DISCUSSION PAPER The views presented hero, are those.of the author, and they should not be interpreted as reflecting those of the World Bank TRADE POLICIES AND PRODUCTIVITY CHANGE IN SEMI-INDUSTRIALIZED COUNTRIES Mieko Nishimizu Sherman Robinson Development Research Department' The World Bank December 1982 Revised, March 1983 Paper presented at the 1982 Meeting of the American Economic Association TRADE POLICIES AND PRODUCTIVITY CHA!!GE I N SEMI-INDUSTRIALIZED COUNTRIES Mieko MSHIMIZU and Sherman ROBINSON* The World Bank, Washington, D. C. 20433, USA The r o l e of trade policies i n increasing growth and efficiency has long been a major focus i n the development l i t e r a t u r e . This paper examines the impact of d i f f e r e n t development s t r a t e g i e s , especially export expansion and import substitution trade policies, on t o t a l factor productivity growth i n t h e manufacturing industries. The analysis is based on recently developed d a t a on s e c t o r a l t o t a l factor productivity i n Korea, Turkey, and Yugoslavia, with Japan a s a comparator. Our r e s u l t s indicate t h a t there a r e important l i n k s between trade policies and i n d u s t r i a l productivity performance. 1. Introduction Total factor productivity (TFP) measures the economic and technical efficiency with which resources are converted i n t o products. The growth of an economy, an industry, o r a firm is determined by the r a t e of expansion of its productive resources and the rate of TI'? growth. I n the case of developins countries, there are a l l s o r t s of constraints and l i m i t s t o how f a s t employable resources ,can grow. Achieving rapid r a t e s of TFP growth is then a r e a l issue i n breaking bottlenecks. Furthermore, an important part of t h e "catching up" process involves exploiting changing comparative advantage. I n which a c t i v i t i e s and how quickly, can international competitiveness be achieved? D i f f e r e n t i a l s e c t o r a l r a t e s of TFP growth are c r u c i a l determinants of evolving comparative advantage and have a major impact on both growth and s t r u c t u r a l change i n the medium t o long run. I @ - *,f - - n,'q32-0' d ~ q r r f h n q- o e - ~ l+ . 2 '?--TI ..3(jj r i d ~ ' - 7 - ?-:I: ::3-J;::I kj;\j,cLt. of Growth and Productivity Change: A ~ o m ~ a r a t i Analysis". v g We would l i k e t o thank Bela Balassa, Hollis Chenery, Z i c h q d Nelson, Simon T e i t e l , 9 Carry Westphal, and J e f f r e y Williamson f o r helpf comments on e a r l i e r d t a f t s , and Mark Gersovitz and John Page for help u l discussions. We wish too thank Deborah Bateman f o r her excellent research assistance. The views and interpretations i n t h i s paper a r e those of thc authors and should not be a t t r i b u t e d t o the World Bank, t o its a f f i l i a t e d organizations, or to any individual acting i n t h e i r behalf. There a r e two issues concerning TFP growth t h a t a r e e s p e c i a l l y relevant for development policy. F i r s t , what is the range of TFP growth r a t e s one can reasonably expect? Confidence i n t e r v a l s f o r TFP growth r a t e s can, i n principle, be obtained from h i s t o r i c a l records of firms, industries, o r economies operating under varying production environments. They provide useful information f o r determining, f o r example, the appropriate duration of infant industry protection o r promotion policies. Is five years too short? Twenty years too long? Should the duration be uniform among i n d u s t r i e s , o r should they d i f f e r from industry t o industry? The second issue of policy r e l e v a x e has to do with the causes o r sources of TFP growth. For example, does protection from competing imports destroy incentives for improving efficiency i n production? Can some policies improve productivity -- f o r exampJ.e, subsidies tailored t o s p e c i f i c factors such a s f i s c a l incentives for accelerated depreciation, employee training subsidies, etc.? The empirical l i t e r a t u r e on TFP change has, over the years, accumulated a substantial body of "stylized facts" about the contribution of productivity change and f a c t o r input growth t o economic performance i n various c o u n t r i e s j ' Perhaps the most s i g n i f i c a n t stylized f a c t that has emerged is the importance of TFP change i n contributing t o growth - a s much a s one-third t o one-half of growth i n output can be a t t r i b u t e d t o TFP change. Until q u i t e recently, much of what we knew was i n terms of macro aggregates. 1 There fs 1: .- : L Z < K ~ L C ~ L bee ~ , - \ r - - o , I . ,I :!ic Aadirl (1970, 1972). For an excellent 4 c r i t i c a l survey of the productivity l i t e r a t u r e , see Nelson (1981). -2/ See, for example, Christensen and Jorgenson (1973), Chrfstensen, Cummings, and Jorgenson (1980), Denison (1967, 1974), Denison andEhung (1976), Ezaki and Jorgenson (1973), and Griliches and Jorgenson'(1967), for studies on developed countries. On developing countries, Christensen, Cummings, and Jorgenson (1980) included Korea i n t h e i r i n t e r n a t i o n a l comparison, B r u t c ~(1967), and Elias (1978) studied Latin American countries. See a l s o Robinbon (1971) and Feder (1982), and s t u d i e s c i t e d by them, a s well as by Nadir: (1972). now, however, a small but growing empirical l i t e r a t u r e on TFP change a t a disaggregated level. 2' The f i r s t objective of t h i s paper is t o add t o t h i s body of "stylized facts" based on recent time-series data developed a t t\e World Bank on TFP growth a t the sectoral l e v e l within manufacturing for three countries: Korea, Turkey, and Yugoslavia. We add Japan to the sample as a comparator based on data developed by Jorgenson and Nishimizu (1981). I n contrast with the growing stock of empirical estimates on TFP growth, we have not accumulated s u f f i c i e n t evidence about causes of productivity change. A s surveyed and discussed extensively by Nelson (1981), the l i t e r a t u r e on productivity change o f f e r s a wide variety of possible causes, but without any c l e a r consensus a s to where one should focus most attention. I n the development l i t e r a t u r e , the r o l e of trade policy i n increasing growth and efficiency has long been a major focus. The second, and more important, objective of t h i s paper is t o examine the impact of d i f f e r e n t development s t r a t e g i e s , especially trade policies, on sectoral TFP growth. 0ur.analysis is e x p l a z a t x y and considers a number of hypotheses that have been suggested over the years. There is c e r t a i n l y no shortage of hypotheses. Indeed, it is d i f f i c u l t t o s o r t out the differences among them and to define the appropriate measures and tests required t o address them. Our analysis does indicate that there are important links between trade . I -3/ See, for example, Kendrick (1961, 1973)' and Gollop and Jorgenson (1980) - - f o r the ~ni&d States; Nishimizu and Hulten (1978) and Kuroda and Imamura w (1981) f o r &pan; and a comparative study of the U.S. and Japan by Jorgenson and Nishimizu (1981). I n addition, there a r e productivity studies of regulated industries o r firms i n the U.S. and Canada. See, for example, Cowing and Stevenson, eds. (1981). A comprehensive study of Indian manufacturing industries was made recently by Ahluwalia (1982). See also Ezaki (1975) on the Philippines, Kuo (1983) on Taiwan, and Kim and Son (1979) on Korea. policies and productivity performance, and r a i s e s a number of issues f o r further research. One hypothesik- which has b2en suggested i n the l i t e r a t u r e is that there is a positive relationship between productivity change and the r a t e of growth of output. Expressed i n terms of labor productivity, t h i s e f f e c t bas been called "Verdoorn's Law" a f t e r P.J. Verdoorn who suggested it i n 1949. A \ number of people have investigated the relationship and Kaldor (1967) has argued that the fundamental explanation for it is scale economies/ He noted f u r t h e r that it is observed most prominently i n ma~ufacturingand other secondary i n d u s t r i a l a c t i v i t i e s . In developing countries, t h e importance of s c a l e economies and "size of market" have long been thought t o be very important i n determining growth and s t r u c t u r a l changel/ The existence of s c a l e economic,-;, o r any other j u s t i f i c a t i o n for t h e Verdoorn's Law, implies t h a t widening the market through t,;ade should lead t o . reduct ions i n product i o n costs. The argument is usually made i n terms of the benefits of expansion i n demand through increased exports. Although it tepends on the s i z e of domestic markets, the argument should i n principle apply t o import s u b s t i t u t i o n a s A quite d i f f e r e n t trade-policy argument is that opening up t o international competition w i l l induce increases i n domestic efficiency. There I I is an implicit*"challenge-response" mechanism induced by competition, forcing domestic industries t o adopt new technologies, t o reduce "X-inefficiency," and generally to reduce costs wherever possible. '&cording t o t h i s argument, C export expansion is good, and so too is i m p o r t ~ i b e r a l i z a t i o n . While a policy w - 41 See a l s o S a l t e r (1960) and Kaldor (1961). -5/ See Chenery and Westphal (1979) and Balassa (1967). of increasing imports may restrict the market for domestic goods, it also increases competition and hence induces greater efficiency. The converse is aLso widely asserted; protectionist policies designed to promote import substitution reduce competitiveness and lead to inefficiency in production. Oce must be careful not to overstate the argument. "Infant industry" protection, by definition, is affordc1 to high cost industries which cannot * compete with imports until they "grow up" and become internationally competitive. Conversely, export promotion policies such as excessive export subsidies may distort incentives and lead to increasing inefficiency. It is important to focus on the causal mechanism assumed to be working: export expansion and import substitution policies may increase or decrease TFP (levels and/or growth rates) depending on their impact on competitive, cost- reducing incentives to producers in the media to long run. The literature on foreign exchange constraints provides yet another argumont for a link between trade and productivity. A stylized fact characterizing developing countries is that intermediate and capital goods imports are not very substitutable with domestically produced goods. In a I sense, these imported inputs embody technologies that are unavailable to 1 domestic producers and can only be attained through imports. Any policies that limit the availability of such imports, or make them more expensive, will a 1 lead to poor productivity performance. Conversely, policies which increase t h e availability of imported inputs, w lower their c o s t , fe.:., # !.ncreaser! ? foreign aid or an export-led development strategy) will lead to cost ' - &* . reductions to domestic kdustries, and hence to better productivity I t I performance. In this view, exports are important only as a source of foreign 1 exchange, permitting industries to buy inputs which can be produced I domestically only at much higher, if not infinite, cost. It is important to note that these hypotheses about possible links between a l t e r n a t i v e development s t r a t e g i e s distingpished by trade policies and Tr'P growth a r e not mutually exclusive. They may a l l be true, and the postulated e f f e c t s a l s o need not be independent of one anothsr. Given the current s t a t e of knowledge, it is simply not possible to discriminate f i n e l y among them. Indeed, it is not even possible t o s t a t e with any r e a l confidence * what is the d i r e c t i o n of causation. It is. just as likely, for example, that "exogenous" TFP growth in a sector generates a s h i f t i n the supply curve and hence, i f domestic demand is limited, a strong incentive t o open up export markets. The possible relationships are myriad, and w i l l probably have to be sorted out on a case-by-case basis. I n t h i s paper, we begin by looking a t the h i s t o r i c a l experience of four countries and t r y t o s o r t out s i m i l a r i t i e s and differences among them a t the sectoral level. We then explore some of the hypotheses discussed above, using additional data on the aature of the development process i n t h e countries. The four countries, a t various times, have pursued a v a r i e t y of development s t r a t e g i e s and supporting trade policy regimes. This v a r i e t y y i e l d s experiments i n which d i f f e r e n t e f f e c t s dominate the r e s u l t s , and enables us to explore the r e l a t i v e importance of f a c t o r s such as import substitution versus export expansion. Before considering the empirical I e r e s u l t s , however, we f i r s t discuss the nature of TFP measures we use. The measures snbody a number of strong assumptions that a f f e c t how they should be interpreted, and what they capture a s productivity change. - 2. The Analytical Framework for TFP Measurement 8 I The a n a l y t i c a l framework f o r TFP measurement is founded on the economic theory of cost and production. In recent decades, developments i n the f i e l d of productivity research have been accompanied by advances i n closely related areas of ecnnomic theory and measurement. They include duality theory, theory of index numbers, and the development of f l e x i b l 2 functional forms which a r e less r e s t r i c t i v e i n representing economic r e l a t i o n - ships such as production functions and cost functions. / Advances i n t h e areas have strengthened trhe t h e o r e t i c a l foundations of TFP measurement.- 7 l Indices of TFP change a r e usually given i n terms of output per unit of t o t a l factor inputs, and are functions of scale elasticities, output a d input elasticities, and q u a n t i t i e s ( o r prices) of outputs and inputs. It is usually assumed t h a t output and input markets a r e competitive, and t h a t firms maximize p r o f i t subject t o a constant returns t o scale production f u n c t i o t and market prices which a r c taken as parameters. Under these assumptions, o u t p u t and input e l a s t i c i t i e s are equivalent t o the observed c o s t shares of f a c t a r inputs and revdnue shares of each output The index of TFP ch-e can then be c:ornp.l:ed using only the prices and quantities of outputs and inputs, and equals the difference between revenue share weighted output growth r a t e s and cost share weighted input growth r a t e s . There is an extensive l i t e r a t u r e on the choice of an "appropriate" index of TFP change- 9/ Essentially, one must specify something about the form of t h e production function ( o r , a l t e r n a t i v e l y , the cost function) i n order t o j u s t i f y a -6/ Caves, Christensen, and Diewert (1981, 1982a,b) provide a good and c o d s e summary of the l i t e r a t u r e and references. ,. See also Gallop and Jotyenss? , ,A;%;. '? -7/ t h i s section, we s h a l l ?tovide a brief exposition of the analytical fgmework. For a more d e t a i l e d and t e c i n i c a l discussion see some ofwthe - r d e r e n c e s c i t e d above. -81 When these assumptions are not tenable - a s i n Yugoslavia - d i r e c t estimates of output and input e l a s t . i c i t i e s and s c a l e e l a s t i c i t i e s must be generated. See Nishimizu and Page (1982). -91 For a survey on the theory of index numbers, see D i e w e r t (1979). particular form of