Actors n a trade process The structure of ths paper s as follows. In secton 2 we wll ntroduce the concept of unrevealed comparatve advantage and propose ts ndcator. In secton 3 ths ndcator s compared wth the Balassa ndcator by applyng both on the same dataset. In secton 4 we wll explan the need of an ndcator that s based on the Heckscher-Ohln theory. In secton 5 we wll propose alternatve ndcators that are based on the Heckscher-Ohln theory. Fnally n secton 6 a concluson and future work s gven. 2. Unrevealed comparatve advantage Davd Rcardo opposed the theory of absolute advantage, by developng the theory of comparatve advantage, whch states that (Mahoney et al 1998 a country should produce and export those goods and servces for whch t s relatvely more productve than other countres and mport those goods and servces for whch other countres are relatvely more productve. The term comparatve advantage refers to the opportunty cost of two producers. The Rcardan model s based on technologcal dfferences across countres, whch result n dfferences n the productvty. The revealed comparatve advantage can be consdered as the frst method to measure comparatve advantage of nternatonal trade specalzaton (Balassa Snce t s one of the oldest TPIs, t has as such been used n many economc analyss. Over all the years some shortcomngs have been notced n ths ndcator and many alternatves have been proposed (Daskapan 2008a. The Revealed Comparatve Advantage (RCA or Balassa ndcator s defned as X X RCA (1 X X

3 where X s the export of sector from country. Here the numerator represents the percentage of share of one gven sector n the total natonal exports of country, whereas the denomnator represents the percentage share of a gven sector n world exports. The RCA ndcator s thus a comparson of natonal export structure wth the world export structure. When RCA equals 1 for a gven sector n a gven country, the percentage share of that sector s dentcal wth the world average. If RCA s above 1 the country s consdered to be specalzed n that sector and vce versa, where RCA s below 1. One of the concerns of RCA and many other current TPIs,.e the lack of an ndcator for non-exported goods (Daskapan 2008b Snce an ndcator, whch s only based on the export numbers, lke the Balassa ndex a.o., gves an uncomplete outcome, t s necessary to derve the comparatve advantage on the bass of other fgures than export data. Ths part of non-exported domestc goods that s not ncluded n the current ndctors s called the unrevealed comparatve advantage. Gven the factor prce equalzaton proposton and StolperSamuelson (see secton 4, the here proposed alternatve s based on the producton costs of a good n a country n relaton to the average costs of the total producton n the world. Ths alternatve ndcator s labelled the factor prces or unrevealed comparatve advantage (FPCA. It s based on the costs of requred resources for the producton of the goods n the consdered sector. We wll call those costs factor prces FP. FPCA FP FP FP FP (2 In order to compare ths ndcator we wll ntroduce n the next secton frst a fcttous case of four countres. The normalzed data wll enable us to compare the FPCA wth the Balassa ndcator. 3. COMPARING INDICATORS IN A FICTIVE WORLD OF FOUR COUNTRIES A common approach n economcs s to use a fctve or hypothetc case (Pasteels and rchbach Our purpose s after all not to measure the actual current trade specalzatons of the dfferent countres, but rather to compare and dscuss the TPIs for trade specalzaton. Assume then a world that comprses four countres: I, E, N and T, where I represents a country wth an economy n transton. Assume further that those countres produce three types of goods, whch can also to some extent be exported, namely a. Food, b. Textle and c. Wood. In order to make comparsons between the countres and ther goods we need also to defne a standard unt for all goods: One trade unt s the standardzed quantty of one good for a certan fxed amount of money (say n $. For example: 1 unt of food s defned as a nutrtous good wth an energy of 1000 cal and one unt of wood s a pece of 1 kg of wood and so on, such that 1 unt food = 1 unt of wood =1 unt of textle =10$. In the next tables the producton and exports of those unts of goods per country are shown. In table 1 the export of goods per country per sector (n unts x10 3 s depcted. It s obvous that Country T has the hghest and Country I the lowest export rate. On a total (global scale all the three goods are exported almost at the same rate (7 x10 3 and 8 x10 3.

