ECONOMIC REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH IN INDIA AND SOUTH KOREA: A COINTEGRATION ANALYSIS

Transcription

1 JOURNAL OF ECONOMIC DEVELOPMENT 39 Volume 36, Number 2, June 2011 ECONOMIC REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH IN INDIA AND SOUTH KOREA: A COINTEGRATION ANALYSIS SVITLANA MAKSYMENKO AND MAHBUB RABBANI * Universiy of Pisburgh and Onario Financing Auhoriy By employing a mulivariae ime series model, his sudy advances heoreical and empirical research on he role of economic reforms and human capial accumulaion in he pos-reform economic growh. We consruc wo indexes -a human capial index and a composie economic reform index- and perform a coinegraion analysis of a long-run equilibrium growh pah in India and Souh Korea welve years afer he implemenaion of reform. The significan posiive effec of human capial accumulaion is revealed in boh India and Souh Korea. The impac of economic reforms is found o be heerogeneous across counries: he effec is posiive, significan, and sizable in Souh Korea, while i is negaive and relaively small in India. This resul is suggesive of differen degrees of efficiency of reform measures implemenaion in wo counries. Keywords: Economic Growh, Human Capial, Economic Reforms, India, Souh Korea JEL classificaion: O10, O15, O47, O53 1. INTRODUCTION I is widely believed ha echnological change and capial accumulaion play a key role in economic growh. A he same ime, growh economiss recognize ha he developmen process usually deceleraes wihou organized markes, and as a resul, sociey ges deprived of a subsanial par of growh benefis. Taking ino accoun he imporance of economic reforms for marke organizaion, we invesigae he role of economic reforms and he role of human capial on economic developmen of wo Asian economies -Souh Korea and India- in heir respecive pos reform period. In he 1950s, Souh Korea was a poor developing counry. Is GDP per capia a he end of Korean War was less han $800. In less han fory years, Souh Korea s GDP per capia increased more han en imes, o $7235. India is a good sudy in conras. In he * Auhors are graeful o he anonymous referees for very helpful commens and suggesions.

2 40 SVITLANA MAKSYMENKO AND MAHBUB RABBANI middle of wenieh cenury, India s GDP per capia was slighly lower han Souh Korea s. India, like Souh Korea, prior o reform implemenaion, was a labor-abundan economy, closed o inernaional rade. Is capial sock per worker was $786, and degree of openness was 10.4%. Today, India is sill a poor, labor-abundan economy, wih a deprived sae of demographic and social developmens. Considering hese wo Asian economies, i is ineresing o explore he facors ha worked behind heir divergen pahs of economic growh and explain he considerable differences beween hem. Expers generally credi Souh Korean economic success o pragmaic marke reforms (World Bank, 1993). However, Rodrik (1996) aribues Souh Korean economic growh o marke-oriened policies and he reduced role of governmen inervenion. He also argues ha he success of reforms in Souh Korea could be explained by a beer-educaed labor force ha migh have simplified he esablishmen of a compeen bureaucracy and have enhanced he produciviy of inervenions aimed a increasing privae invesmen. Furhermore, Edwards (1992), and Levin and Renel (1992) poin ou ha marke reforms are associaed wih growh only in hose economies ha have appropriae human capial o absorb new developmens efficienly. According o Nehru e al. (1995), Souh Korea has he highes average educaion sock and he highes growh rae of educaion sock among developing counries, while India is a a relaively low level in boh caegories. To his end, Harvie and Pahlavani (2006) indicae ha i is he impressive invesmen in human capial (educaion, in paricular) ha has boosed Souh Korea s economic growh far beyond he level of oher Souh and Eas Asian economies. So, which facor -policy reform or human capial- has been more imporan for growh? This quesion has no been addressed ye. We aim o invesigae his issue in deeper deail in his paper. In his paper, we will comparaively evaluae he impac of wo facors -marke reforms and human capial- along wih heir complemenary effecs on economic growh in India and Souh Korea in he pos-reform period during he second half of he wenieh cenury. Our choice of counries is jusified by he fac ha India and Souh Korea are he wo larges economies in Asia. They were a a somewha similar economic level in per capia GDP erms when he reforms were iniiaed. And ye, he impac of reforms has differed significanly beween hese wo economies. In paricular, we have he following objecives for his paper. We perform a comparaive analysis of he economic reforms and he movemen of major economic variables before and afer he reform implemenaion in India and Souh Korea. As no explici daa are available for marke reforms and human capial, we develop a new mehodology for deerminaion and consrucion of wo composie indexes -a reform index and a human capial index- o measure hese variables implicily. We use a muli-variable ime series model o es he hypoheses of he impac of economic reforms and human capial on economic growh in India and Souh Korea in he long run. Our model is based on a modified producion funcion, which along wih convenional facors of producion, incorporaes composie reform and human capial indexes. We perform a coinegraion analysis of he poenial long-run growh funcions of each

