ADB Economics Working Paper Series

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea
Sangho Kim and Donghyun Park No. 295 | December 2011

ADB Economics Working Paper Series No. 295

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea

Sangho Kim and Donghyun Park December 2011

Sangho Kim is Professor at the Department of International Trade, Honam University. Donghyun Park is Principal Economist in the Economics and Research Department, Asian Development Bank. An earlier version of this paper was published in August 2011 as Nanyang Technological University Economic Growth Centre Working Paper No. 2011/1. The authors accept responsibility for any errors in the paper.

Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org/economics 2011 by Asian Development Bank December 2011 ISSN 1655-5252 Publication Stock No. WPS124406 The views expressed in this paper are those of the author(s) and do not QHFHVVDULO\ UHHFW WKH YLHZV RU SROLFLHV of the Asian Development Bank.

The ADB Economics Working Paper Series is a forum for stimulating discussion and eliciting feedback on ongoing and recently completed research and policy studies undertaken by the Asian Development Bank (ADB) staff, consultants, or resource persons. The series deals with key economic and development problems, particularly WKRVH IDFLQJ WKH $VLD DQG 3DFLF UHJLRQ DV ZHOO DV FRQFHSWXDO DQDO\WLFDO RU methodological issues relating to project/program economic analysis, and statistical data and measurement. The series aims to enhance the knowledge on Asias development DQG SROLF\ FKDOOHQJHV VWUHQJWKHQ DQDO\WLFDO ULJRU DQG TXDOLW\ RI $'%V FRXQWU\ SDUWQHUVKLS VWUDWHJLHV DQG LWV VXEUHJLRQDO DQG FRXQWU\ RSHUDWLRQV DQG LPSURYH WKH TXDOLW\ DQG availability of statistical data and development indicators for monitoring development effectiveness. 7KH $'% (FRQRPLFV :RUNLQJ 3DSHU 6HULHV LV D TXLFNGLVVHPLQDWLQJ LQIRUPDO SXEOLFDWLRQ ZKRVH WLWOHV FRXOG VXEVHTXHQWO\ EH UHYLVHG IRU SXEOLFDWLRQ DV DUWLFOHV LQ SURIHVVLRQDO journals or chapters in books. The series is maintained by the Economics and Research Department.

Contents
Abstract I. II. Introduction Ownership Structure and Export Performance: Conceptual Basis and Empirical Literature A. B. III. IV. Conceptual Overview Empirical Literature v 1

2 2 3 4 6 7 10 14 15

Data and Variables Main Empirical Findings A. B. Logit Estimation of Exporters versus Nonexporters Tobit Estimation of Export Propensity

V.

Concluding Observations

References

Abstract
The central objective of our paper is to empirically examine the relationship EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH 7R GR so, we use data from the Republic of Korea, a classic example of successful export-oriented industrialization. While a large and growing empirical literature investigates the relationship between the ownership structure and overall SHUIRUPDQFH RI UPV WKHUH DUH DOPRVW QR VWXGLHV WKDW GHOYH LQWR WKH LVVXH RI whether the concentration of ownership has a positive or negative effect on export performance. The primary contribution of our study is to help remedy this serious gap in the empirical literature on ownership and performance. Our HPSLULFDO UHVXOWV LQGLFDWH WKDW .RUHDQ UPV ZLWK PRUH FRQFHQWUDWHG RZQHUVKLS DUH more likely to be exporters, and to export more.

I. Introduction
$FFRUGLQJ WR WKH FRUSRUDWH JRYHUQDQFH OLWHUDWXUH D UPV RZQHUVKLS VWUXFWXUH DIIHFWV LWV performance. In a seminal paper, Berle and Means (1932) proposed a simple intuitive H[SODQDWLRQ IRU WKH UHODWLRQVKLS EHWZHHQ D UPV RZQHUVKLS VWUXFWXUH DQG LWV SHUIRUPDQFH $ PDQDJHU WHQGV WR SXUVXH KLVKHU RZQ LQWHUHVW UDWKHU WKDQ RZQHUV SURWV LQ WKH DEVHQFH RI DGHTXDWH PRQLWRULQJ :KHQ RZQHUVKLS LV ZLGHO\ GLVSHUVHG RZQHUV GR QRW IDFH VWURQJ incentives to engage in monitoring the management since they incur high monitoring FRVWV EXW FDSWXUH RQO\ D VPDOO VKDUH RI WKH EHQHWV $OO RZQHUV WKXV IDFH DQ LQFHQWLYH WR IUHH ULGH RQ RWKHUV 7KLV LPSOLHV WKDW UPV ZLWK D PRUH FRQFHQWUDWHG RZQHUVKLS ZLOO EH more closely monitored and will perform better. At the same time, there are theoretical reasons for why concentrated ownership may harm corporate performance. For example, D GRPLQDQW VKDUHKROGHU FDQ H[SURSULDWH UP UHVRXUFHV WR KLV EHQHW DW WKH H[SHQVH RI minority shareholders (Joh 2003). In theory, ownership concentration can thus have either a positive or negative impact on D UPV SHUIRUPDQFH 7KHUHIRUH ZKHWKHU RZQHUVKLS FRQFHQWUDWLRQ LV EHQHFLDO RU KDUPIXO IRU FRUSRUDWH SHUIRUPDQFH LV XOWLPDWHO\ DQ HPSLULFDO TXHVWLRQ WKDW QHHGV WR EH UHVROYHG through empirical analysis. A large and growing empirical literature has sprung up to H[DPLQH SUHFLVHO\ WKH LVVXH 7KH OLWHUDWXUH ZKLFK LQFOXGHV 'HPVHW] DQG /HKQ  0RUFN 6KOHLIHU DQG 9LVKQ\  0F&RQQHOO DQG 6HUYDHV  DQG +HUPDOLQ DQG :HLVEDFK  IDLOV WR XQFRYHU DQ\ GHQLWLYH SDWWHUQ $ QXPEHU RI DGGLWLRQDO VWXGLHV have elaborated upon the empirical methodology. For example, in order to account for UP KHWHURJHQHLW\ +LPPHOEHUJ +XEEDUG DQG 3DOLD  LQFOXGHG UPVSHFLF FRQWURO YDULDEOHV DQG *ULOLFKHV DQG +DXVPDQ  DSSOLHG WKH [HGHIIHFWV PRGHO WR SDQHO data. However, the overall evidence on the relationship between ownership structure and corporate performance remains mixed and inconclusive. The central objective of our paper is to empirically examine the relationship between WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH 'XH WR JOREDOL]DWLRQ DQG LQWHJUDWLRQ RI PDUNHWV H[SRUW SHUIRUPDQFH LV DQ LQFUHDVLQJO\ LQXHQWLDO GHWHUPLQDQW RI overall corporate performance. At the same time, there are some conceptual grounds for D UHODWLRQVKLS EHWZHHQ D UPV RZQHUVKLS VWUXFWXUH DQG LWV SHUIRUPDQFH DV H[SODLQHG LQ 6HFWLRQ ,,, ,Q SDUWLFXODU UPV ZLWK PRUH FRQFHQWUDWHG RZQHUVKLS PD\ EH PRUH OLNHO\ WR export due to the higher risk of exporting vis--vis selling in the domestic market. While a large and growing empirical literature investigates the relationship between ownership structure overall corporate performance, there are almost no studies that delve into the relationship between ownership structure and export performance. The main contribution

2 | ADB Economics Working Paper Series No. 295

of our study is to help remedy this serious gap in the empirical literature by investigating WKLV UHODWLRQVKLS RQ WKH EDVLV RI UPOHYHO GDWD IURP WKH 5HSXEOLF RI .RUHD At a broader level, the Republic of Korea is a well-known example of highly successful export-led industrialization and growth. Underlying its export prowess is the success of Korean companies such as Samsung, LG, and Hyundai, which have become global EUDQGV ZLWK RSHUDWLRQV DOO RYHU WKH ZRUOG 0RUH JHQHUDOO\ .RUHDQ UPV WKDW YDU\ ZLGHO\ in terms of structural characteristics, including size and ownership concentration, export a broad range of manufactured products. Given the export success of Korean companies and their structural diversity, it would be interesting to examine whether ownership FRQFHQWUDWLRQ KDV D VLJQLFDQW HIIHFW RQ WKHLU H[SRUW SHUIRUPDQFH ,Q DGGLWLRQ WR LQIRUPLQJ XV DERXW WKH UROH RI RZQHUVKLS VWUXFWXUH LQ .RUHDQ UPV H[SRUW SHUIRUPDQFH WKH VWXG\ PDUNV D UVW VWHS WRZDUG OOLQJ D PDMRU JDS LQ WKH EURDGHU HPSLULFDO OLWHUDWXUH RQ ownership structure and corporate performance. The rest of the paper is organized as follows. Section II outlines a conceptual basis for D UHODWLRQVKLS EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH and reviews the empirical literature on this relationship. Section III describes the data and variables used in the empirical analysis, and Section IV reports and discusses the PDLQ QGLQJV RI WKH DQDO\VLV 6HFWLRQ 9 EULQJV WKH SDSHU WR D FORVH ZLWK VRPH FRQFOXGLQJ observations.