4 Table 1 Export of goods per country per sector (n unts x10 3 Sector a. Food b. Textle c. Wood Total Country I E N T Total Table 2 Producton of goods per country per sector (n unts x10 3 Sector a. Food b. Textle c. Wood Total Country I E N T Total Whereas n Table 1 Country T has the most exported unts (11 and Country I the least (2, n Table 2 we can see that Country I has the hghest producton of unts. The world export of all goods conssts of 23 unts n Table 1 and the world producton of goods conssts of 67 unts n Table 2. Ths means that a total of = 44 x10 3 unts s meant for domestc consumpton. Table 3 Resources (labor + captal for producton of one unt n $ Sector a. Food b. Textle c. Wood Total Country I E N T Total Low trade performance hgh trade performance From Table 3 we see that Country I needs the least resources to produce one unt of any of the three goods. Country E on the other sde has the hghest costs of producton. Accordng to the theory of absolute advantage Country I has compared to the other countres an absolute advantage and Country E has an absolute dsadvantage for all the goods. As such, Country I wll gan when exportng all ts products whereas Country E wll loose. However, ths statement does not hold, snce the theory of absolute advantage has been overruled by the Rcardan comparatve advantage and Heckscher-Ohln theory. When we apply the FPCA ndcator of expresson 3 by usng the data from Table 3, we see n table Table 4 some dfferences wth the conventonal Balassa ndcator (from table 1. For example, RCA(I,a = (12 (823 = 1.44 FPCA(I,a = (13(1033 = 1,10 Table 4 The FPCA ndcator compared wth Balassa Sector Country a. Food b. Textle c. Wood a. Food b. Textle c. Wood I. 1,10 0,92 1,00 1,44 0,00 1,64 E. 0,88 1,10 1,00 0,48 1,44 1,10 N. 1,10 0,92 1,00 1,44 0,72 0,82 T. 1,10 0,92 1,00 1,05 1,05 0,90 a. FPCA ndcator b. Balassa ndcator In contradcton to the Balassa ndcator we see now that Country I wth a transton economy has also a (unrevealed comparatve advantage for textle (I,b. The Balassa ndex would not have revealed ths

5 opportunty. Also the other combnatons dffer somehow: some, lke (N,c and (T,c mproved and some, lke (I,c and (N,a, deterorated. Lemma 1: The unrevealed comparatve advantage ndcator ndulges to the non-exportng country problem for transton economes. 4. Consequences of the HO model for transton economes The Heckscher-Ohln (H-O theory presents the ssue that nterregonal dfferences n producton costs occur because of the dfferences n the supply of producton factors (Mcculloch Those goods that requre much of the abundant, thus less costly factor, wll have lower producton costs, so that they can be sold for less n nternatonal markets (Salvatore and Domnc Ths theory of factor endowments states thus that: a regon should export products that use ntensvely ts relatvely abundant factors and mport products that use ntensvely ts scarce factors. We defne abundance as the postve dfference when the supply exceeds the demand of factors. A fundamental dstncton between the H-O model and the Rcardan model s that the Rcardan model assumes that producton technologes (only labor dffer between countres, whereas the H-O model assumes that producton technologes (labor and captal between countres are the same. Trade performance ndcators should embrace both theores. However, that s currently not the case, snce all TPIs are now based on the Rcardan model. Ths s a defcency for transton economes, snce many transton economes, a.o. many of the new countres of the former Sovet Republc, are already technologcally advanced. Accordng to the Rcardan model, the technologcal small dfference between them and the developed countres would not gve them much mpulse to trade. On the other hand, transton economes do have abundance n several (natural resources that would cause accordng to the H-O theory a trade mpulse. Unfortunately, a TPI based on H-O theory s lackng. Before proposng an H-O TPI n the next secton, we wll show n ths secton the effect of H-O on trade flow between transton economes and developed economes. The H-O model has led to the formulaton of certan named conclusons arsng from the assumptons nherent to the H-O model: - The Rybczynsk theorem says that an ncrease n a country's endowment of a factor wll cause an ncrease n output of the good, whch uses that factor ntensvely, and a decrease n the output of the other good (Amano The Stolper-Samuelson theorem says that relatve changes n output goods prces wll drve the relatve prces of the factors used to produce them. If the world prce of captal-ntensve goods ncreases relatve to the prce of labour ntensve goods the rental rate wll ncrease relatve to the wage rate (Stolper The factor-prce equalzaton theorem says that when the prces of the output goods are equalzed between countres, as when countres move to free trade, then the prces of the factors (captal and labor wll also be equalzed between countres (uga Based on those theores we can pose two lemmas: Lemma 2: prce of factors and of ther produced goods decreases when abundance of the factors ncreases. Lemma 3: Gven a one country closed economy, ths change n prce s lnearly related to the change of the abundance of factors. The basc assumptons for those lemma s are that n ths economy there s one country n the world, labour and captal are the only factors of producton, only one good s produced n each country, the supply of factors s fxed n each country, the productvty s fxed, perfect competton prevals n all markets, factors are not moble across the two countres, but wthn each country, there are no other costs nvolved (transportaton costs, etc..