3 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 41 counry. By finding elasiciies of counries GDP o economic reforms and human capial augmened labor, we race he effecs of differen reforms and human capial accumulaion on balanced growh pah of each economy. Based on he conclusions of our empirical analysis, we discuss policy implicaions and assess he soundness of economic reform policies. The res of he paper is srucured as follows. Secion 2 reviews he lieraure. Secion 3 develops a simple model o quanify empirically he impac of economic reforms and human capial accumulaion on economic growh. Secion 4 briefly inroduces readers o he recen economic developmens in India and Souh Korea wih an assessmen of marke reforms oucomes, describes daa, and performs empirical analysis based on a muli-variable ime series model. Secion 5 provides policy recommendaions, explores avenues for furher research, and offers conclusions. 2. LITERATURE REVIEW Wha is he key o he economic success of Souh Korea? Among professional economiss, he view is ha he Eas Asian miracle could be aribued o marke-oriened policies and he reduced role of governmen inervenion. 1 In paricular, Krueger (1993) argues ha in conras o he overconrolled, overregulaed, and highly disored oher Eas Asian economies, he Korean economy has been characerized by diminishing inervenion in mos spheres of economic aciviy, and much smaller degree of disorion. Rodrik (1996) also confirms ha he success of Souh Korea and Taiwan a he end of he wenieh cenury could acually be explained by an exensive se of reforms during he lae 1950s and early 1960s. Harvie and Pahlavani (2006) examine he major deerminans of GDP growh in Souh Korea during , allowing for he presence of poenial srucural breaks, which coincide in ime wih he effecs of he Asian crisis on he Korean economy. The auhors argue ha here is a coinegraion relaionship among he Korean GDP growh, invesmen, rade, and human capial. Employing an error correcion procedure, Harvie and Pahlavani (2006) conclude ha, in he long run, policies aimed a promoing various ypes of physical and human capial, and rade openness improved Korea s economic growh in The impac of oal impors is found o be insignifican, primarily due o capial vs consumer goods composiional changes. These are in line wih Ahmed (2008) and Zhuang and Koo (2008). Ahmed (2008) claims ha Souh Korean growh is inpu driven. Promoion of FDI helped he manufacuring secor o become he engine of economic growh, and as a resul companies such as Daewoo, Samsung and LG were o compee globally. Zhuang and Koo (2008) empirically prove ha he prospecive U.S. free rade agreemen wih Souh Korea would increase he Korean GDP by abou %. 1 For addiional discussion, see Ranis and Mahmood (1992).

4 42 SVITLANA MAKSYMENKO AND MAHBUB RABBANI However, sudies abou Indian growh are conroversial wih regard o reforms. Rodrik and Subramanian (2004) explore he causes of India s produciviy surge since 1980, more han a decade before serious economic reforms were iniiaed. They find ha radiional reforms -rade liberalizaion, expansionary demand, a favorable exernal environmen, and improved agriculural performance- did no play a role. Insead, hey aribue he growh o earlier pre-reform environmen ha played an imporan role in deermining which sae ook advanage of furher policy changes. In anoher sudy on Indian growh, Hausmann e al. (2005) argue ha growh ends o be correlaed wih increases in invesmen, rade, and real exchange rae depreciaion. However, as all hese growh acceleraions are highly unpredicable, hey conclude ha in mos insances economic reforms in India do no necessarily produce growh. Acharyya (2009) finds a saisically significan long run posiive impac of FDI inflow on GDP growh in India during On he oher hand, according o Acharyya (2009), his is also associaed wih a large negaive impac on environmen, via increased emission of CO 2. In conras, Zhang and Fan (2004) uses generalized mehod of momens o illusrae ha regardless of he declining invesmen rend in rural areas, infrasrucure reform in India urned ou o be producive. Oher sudies on Indian growh also repor ambiguous success of reforms, allowing Virmani (2009) o claim ha India s growh is an example of economic Sudoku. Boh sudies -of Indian and Souh Korean developmen- have some shorcomings. For example, Harvie and Pahlavani (2006) explain he growh of Korean GDP employing an unusual specificaion of producion funcion, 2 which makes he resuls quesionable. Invesigaion of he Indian growh in conrary overlooks he labor and human capial componen of developmen, wih he excepion of Venkaaramani e al. (2010) ha analyzes how improvemens in healh impac economic performance in agriculural secor. In his paper, we will address his gap in he lieraure and comparaively evaluae he impac of wo facors -combined marke reforms measures and human capial- along wih heir complemenary effecs on each oher, on economic growh in India and Souh Korea in he pos-reform period based on a coinegraion analysis. 3. THE MODEL In his secion, we develop an empirical model o ascerain he impac of hree major ypes of marke reforms and human capial developmen on economic growh in India and Souh Korea in he pos-reform periods. The hree major marke reforms considered 2 i.e., in Harvie and Pahlavani (2006) framework, he Cobb-Douglas producion funcion incorporaes along wih radiional deerminans -physical and human capial- wo addiional facors (expors and impors) ha come from expendiure approach o measuremen of GDP raher han producion process.