II. Ownership Structure and Export Performance: Conceptual Basis and Empirical Literature
,Q WKLV VHFWLRQ ZH UVW RXWOLQH WKH FRQFHSWXDO EDVLV IRU D UHODWLRQVKLS EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI D UP DQG LWV H[SRUW SHUIRUPDQFH DQG WKHQ EULH\ UHYLHZ WKH empirical literature.

A.

Conceptual Overview

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concentrated ownership will try to minimize risk and thus concentrate on the domestic

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 3

market. This manager wants to achieve concrete results in a short period by focusing on the less risky domestic market rather than the more risky foreign market. The implication LV WKDW UPV ZLWK PRUH FRQFHQWUDWHG RZQHUVKLS DUH PRUH OLNHO\ WR H[SRUW This study will hypothesize that managers preferences toward risk are related to its RZQHUVKLS VWUXFWXUH ,Q D UP ZLWK D GLIIXVH RZQHUVKLS VWUXFWXUH WKH PDQDJHU PD[LPL]HV KLV RZQ LQWHUHVWV UDWKHU WKDQ VKDUHKROGHUV SURWV E\ YHQWXULQJ LQWR WKH IRUHLJQ PDUNHW 7KH PDQDJHU ZLOO PD[LPL]H KLV H[SHFWHG XWLOLW\ IURP WKH UPV SURWV E\ DYRLGLQJ ULVN +RZHYHU LQ D UP ZLWK FRQFHQWUDWHG RZQHUVKLS WKH PDQDJHU UHSUHVHQWV WKH GRPLQDQW VKDUHKROGHUV LQWHUHVW DQG WKXV PD[LPL]HV WKH UPV H[SHFWHG SURWV UDWKHU WKDQ KLV RZQ H[SHFWHG XWLOLW\ 7KHUHIRUH WKH PDQDJHU RI VXFK UPV EHDUV WKH ULVNV DVVRFLDWHG ZLWK IRUHLJQ PDUNHWV 5LVN SUHIHUHQFH KDV FUXFLDO LPSOLFDWLRQV IRU D UPV H[SRUW GHFLVLRQ VLQFH H[SRUWLQJ LV D IXQGDPHQWDOO\ ULVN\ DFWLYLW\ %UHDNLQJ LQWR WKH H[SRUW PDUNHW UHTXLUHV D ODUJH sum of upfront sunk costs, including adapting products to foreign consumer preferences, complying with foreign government regulations, and building distribution networks. ([SHFWHG SURWV DUH VXEMHFW WR KLJK ULVNV GXH WR OLPLWHG LQIRUPDWLRQ DERXW IRUHLJQ GHPDQG DQG H[FKDQJH UDWH XFWXDWLRQV ,Q WKHVH FLUFXPVWDQFHV WKH GHFLVLRQ WR H[SRUW GHSHQGV PRVWO\ RQ FRVWV DQG H[SHFWHG SURWV IRU UPV ZLWK FRQFHQWUDWHG RZQHUVKLS VWUXFWXUH EXW ULVN DWWLWXGHV IRU UPV ZLWK GLVSHUVHG RZQHUVKLS

B.

Empirical Literature

As noted earlier, there is a large and growing empirical literature that delves into the UHODWLRQVKLS EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU RYHUDOO SHUIRUPDQFH 7KLV literature looks at the relationship between ownership concentration and measures of RYHUDOO FRUSRUDWH SHUIRUPDQFH VXFK DV SURW UDWH ,Q PDUNHG FRQWUDVW WR WKLV ULFK OLWHUDWXUH there are only a few empirical studies investigating the nexus between ownership and exports. Most of the few studies look at the impact of foreign ownership rather than ownership concentration on export performance. Cole, Elliot, and Virakul (2010) investigate the relationship between foreign ownership DQG D UPV GHFLVLRQ WR H[SRUW XVLQJ WKH DQQXDO VXUYH\ RI 7KDL PDQXIDFWXULQJ UPV IURP  WR  7KH\ QG WKDW IRUHLJQRZQHG UPV DUH PRUH OLNHO\ WR H[SRUW WKDQ GRPHVWLF UPV 7KH\ IXUWKHU QG WKDW WKH SURSHQVLW\ WR H[SRUW GLIIHUV DFFRUGLQJ WR WKH FRXQWU\ RI RZQHUVKLS 1JRF DQG 5DPVHWWHU  DQDO\]H GDWD RQ PXOWLQDWLRQDO UPV LQ 9LHW 1DP to examine the relationship between foreign ownership and exports in the Vietnamese PDQXIDFWXULQJ VHFWRU 7KH\ QG WKDW FRPSDQLHV ZLWK D YHU\ KLJK VKDUH RI IRUHLJQ ownership, i.e., 90% or more foreign-owned, make a disproportionate contribution to Viet Nams manufacturing exports. A number of additional empirical studies examine the relationship between foreign ownership and export performance in other countries. These LQFOXGH )LODWRWFKHY 6WHSKDQ DQG -LQGUD  :LJQDUDMD  5RMHF 'DPLMDQ DQG 0DMFHQ  DQG 5DVLDK   DQG 

4 | ADB Economics Working Paper Series No. 295

To summarize our literature review, it should be clear that the empirical literature on WKH UHODWLRQVKLS EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH LV TXLWH OLPLWHG 0RUH LPSRUWDQWO\ WKLV OLPLWHG OLWHUDWXUH GHDOV SULPDULO\ ZLWK WKH HIIHFW RI foreign ownership on export performance. Some studies compare the export performance RI IRUHLJQRZQHG UPV YHUVXV GRPHVWLF UPV ZKHUHDV RWKHUV FRPSDUH WKH H[SRUW SHUIRUPDQFH RI UPV ZLWK GLIIHUHQW VKDUHV RI IRUHLJQ RZQHUVKLS 1RQH RI WKH VWXGLHV delve into dimensions of ownership structure other than foreign ownership. In particular, the studies do not look at the effect of ownership concentration on export performance. +RZHYHU WKHUH DUH FRQFHSWXDO JURXQGV IRU EHOLHYLQJ WKDW ZKHWKHU D UPV RZQHUVKLS VWUXFWXUH LV UHODWLYHO\ FRQFHQWUDWHG RU GLIIXVH ZLOO LQXHQFH LWV VXFFHVV DV DQ H[SRUWHU 7KH underlying intuition is that exporting is fundamentally risky activity, as explained above.