an index. We chose the translog production function and the resulting translog index number in our methodology- 10/ This framework for TFP measurement has a number of shortcomings, The simple stylization of production and markets ignores a number of factors and constraints that nay be important. Nelson (1981), in his recent review of the productivity literature, provides a detailed criticism and evaluation of the approach. A couple of issues he raises are worth emphasizing since they affect the interpretation of our empirical results. A production proceps can be seen as the application of "technology" to the production of goods and services. Technology, however, represents -re than simply machines, tools, and equipment. It may be embodied in workers and managers, physical characteristics of material inputs, or in procedures and: . organizational principles that determine how various inputs are combined. It may also be embodied in produced outputs themselves. As Nelson discusses at length, TFP changes may therefore result from all sorts of changes in :his broadly interpreted "technology" applied to produc;tion. Nelson also points out that production takes place within "production environments," defined by the nature of markets for inputs and outputs, and by a set of market and nonmarket ::onstraints such as government For a detailed exposition of this approach, see Diewert (1976) for a theoretical discussion, and Gollop :and Jorgenson (1980) for . q - - ~ t.-:?- ., -. '-55,z -; ..2::.2, j,tuu;G a vaiue 7 ? . : i-- G L , ~ J U L ~dicit added production function (e&luding intermediate input) or with a gross production function (including intermediate input). We chose the gross production function approachrn this paper, since we believe that intermediate inputs "matter'fn sectoral TFP change, and that it is misleading to assume that intermediate inputs are separable from capital and labor. There is an extensive literature on this issue, but the mose comprehensive treatment can be found in Gollop and Jorgenson (1979). Gollop and Jorgenson (1979) also provide a comprehensive treatment and survey of the literature on aggregation over sectoral TFP estimates and the impact of interse2toral resource shifts on TFP change at the macro level. policies. Changes in production environments ultimately affect productivity performance by altering production constraints via changes in prices, quantities, or qualities of inputs and outputs. They may also haqe important shorter-run impacts on TFP changes during the process of adj*~stmentto neu conditions in production environments. Thus, our empirical results on TFP change should not be interpreted as measuring only "technical change" in the sense of a shift in the production possibility frontier due to the implementation of ' iiew generation of technicaX knowledge." Instead, the measures must be interpreted quite broadly, and include such issues as industrial and plant organization, eaginrering know- how, or changes due to disruptions in the production procees that affect capacity utilization in the relatively short run. The measures really treat production units as a black box. We measure the inputs and the outputs, but -make no real attempt to describe exactly what is going on inside the plant gate. Figuring out how the black box works is important, but is beyond the scope of the present pap.r.&/ We seek to delineate the stylized facts at a fairly aggregate level, and will necessarily be modest in our attempts to generalize and to discern causal links. Growth and Productivity Change in the Manufzctilring Industries In this section, we attempt to distinguish systematic patterc~of I I a output, input, and TFP growth in mantlEacturing industries of Japan, Korea, Turkey, and Yugoslavia. The historical period ve consibr is from the 3atp .* - 1950s/early 1960s to the early/late 1970s. One common characteristic t!lat - * e - 4 -11/ See Nelson (1981) for a survey of the relevant literature. Research towards this objective is under way at the World Bank, in two research projects: "Acquisition of Technological Capability" (RPO 672-48) by Carl Dahlman and Larry Westphal. and "Productivity Change in Infant Industries" (RPC 672-86) by Mieko Nishimizu and John M. Page, Jr. unites the development experiences of these focr -0untrCes during the peerod is that they are all semi-industrialized countries, with Japan graduating to industrialized status sometime during che 1960s. Anan: the four countries, however, there is a variety of different development strategies and supporting trade policy regimes. If factors related to stage of developmect or trade policy have a signj-ficant impact on productivity performance, we should be able to see systematic similarities and differences among the four conntriee which are indicatiw of their influence. Out empirical analysis is based on Japanese data developed by Jorgensoe and Nishimizu (1981), and data recently developed at the World Bank on TFP growth in manufacturing industriss in Korea, Turkey, and Yugoslavia. 121 To sumnarire, data on gross output, labor, capital, and material input in current and constant prlces were aqsembled for the -- I ?p - -. - - manufacturing industries in these four c o u n t r i e s Cuttceptually similar methodologies were used in defining the variables and in aggregating to achieve comparabilityH Gross output and material input by industry are in constant 1970 prices in each country. Capital is defined as net capital stock at replacement cost in 1970 nrices and includes all -7onresidential structures - 121,The data come from two World Bank research projects: "Sources of Growth and Productivity Change: A Comparative Analysis," and "Productivity Change in Yugoslavia." More detailed discussions of the results are available in separate papers. See Krueger and Tuncer (1980) on Turkey, - Ohn,-I : ??,I\ . ,- , -.- - -- ..? --a, .~-=,.,ar~cl .I. 1 . La ~ ei 2 3 i j on tugosiavla, and 2 A\-- Ni,himizu and Robinson (1983) for a more compre\ensive treatment of the four country comparisons. a - * - s - 131 Data for Korea include 52 manufacturing industries; for Turkey 33 - industries; for Yugoslavia 19 industries; and for Japan 21 industries- For the purpose of this paper, we aggregated these data for each country to sixteen comparable sectors (roughly the ISIC two-digit classiflcatloa). - 14/ The methodology for Yugoslavia differs from the others because one could not assume that cast-share data reflected the workings of a competitive market (see Nishimlzu and Page (1982)). and producers' durabies. Land and im~entoriesunfortunately could not be included due to data unavailability in the developing countries. Labor is defiued as persons employed-data on hours worked are not readily available in the three developing countries. The summary of industry estimates on which our analyses are based are presented in Tables A . l and A.2 in the Appendix. - i51 In an essay on cconomic growth, Kaldor stated (with empirical support) that "fast rates of growth are almost invariably associated with the fast rate of growth of the secondary sector, mainly manufacturing, and -,. this is an attribute of an intermediate stage of development."161 Table 1 presents average annual gr9wth rates of TFP, gross output, and capital, labor, and material input, as weil as the standard "sources of growth" decomposition, for aggregate manufacturing in Korea, Turkey, Yugoslavia and Japan. These countries all demonstrate rapid growth in manufacturing, and seem to fit Kaldor's stylization of being at an "intermediate stage" of development, Figure 1 plots output growth rates of different industries for each country (frim Table A.l), along with the sample mean and sample standard deviaeion. The figure shows that the rapid manufacturing growth in these countries is the result of many industrit-sgrowing "uniformly" fast. Over two thirds of all industries have growth