5 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 43 are rade liberalizaion, financial reform, and enerprise resrucuring. We make a convenional assumpion ha economic reforms, if properly implemened, boos oal facor produciviy (TFP). Insiuional changes ha occur in he process are considered exogenous. We assume a coninuous ime infinie horizon economy wih idenical, raional agens. A each ime, producion of a single homogenous good is represened by: Q A( R, O) F( K, hl), (1) where Q is he quaniy of oupu produced per period of ime, A is he oal facor produciviy, R is he reform measure, O is a cach-all facor for all oher effecs no explained by R, F is a general consan elasiciy of subsiuion producion funcion, K is a capial sock, L is he number of workers, h is he measure of human capial per worker, and hl is he oal labor inpu in he economy. According o Weil (2004), produciviy A can be deermined by wo facors: echnology, ha represens he knowledge abou how facors of producion are combined o produce oupu, and efficiency, ha measures how effecively given echnology and facors of producion are employed. The posiive impac of reform measures such as rade liberalizaion, financial reform, and enerprise resrucuring on boh echnology and efficiency componens of he oal facor produciviy A is well suppored by Edwards (1992). To his end, we esablish he following propery of a produciviy erm in our model A R 0. A shock o he economy from marke reforms will have a posiive effec on oal facor produciviy A. Moreover, a higher produciviy will accelerae economic growh, and oal producion Q will reach a new higher level a he end of ransiion process. The effecs of rade liberalizaion, openness o foreign invesmens, and enerprise resrucuring on economic growh are easily jusified. Liberalizaion of rade, wih reducion of ariffs, subsidies, and quoas on impors and expors increases compeiion in he domesic marke. In order o compee wih impored producs, domesic producers would have o increase heir produciviy. Similarly, exporers would need o increase qualiy and produciviy of heir oupu in order o compee in he world marke. Thus, by engendering efficiency, rade liberalizaion will boos economic growh and overall producion in he economy. When he economy reduces resricions on foreign invesmen, he level of capial inflows ino he economy is likely o increase. In he absence of disorions, and wih a raional maximizers behavior, he hihero inefficien economy should become more efficien. 3 Increased invesmen and enerprise resrucuring will boos producion direcly by increasing he capial level, and indirecly by improving efficiencies in he 3 However, if rade disorions are presen in he economy, foreign invesmens may furher increase producion of a wrong good and make he economy even more inefficien.

6 44 SVITLANA MAKSYMENKO AND MAHBUB RABBANI producion processes. New echnologies and more efficien producion mehods associaed wih foreign invesmen will speed up he economic growh. In a similar way, he effec of human capial on economic growh can be assessed. Improvemen in human capial will make labor more efficien. According o Equaion (1), his will have a direc posiive impac on he counry s producion. We specify he producion funcion similar o Kushnirsky (2001), and use his producion funcion as a benchmark funcion in our esimaions: Q OK ( hl R, (2) ) where a consan erm O and all he exponens,, and are independen of a counry choice. The funcion above will be esimaed wih coinegraion approach o evaluae how differen pahs of economic reforms and human capial accumulaion affec he growh of a counry. 4. EMPIRICAL STUDY The main objecive of his secion is o comparaively evaluae economic reforms in India and Souh Korea, and o explore he long-run relaionship among economic reforms, human capial accumulaion, and economic growh in hese wo counries in he pos-reform period by employing a coinegraion approach Daa Descripion For empirical sudy, we confine our daase o he periods when he major economic reforms were implemened in boh counries. Arguably, he defining iniial year of a major economic reform in Souh Korea is 1965, wih a second wave of changes iniiaed in he lae 1970s. As for India, Rodrik and Subramanyam (2004) poin ou ha even hough he firs hesian marke-oriened reform ook place in he mid-1980s, he more decisive reform occurred only in Driven by a relaively shorer pos-reform period in India, we consider a welve-year pos-reform period. To mainain symmery in daa, we confine our daase o in case of Souh Korea, and o in case of India. 4 The variables in our empirical analysis for boh India and Korea are as follows: GDP (Q): GDP in 2000 consan US dollars comes from he World Developmen 4 Even hough we recognize ha welve years imeframe is a relaively shor period for coinegraion analysis, we focus on his inerval in order o discuss an immediae impac of reform measures implemenaion in each counry.

7 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 45 Indicaors (WDI) The World Bank defines his as he sum of gross value added of all residen producers in he economy adjused for axes and subsidies. Labor (L): Toal labor force comes from he WDI The World Bank uses Inernaional Labor Organizaion (ILO) definiion of economically acive populaion o deermine oal labor force. I includes boh employed and unemployed, bu excludes homemakers, oher unpaid caregivers, and workers in he informal secor. Taking ino accoun a hriving informal secor in India, 5 and a relaively undeveloped daa collecion procedures in Korea in he 1960s, we acknowledge ha labor force daa in boh counries migh be biased and should be reaed wih cauion. Capial Sock (K): As no explici daa available for he capial sock, we conver Larson e al. (2000) capial sock daabase in 1990 US dollars o 2000 consan US dollars. We furher exend he daase o he remaining years of he period. Specifically, we use he Larson e al. (2000) fixed capial deflaor o conver oal capial sock in 1990 consan US dollars o curren US dollars. We hen employ he GDP deflaor from he WDI 2005 o conver he oal capial sock in curren US dollars o consan 2000 US dollars. We expand he capial sock daa over he period by applying a perpeual invenory mehod (PIM): K K 1 I D, (3) where K is a capial sock a ime, K 1 is a capial sock a ime (-1), I is invesmen a ime, and D is a depreciaion a ime. We use gross capial formaion daa from WDI 2005 as a proxy for invesmen (I) and assume ha he depreciaion rae is 5 percen, which according o he recen economic lieraure is close o realiy. Reform Index (R): We consruc a composie reform index R in he syle of De Melo e al. (1996). 6 Taking ino accoun he specifics of economic reforms in India and Souh Korea, our index is an average of hree reform indicaors -rade reform, financial reform, and enerprise resrucuring. Trade Reform: We define rade reform as a poliical process ha employs various ools o enhance counry s opporuniies and increase he volume of inernaional rade. As boh India and Souh Korea followed an impor-subsiuion developmen policy prior o reform implemenaion, liberalizaion of rade did no play a significan role in heir economies. In order o increase compeiiveness of expors, India implemened a major exchange rae reform in he early 1990s. Impors were also liberalized by removing quaniaive resricions and by gradually decreasing ariff raes. Ye, ariff 5 For deailed esimaions of he informal secor in India- refer o Chakrabari and Kundu (2009). 6 The De Melo e al. (1996) reform index is a weighed average of hree reform indicaors: 1) price liberalizaion and compeiion, 2) rade and foreign exchange regime, and 3) privaizaion and banking reform.