6 So, when the quantty Q of factors ncreases by Q = m, whle demand remans the same, ts prce P wll decrease by P = n = am+b, wth a and b as constants, such that the natonal value V of all those goods remans the same: V = (PxQ before = (PxQ after. Assume for example that before the change the total quantty of apples n a country s 100 whereas the prce per apple s 10, so that the total natonal value s 100 x10 = 500. If n the next perod the number of harvested apples n the country ncreases by 25 (.e , whereas the number of consumers remans the same, then lemma 2 says that the prce wll decrease by n (.e n, such that (10-n x 125 = 500. In ths case ths wll result n n = 6. Gven the prevous lemmas and the prce equalzaton theory we pose the next lemma. Lemma 4: Assumng dfferent but constant demands n dfferent countres, perfect competton and free trade of goods n and between the countres, dfferences n factor endowments wll lead to dfferences n prces between the countres. Gven then the prce equalzaton theory countres trade untl a new equlbrum s reached. Dfferences n factor endowments between the transton economes and mature economes wll thus lead to trade untl a certan equlbrum s reached. Ths can be explaned as follows. Gven those 3 lemmas ths means that any dstorton n prces or quantty n a free trade relaton between countres wll lead to a new equlbrum. For example, n a two country economy wth only Amerca (A and Moldava (M that trade wth each other ths equlbrum wll be derved. There s free trade of goods, but not of consumers, so that both demands reman constant. V A = (P A,before x Q A, before = (P A,after x Q A,after (3 V M = (P M,before x Q M, before = (P M,after x Q M,after (4 We know now that P A,before = P M,before = P before, snce there was a trade balance before the dstorton. The dstorton of the equlbrum wll lead to dfferent prces and thus ntates trade. Ths trade wll contnue untl P A,after = P M,after = P after. V A = (P before x Q A, before = (P after x Q A,after => P before P after = Q A,after Q A, before (5 V M = (P before x Q M, before = (P after x Q M,after => P before P after = Q M,after Q M, before (6 The prce equalzaton theory between both countres enables (5 n (6, whch gves Q A,after Q A, before = Q M,after Q M, before (qed. (7 Ths means that trade wll stablze when the relatve ncrease of factors between the developed economy and the economy n transton becomes equal. 5. The Heckscher-Ohln ndcator Another concern s the lack of an approprate TPI based on the Heckscher-Ohln theory. Whereas the other trade theores can lean on well developed and used ndcators, as we have seen n the prevous sectons, measurng trade dfferentaton accordng to the Heckscher-Ohln theory lacks an approprate ndcator. As such we propose alternatve Heckscher-Ohln ndcators (HOI to measure factor endowments. Instead of only presentng the fnal expressons, we wll descrbe our consderatons durng ths quest for HOIs Heckscher-Ohln Indcator based drectly on factor prces A country s abundant n a factor relatve to another country f t has more endowment n that factor than the other country. If we want to capture ths abundance n an ndcator such that a comparson between countressectors becomes possble, one way s to express t relatvely to those other sectors and countres. We assume further that lower factor prces represent hgher factor endowments. As such, f we could measure an aggregated factor prce drectly the ntal composton of HOI s

7 F F F F HOI (8 where the umlaut represents the average value. F depcts the average factor prce n sector of country, F gves the average factor prce n country. Here the numerator represents the rato between the average factor prce n sector of country and the average factor prce for all sectors n country. The denomnator represents the rato between the average factor prce of sector n the world and the average factor prce n the world. The HOI s thus here a comparson of the average natonal factor prce compared wth the average world the factor prce. When HOI equals 1 for a gven sector n a gven country, the endowment of that sector s dentcal wth the world average. Where HOI s above 1 the country s consdered to be factor abundant n that sector and vce versa where HOI s below 1. Although ths ndcator measures drectly