rates in the range 9 to 15 percent in Japan, 17 to 27 percent in Korea, 9 to 18 percent in Turkey, and 6 to 13 percent in I I - - 151 Table A . lshows the average annual growth rates of T F ~ ,gross output, and capital, labor, aEd material input. Table A.2 gives the standard "sources of growth" decomposition in terms of the share of TF? change and growth of eacli input in gross output growth. - 16/ Kaldor (1967), p. 7. JAPAN - x = 12.41 s = 3.05 KOREA - x = 22.01 s = 5.38 TURKEY - x = 13.66 s 4.64 YUGOSLAVIA - X 9.20 M . M * M4M M S M M M s = 3.68 Yugoslavia. In Japan, Korea, and Turkey, there are r e a l l y no D1slov growth" industries; and i n Yugoslavia, only two i n a u s t r i e s have growth rates under s i x percent a year. E a r l i e r . we noted that, s i z e of market may be an important f a c t = fn determining growth and productivity change. One way i n which t h i s s c a l e e f f e c t comes about is through interindustry linkages. Balassa has argued that "cost reductions tend t o have a cumulativs effect: improvements i n p a r t i d a r industries are transmitted t o other sectors through input-output r e l a t l o a s i x i p s and through the e f f e c t s of higher incomes on the demand for consumer goods. **l8/ These inter-sectoral links, while also signif icant i n developed countries, a r e especially important i n developif& countries which a r e undergoing major changes i n input-output structure and i n the composition of f i n a l demand as part of the process of d e v e l o p m e n t g One inportant c h a r a c t e r i s t i c of the intermediate stage of development is t h a t the share of intermediate demand is t o t a l gross production increases signfficantly o v e r time. This trend would lead one t o expect consistently higher output g r o v t h in the '!producer goods" sectors across a l l of the four countries. And, w i t h o a t prejudging causation, where one sees high growth r a t e s , one a l s o expects to see high TFP growth. - 171 Note that the sample means of F i g ~ r e1 d i f f e r from aggregate rnanufacturlnR orit?ut growth z ? v c n i n T 2 5 l s "" " '. , ' : - * > , J ;Ir:cp ; L ; ; c ;is iiiiipu;C.$ weighted average. & - 181 Balassa (1967), p. 97. 8 - - 191 Rapid s t r u c t u r a l c h h g e , especially i f it a l s o leads t o sustained disequilibrium i n the factor markets (i.e., d i f f e r e n t marginal prodccticities across sectors), a l s o has a profound e f f e c t on aggregate growth. We w i l l not pursue t h i s issue further, but see for example Feditr (1982), and Robinson (1971). * Although it is d i f f i c u l t t o map our industry c l a s s i f i c a t i o n s t r i c t . 1 ~ i n t o producer goods sectors, we can group the i n d u s t r i e s i n t o the f o l l o w h g four groups ( a s shown on Table 2): (1) consumer goods, (2) light intermediates, (3) heavy intermediates, and (4) investment goods. We comsSder industry rankings within countries of both output and TFP growth. Using such rank'ings a b s t r a c t s from country differences i n the average growth rates. Table 2 presents these industry rankings, which a r e grouped i n q u a r t i l e s aPd then c i r c l e d where there is a similar industry ranking i n more than one country within each quartile. Table 3 summarizes the r e s u l t s f u r t h e r by giving a frequency count of i n d u s t r i e s i n each ranking across countries within the four i n d u s t r i a l groups. The industry ranking of output and TFP growth arranged i n t h i s manner shows a s t r i k i n g l y s i m i l a r pattern of f a s t e r growth i n heavy i n d u s t r i e s and slower growth i n l i g h t i n d u s t r i e s across the four countries. Inves-t goods industries a r e the f a s t e s t growing, followed by heavy intermediate @s industries and then the two l i g h t industry groups. Kendall's (multiple, ramk correlation coefficient, which measures the s i m i l a r i t y i n rankings a c r o s s a l l four countries together, is 0.75 f o r output growth excluding Yugoslavia Chere the industry c l a s s i f i c a t i o n d i f f e r s somewhat from the other three c o u n t r k ] , 4 and 0.52 when Yugoslavia is included. A similar but weaker c o r r e l a t i o n Es I t observed i n t h e industry ranking of TFP growth r a t e s among the four countries. Similarly, a chi-square t e s t t r e a t i n g Table 3 a s a contingency t a b l e yields values of 25.4 and 15.7 f o r the output g r w t h and TFP growth - tables. These values indicate a s i g n i f i c a n t fissociation between the f o u r . i n d u s t r i a l groups and rank according to both output and TFP growth, alt- the l a t t e r is s i g n i f i c a n t a t only the 90-95 percent confidence level. Tabl. 2 - Indumtry Raakiaa of Clowm Output 8md Total Factor Produetiticy CrovCb Total Factor hoductirlty hovtb - - - - - k o k 1 b Rnak 5 8 1 h s k 9 12 Rank 13 16 h n k l - b R a n k s - 8 R a n k * - U h a k n - X 6 J K T y l J K T T : J K T T J K T T J K T T J K T 1 J K T T J K r Y - - - *Am RmlVeDUTES . QI.oiealm . ~ a t r o ~L w bd' 4 I. Rubkr a3 L.Stom Clay lama:' :. ~ a m i ckta1d.f 3 - em: Cxcludln: Tobacco in lusolmavim Tmxtilm a d Apparal arm conmiderd a ti8 raaUa8 for Tu~omlav18 L u k r b Wood mad Furaitur8 8- emsid8r.d a ti. r a a t l u f o r lugomlavia !Zxeludias Coal for Tugonlavia Av8ragm of Buildla8 Utmrinlm aad A o ~ a t a l l i cMinerals for Tuunmlmrla . . . - Av.rmr. ?-r-n-lr \:on-<---, . - w e - , , , % * , : r > . i - r r r d n a r r ~ aa d .achloeq are conaid8red a t i 8 r a d - for Tugoal.via shlpbusldiw 0.11 for jq0a18via Table 3 Sectoral Frequencies Across Countries of Ranks of Out~utand TFP Growth Gross Out~utGrowth Rank: Rou Aggregate sector 1-4 5-8 9-12 13-16 sum Consumer goods 2 3 4 7 16 Light intermediates 0 2 Heavy intermediates 4 7 Investment goods LO 4 Column sum 16 16 16 16 64 TFP Growth Rank: Rov Aggregate sector 1-4 5-8 9-12 13-16 SUP - Consumer goods 1 6 5 4 16 Light intermediates 1 3 4 4 . 12 Heavy intermediates 5 4 4 7 20 Investment goods 9 3 3 1 16 Column sum 16 16 16 16 64 Notes : Consumer goods = sectors 1 to 4 Light intermediates = sectors 5 to 7 Heavy intermediates.= sectors 8 to 12 ~nvestmentgoods = sectors 13 to 16 Given these broad similarities, what are the major differences among the four countries? In particular, are there systematic differences in productivity performance by industries among the four countries? The country differences in TFP growth can be summarized statistically. For this purpose, we estimate a log-linear time-trend equation for TFP change over the . individual industry's annual time series pooled across countries. The ordinary least squares regression (with standard errors given in parentheses) 4 is: lZ2 = 0.475 sample size = 1054 The variable t is time and the variable K, T, and Y are country dy*mmiesset to one for Korea, Turkey, and Yugoslavia respectively, and zero otherwise. All estimated coefficients are significant at the 99 percent level, other than the intercept term (as should be expected since the level index of TFP i? unity in the base year). These results indicate that sectoral TFP growth rates in Korea, Turkey and Yugoslavia differ significantly from Japan. In Korea, TFP growth is 1.77 percent above Japan, Turkey is 1.05 percent below Japan, and Yugoslavia is 1.95 percent below Japan. The difference in TFP gro~thrates is also statistically significant between Korea and Turkey, Korea and Yugoslavia, I I and Turkey and Yugoslavia. F;r tt.rrmore, Yugoslavia's TFP growth rate is the only one which is not significant.1~different from zero. - Returning briefly to the aggregate manufacturing esfirnatas in Table *-- 1, we note that these country differences in TFP growth reflee another marked difference in .the manufacturing growth process of these four countries betwee., the relative importance of TFP growth in output growth compared to