III. Data and Variables

In this section, we describe the data and variables used in our empirical analysis of the UHODWLRQVKLS EHWZHHQ WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH 7KH data set used in this paper is an unbalanced panel consisting of annual time-series for  .RUHDQ PDQXIDFWXULQJ UPV GXULQJ  ZLWK D WRWDO RI  REVHUYDWLRQV 7KH VDPSOH FRYHUV DOO PDQXIDFWXULQJ UPV ZKRVH VWRFNV DUH OLVWHG RQ WKH .RUHDQ 6WRFN ([FKDQJH 7KH HQOLVWHG UPV DUH UHTXLUHG WR UHSRUW WKHLU QDQFLDO VWDWXV $OO UPV GDWD DUH WDNHQ IURP WKHLU QDQFLDO UHSRUWV Exports are observed in 1,640 observations, which is about 29.5% of total observations. *LYHQ WKDW D ODUJH SRUWLRQ RI REVHUYDWLRQV DUH GRPHVWLF UPV D ELQDU\ YDULDEOH RI H[SRUWHUVQRQH[SRUWHUV LV UVW FRQVWUXFWHG WR LQYHVWLJDWH WKH LPSDFW RI RZQHUVKLS FRQFHQWUDWLRQ RQ D UPV H[SRUW GHFLVLRQ :H XVH WKH ORJLW PRGHO ZKLFK LV ZLGHO\ XVHG LQ WKH OLWHUDWXUH RQ UPVSHFLF HIIHFWV RQ H[SRUW DFWLYLW\ 7KLV PRGHO UHJUHVVHV WKH binary variable on the set of explanatory variables that include concentration rate and RWKHU UHOHYDQW YDULDEOHV ,Q WKH ORJLW PRGHO FRHIFLHQW HVWLPDWHV UHSUHVHQW WKH LPSDFW RI H[SODQDWRU\ YDULDEOHV RQ WKH SUREDELOLW\ RI UPV EHLQJ H[SRUWHUV ,Q DGGLWLRQ ZH HVWLPDWH D 7RELW PRGHO WR VWXG\ UPV H[SRUW SURSHQVLW\ LH UDWLR RI H[SRUWV WR WRWDO VDOHV 7KH SURSHQVLW\ WR H[SRUW LV GHQHG RQ > @ ZKLFK VXJJHVWV WKDW D 7RELW PRGHO GHVLJQHG IRU censored data will be useful.1 Our key variable of interest is the ownership concentration rate (CR) since the central objective of our empirical analysis is to investigate the effect of ownership structure on H[SRUW SHUIRUPDQFH &5 LV GHQHG DV WKH UDWLR RI WKH GRPLQDQW VKDUHKROGHUV VKDUH WR WRWDO VKDUHV DQG UHHFWV WKH H[WHQW WR ZKLFK WKH GRPLQDQW VKDUHKROGHU FRQWUROV WKH management. CR is a good proxy for ownership structure since dominant shareholders WHQG WR H[HUFLVH D JUHDW GHDO RI LQXHQFH RQ WKH PDQDJHPHQW RI .RUHDQ UPV
1

Applying ordinary lease square regression method will produce biased estimates.

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 5

To accurately estimate the impact of ownership concentration on exports, we have to FRQWURO IRU RWKHU UPVSHFLF IDFWRUV WKDW LQXHQFH H[SRUWV 2XU FRQWURO YDULDEOHV LQFOXGH standard variables such as wage rate, capital intensity, research and development 5 ' VWRFN UP VL]H SURGXFWLYLW\ DQG UP DJH $PRQJ WKHVH YDULDEOHV ZDJH UDWH DQG FDSLWDO LQWHQVLW\ UHSUHVHQW WKH WUDGLWLRQDO IDFWRU HQGRZPHQW WKHRU\ 5 ' VWRFN UHHFWV WKH technology gap theory of trade (see Posner 1961, Krugman 1979) or the product cycle theory (see Vernon 1966) in which technological innovation plays a central role in shaping international trade structure. Many empirical studies have analyzed the impact of R&D on UPV H[SRUW DFWLYLW\ VHH Kumar and Siddharthan 1994, Hirsch and Bijaoui 1985, Wakelin 1998). 7KHUH DUH [HG FRVWV DVVRFLDWHG ZLWK HQWHULQJ H[SRUW PDUNHWV 7KHVH LQFOXGH FROOHFWLQJ information, establishing a distribution network, and adapting products to foreign tastes and regulations. Since these costs are sunk costs, uncertainty may cause persistence in export participation. Firms may continue to export even though it is temporarily XQSURWDEOH WR GR VR RU KHVLWDWH WR H[SRUW GXH WR WKH RSWLRQ YDOXH RI ZDLWLQJ IRU PRUH information (see Roberts and Tybout 1997). For this reason, a number of studies point out that exporters are large and productive enough to absorb the costs of waiting (see Richardson and Rindal 1995, Bernard and Jensen 1997a and 1997b, Bleaney and Wakelin 1999, Roberts and Tybout 1997). Firm size affects export performance through economies of scale in production and export marketing, higher capacity for taking risks, EHWWHU DFFHVV WR QDQFLQJ DQG VXIFLHQW PDQDJHULDO 5 ' DQG PDUNHWLQJ UHVRXUFHV %RWK SURGXFWLYLW\ DQG DJH DUH DGGLWLRQDO UPVSHFLF YDULDEOHV WKDW PD\ LQXHQFH D UPV decision to export. The effect of productivity on exports is intuitively straightforward since global markets are typically much more competitive than the domestic market. Therefore, PRUH SURGXFWLYH DQG HIFLHQW UPV DUH PRUH OLNHO\ WR EH LQWHUQDWLRQDOO\ FRPSHWLWLYH DQG PRUH OLNHO\ WR H[SRUW 7KH LPSDFW RI UP DJH RQ H[SRUWV LV DPELJXRXV 2Q RQH KDQG ROGHU UPV PLJKW KDYH KLJKHU H[SRUW SURSHQVLWLHV EHFDXVH WKH\ DUH PRUH H[SHULHQFHG LQ LQWHUQDWLRQDO WUDGH 2Q WKH RWKHU KDQG PDQ\ QHZHU UPV DUH PRUH VXFFHVVIXO ZLWK QHZ technology, which can be an important tool for exports. The positive impact of age might GLPLQLVK EH\RQG D FHUWDLQ WKUHVKROG DV D UPV OHDUQLQJ FXUYH ULVHV DW D GHFUHDVLQJ UDWH ,Q OLJKW RI WKLV SRVVLELOLW\ ZH LQFOXGH ERWK DJH DQG DJH VTXDUHG LQ RXU HVWLPDWLRQ We compute the wage rate by dividing total labor costs by the number of employees (L). Total labor costs consist of wages, bonuses, retirement compensation, and all other costs associated with employee remuneration. The capital stock (K) is the real amount RI WDQJLEOH [HG DVVHWV 'LYLGLQJ . E\ / JLYHV XV WKH FDSLWDOODERU UDWLR ./ /DERU SURGXFWLYLW\ 9$/ LV SHU FDSLWD YDOXH DGGHG 9$ DQG FRPSULVHV QHW SURWV ODERU FRVWV net interest payments, rents, taxes other than corporate tax, and depreciation costs. 5 ' VWRFN LV HVWLPDWHG E\ SHUSHWXDO LQYHQWRU\ PHWKRGV EDVHG RQ UPV 5 ' LQYHVWPHQW Following much of the literature, we apply a depreciation rate of 10%. Firm age is calculated from the founding year. All variables are converted into constant 2000 prices.

6 | ADB Economics Working Paper Series No. 295

Table 1 presents sample means and standard deviations. We perform the t-test on the QXOO K\SRWKHVLV WKDW PHDQ YDOXHV DUH HTXDO EHWZHHQ H[SRUWHUV DQG QRQH[SRUWHUV 7KH QXOO ZDV UHMHFWHG DW WKH  VLJQLFDQFH OHYHO IRU HYHU\ YDULDEOH H[FHSW DJH 0RVW VLJQLFDQWO\ this implies that exporters have higher ownership concentration than nonexporters. In addition, relative to nonexporters, exporters pay higher wages and have greater capital intensity ratio, R&D stock, labor productivity, and sales. Table 1: Variables for Ownership Concentration and Exports of Korean Manufacturing Firms: Means (standard deviations)
Variable (Abbreviation) Exporters Export propensity Ownership concentration rate (OC) Wage rate (Wage) Capital intensity (K/L) Research and Development Stock (R&D) Labor Productivity (Prod.) Firm age (Age) Sales
Note:

Definition Exporters if exports>0 Exports/sales Largest shareholders share rate Labor costs/employment Fixed capital (K)/employment (L) R&D stock Value added (VA)/employment (L) Years elapsed from foundation Total sales

Exporters 1640+ 0.410(0.291) 29.74(19.16) 14.69(19.94) 1.875(2.410) 264.76(2008.7) 17.42(30.44) 28.65(11.91) 10927(54940)

Nonexporters 3916 25.76(19.51) 11.83(17.78) 1.581(1.845) 77.67(933.0) 13.28(29.94) 28.74(16.73) 2621(12434)

+ denotes the number of exporting firms. All the other variables are in 10 million Korean won in 2000 constant prices, except largest shareholders share rate and firm age. To convert into approximate US dollars, divide by 1,000.