8 46 SVITLANA MAKSYMENKO AND MAHBUB RABBANI raes in India were relaively high compared o oher Eas Asian counries unil One of he noable feaures of Indian expors dynamics was a sharp increase in is service expors. A less repressive regulaion and an inflow of foreign invesmen were he wo key facors ha sood behind he success of service secor in India. In conras, he Korean governmen implemened a major exchange rae reform much earlier, in he mid-1960s. Reforms in he Korean impors focused mainly on he removal of quaniaive barriers raher han decreasing ariffs. A rade reform in Souh Korea was characerized by a consisen governmen suppor in expanding expor marke and broad incenives in he form of ax reducion for expor income. As a resul, in Souh Korea, he share of he manufacuring secor in GDP increased significanly in he pos-reform period. To measure rade reform, we use he volume of rade, or sum of impors and expors as a percenage of GDP in 2000 consan US dollars. Figure 1 shows his index for India and Souh Korea in he period of ineres. Figure 1. Volume of Trade as a Percenage of GDP in India and Souh Korea, The volume of rade as a percenage of GDP in 2000 consan US dollars is available from he WDI Assuming ha changes in his index are driven by policy changes, we use oucome-based measures o evaluae effeciveness of rade reforms. Specifically, he effeciveness of rade reforms is judged meaningfully by an increase in he rade volume i generaes. 7 7 Following Edwards (1992), we also performed evaluaion of he Srucure Adjused Trade Inensiy (SATI) o conrol for counry s size, GDP, ranspor cos and oher relevan variables. This alernaive measure of rade reform effeciveness produced no significan resuls ha migh have changed our choice.

9 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 47 Financial Reform: We look a he financial reform in India and Souh Korea as a broad se of policy measures ha leads o he furher secor developmen, liberalizaion, and deregulaion, wih paricular focus on allocaing credi o marke forces via assigning greaer flexibiliy in deermining ineres raes. India implemened a wide range of financial reform measures in early 1990s. These reforms included ineres rae deregulaion, opening up he banking secor o privae and foreign banks, and reducion of governmen inervenions in credi allocaion. Prudenial regulaions following Basel Commiee recommendaions significanly improved bank supervision. New rules were enaced o manage he securiies marke. Similarly, Souh Korea inroduced he firs sep of financial secor reforms by deregulaing ineres raes. Conrol over bank credis was reduced, and he banking secor was parially opened o foreign and privae banks in he lae 1960s - early 1970s. Similar o rade reform measure, we jusify variables measuring financial reform effeciveness by he oucome-based approach. Our financial reform index consiss of he following wo variables, wih equal weighs: 1) a raio of broad money o GDP (M2/GDP) ha measures a level of financial expansion in he economy: 8 2) a raio of domesic credi o privae secor o GDP ha approximaes a reducion of governmen inervenion in bank lending. All variables employed in he consrucion of his index come from he WDI Figure 2 shows dynamics of he financial reform indices in India and Souh Korea in Figure 2. Financial Reform Index in India and Souh Korea, Enerprise Resrucuring: Under enerprise resrucuring reform we look a he naional governmen pursues o resrucure enerprises from public o privae ownership, by enacing new laws ha encourage privae secor paricipaion in various economic 8 M2/GDP is assumed o move upward wih reforms in financial secor.

10 48 SVITLANA MAKSYMENKO AND MAHBUB RABBANI aciviies, and an increase in he privae secor share of value added and employmen. Since independence, India s economy was dominaed by huge public secor enerprises. In he 1990s, he governmen removed subsidies and preferenial access o bank loans for hese enerprises. A sharp reducion in he number of areas reserved for he public secor enerprises improved incenives for privae paricipaion and foreign direc invesmen (FDI). Reduced governmen regulaion and bureaucraic red ape along wih a larger amoun of FDI conribued o a surprising growh of he Indian service secor. Unlike India, he Souh Korean economy was no dominaed by public secor enerprises a he beginning of economic reforms. The Korean governmen resrucured he public secor in he lae 1960s o he early 1970s by mainly selling unprofiable public enerprises, and reorganizing oher enerprises geared owards economic developmen. Imporan seps were aken in order o remove favorable reamen of large conglomeraes and o enac legislaion regulaing monopolies. Considering daa availabiliy, we employ wo variables ha capure enerprise resrucuring effec on he economy: 1) privae secor share of oal employmen (in India), and 2) privae secor share of value added (in Souh Korea). The privae secor share of employmen daa were obained from he Reserve Bank of India; he daa on privae secor share of oal value added in Souh Korea come from Kang (1989) and he Bank of Korea. We consruc an overall reform index as an average of he hree reform measures - rade reform index, financial reform index, and enerprise resrucuring index. 9 Figure 3 depics dynamics of composie reform index in India and Souh Korea in Figure 3. Overall Reform Index R in India and Souh Korea, In our coinegraion analysis, we follow a benchmark model of Rabbani and Maksymenko (2011) o assign equal weighs o hree reform measures. Wihin he sensiiviy check in he panel daa analysis, Rabbani and Maksymenko (2011) also show ha assigning differen weighs o reform measures does no effecively change he resuls.