factor input growth ( ~ e erows 5 through 8 in Table 1). Japan and Korea are similar \ in that TFP change is as important as capital and labor input growth x' combined. For Turkey, although the rate of TFP growth is respectable. its contribution to output growth is significantly less important ~ h a nthat of capital and labor combined. In a sharp contrast, Yugoslavia's manufacturing growth involves very little TFP growth -virtually all growth is derived frrnn increases in the quantity of inputs. t. The relative importance of TFP growth on the one hand and capital and labor growth on the other at the disaggregated industry level can be examined in Figure 2. It plots the share of TFP growth in output growth against the share of capital plus 12bor growth in output growth in each industry by country from Table A.2, with the 45-degree line indicating equal contributions. These industry results also show the contrast between Japan and Korea vs. Turkey and Yugoslavia noted above. In Turkey and Yugoslavia, the individual industry results mirror the aggregate pattern. In both countries, with only one exception, the contribution of sectoral TFP growth is less important than that of capital and labor input combined. In Yugoslavka, almost all industries derive their growth in output from increases in factor inputs, with zero or negative contribution from TFP growth. The differences in productivity performance among countries might b~ at least partly caused by the nature of economic policies pursued by each a country. One importrst dimension which distinguishes Japan, Korea, Turkey, and Yugoslavia from each other is their choice of trade policies in their development strategies. Korea and Yugoslavia (in that order) have - . manufacturing sectors which are relvively more open to trade; while Turkey - 20/ Since the "contribution" of material input growth to gross output growth is always the dominating factor in manufacturing, we shall focus on the relative importance of TFP, capital, and labor growth in our discussion- FIGURE 2 RELATIVE CONTRIBUTION OF TFP AND P R I W Y IiVUTS TO GROWTH Note: TFP axis shows the ratio of TFP growth t o output growth in perccnt- Capital and labor axis shows the ratio of capital growth plus labar growth t o output growth in' peremr. and Japan are relatively more closed, the former by policy design and the latter mainly by the size of the domestic market. As we shall discuss in the next section, Korea's development strategy was distinguished by strong export promotion policies, often applied to selected industries. Turkey h3s long pursued import substitution policies across much of its manufacturing industries, many o' which are dcminated by state enterprises. Yugoslavia is distinguished by strong import liberalization accompanied by export I expamion. In addition, it has long sought regional and sectoral equalizatfon of productive performance with rage, emplo'yment,and investment policies designed to affect different industries similarly. Japan has made use of mixed export promotion and import substitution policies at different times, In the next section, we shall examine the relationship between growth and productivity performance of manufacturing industries, and the effect of the choice of an open, export-led development strategy versus a closed, import- substitution strategy. 4. Trade Strategies and TFP Growth In the introduction, we discussed some hypotheses linking TFP grovch and trade policies. The liat of hypotheses includes: (1) a positive link between higher exports or (depending on the size of the domestic market) increased import substitution and TFP growth, arising from "Verdoorn's Law" 8 I and the role of export expansion and import substitution policies in I n c t e a r i n g t h e size-of the market: (2) a p c s f t i v r ? l i n k F ~ _ t w - e54~y."-re x n q ~ r q ~ '3 and TFP growth, and4& negative (positive) link with import substitution e - (liq.eralization), aesing from competitive cost-reducing incentives or lack 0 thereof, and (3) a positive link between export expansion, import liberalization, and TFP growth, arising from the importme of foreign exchange constraints and non-substitutable imports of intermediate inputs and capital goods. It is likely that one observes the net effect of all these hypotheses simultaneously. They are certainly not rcucually exclusive, and distinguishing among them can be quite difficult. All these hypotheses can be seen as involving a supply response in terms of TFP change to changes in two components of demand: export expansion an4 import substitution. Taking these components as exogenous, or determined by exogenous policy regimes, we can then relate TFP growth to changes in the sources of demand growth. One must be very cautious, however, in implying the direction of causality in the relationship. For example, it may be that higher rates of exogenous TFP change lead to rapid growth in demand through lower cost!, and prices. Regardless of causality, however, the existence of any statistically significarit relationship will provide an ..nteresting starting point for further investigation. The single equation m d e l to be estimated is: where TFPG, x and x are, respectively, annual rates of TFP growth, output E&' IS . * growth allocated to export zxpansion, and output growth allocated to import - * P 8 r, '211 We assume that the random disturbances havegthe same distribution in all - time periods with zero expected value, finite variance, and nonsingulat covariance matrix. We can then employ ordinary least squares to estimate the model. Since the model consists of variables which are in growth rates, the random disturbances are not necessarily serially independent. When significznt serial correlation was detected, we employed the Cochrane-Orcutt method in our estimation. For ezch industry i n Japan, Korea, Turkey and Yugoslavia, we u s e as t h e dependent v a r i a b l e our estimate of annual r a t e of TFP c h a n g e s For the explanatory v a r i a b l e s , w e combine our e s t i m a t e s of annual o u r ~ u cgrowth rates with demand-side sources of grcwth decomposition m e a s u r e s 2 For each industry, t o t a l demand c . n be decomposed i n t o the following components: n where u is the diagonal matrix of domestic demand r a t i o s ( r a t i o s of domestic demand t o domestic plus import demand); D, W , and E a r e f i n a l demand, intermediate demand, and export demand respectively; and t h e s u b s c r i p t refers LO the t i m e p e r i o d E / ~ h &t h i r d and f s u r t h terms i n t h e decomposition give. respectively, the export expansion and import s u b s t i t u t i o n components of demand changes. Dividing each of these two terms by AX, we o b t a i n s h a r e mzasures of export expansion and import s u b s t i t u t i o n . i n gross output changes - 22/ Note t h a t i n t h i s s e c t i o n we aggregate. t o t h i r t e e n i n d u s t r i e s (from those appearing i n Table 2) to achieve aonsistency among a l l c o u n t r i e s w i t h t h e . I C d a t a on demand components. - 231 See Kubo, Robinson and Syrquin (1981) f o r a d e s c r i p t i - q of t h t - ulrir~uud~ogj.*lie udia w e u s e are described LLI uervls, ue ielo, dw Robinson (1982), Chapter 4. - d r) - * 241 There is a l s o a " p t a l " decomposition equation which uses the inpyF-output matrix and t h e r e f o r e incorporates i n d i r e c t linkages i n the decompokition, Since we a r e concerned with the supply response of i n d i v i d u a l s e c t o r s to changes i n demand, the d i r e c t decomposition equation is more approp-fate f o r our purpose than the t o t a l decompoqition measure. There is a l s o an index-number problem a r i s i n g from the choice of i n i t i a l o r terminal weights. W e use an average of the analogous Paasche and Laspeyres indices. for each industry in each country. We then multiply these st;are measures by 25/ annual growth rates of gross output of each industry.