IV. Main Empirical Findings

In this section, we report and discuss the main results from our empirical analysis, which consists of two parts (i) logit estimation of the decision to export or not and (ii) tobit estimation of export propensity. Before estimating the logit and tobit models, we examine the correlation between explanatory variables. Table 2 reports the Pearson correlation FRHIFLHQWV 7KHUH H[LVWV D VWURQJ FRUUHODWLRQ EHWZHHQ ZDJH UDWH DQG ODERU SURGXFWLYLW\ R&D stock and both sales and employment, and sales and employment. We do not use these pairs of variables together in regressions to avoid multicollinearity.2

Putting these correlated variables together changed the significance and sign of coefficient estimates, which is apparent symptom of multicollinearity.

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 7

Table 2: Correlation between the Variables for Ownership Concentration and Exports of Korean Manufacturing Firms: Pierson Correlation Coefficient
Variable Wage K/L R&D Prod. Age Sales L OC 0.0615 0.0930 0.0617 0.0355 0.0817 0.0531 0.0771 Wage 0.2273 0.1099 0.855 0.0427 0.176 0.0777 K/L R&D Prod. Age Sales

0.0223 0.1803 0.1526 0.1113 0.0313

0.1496 0.0135 0.7279 0.6499

0.0328 0.2372 0.1087

0.0162 0.0566

0.7775

Age = firm age, K/L = capital intensity, OC = ownership concentration rate, Prod. = labor productivity (=VA/L), R&D = research and development, Sales = total sales, Wage = wage rate. Source: Authors estimates.

A.

Logit Estimation of Exporters versus Nonexporters

Table 3 reports the results of our logit estimation of the export decision, i.e., whether or QRW WR H[SRUW RI RXU VDPSOH RI .RUHDQ PDQXIDFWXULQJ UPV 7KH UHJUHVVLRQ XVHV D ELQDU\ variable of exporter or nonexporters as the dependent variable. For our purposes, the key explanatory variables is the ownership concentration rate. Additional explanatory variables include variables widely used in the trade literature, such as wage rate, capital intensity, and R&D stock. We experiment with various permutations of explanatory variables to estimate four different models. Table 3: Logit Regression for Ownership Concentration and Exports of Korean Manufacturing Firms: Dependent Variable (exporters/nonexporters)
Variable OC Wage K/L R&D Prod. Age Age2 Sales Constant LLR 1.181*** (0.064) 2937.85 65.47*** 4837 1.167*** (0.063) 2938.46 64.25*** 4837 0.00001*** (0.00000) 1.131*** (0.060) 3212.95 101.16*** 5261 Model (i) 0.005*** (0.001) 0.005*** (0.001) 0.055*** (0.014) 0.0001*** (0.00004) (ii) 0.005*** (0.001) (iii) 0.005*** (0.001) 0.004*** (0.001) 0.034** (0.014) (iv) 0.006*** (0.001) 0.004*** (0.001) 0.073*** (0.015) 0.0001*** (0.0000)

0.057*** (0.014) 0.0001*** (0.0000) 0.003*** (0.001)

0.005 (0.008) 0.0003*** (0.0001)

1.007*** (0.139) 2906.21 128.75*** 4837

x 2 (4)
No. of Obs.

***, ** and * denote statistical significance at the 1%, 5% and 10% level, respectively. Age = firm age, K/L = capital intensity, OC = ownership concentration rate, Prod. = labor productivity (=VA/L), R&D = research and development, Sales = total sales, Wage = wage rate. Note: Standard errors are in parentheses. Source: Authors estimates.

8 | ADB Economics Working Paper Series No. 295

7KH FRHIFLHQW HVWLPDWHV RI RZQHUVKLS FRQFHQWUDWLRQ UDWH ZDJH UDWH FDSLWDO LQWHQVLW\ DQG 5 ' VWRFN DUH DOO SRVLWLYH DQG VLJQLFDQW LQ 0RGHO L (VWLPDWLRQ UHVXOWV VKRZ WKDW D UPV SUREDELOLW\ RI HQWHULQJ WKH IRUHLJQ PDUNHW LQFUHDVHV ZLWK RZQHUVKLS FRQFHQWUDWLRQ UDWH ZDJHV UDWH FDSLWDO LQWHQVLW\ DQG 5 ' VWRFN 7KLV LPSOLHV WKDW H[SRUWLQJ UPV SD\ higher wages and have greater capital intensity and R&D stock. Model (ii) substitutes wage rate with labor productivity as explanatory variable. These two variables are closely correlated.3 7KH FRHIFLHQW HVWLPDWHV RI DOO WKH H[SODQDWRU\ YDULDEOHV UHPDLQ SRVLWLYH DQG VLJQLFDQW 7KLV LV VWLOO WKH FDVH ZKHQ 5 ' VWRFN LV UHSODFHG ZLWK VDOHV LQ 0RGHO LLL4 0RGHO LY DGGHG ERWK UP DJH DQG UP DJH VTXDUHG DV DGGLWLRQDO H[SODQDWRU\ YDULDEOHV WR 0RGHO L 7KH FRHIFLHQW HVWLPDWHV RI WKH EDVLF H[SODQDWRU\ YDULDEOHV UHPDLQ WKH VDPH +RZHYHU WKH FRHIFLHQW RI DJH LV SRVLWLYHO\ LQVLJQLFDQW DQG LWV VTXDUHG WHUP LV QHJDWLYHO\ VLJQLFDQW5 This suggests that the impact of business experience on entering H[SRUW PDUNHW GLPLQLVKHV DV UPV JURZ ROGHU HYHQ WKRXJK LW PLJKW LQLWLDOO\ KDYH D SRVLWLYH LQXHQFH 0RVW VLJQLFDQWO\ RXU HVWLPDWLRQ UHVXOWV VKRZ WKDW KLJKHU RZQHUVKLS FRQFHQWUDWLRQ UDWH LQFUHDVHV WKH SUREDELOLW\ RI UPV HQWHULQJ IRUHLJQ PDUNHWV 7KH UHVXOWV WKXV VXSSRUW RXU FHQWUDO K\SRWKHVLV WKDW UPV ZLWK FRQFHQWUDWHG RZQHUVKLS YHQWXUH LQWR ULVN\ H[SRUW PDUNHWV WR PD[LPL]H H[SHFWHG SURWV ZKHUHDV UPV ZLWK GLVSHUVHG RZQHUVKLS WHQG WR stay home to avoid the risk of incurring the large sunk costs associated with exporting. 0DQDJHUV LQ UPV ZLWK GLVSHUVHG RZQHUVKLS SUHIHU WR DYRLG ULVN DQG DFKLHYH FRQFUHWH business outcomes in a short period. This discourages them from entering risky foreign PDUNHWV ZKLFK UHTXLUH ODUJH VXQN FRVWV WKDW FDQ RQO\ EH UHFRYHUHG LQ WKH ORQJ UXQ 2XU HPSLULFDO UHVXOWV FRQUP WKH H[LVWHQFH RI DQ DJHQF\ SUREOHPULVNDYHUVH PDQDJHUV RI UPV ZLWK GLIIXVH RZQHUVKLS GR QRW H[SRUW HYHQ WKRXJK H[SRUWLQJ LV SURWDEOH DQG WKXV EHQHFLDO IRU VKDUHKROGHUV :LWK UHVSHFW WR RXU FRQWURO YDULDEOHV RXU UHVXOWV VKRZ WKDW UPV DUH PRUH OLNHO\ WR EH exporters as wage rate, capital intensity, R&D stock, productivity, and sales increase. Our UHVXOWV FRQUP WKDW ZLWKLQ WKH .RUHDQ PDQXIDFWXULQJ VHFWRU H[SRUWHUV DUH ODUJHU PRUH productive, more capital- and R&D-intensive, and pay higher wages than nonexporters. 2XU HYLGHQFH LV FRQVLVWHQW ZLWK D ODUJH ERG\ RI HPSLULFDO OLWHUDWXUH WKDW QG VLPLODU differences between exporters and nonexporters (see Bernard and Jensen 1997a and E Aitken, Hanson, and Harrison  $Z DQG +ZDQJ  &OHULGHV /DFK DQG 7\ERXW  DQG 5REHUWV DQG 7\ERXW  6XFK HYLGHQFH LV LQWXLWLYHO\ SODXVLEOH VLQFH PRUH HIFLHQW DQG ODUJHU UPV ZLWK DGHTXDWH UHVRXUFHV DUH EHWWHU DEOH WKH ODUJH VXQN FRVWV UHTXLUHG WR HQWHU IRUHLJQ PDUNHWV

Coefficient estimates of these two variables become insignificant due to multicollinearity when they are both included. 4 Multicollinearity causes coefficient estimates of R&D to be insignificant when these two variables are both included. 5 Coefficient estimates of wage rate, labor productivity, and R&D stock become insignificant due to multicollinearity when labor productivity and sales are added to Model (iv).