11 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 49 Human Capial (h): I has been esablished in economic heory ha human capial is one of he mos significan sources of economic growh. Neverheless, he empirical research has no ye produced convincing resuls o ascerain he imporance of i for economic growh. The main problem lies in he consrucion of a human capial variable, which is no direcly measurable. In order o proxy a human capial for he Asian region, mos empirical sudies (Harvie and Pahlavani, 2006; Song, 1990; Guesan, 2004) rely more on daa availabiliy raher han on a heoreical definiion. The variables commonly used as a proxy for human capial are invesmen in educaion, secondary or oal school enrollmen raio, lieracy rae, and average years of schooling. However, all hese raios have several disadvanages. 10 In addiion, he above menioned sudies omi healh componen of he human capial variable. In conras, Venkaaramani e al. (2010) focus on healh (proxied by disric infan moraliy rae) only. In our judgmen, a plausible variable for human capial should ake ino accoun reurns on all ypes of invesmens ha human beings underake in order o increase heir fuure well-being and producion poenial. Therefore, using concepual foundaions of he erm, we consruc a composie human capial index, which capures boh major componens of human capial -educaion and healh. The composie human capial variable is creaed as a weighed average of he wo indices- average years of schooling and life expecancy a birh - based on a principal componen analysis. Daa on he years of formal schooling received, on average, by aduls over age 15, defined as average years of schooling, are available from Barro and Lee (2000) for We use a linear inerpolaion mehod o esimae missing observaions. Daa on life expecancy a birh, or he number of years a newborn infan would expec o live if prevailing paerns of moraliy a he ime of her birh were o say he same hroughou her life, come from he WDI To address he issue of comparabiliy of indices, we se India s average years of schooling and life expecancy a birh in 1960 o uniy, and normalize he res of schooling and life expecancy daa respecively. Figure 4 depics he dynamics of human capial indices in wo counries in (2003). 10 See Barro (1991), Nehru e al. (1995), Gemmel (1996), Hanushek and Kimko (2000), and Wobmann

12 50 SVITLANA MAKSYMENKO AND MAHBUB RABBANI Figure 4. Human Capial Index in India and Souh Korea, Table 1 provides a summary of definiions and sources of variables used in his sudy. Table 1. Variables for he Empirical Analysis and Index Composiion Variable Definiion Source Q Gross Domesic Produc, in consan World Developmen Indicaors 2000 US dollars 2005 hl Human capial augmened labor Auhors calculaion L h K Toal labor force, or economically acive World Developmen Indicaors populaion 2005 Human capial index, assigns weighs a) Barro and Lee (2000) based on PCA o: b) World Developmen Indicaors 2005 a) average years of schooling b) life expecancy a birh Physical capial, in consan 2000 US dollars World Developmen Indicaors 2005 and Larson e al. (2000) R Reform index, assigns equal weighs o Auhors calculaions rade reform index, financial reform index, and enerprise resrucuring index Trade Reform (Expors +Impors)/GDP World Developmen Indicaors 2005 Index Financial Assigns equal weighs o he M2/GDP, World Developmen Indicaors 2005 Reform Index and he raio of domesic credi o privae secor o GDP Enerprise Resrucuring Index a) Privae secor share of oal employmen (India) and b) privae secor share of value added (Souh Korea) a) Handbook of Indian Economy by he Reserve Bank of India b) Bank of Korea, Naional Accouns ; Kang (1989)

13 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH Mehodology When differen ses of economic reforms are imposed on wo economies wih similar iniial condiions, economic forces migh drive hese economies owards differen long run equilibrium condiions. As discussed earlier, Souh Korea and India implemened economic reforms in a differen way: he Korean reform program was much more profound compared o he gradual Indian reform program. To explore he long-run relaionship beween economic growh and various economic reforms, we propose o use a coinegraion analysis. I will enable us o examine he long-run equilibrium condiions for a growh pah in each economy, and o race he effecs of differen ypes of reforms on his balanced growh pah over ime. The basic idea behind coinegraion is ha each of he componens of a ( p 1) vecor ime series Z may follow a non-saionary uni roo I(1) process, ye here may exis some linear combinaion of T Z, which ies he individual componens ogeher and is saionary (wihou a uni roo). 11 This linear combinaion can be inerpreed in economic erms as a long-run equilibrium relaionship among variables of a vecor Z. We will use a mulivariae maximum likelihood coinegraion procedure proposed by Johansen (1988), ha allows for he esimaion and esing for a number of coinegraing relaionships in he sysem. 12 All he hypohesis ess may be conduced in his case o using sandard asympoic chi-square ess Applicaions o India and Souh Korea In his secion, we conduc a coinegraion analysis of he variables in he growh model for India and Souh Korea. Specifically, he funcional relaionship in growh Equaion (2) Q F( K, hl, R ) will be analyzed. Every variable in he equaion is measured in logarihms. The coinegraion analysis for boh counries is conduced for pos-reform periods of 12 years, namely, for India and for Souh Korea. To apply Johansen s mehod for coinegraion analysis, every variable should be non-saionary, i.e., inegraed of some order greaer han zero. We perform he augmened Dickey-Fuller (ADF) es o deermine he inegraion properies of each variable series. The ADF es consrucs a parameric correcion for higher-order correlaion by adding p lagged difference erms of dependen variable z o he righ-hand side of he es regression under assumpion ha z series follow an AR(p) process: 11 If his is a case, is called a coinegraing vecor. 12 Phillips (1991) showed ha he mos preferred approach o he esimaion of coinegraed sysems is he Johansen s mehod, as he coefficien esimaes obained hrough his procedure are symmerically disribued, median unbiased, and asympoically efficien.