- Table 4 provides a summary of the decomposition results at the aggregate level for the four countries. Note first that thela is a great deal of variation in the relative roles of domestic demand expansion, export expansion, and import substitution, both over time and across countries. In every country, the role of export expansion increases over time--dramatically so in Korea and Yugoslavia--and .in every country but Japan the role of import substitution decreases. Yugoslavia actually shows significant import liberalization (i.e., negative import substitution). Korea and Turkey appear to have distinct phases, with a period characterized by significant import substitution followed by a period of exporr expansion--although the short export expansion phase in Turkey was hardly dramatic, especially compared to Korea. In Japan, although export expansion is significant in all three periods, the country is very large and the domestic market is the dominant component of demand. All in all, the sample of countries represents z vadety of development experiences and, rith the exception of Japan, they each underwent a significant shift in development strategy during the period under study. They should thus provide a good sample for statistical an.ilysis. Whether or not trade po1ici.e~are tailored to partiwlar' industries, there is no a priori reason to expect that the manner in which they affect - w Since the shre measures are based on data-for a few benchmark periods, w e apply the nearest shares to output growth rates for intervening years in the bench~arkperiods. The effect is to assume that the share measures reflect a t- that is uniform for each period. While clearly not ideal, this procedure does provide measures of the :wo explanatory variables. The benchmark periods for each country are as follows: Japan (1955-60, 1960-.65,1965-70), Korea (1963-70, 1970-73), Turkey (1963-68, 1968-73), and Yugoslavia (1966-72, ,1972-78). Decomposition of Growth of Manufacturing Demand - (units : percent) - - cl Manufacturing: Growth Decomposition: Outpu&/ Growth demand bl E x p ~ r t -PO- sharo, r a t e expansion expansion s u b s t i t a t l o o Rorea : 1955-1963 32.1 10.4 64.3 7.2 28-5 1963-1970 41.9 18.9 81.8 18.0 0-2 1970-1973 49.6 23.8 62.9 38.1 -1 -0 Turkey: 1953-1963 27.9 6.4 90.6 1.3 8-1 1963-1968 31-8 9-9 89.6 3.2 7-2 1968-1973 36-5 9.4 94.2 6.7 Q-9 Yugoslavia: 1962-1966 39.O 16.6 90.0 12.7 -2 -7 1966-1972 45.O 9.L 91.5 21.2 -12-7 Notes: a/ - Average share of manufacturing i n aggregate gross output durlag the period. I I -b/ Average annual rate of gzowth of gross output. c/ Decomposition methodolony is described i n the t e x t . The thtec comporieiltu sun to i u O percent. The f i r s t , domestic demand .? expansion, includes both intermediate and f i n a l demanc!. - * industries. Applying covariance analyses to our panel data indicates thar there a r e significant differences i n the estimated regressions among industries i n each country, and across countries i n each industry. Therefore, we report a separate regression for each industry i n each country i n Table 5. The regressions reported i n Table 5 indicate t h a t , overall, substantial portions of the v a r i a t i o n i n TFP growth r a t e s a r e "explained" by output growth allocated t o export expansion and import substitution i n Korea, Turkey and Yugoslavia, but (interestingly) cot i n Japan. There a r e a l s o significant Ltfferences among manufacturing industries. I n Korea, 13 percenc t o 83 percent of the variance i n TFP change is "explained" by the export expansion and import substitution growth r a t e s i n gross output. Only three industries show l e o s ' t h a n 30 percent of the variance explained. I n Turkey, the range is 13 percent to 95 percent, with only three industries below 30 percent. I n Yugoslavia, the low is 3 percent t o a high of 93 percent, with fcur industries below 30 percent. In Japan, on the other hand, the range is 2 percent t o 41 percent, with a l l industries other than t e x t i l e s and apparel showing l e s s than 30 percent of variance i n TFP change explained. Note a l s o A i n Table 5 that a l l the s t a t i s t i c a l l y significant constant terms (B,) f o r Korea, Turkey and Yugoslavia a r e negative, while they a r e a l l positive for Japan. P7:qative constant terms i n the three countries imply reductions i n TFP e I A levels (i.e., increases i n unit cost of production), unless o f f s e t by SuffiCientlv ~ o s i t i v ec o n t r i b u t i o n s from tho e;rol~t4i.1 n.itn31+ !I)- r r : -.unn.-t .* expansion and/or import subst itufion. These s t r i k i n g contrasts between Korea, 6 0 Turkey and Yugoslavia, on the onehand, and Japan on the other points to the I) r e l a t i v e importance of trade and trade policies i n the three developing countries. The r e s u l t s a r e a l s o consistent with ,he view that domestic demand has been the prime source of growth i n Japan. -- a .- .-.-.-.-.- L t - . ... ,, 38 t - 9 . :- z 0 - 2% -. = an r.. n- 4;: - ..". -. 15--;;b; c: ?? 3; :q i i! -t !! ?;i? 4: ... - i~'2 22 ..-: .- " 2" it nz ji ? I = .- 1 - t - * " 9 83 - 5: 3; :aI'zqq $3 $9 $j p$ :'=. ;::a I- 33. iij 4; I.. I.. I- I- I- I- I- I- o w Ve also observe in Table 5 that the estimated elasticities of TFP change with respect to export expansion and import substitution growth rates are distinctly larger (in absolute values) in Turkey compared with the other countries. Furthermore in Turkey, in a11 rndustries except paper products, the elasticities with respect to export expansion are greater than those viah respect to import substitution. Turkey is probably the most closed economy among the four countries, and these results emphasize the importance of trade ai the margin for such an economy. -261 Table 6 presents a summary of the regression results, indicating only the signs of the statistically significant estimated coefficients, and also provides an indication of export-oriented and import-competing industries in each country. Export-oriented industric~are defined as those with exports greater than 10 percent of total production and import-competing industries as those with imports greater than 10 percent of total domestic supply. Table 6 also gives aggregate export and import shares to indicate the relative "openess" of the manufacturing sector in these countries. The summary table reveals some interesting results which are difficult to obserl-e in Table 5 . Of the 28 cases where statistically significpn? elasticities with respect to export expansion are estimated, only two ,are negative. In contrast, 13 out of 21 sikificant elasticities with respect to import substitution are negative. Import substitution regimes thus s d <,- . . ,. FLcg4Li-tei, L"i.Le;dLdti wr*c;i L CC'~lldllge,wirertias exporr expansion '2 regimes are positively correlatd with TFP change. In Korea, no industry - & m r "suffers" from export expansion,-and those industries that "benefit" from ft are concentrated in light manufacturing and heavy-intermediate industries. In - 261 See Celasun (1983) for an analysis of the structure of Turkish growth daring the period. Table 6 sumnary of Regression Rerultr Export Expaaaion Import Subrtitution Korea Turkey Yugaolavia Japan Korea Turkey Yugorlavia Japan - a 2. Textilea, Apparel + EO + EO + EO HC ( 4. Lumbar 6 Wood, Furniture + EO EO 5. - Papar + 7. Petroleum 6 Coal t t - - - HC 8. Rubbar + E O + + e0 + HC 9. Stona, Clay, b Claer + EO + + EO + - - m: 10. Baric n e t a l s EO + EO nc + n~ nc 11. Pab. Uatalr 6 Hachinary + EO + + EO nc !4C m: - 12. Electrical Uachiaary + + EO HC HC - #C - 13. Tranrportatioa Equipment + EO EO + YC + HC + HC 14. Total Ufg. Share of Export/Production .254 .037 -164 .061 15. Total Hfg. Share of Inport/~omarti c supply .27e el12 .237 .044 - - Note: E0 - &port-ortanted