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 9

6LQFH WKH SUREDELOLW\ RI D UPV EHLQJ DQ H[SRUWHU PLJKW GLIIHU DFURVV LQGXVWULHV ZH DOVR UHUXQ 0RGHO LY LQ 7DEOH  IRU UPV LQ GLIIHUHQW LQGXVWULHV 7DEOH  UHSRUWV WKH UHVXOWV 7KH FRHIFLHQW HVWLPDWHV RI FRQFHQWUDWLRQ UDWH DUH SRVLWLYHO\ VLJQLFDQW LQ WKH WH[WLOHV SDSHU EDVLF PHWDO DQG IDEULFDWLRQ LQGXVWULHV EXW QHJDWLYHO\ VLJQLFDQW LQ WKH FKHPLFDO LQGXVWU\ DQG LQVLJQLFDQW LQ WKH RWKHU LQGXVWULHV $Q LQFUHDVH LQ FRQFHQWUDWLRQ UDWH LQFUHDVHV WKH probability of entering the export market in more industries (four industries out of seven) than not. However, the correlation is negative in the chemical industry, which consists of ODUJH UPV SURGXFLQJ VWDQGDUGL]HG UDZ PDWHULDOV DQG SURGXFWV IRU WKH 3HRSOHV 5HSXEOLF RI &KLQD PDUNHW 7KH RZQHUVKLS RI WKHVH UPV LV GLVSHUVHG VLQFH WKLV LQGXVWU\ LV PDWXUH enough to attract a large number of investors.6 Moreover, entering the Chinese market PLJKW QRW SRVH PXFK ULVN WR WKH UPV LQ WKLV LQGXVWU\ GXH WR ERWK SURGXFW VWDQGDUGL]DWLRQ and robust structural demand. Table 4: Logit Regression for Ownership Concentration and Exports of Korean Manufacturing Firms by Industry: Dependent Variable (exporters/nonexporters)
Variable OC Wage K/L R&D Age Age2 Constant LLR Industry Food 0.005 (0.009) 0.033* (0.008) 0.347** (0.162) 0.0009 (0.0006) 0.088* (0.033) 0.0006** (0.0003) 4.157* (0.908) 153.64 27.07* 399 Textiles 0.013** (0.006) 0.040*** (0.022) 0.317** (0.127) -0.008 (0.013) 0.262* (0.049) 0.002* (0.0005) 7.265* (1.064) 202.41 69.20* 412 Paper 0.020** (0.008) 0.014 (0.015) 0.282* (0.077) 0.017** (0.008) 0.058 (0.074) -0.001 (0.001) 3.285* (1.209) 117.53 38.25* 224 Chemical 0.008** (0.003) 0.005 (0.003) 0.218* (0.033) 0.001* (0.0003) 0.011 (0.016) 0.0002 (0.0002) 0.542*** (0.284) 699.52 137.37* 1281 Nonmetal 0.012 (0.010) 0.013 (0.013) 0.102* (0.039) 0.001 (0.001) 0.084 (0.140) 0.001 (0.001) 0.862 (2.466) 112.80 15.85** 274 Basic Metal 0.023* (0.005) 0.012 (0.008) -0.031 (0.041) 0.0002 (0.0002) 0.052 (0.039) 0.0006 (0.0006) 2.131* (0.624) 286.07 32.43* 446 Fabrication 0.018* (0.002) 0.002 (0.002) 0.167* (0.060) 0.0001* (0.0000) 0.035** (0.016) 0.001* (0.0002) 1.001* (0.240) 1051.73 105.49* 1706

x 2 (8)
Obs.

***, ** and * denote statistical significance at the 1%, 5% and 10% level, respectively. Age = firm age, K/L = capital intensity, OC = ownership concentration rate, R&D = research and development, Wage = wage rate. Note: Standard errors are in parentheses. For others, see notes to Table 3. Source: Authors estimates.

7KH FRHIFLHQW HVWLPDWH RI WKH ZDJH UDWH LV SRVLWLYH DQG VLJQLFDQW LQ WKH IRRG DQG WH[WLOHV LQGXVWULHV &DSLWDO LQWHQVLW\ LV SRVLWLYH DQG VLJQLFDQW LQ WKH SDSHU FKHPLFDO DQG QRQPHWDO LQGXVWULHV EXW QHJDWLYH DQG VLJQLFDQW LQ WKH IRRG WH[WLOHV DQG IDEULFDWLRQ LQGXVWULHV 5 ' VWRFN LV SRVLWLYH DQG VLJQLFDQW LQ WKH SDSHU FKHPLFDO DQG IDEULFDWLRQ PHWDO LQGXVWULHV $JH LV SRVLWLYH DQG VLJQLFDQW LQ WKH IRRG WH[WLOHV DQG IDEULFDWLRQ LQGXVWULHV DQG DJH VTXDUHG LV QHJDWLYH DQG VLJQLFDQW LQ WKH VDPH LQGXVWULHV 2XU UHVXOWV
6

The foundation of Korean chemical industry was built up in the early 1970s.

10 | ADB Economics Working Paper Series No. 295

FRQUP WKDW WKH LQXHQFH RI WKH H[SODQDWRU\ YDULDEOHV GLIIHUV DFURVV LQGXVWU\ 5HODWLYHO\ ODERULQWHQVLYH UPV DUH PRUH OLNHO\ WR EH H[SRUWHUV LQ WKH IRRG WH[WLOHV DQG IDEULFDWLRQ LQGXVWULHV ZKLOH UHODWLYHO\ FDSLWDOLQWHQVLYH UPV DUH PRUH OLNHO\ LQ WKH SDSHU FKHPLFDO DQG QRQPHWDO LQGXVWULHV 7KH UHVXOWV VKRZ WKDW H[SRUWLQJ UPV SD\ VLJQLFDQWO\ KLJKHU wages only in the food and textiles industries. Our results suggest a nonlinear (initially SRVLWLYH EXW VXEVHTXHQWO\ QHJDWLYH LPSDFW RI UP DJH LQ WKH IRRG WH[WLOHV DQG IDEULFDWLRQ LQGXVWULHV 7KH VLJQV DQG VLJQLFDQFH RI WKH FRHIFLHQW HVWLPDWHV UHPDLQ WKH VDPH LQ every industry when we substitute labor productivity for wage rate.7

B.