the aggregated factor endowment per sector and per country, t s not practcal snce t does not dstngush between the dfferent specfc factors. As such t s then not possble to draw conclusons about a country s comparatve advantage per factor per sector. The 2x2x2 model only assumes the factors labour (L and captal (. The factor land or property (D s not ncluded, although ths also an accepted factor n economcs (Fetter 1977; Ryan Recall that n the neo-classcal model country A s consdered captal abundant relatve to B f: A B A LB. 1 LA LB B LA That means that wth a 2x3x2 model A s consdered captal abundant relatve to B f: A B A DB. LB. 1 D A. L A DB. LB B DA. LA Therefore, we can substtute the aggregated factor F wth the specfc factors F l, F r, F t : F = Fl +Fr +Ft, (9 where F l s the factor prce of labor, F r the factor prce of land and F t the factor prce of captal. Ths gves after substtuton n expresson (8: (F F F F F w Fr Ft (Fw (Fr (Ft (10 HOI F F F F F F F F where each term represents one of the followng specfc Heckscher-Ohln ndcators. Labor endowment s represented n the frst term of the prevous equaton: (F F w HOI (w F F (11 However, unlke the other specfc HOI, labor endowment can be measured drectly, and not necessarly as a dervatve from the factor prces, when usng the unemployment rate. The assumpton here s that unemployment s the result of a negatve balance between demand and supply of the factor labor. (w HOI U U U U where U s the unemployment rate n sector of country. Here the numerator represents the rato between the unemployment rate n sector of country and the unemployment rate for all sectors n country. The denomnator represents the rato between the unemployment rate of sector n the world and the unemployment rate n the world. The HOI s thus here a comparson of the natonal unemployment rate compared wth the world unemployment rate. When HOI equals 1 for a gven (12

8 sector n a gven country, the endowment of that sector s dentcal wth the world average. Where HOI s above 1 the country s consdered to be factor abundant n that sector and vce versa where HOI s below 1. HOI HOI For captal endowment we suggest the second term of the prevous equaton. (r (F F r F F and for land endowment (t (F F t F F (13 (14 Lemma 5: The Heckscher-Ohln ndcator to measure the factor endowments can be based drectly on factor prces 5.2. Heckscher-Ohln Indcator based ndrectly on fnal goods Based on Stolper-Samuelson theorem and the Factor-Prce Equalzaton theorem we assume that factor abundance can also be measured by the prces of the fnal goods. We assume that lower factor prces represent hgher factor endowments. As such, factor endowments lead to lower factor prces and thus to lower prces of the fnal goods. The HOI for the average aggregated prce P of a good n sector of country s then: HOI P P P P where the umlaut represents the average value. P, P, P and P represent then respectvely the average prce of a good n country and n sector, of all goods n country, of a good n all sectors, of all goods n the world. Lke the prevous ndcator ths one does not dstngush between the dfferent factor endowments. It s also not a workable ndcator, snce P, P and P are unrealstc varables to calculate. As such t s then also not possble to draw conclusons about a country s comparatve advantage per factor per sector. The frst attempt to reach a workable HOI s by usng the CPI for P and P, so that HOI P CPI P C P I (15, (16 where C PI represents the consumer prce ndcator n country and C P I the average consumer prce ndcator n the world. Another opton s to decompose the aggregated prce P n to the specfc prces P l, P r, P t, whch are expressed n terms of factor prces. As for labor, we use the ndcator real wage (w. For land we propose rent (r and for captal nterest (t. The expresson for the aggregate prce P of a good n sector of country s then: P = (a.t+b.w+c.r.(1+m