induntry (axports graatar than 10%of t o t a l production) nC Import competing indurtxy (importr greater than 10%of t o t a l doaaetic supply, ,.a. importr plur t o t a l orodoction l e r r exportr) J"here export and import aharer were computed f o r 1973 i n Korea and Turkey, 1972 i= a Yugorlavia, and 1970 i n Japan. Turkey, the concentration shifts down towards all heavy industries. Yugoslavia shows no clear pattern of concentration. Paper products in Turkey and petroleum and coal products in Yugoslavia are the only two industries that show an adverse impact on productivity from export expansion. In Japan, only two industries benefit significantly from export expansion. Industries that experience a significant impact from import substitution are concentrated in Turkey and Korea. More of these are heavy-industriesin Turkey compared with Korea. Only four industries in Yugoslavia and one in Japan show a significant impact from import substitution. These results support some of the hypotheses outlined above and raise new hypotheses that are worth examining in future work. First, the results do not support the simple version of "Verdoon's Law" which implies that any expansion of the market, regardless of sol:rce, should improve productivity performance. There are significant and strong differences in the impact of export expansion versus import substitution. Second, the results are consistent with the hypothesis that export expansion leads to higher TFP growth, through economies of scale and/or through competitive incentives. Third, the results are also consistent with the converse hypothesis that increased import substitution (import liberalization) leads to lower (higher) TFP growth, perhaps through reducing (increasing) competitive cost-reduction incentives. Finally, the results are also consistent with the hypothesis that 4 L . L f V L C . i ~ d ~ i s djuu~ LiI I I P O S L ~ A O ~ C ~ A A L ~ L I O L ~ ~ lnccedse ir'P growth through relaxing the foreign exchange constraint and imports of non-substitutable C d intermediate and capital goods. The results provide some interesting mzterial for debates on infant industy protection policks, In every case but one in Korea and Yugoslavia, and in every case in Turkey, sectors with a statistically significant negatlve impact of import s u b s t i t u t i o n on TFP growth a r e a l s o s e c t o r s with a s i g n i f i c a n t p o s i t i v e impact of export expansion. Westphal (1981) has r e c e n t l y revived the infant-industry argument f o r s e l e c t i v e p r o t e c t i o n by noting a strong l i n k a t the micro l e v e l between p r o t e c t i o n and export performance. He concludes (p. 35) t h a t he has i d e n t i f i e d "...one possible reason why t h e i n d u s t r i a l s e c t o r i n a country l i k e Korea, following an outward-looking s t r a t e g y , p e r f o m s so well; namely, the p o s s i b i l i t y t h a t its s e l e c t i v e l y ! promoted i n f a n t i n d u s t r i e s e x h i b i t superior performance as a r e s u l t of t h e i r export a c t i v i t y . " Our .results f o r Korea a r e c e r t a i n l y c o n s i s t e n t with t h i s - argument 271 Krueger and Tuncer (1982) consider t h e standard i n f a n t i n d u s t r y argument i n Turkey, and conclude (p. 1149): "...input per u n i t of output must f a l l more r a p i d l y i n more protected i n d u s t r i e s i f t h e r e is t o be any r a t i o n a l e f o r i n f a n t i n d u s t r y protection. In the Turkish case, t h e r e was no such tendency over t h e period covered." I n Turkey, i n c o n t r a s t t o Korea, the export expansion phase was very s h o r t and not t h a t strong. It can be stated, t h e r e f o r e , t h a t t h e p o s i t i v e impact of export expansion on TFP growth which we found did not o f f s e t the negative impact of import s u b s t i t u t i o n . Our r e s u l t s a r e thus c o n s i s t e n t with those of Krueger and Tuncer, but we would be more d i f f i d e n t i n concluding t h a t p r o t e c t i o n was not j u s t i f i e d . The p o s i t i v e * m s r e l a t i o n s h i p between TFP growth and export performance i n Turkey indicates t h e n n g ~ f h l l i t v t h a t they ro~rlrfb n * ~ eF ? ? ? ~ ~ wCc ~~ "nrnan C n---2-g!,- T O ~ . - , - V : . , , - 3 - +- protection, with export, performance providing a t e s t of success. Indeed, they * still might do SO. 0 --27/ One should note, however, t h a t our r e s u l t s may at l e a s t p a r t l y r e f l e c t aggregation problems. Within any one of our "sectors," e x p o r t s and import s u b s t i t u t e s may be very d i f f e r e n t products. One f i n a l point is worth noting. I n both Korea and Turkey, an import substitution phase was followed by a phase with s i g n i f i c a n t export contribution t o growth. iJhile the Turkish export phase from 1970-73 turned out t o be abortive, largely because the government allowed incentives t o move against exports, the country i.s curren+ly entering a new period of rapid and successful export promotion. The observed phading leads n a t u r a l l y t o the hypothesis that a period of protected import substitution is useful--perhaps even necessary--to build a base from which a successful export drive, with associated positive TFP growth, can be launched. 28' IJestphal-s argument holds out the hope t h a t the benefits of export expansion on TFP growth can be -. realized simultaneously with the protection phase, but only i f the i n c e n t i v e s a r e t i e d t o export performance. Such was not the case i n Turkey, nor was there such an intention on the part of the policy makers. However, the question of whether a period of protected import substitution, with a s s o c i a t e d negative impact on TFP growth, is "worth" the costs is not so e a s i l y answered. The crucial policy question is t h a t of timing. How long must one wait for an infant t o mature? And is it possible t o devise a policy mix t h a t hastens the maturation process by tying policy incentives t o performance (especially exports)? 5. Conclusion . o The r e s u l t s we have presented have raised as many questions a s they - - t;,? . 7 p ,,. n.JC r,7 + - , ;> , - ' , . . ! . .. . .q - d - '--, I - . a sracq 01 Y . . 4 - L A iCI L ~ S C ? ~ Z C S U C ~ I I ~ , 6 a f f a i r s is probably-desirable. There !.s a r e a l need to coordinate gesearch a t r the micro and "secttral" and/or aggregate levels. The s o r t s of "stflized facts" that w e a r e considering must be tested against work a t the micro l e v e l -28/ See a l s o Kubo and Robinson (1979) who present data on such phasing i n other countries. See a l s o Balassa (1979). t o see i f t h e - s W m kes sense. Similarly, the micro work must be t e s t e d a g a i n s t z r ~ ~ s % L & a tata more aggregate l e v e l s t o see what s o r t s of g e n e r a l i z a t i o n s.r; -1e. There= - m e s o l v e d questions about t h e interdependence of d i f f e r e n t polic&xm; , A ~&smilopment strategy" implies a coordinated ef f e r t t o devise a consis- n-fpmlicies i n a number of areas. By d e f i n i t i o n , s u c h a strategy h a s l l r - a - b r g e p a r t of t o t a l economic a c t i v i t y i n a country. The e- si'JX5nkages and e x t e r n a l i t i e s implies t h c t it w i l l be d i f f i c u l t i f n0-m consider t h e impact of such s t r a t e g i e s i n a -- partial-equilib- The work t o d a t e i n d i c a t e s t h a t the r e l a t i o n s h i p s b-- development s t r a t e g i e s and TFP growth a r e v e r y important, i f n e m s a x C h & m gaining an understanding of what makes a "successful" dtf* 2-tegy. - - -. .., , m ? ? . ? : r n r ? ? -. n I n I d o - C: r ~ n ~ C I 1 - ? 