Tobit Estimation of Export Propensity

The previous section treated exporting as a zero-one binary variable by dividing the VDPSOH UPV LQWR H[SRUWHUV DQG QRQH[SRUWHUV ,I WKHUH DUH PDQ\ GLIIHUHQW H[SRUW PDUNHWV WKDW UHTXLUH VHSDUDWH [HG FRVWV WR HQWHU ULVNDYHUVH UPV DUH OLNHO\ WR H[SRUW WR D VPDOOHU QXPEHU RI PDUNHWV VLQFH HQWHULQJ PRUH PDUNHWV HQWDLOV D ODUJHU WRWDO [HG FRVW $V GLVFXVVHG HDUOLHU D ORW RI WKH [HG FRVWV DUH VXQN FRVWV :H QRZ WUHDW D UPV H[SRUWV DV D FRQWLQXRXV YDULDEOH UDWKHU WKDQ D ELQDU\ YDULDEOH 0RUH SUHFLVHO\ ZH PHDVXUH D UPV export performance as export propensity, or the ratio of export revenues to total sales. 7DEOH  UHSUHVHQWV WKH FRHIFLHQW HVWLPDWHV RI WKH WRELW HVWLPDWLRQ RI H[SRUW SURSHQVLW\ IRU .RUHDQ PDQXIDFWXULQJ UPV ([SRUW SURSHQVLW\ GHQHG RQ > @ LV WKH GHSHQGHQW YDULDEOH ZKHUH  VWDQGV IRU QRQH[SRUWHUV DQG  VWDQGV IRU UPV H[SRUWLQJ WKHLU HQWLUH RXWSXW $SSO\LQJ RUGLQDU\ OHDVW VTXDUHV HVWLPDWLRQ PHWKRG WR WKLV FHQVRUHG GDWD ZLOO FDXVH FRHIFLHQW HVWLPDWHV WR EH ELDVHG VLQFH WKLV PHWKRG FDQ JHQHUDWH SUHGLFWHG YDOXHV of the dependent variable that lie outside the feasible range. To deal with the problem, we use a tobit model censored at both right and left ends. As for the logit estimation, our key explanatory variable of interest is the ownership concentration rate. In addition, we include explanatory variables widely used in the trade literature such as wage rate, capital intensity, and R&D stock. In estimation, a semi-log model is utilized, which transforms DOO H[SODQDWRU\ YDULDEOHV LQWR ORJDULWKPV WR FRQWURO IRU KHWHURVFHGDVWLFLW\ DULVLQJ IURP UP scale. However, we use the dependent variable in its original form to keep its censored characteristics.8 We estimate four models, which represent various permutations of the explanatory variables.

The results we obtain when we substitute labor productivity for wage are available upon request. Labor productivity has a positive and significant effect on a firms decision to export. 8 However, age and its square term is used in their original form because the logs of these variables are perfectly correlated.

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 11

Table 5:Tobit Regression for Ownership Concentration and Exports of Korean Manufacturing Firms: Dependent Variable (exports/sales)
Variable log(OC) log(Wage) log(K/L) log(R&D) log(Prod) Age Age2 log(L) Constant LLR 1.455*** (0.174) 2471.92 117.90*** 1.964*** (0.200) 2473.02 115.72*** 1.620*** (0.180) 2418.36 225.02*** Model (i) 0.083*** (0.018) 0.078*** (0.011) 0.021* (0.012) 0.009** (0.003) (ii) 0.086*** (0.018) (iii) 0.076*** (0.018) 0.096*** (0.011) 0.048*** (0.012) 0.006* (0.003) (iv) 0.079*** (0.018) 0.147 (0.097) 0.005 (0.013) 0.002 (0.003) 0.097*** (0.012) 0.0008 (0.003) 0.0001*** (0.0000) 0.022 (0.097) 2.574*** (0.217) 2377.50 306.75*** 3947 (2731, 32)

0.011 (0.013) 0.021*** (0.003) 0.071*** (0.010)

0.001 (0.003) 0.0001*** (0.0000)

x 2 (4)
No. of total Obs. (left, right censored)

. ***, ** and * denote statistical significance at the 1%, 5% and 10% level, respectively. Age = firm age, K/L = capital intensity, OC = ownership concentration rate, Prod. = labor productivity (=VA/L), R&D = research and development, Wage = wage rate. Notes: Standard errors are in parentheses. For others, see notes to Table 3. Source: Authors estimates.

7KH FRHIFLHQW HVWLPDWHV RI RZQHUVKLS FRQFHQWUDWLRQ UDWH FDSLWDO LQWHQVLW\ DQG 5 ' VWRFN DUH DOO SRVLWLYHO\ VLJQLFDQW EXW WKDW RI ZDJH UDWH LV QHJDWLYHO\ VLJQLFDQW LQ 0RGHO L (VWLPDWLRQ UHVXOWV VKRZ WKDW UPV H[SRUW SURSHQVLW\ ULVHV ZLWK WKH RZQHUVKLS concentration rate, capital intensity, and R&D stock, but falls with the wage rate. Model (ii) substitutes wage rate with labor productivity as explanatory variable. These two variables are closely correlated.9 7KH FRHIFLHQW HVWLPDWHV RI FRQFHQWUDWLRQ UDWH 5 ' VWRFN DQG ODERU SURGXFWLYLW\ DUH SRVLWLYH DQG VLJQLFDQW EXW WKDW RI FDSLWDO LQWHQVLW\ EHFRPH LQVLJQLFDQW :KHQ ERWK DJH DQG VTXDUHG WHUP RI DJH DUH DGGHG DV H[SODQDWRU\ YDULDEOHV LQ 0RGHO LLL WKH FRHIFLHQW HVWLPDWHV RI RULJLQDO H[SODQDWRU\ YDULDEOHV DUH SRVLWLYHO\ VLJQLFDQW EXW WKDW RI DJH LV SRVLWLYHO\ LQVLJQLFDQW DQG LWV VTXDUH WHUP QHJDWLYHO\ VLJQLFDQW :KHQ ODERU SURGXFWLYLW\ DQG HPSOR\PHQW DUH DGGHG LQ 0RGHO LY WKH cRHIFLHQW HVWLPDWHV RI ZDJH UDWH FDSLWDO LQWHQVLW\ DQG 5 ' VWRFN EHFRPH LQVLJQLFDQW due to multicollinearity.
9

Coefficient estimates of these two variables become both insignificant due to multicollinearity when both are included.

12 | ADB Economics Working Paper Series No. 295

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hey may prefer instead to concentrate on fast-growing export markets that they are more familiar with due to geographical proximity. This type of export strategy is less risky and involves lower sunk costs than exporting to many different countries. For example, the Peoples Republic of China is a highly promising market for a large number RI .RUHDQ UPV UHJDUGOHVV RI WKHLU RZQHUVKLS VWUXFWXUH 2Q WKH RWKHU KDQG RXU UHVXOWV LQGLFDWH WKDW UPV ZLWK PRUH FRQFHQWUDWHG RZQHUVKLS DUH PRUH OLNHO\ WR EHDU WKH KLJKHU risk associated with exporting to slower growing and less familiar markets. Our evidence VXSSRUWV WKH H[LVWHQFH RI DJHQF\ SUREOHP LQ D UPV GHFLVLRQ PDNLQJ DERXW WKH QXPEHU RI H[SRUW PDUNHWV 5LVN DYHUVLRQ GHWHUV WKH PDQDJHUV RI UPV ZLWK GLVSHUVHG RZQHUVKLS IURP H[SRUWLQJ WR PRUH PDUNHWV HYHQ WKRXJK GRLQJ VR PD\ UDLVH SURWV DQG WKXV EHQHW shareholders. With respect to our control variables, much of our evidence mirrors the results of our logit DQDO\VLV 0RUH VSHFLFDOO\ RXU WRELW UHVXOWV LQGLFDWH WKDW UPV ZLWK KLJKHU FDSLWDO LQWHQVLW\ R&D stock, and productivity export to more markets. One departure from the logit results LV WKDW UPV SD\LQJ ORZHU ZDJHV DUH OLNHO\ WR HQWHU PRUH H[SRUW PDUNHWV WKDQ UPV SD\LQJ higher wages. Our tobit results are based on comparing the wage rates of exporters with those of other exporters while our logit results are based on comparing the wage rates of exporters and nonexporters. Our tobit results suggest that exporters paying lower wages export to more countries than exporters paying higher wages, whereas our logit results suggests that exporters pay higher wages than nonexporters. 6LQFH UPV H[SRUW SURSHQVLW\ PLJKW GLIIHU DFURVV LQGXVWULHV ZH DOVR UHUXQ 0RGHO LLL LQ 7DEOH  IRU UPV LQ GLIIHUHQW LQGXVWULHV 7DEOH  UHSRUWV WKH UHVXOWV &RHIFLHQW HVWLPDWHV RI FRQFHQWUDWLRQ UDWH DUH SRVLWLYHO\ VLJQLFDQW LQ WKH WH[WLOHV EDVLF PHWDO DQG IDEULFDWLRQ LQGXVWULHV EXW QHJDWLYHO\ VLJQLFDQW LQ WKH FKHPLFDO LQGXVWU\ DQG LQVLJQLFDQW LQ WKH other industries. Empirical results show that increase in concentration rate increases the probability of entering more export markets in three industries out of seven, but decreases the probability in the chemical industry. The results are very similar with industry-level estimation results from logit analysis, except in the paper industry, where WKH UHODWLRQVKLS WXUQV LQVLJQLFDQW 2XU UHVXOWV DUH WKXV IDLUO\ UREXVW DFURVV GHSHQGHQW variable and estimation models. 7KH FRHIFLHQW HVWLPDWHV RI ZDJH UDWH DUH SRVLWLYH DQG VLJQLFDQW LQ WKH WH[WLOHV DQG QRQPHWDO LQGXVWULHV EXW QHJDWLYH DQG VLJQLFDQW LQ WKH SDSHU EDVLF PHWDO DQG IDEULFDWLRQ LQGXVWULHV &DSLWDO LQWHQVLW\ LV SRVLWLYH DQG VLJQLFDQW LQ WKH SDSHU FKHPLFDO DQG QRQPHWDO LQGXVWULHV EXW QHJDWLYH DQG VLJQLFDQW LQ WKH WH[WLOHV EDVLF PHWDO DQG IDEULFDWLRQ