14 52 SVITLANA MAKSYMENKO AND MAHBUB RABBANI z z x z... z p p, (4) where z is a non-saionary series whose variance increases wih ime and approaches infiniy, x are exogenous regressors, which may conain a consan, or a consan and a rend, and are parameers o be esimaed, and is a whie noise. The null and alernaive hypoheses are as below: H H 0 1 : : 0, 0. (5) The above hypoheses are esed using a convenional -raio for : ˆ. (6) se( ˆ) As indicaed, we conduc he ADF es for he 12-year pos-reform period for boh India and Souh Korea. The iniial resuls of ADF saisics for ln(q), ln(hl), ln(k), and ln(r) are given in Table 2. Table 2. ADF Uni Roo Tes for Level Daa India Souh Korea Variables ADF Tes Saisics P Value ADF Tes Saisics P Value ln (Q) ln (hl) ln (K) ln (R) Noe: The p-values for he ADF es in he able above are based on McKinnon (1996). A 10% level of significance, all four variables are non-saionary for India and Souh Korea. In order o deermine he order of inegraion, we run he ADF es for he series in firs differences. The resuls in Table 3 show ha all differenced variables are saionary a leas a he 8 percen level.

15 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 53 Table 3. ADF Uni Roo Tes for Firs-Differenced Daa India Souh Korea Variables ADF Tes Saisics P Value ADF Tes Saisics P Value D(ln (Q)) D(ln (hl)) D(ln (K)) D(ln (R)) Noes: The p-values for he ADF es in he able above are based on McKinnon (1996). D( ) refers o he firs-differenced variable. Knowing ha all he variables are I(1) in he pos-reform period, we proceed wih coinegraion analysis. The iniial sep in coinegraion analysis is o deermine he number of lags needed in he VAR(p) model. We use Sims (1980) modified lag es o deermine he appropriae number of lags. We begin wih he maximum number of lags m and es he hypohesis ha coefficiens of lag m are joinly equal o zero by employing he likelihood raio (LHR) es: LHR T C)(ln ln ), (7) ( m 1 m where T is he number of observaions, C is he number of parameers esimaed in each equaion, m is he variance/covariance marix of he residuals from he VAR(p) sysem. The above LHR saisic has an asympoic chi-square disribuion wih degrees of freedom equal o he number of resricions in he sysem. In order o choose he appropriae lag lengh for he sysem, we sequenially es he significance of lags by comparing he LHR saisic o he 5 percen criical value saring from he maximum lag and decreasing one lag a a ime unil we firs ge a rejecion. The LHR saisic indicaes ha one lag would be appropriae o capure he dynamics in he pos-reform period for boh India and Korea. Table 4 presens he resuls of Johansen s coinegraion es ha deermines he coinegraing rank of he model. Boh he race and he maximum eigenvalue es saisics indicae a coinegraing rank of one for India and Souh Korea. The esimaed normalized coinegraing vecors of Z ( Q, K, hl, R ) are repored in Table 5. Adjusmens coefficiens usually inerpreed as he average speed of adjusmen owards he esimaed long run equilibriums sae. The adjusmen coefficiens derived for India and Korea are all significan a he 5% level. These significan adjusmen coefficiens imply ha all variables are converging owards he long run equilibrium.

16 54 SVITLANA MAKSYMENKO AND MAHBUB RABBANI Table 4. Johansen Coinegraion Tes Resul for Coinegraion Rank India s Saisics Souh Korea s Saisics Rank Trace Maximum Eigenvalue Trace Maximum Eigenvalue None * * * * A mos A mos A mos Noe: * denoes rejecion of he hypohesis a he 5% level. Table 5. Coinegraed Vecors from Johansen s Coinegraion Tes Unresriced Normalized Variable India Souh Korea India Souh Korea ln(q) ln(k) ln(hl) ln(r) Noe: all unresriced correlaion coefficien are saisically significan a he 5% level. By rearranging he coinegraing relaionship T Z o capure he long-run equilibrium condiions, we obain he following equaions: India: ln( Q ) 0.245ln( K ) 1.719ln( hl ) ln( R ), (8) Souh Korea: ln( Q ) ln( K ) 0.511ln( hl ) ln( R ). (9) In he economic sense, we can inerpre Equaions (8)-(9) as poenial long-run growh funcions. Several very imporan resuls can be observed. Firsly, he posiive elasiciy of GDP wih respec o human capial augmened labor in Equaions (8)-(9) confirms ha in boh counries -India and Souh Korea- human capial accumulaion played an imporan role in he pos-reform growh, in compliance wih Nehru e al. (1995), Rodrik (1996), Venkaaramani e al. (2008) and Rabbani and Maksymenko (2011). The impac of human capial augmened labor on Indian growh is found o be much larger in magniude compared o is Souh Korean equaion counerpars. No surprisingly, being a labor-abundan economy wih relaively low levels of schooling and life expecancy a he beginning of reform implemenaion, India experienced a much sronger inpu o is developmen process from he accumulaion of human capial raher han from is scarce facor of producion-capial.