9 wth a,b,c as the quanttes and m as a prefxed constant for proft or loss margn. We wll make a smplfcaton by assumng that for ths case the proft margn for each product n each country s equal to 1+m: P = (a.t +b.w +c.r.(1+m (17 The HOI becomes now P P ( a. t b. w c. r ( a. t b. w c. r (18 HOI ( x, y {[ 1, ], [1, ]} P ( a. t b. w c. r ( a. t b. w c. r P Furthermore, we defne three categores of goods. In the frst category, b and c are that relatvely small compared to a that we assume c = b 0. In the second category, we assume a and c to be zero and n the thrd category, a and b are assumed to be equal to zero, so that P (cat1 = (a.t.(1+m Λ c = b 0, P (cat2 = (b.w.(1+m Λ a = c 0, P (cat3 = (c.r.(1+m Λ a = b 0. The HOI can now be defned separately for each factor now. We suggest for captal endowment: P P ( a. t ( a. t HOI (t ( x, y {[ 1, ], [1. ]} P ( a. t ( a. t P (19 for labor endowment: P P HOI (w P P ( b. w ( b. w ( b. w ( b. w ( x, y {[ 1, ], [1, ]} (20 for land endowment: P P ( c. r ( c. r (21 HOI (r ( x, y {[ 1, ], [1, ]} P ( c. r ( c. r P If we also assume that the producton functon s equal, so that the productvty s equal, the HOI s can be smplfed nto: t t HOI (t ( x, y {[ 1, ], [1. ]} t t w w HOI (w ( x, y {[ 1, ], [1. ]} w w r r HOI (r ( x, y {[ 1, ], [1. ]} r r Ths enables us to dstngush the effect of the dfferent factor endowments for a specfc good. Lemma 6: The Heckscher-Ohln ndcator to measure the factor endowments can be based ndrectly on fnal goods 6. Conclusons and Future work In ths paper we proposed alternatve ndcators to overcome the shortcomngs of the current trade performance ndcators (TPIs to measure the trade potental of transton economes properly. We have focussed on three types of shortcomngs. Frst, many potental trade sectors are excluded n the current calculatons, snce no ex-post trade data exsts yet about them. Second, the gven TPIs do not

10 represent trade specalzaton or comparatve advantage, but rather the relatve export share. Thrd, a TPI based on Heckscher-Ohln (H-O theory s lackng to measure the trade potental based on factor abundance. To overcome the frst and second shortcomngs the unrevealed comparatve advantage ndcator that s based on factor prces s ntroduced. To meet the thrd shortcomng two alternatve H- O ndcators are derved: the H-O ndcator that s based drectly on factor prces and the H-O ndcator that s based ndrectly on factor prces. In future work we ntend to test both H-O ndcators and the unrevealed comparatve advantage ndcator wth real data. In ths test we wll use data from developed countres and data from transton economes. Both datasets wll be used accordng to the old TPIs and accordng to the here new proposed TPIs. Our expectaton s that when the results of both,.e. old vs. new TPIs, are compared the new TPIs wll reveal more sectors wth trade potental than the old TPIs. REFERENCES - Amano, Akhro, Factor Endowment and Relatve Prces: A Generalzaton of Rybczynsk's Theorem, Economca, New Seres, Vol. 30 (120, pp , Daskapan, Semr, Measurng trade advantage of natons: a study on the relablty of trade performance ndcators, Internatonal Academy of Busness and Economcs Conference, Stockholm, 2008b - Daskapan, Semr, Trade performance ndcators n the new economy, Internatonal Telecommuncatons Socety 17th Bennal Conference, Montreal, 2008a - Heckscher, E, The Effect of Foregn Trade on Dstrbuton of Income, Readngs n the Theory of Internatonal Trade, Vol 4, Ohln, Bertl, Interregonal and Internatonal Trade, Harvard Unversty Press, uga, yosh, The Factor-Prce Equalzaton Theorem, Econometrca, Vol. 40 (4, pp , Stolper, W.F. and P.A. Samuelson, "Protecton and Real Wages", Revew of Economc Studes, Vol. (9, pp , Balassa, Bela. Traded Lberalzaton and Revealed Comparatve Advantage, The Manchester School of Economc and Socal Studes, Vol. 33, pp , Fetter, Frank A., Captal, Interest and Rent: Essays n the Theory of Dstrbuton, Sheed Andrews and McMeel Inc, ansas Cty, Mahoney, D., M. Trgg, R Grffn, M Pustay, Internatonal Busness: A Manageral Perspectve, Addson Wesley Longman, Melbourne, Marsh, Ian and Stephen Tokarck, Compettveness Indcators: A Theoretcal and Emprcal Assessment, Workng Paper 9429, Internatonal Monetary Fund, Washngton, Mcculloch, Ball, Internatonal Busness: The Challenge of Global Competton, IrwnMcGraw-Hll, Pasteels, Jean-Mchel and Fredrch von rchbach, Relablty of trade statstcs, UNCTAD, Ryan,., Land as a factor of producton, The Amercan Journal of Economcs and Socology, Vol. 61(5, pp. 7-25, Salvatore, Domnc, Theory and Problems of Internatonal Economcs, McGraw-Hll, Scott, Chrstopher, Measurng Up to the Measurement Problem, London School of Economcs