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 13

LQGXVWULHV 5 ' VWRFN LV SRVLWLYH DQG VLJQLFDQW LQ WKH IRRG DQG QRQPHWDO LQGXVWULHV EXW QHJDWLYH DQG VLJQLFDQW LQ WKH WH[WLOHV LQGXVWU\ Table 6: Tobit Regression for Ownership Concentration and Exports of Korean Manufacturing Firms by Industry: Dependent Variable (exports/sales)
Variable log(OC) log(Wage) log(K/L) log(R&D) Age Age2 Constant LLR Industry Food 0.017 (0.031) 0.006 (0.016) 0.018 (0.024) 0.021** (0.009) 0.005*** (0.003) 0.00005*** (0.00003) 0.842** (0.333) 80.14 18.09* 364 (307, 0) Textiles 0.253* (0.096) 0.245* (0.078) 0.284* (0.086) 0.022*** (0.013) 0.111* (0.022) 0.0009* (0.0002) 1.122 (1.197) 102.94 70.84* 228 (166, 0) Paper 0.094 (0.059) 0.302* (0.082) 0.152** (0.059) 0.002 (0.009) 0.042 (0.026) 0.0004 (0.0003) 3.429* (0.883) 60.31 29.11* 166 (124, 0) Chemical 0.086* (0.030) 0.025 (0.025) 0.209* (0.023) 0.009 (0.008) 0.0002 (0.006) 0.0001** (0.0000) 2.533* (0.336) 608.66 157.11* 1125 (807, 1) Nonmetal Basic Metal Fabrication 0.025 0.330* 0.182* (0.065) (0.047) (0.032) 0.157** 0.126* 0.155* (0.069) (0.026) (0.019) 0.090*** 0.093* 0.054** (0.053) (0.030) (0.028) 0.073** 0.007 0.004 (0.032) (0.005) (0.008) 0.071** 0.002 0.014** (0.032) (0.010) (0.006) 0.0009** 0.0000 0.0004* (0.0004) (0.0001) (0.0001) 0.186 0.994** 1.279* (0.756) (0.400) (0.349) 83.95 168.37 961.14 12.79** 239 (199, 2) 63.77* 332 (181. 0) 187.83* 1415 (904, 28)

x 2 (8)
No. of Obs. (left, right censored)

***, ** and * denote statistical significance at the 1%, 5% and 10% level, respectively. Age = firm age, K/L = capital intensity, OC = ownership concentration rate, R&D = research and development, Wage = wage rate. Note: Standard errors are in parentheses. For others, see notes to Table 3. Source: Authors estimates.

(VWLPDWLRQ UHVXOWV VKRZ UPVSHFLF IDFWRUV KDYH D GLIIHUHQW LQXHQFH RQ H[SRUW DFWLYLW\ DFURVV GLIIHUHQW LQGXVWULHV 5HODWLYHO\ OHVV FDSLWDOLQWHQVLYH UPV SD\LQJ ORZHU ZDJHV KDYH higher export propensity in the basic metal and fabrication industries, while the reverse KROGV LQ WKH QRQPHWDO LQGXVWU\ /DERULQWHQVLYH UPV KDYH KLJKHU H[SRUW SURSHQVLW\ LQ WKH EDVLF PHWDO DQG IDEULFDWLRQ LQGXVWULHV EXW FDSLWDOLQWHQVLYH UPV GR VR LQ WKH QRQPHWDO industry. As wage rate rises and capital intensity falls, export propensity rises in the WH[WLOHV LQGXVWU\ EXW IDOOV LQ WKH SDSHU LQGXVWU\ 7KHVH UHVXOWV LPSO\ WKDW UPV SURGXFLQJ human capital-intensive clothes and materials perform better in the textiles industry. The FRPSHWLWLYH DGYDQWDJH RI .RUHDQ UPV LQ WKLV PDWXUH LQGXVWU\ PLJKW WKXV OLH LQ GHVLJQHU clothing, high-end functional clothes, and differentiated textile materials. These products DUH FXVWRPPDGH DQG SURGXFHG LQ VPDOO TXDQWLWLHV +RZHYHU UPV PDVV SURGXFLQJ standardized products in large factories tend to export more in the paper industry, where competitive advantage is based mainly on production costs. Furthermore, the scope for product differentiation is limited in the paper industry.

14 | ADB Economics Working Paper Series No. 295

&RHIFLHQW HVWLPDWHV RI UP DJH DUH SRVLWLYH DQG VLJQLFDQW LQ WKH IRRG WH[WLOHV DQG IDEULFDWLRQ LQGXVWULHV EXW QHJDWLYH DQG VLJQLFDQW LQ WKH QRQPHWDO LQGXVWU\ $JH VTXDUHG LV QHJDWLYH DQG VLJQLFDQW LQ WKH IRRG WH[WLOHV FKHPLFDO DQG IDEULFDWLRQ LQGXVWULHV 7KHVH results imply that the positive impact of experience on export performance generally GHFOLQHV DV UPV JHW ROGHU 6LJQ DQG VLJQLFDQFH RI FRHIFLHQW HVWLPDWHV UHPDLQV PRVWO\ unchanged when we substitute labor productivity for wage rate.10

V. Concluding Observations
The central objective of our study was to empirically examine the relationship between WKH RZQHUVKLS VWUXFWXUH RI UPV DQG WKHLU H[SRUW SHUIRUPDQFH 0RUH VSHFLFDOO\ XVLQJ data from the Korean manufacturing sector, we investigate the relationship between ownership concentration and export performance. In contrast to the large and growing empirical literature that delves into the relationship between the ownership structure of UPV DQG WKHLU RYHUDOO SHUIRUPDQFH WKHUH DUH DOPRVW QR VWXGLHV WKDW H[SORUH WKH LPSDFW RI UPV RZQHUVKLS FRQFHQWUDWLRQ RQ WKHLU H[SRUW SHUIRUPDQFH 7KH IHZ HPSLULFDO VWXGLHV WKDW WRXFK XSRQ WKH RZQHUVKLSH[SRUWV QH[XV ORRN DW WKH UHODWLYH H[SRUW SHUIRUPDQFH RI IRUHLJQRZQHG UPV YHUVXV GRPHVWLFRZQHG UPV 7KHUHIRUH WKH SULPDU\ FRQWULEXWLRQ RI RXU VWXG\ LV WR KHOS UHPHG\ WKLV VHULRXV VKRUWFRPLQJ RI WKH OLWHUDWXUH RQ WKH EDVLV RI UP level evidence from the Republic of Korea. :H HVWLPDWH WZR W\SHV RI HPSLULFDO PRGHOVORJLW PRGHOV DQG FHQVRUHG WRELW PRGHOV to examine the relationship between ownership concentration and export performance. 2XU PRVW VLJQLFDQW QGLQJ LV WKDW UPV ZLWK FRQFHQWUDWHG RZQHUVKLS DUH OLNHO\ WR HQMR\ VWURQJHU H[SRUW SHUIRUPDQFH WKDQ UPV ZLWK GLIIXVH RZQHUVKLS 7KH SULPDU\ LPSOLFDWLRQ of our empirical results for policy makers is that the positive effect of concentrated ownership on exports is an important additional factor that must be factored into SROLFLHV LQXHQFLQJ WKH RZQHUVKLS FRQFHQWUDWLRQ RI UPV DQG PRUH JHQHUDOO\ FRUSRUDWH governance. Our evidence lends support to the conventional wisdom that originally IDPLO\RZQHG UPV ZLWK KLJKO\ FRQFHQWUDWHG RZQHUVKLS NQRZQ DV chaebols, made a big contribution to the Republic of Koreas export success. Given the importance of exports in the Korean economy, our analysis suggests a need to exercise caution in policies that seek to regulate and control the chaebols.11 While our study empirically investigates an important but previously underexplored relationship between ownership structure and export performance, it is by no means the
10