7 ? f ? ? n n 5 Z O, O f f = f c : n . n - d . : S t - . . m C CI a n n u n - 2h g . . .. Ou 0 0 z > = = I * . u ;* : I z m r r 2 g i ; + ; < " ' I - . . ; $ i s :I 3 2 5 3 2 - 36 - References Ahluwalia, I.J., 1982, Industrial Performance in India: 1953-60 to 1978- 79. An Analysis of Deceleration in Growth Since the mid-Sixties (Indian Council for Research ori International Ecoaomic Relations, New . Delhi) Balassa, B., 1967, Trade Liberation Among Industrialized Countries: Objectives and Alternatives (McGraw-Hill, New York). Balassa, B., 1979, A Stages Approach to Comparative Advantage, in: I. Adelman, ed., Economic Growth and Resources (MacMillan, London), Bruton, H. J., 1967, Productivity Growth in Latin America, The American Ec.onomic Review LVII, no. 5 , 1099-1116. Caves, D.W., Christensen, L.R., and Diewert, W.E., 1981, A New Approach to Index Number Theory and the Measurement of Input, Output, and Productivity, Journal of Political Economy 88, no. 5 , 958-976. Caves, D.W., Christensen, L.R., and Diewert, W.E., 1982a, Multilateral Comparisons of Output, Input, and Productivity Using Superlative Index Numbers, The Economic Journal 92, no. 365, 73-86. Caves, D.W., Christensen, L.R., and Diewert, W.E., 1982b, The Economic Theory of Index Numbers and the Measurement of Input, Output, and Productivity, Econornetrica 50, no. 6, 1393-1414. Celasun, M., 1983, Sources of Industrial Growth and Structural Change: The Case of Turkey, mimeo (Development Research Deoartment, World Bank, Washington, D. C.). Chenery, H.B. and Westphal, L.E., 1939, Economies of Scale and Investment Over Time, in: H. B. Chenery, ed., Structural. Change and Development Policy (Oxford University Press, Oxford). I I Christenson, L.R., Cummings, D. and Jorgenson, D.W., 1980, Economic Growth, 1947-73: An International Comparison, in: J.W. Kendrick and B.N. Vaccara ens., New Developments in Productivity Measurement and . . ,... .. . - .,. . ~'iA:d~= A L j - \A,,< "r,i . . L C = ~ L ~ u i ~ . l i . c ~ & v-LSS, L G, 1iLd&0,. ' 1 Chrissensen, L.R., and Jorgenson, D.W., 1973, Measuring the performance of the private sector of the U.S. economy, 1929a969, in: M. Moss, ed., ? Measuring Economic and Social Performance (N&iorul Bureau of Economic Research, New York). Cowing, T.G. and Stevenson, R.E. eds., 1981, Productivity Measurement in Regulated Industries (Academic Press, London). Denison, E.F., 1967, Why Growth Rates Differ (The Brookings Institution, Washington, D.C.). Denison, E.F., 1374, Accounting for United States Economic Growth 1929- 1969 (The Brookings Institution, Washicgton, D.C.). Denison, E.F., and Chung, W., 1976, Economic Growth and Its Sources, in: H. Patrick and'H. Rosovsky, eds., Asia's New Giant, How the Japanese Economy Works (The Brookings Institution, Washington, D.C.). Dervis, K., de Helo, J., and Robinson, S., 1982, General Equilibrim Models for Development Policy (Cambridge University Press, Cambridge). Diewert, M.E., 1976, Exact and Superlative Index Numbers, Journal of Econometrics 4, no. 2, 115-146. Diewert, W.E., 1979, The Economic Theory of Index Numbers: A Survey, Discussion Paper 79-09 (Department of Economics, University of British Columbia). Elias, V.J., 1978, Sources of E?onornic Growth in,LatinAmerican Countries, The Review of Economics anL Statistics 60, 363-370. Ezaki, M., and Jorgenson, C.W., 1973, Geasurement of Macroeconomic Performance in Japan, 1951-1968, in: K. Ohkawa and Y. Hayami, eds., Economic Growth: The Japanese Experience Since the Meiji E7a (Japan Economic Research Center, Tokyo). Ezaki, M., 1975, Growth Accounting of the Philippines: A Comparative Study of the 1965 and 1969 Input-Output Tables, The Philippine Economic Journal XIV, no. 4, 399-435. Feder. G., 1983, On Exports and Economic Growth, Journal of Development Economics, (forthcoming). 1, Gollop, F.M., and Jorgenson, D.W., 1979, U.S. Economic Growth: 1948-1973, (unpublished manuscript). . . - ., .. - ., . , . ', _.._A_._p , G L O W C ~by L ' Y L x U di~;t;;:.,;~3;iy L.0. L J U V ,CI. L i i ~ i d ~ i j ~ i i ~ J. '3 Industry, 1947-73, in: J.Y. Kendrick and B.N. Vaccara, eds., New * Developments in Productivity Measurement and Analysis (The - i . University of Chicaqo gess, Chicago). r * Griliches, Z., and Jorgenson, D.W., 1967, The Explanation of Productivity Change, Review of Economic Studies 34, no. 99, 255-283. Jorgenson, D.W., and Nishid.zu, M., 1978, U.S. and Japanese Economic Growth, 1952-1974: An International Comparison, The Economic Journal 88, no. 352, 707-726. Jorge,~son,D.W. and Nishimizu, M., 1981, International Differences in 'Levels of Technology: A Comparison Bet2een U.S. and Japanese Industries, in; International Roundtable Congress Proceedings (The Institute of Statistical Xathenatics, Tokyo). Kaldor, N., 1961, Capital Accumulation and Economic Growth, in: F.A. Lutz and D.C. Hague, eds., The Theory of Capital (MacMLllan, London). Kaldor, N., 1967, Strategic Factors in Eronornic Development (W.F. Hwuphrey Preea, New York). Kendrick, J.W.., 1961, Productivity Trends in the United States. (Princeton University Press, Princeton). Kendrick, J.W., 1973, Postwar Productivity Trends in the United States, 1948-1969 (National Bureau of Econornic Research, New York). Kim, C.K. and Son, C.H., 1979, Productivity Analysis of Korean Manufacturing, 1966-75, (in Korean), Research Paper no. 79-01 ( ~ o k e a Development Institute, Seoul). Krueger, A.O., and Tuncer, B., 1980, Estimating Total Factor Productivity Srowth in a Developing Country." World Bank Staff Working Paper no. 422 (World Bank, Washington, D.C.). Krueger, A.O., and Tuncer, B., 1982, An Empirical Test of the Infant Industry Argument, American Economic Review no. 5, 1142-1152. Kubo, Y. and Robinson, S., 1979, Sources of Industrial Growth and Structural Change* A Comparative Analysis of Eight Countries, mi- (Developuent Research Department, World Bank, Washington, D.C.). . t Kubo, Y., Robinson, S. and Syrquin, M., 1981, The Nethodology of Multisector Comparative Analysis, aimeo (Development Research Department, World Bank, Washington, D.C.) Kuo, S.W.Y., 1983, The Taiwan Economy in Transition (Westview Press, Boulder). Kuroda, E:. and, Imamura, H., 1981, Productivity and ~ a r k e iPerformance, Time-series Analysis (1960-1977) in the ~a~anese?Esonom~,in: International Roundtable Congress Proceedings (The Institute of Statistical Mathematrcs, Tokyo). - Nadiri, M.I., 1970, Some Approaches to the T h e o w and Measurement of Total Factor Productivity: A Survey, Journal of Economic Literature 8, no. 4, 1137-1177. Nadiri, M.I., 1972, International Studies of Factor Inputs and Total Factor Productivity: A Brief Survey, Review of Income and Wealth 18, no, 2 194-254. .. Nelson, R.R., 1981, Research on Productivity Growth and Differences, Journal of Economic Literature XIX, no. 3, 1029-1064. Nishimizu, M. and Ilulten, C.R., 1978, The Sources of Japanese Economic Growth: 1955-1971, The Review of Economics and Statistics LX, no. 3 351-361. K!.sliimizu, M. and Page, J.M.: 1982, Total Factor Productivity Growth, Technological Progress and Technical Efficiency Change: Dimensions of Productivity Change in Yugoslavia, 1965-1978, Economic Journal 92, no. 368, 920-936. ~ishimizu,M. azd Robicson, S., 1983, Sectoral Productivity Growth in Semi- Industrial Countries: A Comparative Anelysis, in: 3.B. Chenery, s. Robinson, and M. Syrquin, eds., Industrialization and Growth: A Comparative Analysis (forthcoming). Rhee, S.Y., 1980, Total Factor Productivity Growth in Korean Mining and Manufacturing Industries, mimeo (Development Research Department, World Ba~k,Washington, D.C.). Robinson, S., 1971, Sources of Growth in Le-s Developed Countries: A Cross ,SectionStu.dy,Quarterly Journal of Economics 8.5,391-408. Salter, W.E.G., 1960, Productivity and Technical Change (Cambridge University Press, Cambridge). Verdaarn, P.J., 1949, Fattori che regolano lo sviluppo della producttivita del lavoro, LOIndustria. Westphal, L.E., 1981, Empirical Justification for Inzant Industry Protection, World Bank Staff Working Paper no. 445 (World sank, Washington, D.C.).