17 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 55 Secondly, long-run equilibrium growh funcions indicae ha he impac of he reform measures is heerogeneous across counries. Namely, he elasiciy of GDP wih respec o economic reform in Souh Korea is posiive, whereas i is negaive in India. This resul is suggesive of a differen degree of efficiency in reform measures implemenaion in wo counries. Equaion (9) implies ha economic reforms as measured by a composie reform index were much more sizable and significan in he Souh Korean pos-reform growh relaive o India. In he case of Souh Korea, our findings agree wih he empirical lieraure, i.e., Harvie and Pahlavani (2006), on he impac of reform measures. The negaive long-run effec of he reform facor on economic growh in India, found by employing a coinegraion echnique, migh be surprising a he firs sigh. Ye, here are several facors ha may explain our resuls. Firsly, an iniial hesian approach o reform implemenaion in India migh parly explain India s less effecive reform oucomes during he firs welve years. Secondly, i can be jusified by unfavourable ime -invarian counry- specific facors (such as iniial sae of he economy, income disribuion, demographic ransiion, legal sysem, democracy, culure and radiions ec.), as well as persisen shor-erm complicaions in reform measures implemenaion in India. Tradiional correlaion analysis is no able o capure hese facors. Thirdly, he economic lieraure suggess ha having appropriae human capial is imporan for advancing he opporuniies opened by reforms. A higher level of human capial in Souh Korea a he beginning of is reform implemenaion o a cerain exen explains Korea s economic success. Rabbani and Maksymenko (2011) confirm a posiive ineracion beween economic reforms and human capial augmened labor in Souh Korea. Ye, in India, where he human capial dividend was a a relaively lower level prior o reform implemenaion, we do no observe his ineracion. To his end, he negaive sign of he reform measures on growh in India can be parly explained by inabiliy of he Indian labor o accommodae he pace of reforms. This leaves a broad avenue for he policy makers o address he issues of human capial, by expanding economic reforms o labor markes, educaion, and healhcare. While India is unique in having a very successful eriary educaion sysem, is primary educaion is in a poor sae, which is evidenced by a low lieracy rae. Therefore, primary educaion sysem is vial for he Indian effor o improve human capial. More funding, adequae eacher raining, and improved infrasrucure should lead owards his goal. Vocaional educaion can be provided o people who do no inend o pursue higher educaion. Finally, he inefficien enerprise-resrucuring and corrup paern of privaizaion may also have conribued o he negaive coefficien by Reform Index R. Prices for many naionalized companies were fixed during he privaizaion process, and he mehods of privaizaion favoured wealhy sraa. Privaizaion procedures and prices in many cases were no deermined by he marke, bu raher by he hierarchal allocaion process. This conribued o inefficiency of reform implemenaion as measured by enerprise resrucuring index, and in urn affeced he sign of coinegraion coefficien in our growh analysis. Wih all hese in mind, we can expec ha beyond he welve-year

18 56 SVITLANA MAKSYMENKO AND MAHBUB RABBANI horizon, when he economy moves furher owards he marke, he negaive impac of reform measures implemenaion in India would fade, and slowly change owards he posiive effec. We will leave his opic for he furher sudies. 5. CONCLUSIONS In his paper, we analyzed he impac of economic reforms and human capial on he economic growh of India and Souh Korea in he pos-reform period. We consruced a modified producion funcion which along wih a convenional facor of producion - physical capial- incorporaes a composie reform index and augmened labor in order show he effec of various reform measures and human capial accumulaion on economic growh. By consrucing wo unique indices -a composie economic reform index and a human capial index- we explored he ransiional dynamics of a growh in oal facor produciviy generaed by rade reform, financial reform, enerprise resrucuring, schooling, and improvemens in life expecancy. We employed a coinegraion analysis based on a modificaion of our baseline producion funcion o assess he effecs of economic reforms on he economic growh in India and Souh Korea. Our analysis suggess ha economic reforms and human capial accumulaion produce a significan long-run effec on economic growh in he counries under consideraion. The significan posiive effec of human capial augmened labor is revealed boh in India and Souh Korea in he welve-year pos-reform period. The impac of marke reforms is found o be heerogeneous across hese economies. The effec is posiive, significan, and sizable in Souh Korea, while i is negaive and small in India. This resul is suggesive of differen degree of efficacy of reform measures implemenaion in wo counries. The negaive long-run effec of he reform facor on economic growh in India is jusified by several facors, including an original hesian approach o reform implemenaion, unfavourable ime -invarian counry- specific facors (such as iniial sae of he economy, income disribuion, demographic ransiion, legal sysem, democracy, culure and radiions ec), and he absence of a complemenary effec of human capial augmened labor o marke reforms. In conclusion, our sudy opens a broad avenue for furher research. We recognize ha economic reforms and human capial accumulaion are imporan facors for explaining growh in India and Souh Korea. As such, i will be ineresing o employ a mulivariae maximum likelihood coinegraion echnique in order o explore he long-run ineracion beween economic growh, reform measures, and human capial developmen in a broader range of Asian counries ha recenly implemened marke reforms and significanly improved heir human capial.