The results we obtain when we substitute labor productivity for wage are available from authors upon request. R&D stock has positive significant impact on firms exports only in the chemical, basic metal, and fabrication industries. Labor productivity has a positive significant effect in every industry except the paper and nonmetal industries. 11 The chaebols have been blamed for contributing to the Republic of Koreas collapse during the Asian financial crisis of 19971998 through excessive investment and expansion. More recently, they are widely blamed for the growing concentration of economic power and the consequent lack of a dynamic small and medium enterprises sector.

Ownership Structure and Export Performance: Firm-Level Evidence from the Republic of Korea | 15

GHQLWLYH QDO ZRUG ,Q IDFW RXU VWXG\ PDUNV D UVW VWHS WRZDUG EHWWHU XQGHUVWDQGLQJ WKH RZQHUVKLSH[SRUW QH[XV WKDW ZLOO KRSHIXOO\ HQFRXUDJH RWKHU UHVHDUFKHUV WR GHOYH LQWR the issue. There are several promising directions for future research. Perhaps the most promising research area is to examine the relationship between ownership structure and export performance in other successful export-led East Asian economies. It would also be LQWHUHVWLQJ WR WDNH D ORRN DW WKH RZQHUVKLSH[SRUWV UHODWLRQVKLS LQ PRUH PDWXUH DGYDQFHG economies such as the US, which have different corporate governance environments. Finally, another potential extension of our research would be to look at the relationship between ownership structure and foreign direct investment.

References
Aitken, B., G. H. Hanson, and A. E. Harrison. 1997. Spillovers, Foreign Investment, and Exports Behavior. Journal of International Economics  Aw, B., and A. R. Hwang. 1995. Productivity and the Export Market: A Firm-Level Analysis. Journal of Development Economics  Berle, A., and G. Means. 1932. The Modern Corporation and Private Property. New York: MacMillan. Bernard, A. B., and J. B. Jensen. 1997a. Exporters, Skill-Upgrading, and the Wage Gap. Journal of International Economics   E ([SRUWHUV DQG 6XFFHVV LQ *HUPDQ 0DQXIDFWXULQJ Weltwirtschaftliches Archiv  %OHDQH\ 0 DQG . :DNHOLQ  6HFWRUDO DQG )LUP6SHFLF 'HWHUPLQDQWV RI ([SRUW Performance: Evidence from the United Kingdom. Research Paper 99/12, Center for Research on Globalization and Labour Markets, University of Nottingham. Clerides, S. K., S. Lach, and J. R. Tybout. 1998. Is Learning by Exporting Important? Microdynamic Evidence from Colombia, Mexico and Morocco. Quarterly Journal of Economics  Cole, M., R. Elliott, and S. Virakul. 2010. Firm Heterogeneity, Origin of Ownership and Export Participation. World Economy  Demsetz, H., and K. Lehn. 1985. The Structure of Corporate Ownership: Causes and &RQVHTXHQFHV Journal of Political Economy  Filatotchev, I., J. Stephan, and B. Jindra. 2008. Ownership Structure, Strategic Controls and Export Intensity of Foreign-Invested Firms in Transition Economies. Journal of International Business Studies  Griliches, Z., and J. Hausman. 1986. Errors in Variables in Panel Data. Journal of Econometrics  Hermalin, B., and M. Weisbach. 1991. The Effects of Board Compensation and Direct Incentives on Firm Performance. Financial Management  Himmelberg, C., R. Hubbard, and D. Palia. 1999. Understanding the Determinants of Managerial Ownership and the Link between Ownership and Performance. Journal of Financial Economics  Hirsch, S., and I. Bijaoui. 1985. R&D Intensity and Export Performance: A Micro View. Weltwirtschaftliches Archiv 

16 | ADB Economics Working Paper Series No. 295

-RK 6 :  &RUSRUDWH *RYHUQDQFH DQG )LUP 3URWDELOLW\ Journal of Financial Economics  Krugman, P. 1979. Increasing Returns, Monopolistic Competition, and International Trade. Journal of International Economics  Kumar, N., and N. S. Siddharthan. 1994. Technology, Firm Size and Export Behavior in Developing Countries: The Case of Indian Enterprise. Journal of Development Studies  0F&RQQHO - - DQG + 6HUYDHV  $GGLWLRQDO (YLGHQFH RQ (TXLW\ 2ZQHUVKLS DQG &RUSRUDWH Value. Journal of Financial Economics  Morck, R., A. Shleifer, and R. W. Vishny. 1988. Management Ownership and Market Valuation: An Empirical Analysis. Journal of Financial Economics  Ngoc, P., and E. Ramsetter. 2009. Foreign Ownership and Exports in Vietnamese Manufacturing. Singapore Economic Review  Posner, M. V. 1961. International Trade and Technical Change. Oxford Economic Papers  41. Rasiah, R. 2003. Foreign Ownership, Technology and Electronics Exports from Malaysia and Thailand. Journal of Asian Economics    )RUHLJQ 2ZQHUVKLS 7HFKQRORJLFDO ,QWHQVLW\ DQG ([SRUW ,QFLGHQFH $ 6WXG\ RI $XWR Parts, Electronics and Garment Firms in Indonesia. International Journal of Technology and Globalisation    ([SRUW 2ULHQWDWLRQ DQG 7HFKQRORJLFDO ,QWHQVLWLHV LQ $XWR 3DUW )LUPV LQ (DVW DQG Southeast Asia: Does Ownership Matter?Asian Economic Papers  Richardson, J. D., and K. Rindal. 1995. Why Exports Really Matter! The Institute for International Economics and the Manufacturing Institute, Washington, DC. Roberts, M. J., and J. R. Tybout. 1997. The Decision to Export in Colombia: An Empirical Model of Entry with Sunk Costs. American Economic Review  Rojec, M., J. Damijan, and B. Majcen. 2004. Export Propensity of Estonian and Slovenian Manufacturing Firms: Does Foreign Ownership Matter? Eastern European Economics  Vernon, R. 1966. International Investment and International Trade in the Product Cycle. Quarterly Journal of Economics  Wakelin, K. 1998. Innovation and Export Behavior at the Firm Level. Research Policy  Wignaraja, G. 2008. Foreign Ownership, Technological Capabilities and Clothing Exports in Sri Lanka. Journal of Asian Economics 

About the Paper Sangho Kim and Donghyun Park empirically examine the relationship between the ownership structure of firms and their export performance. To do so, they use firm-level data from the Republic of Korea, a classical example of successful exportoriented industrialization. Their empirical results indicate that Korean firms with more concentrated ownership are more likely to be exporters and export more.

About the Asian Development Bank ADBs vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries reduce poverty and improve the quality of life of their people. Despite the regions many successes, it remains home to two-thirds of the worlds poor: 1.8 billion people who live on less than $2 a day, with 903 million struggling on less than $1.25 a day. ADB is committed to reducing poverty through inclusive economic growth, environmentally sustainable growth, and regional integration. Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance.

Asian Development Bank 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org/economics ISSN: 1655-5252 Publication Stock No. WPS124406
December 2011

< 0124 4069 >

Printed on recycled paper Printed in the Philippines