19 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 57 REFERENCES Acharyya, J. (2009), FDI, Growh and he Environmen: Evidence from India on CO 2 Emission during he Las Two Decades, Journal of Economic Developmen, 34(1), Ahmed, E. (2008), Foreign Direc Invesmen Inensiy Effecs on he TFP Inensiy of ASEAN-5 plus 2, Journal of Economic Developmen, 33(2), Barro, R. (1991), Economic Growh in a Cross Secion of Counries, Quarerly Journal of Economics, 106, Barro, R., and J. Lee (2000), Inernaional Daa on Educaional Aainmen: Updaes and Implicaions, Naional Bureau of Economic Research Working Paper, Chakrabari, S., and A. Kundu (2009), Formal-Informal Secors Conflic: A Srucuralis Framework for India, Journal of Economic Developmen, 34(2), De Melo, M., C. Denizer, and A. Gelb (1996), From Plan o Marke: Paerns of Transiion, The World Bank Working Paper, Dickey, D., and W. Fuller (1981), Likelihood Raio Tes Saisics for Auoregressive Time Series wih a Uni Roo, Economerica, 49, Edwards, S. (1992), Trade Orienaion, Disorions, and Growh in Developing Counries, Journal of Developmen Economics, 39, Gemmel, N. (1996), Evaluaing he Impacs of Human Capial Socks and Accumulaion on Economic Growh: Some New Evidence, Oxford Bullein of Economics and Saisics, 58, Guesan, M. (2004), Human Capial, Trade and Developmen in India, China, Japan and Oher Asian counries, : Economeric Models and Causaliy Tess, Applied Economerics and Inernaional Developmen, 4, Hanushek, E., and D. Kimko (2000), Schooling, Labor-Force Qualiy, and Growh of Naions, American Economic Review, 90(5), Harvie, C., and M. Pahlavani (2006), Sources of Economic Growh in Souh Korea: An Applicaion of he ARDL Analysis in he Presence of Srucural Breaks, , Universiy of Wollongong Economics Working Paper, Hausmann, R., L. Priche, and D. Rodrik (2005), Growh Acceleraions, Journal of Economic Growh, 10(4), Johansen, S. (1988), Saisical Analysis of Coinegraion Vecors, Journal of Economic Dynamics and Conrol, 12, Kang, S. (1989), Korea s Privaizaion Plans and Pas Experiences, Korea Developmen Insiue Working Paper Series, Krueger, A. (1993), Poliical Economy of Policy Reform in Developing Counries, Cambridge, MA and London: MIT Press. Kushnirsky, F. (2001), A Modificaion of he Producion Funcion for Transiion Economies: Reflecing he Role of Insiuional Facors, Comparaive Economic Sudies, 43, 1-30.

20 58 SVITLANA MAKSYMENKO AND MAHBUB RABBANI Larson, D.F., R. Buzer, Y. Mundlak, and A. Crego (2000), A Cross-Counry Daabase for Secor Invesmen and Capial, The World Bank Economic Review, 14, Levin, R., and D. Renel (1992), A Sensiiviy Analysis of Cross-Counry Growh Regressions, American Economic Review, 82, McKinnon, J. (1996), Numerical Disribuion Funcions for Uni Roo and Coinegraion Tess, Journal of Applied Economerics, 11, Nehru, V., E. Swanson, and A. Dubey (1995), A New Daabase on Human Capial Sock in Developing and Indusrial Counries: Sources, Mehodology, and Resuls, Journal of Developmen Economics, 46, Phillips, P. (1991), Opimal Inference in Coinegraed Sysems, Economerica, 59, Rabbani, M., and S. Maksymenko (2011), Do Economic Reforms and Human Capial Explain Pos-Reform Growh? Macroeconomics and Finance in Emerging Marke Economies, 4(1), Ranis, G., and S. Mahmood (1992), The Poliical Economy of Developmen Policy Change, Cambridge, MA and Oxford: Basil Blackwell. Rodrik, D. (1996), Undersanding Economic Policy Reform, Journal of Economic Lieraure, 34(1), Rodrik, D., and A. Subramanian (2004), Hindu Growh o Produciviy Surge: The Mysery of he Indian Growh Transiion, Inernaional Moneary Fund Working Paper, 04/77. Sims, C. (1980), Macroeconomics and Realiy, Economerica, 48, Song, B. (1990), The Rise of he Korean Economy, Oxford Universiy Press: Hong Kong. Venkaaramani, A., K.R. Shanmugam, and J.P. Ruger (2010), Healh, Technical Efficiency and Agriculural Producion in Indian Disrics, Journal of Economic Developmen, 35(4), Virmani, A. (2009), The Sudoku of India s Growh, Business Sandard Ld.: New Delhi. Weil, D. (2004), Economic Growh, Boson, MA: Addison Wesley. Wobmann, L. (2003), Specifying Human Capial, Journal of Economic Surveys, 17(3), World Bank Sudy (1993), Eas Asian Miracle, Washingon, DC: World Bank. World Developmen Indicaors (2005), Washingon, DC: World Bank. Zhang, X., and S.H. Fan (2004), How Producive is Infrasrucure? A New Approach and Evidence from Rural India, American Journal of Agriculural Economics, 86(2), Zhuang, R., and W. Koo (2008), Implicaion of he U.S. Free Trade Agreemen wih Souh Korea, Journal of Economic Developmen, 33(1),

21 ECONOMIN REFORMS, HUMAN CAPITAL, AND ECONOMIC GROWTH 59 Mailing Address: Svilana Maksymenko, Deparmen of Economics, Universiy of Pisburgh, 4703 Wesley W. Posvar Hall, Pisburgh, PA 15260, U.S.A. Tel: Fax: Received Augus 18, 2008, Revised December 29, 2010, Acceped June 15, 2011.