The Impact of Corruption and Transparency on Foreign Direct Investment: An Empirical Analysis (1)
Zhao, John Hongxin, Kim, Seung H., Du, Jianjun, Management International Review
* Both corruption and transparency are often perceived to damage the investment environment of a country. However, little research has been undertaken to empirically examine their impacts on foreign direct investments.
* This paper examines the influences of corruption and transparency on the level of foreign direct investment (FDI). Based on a cross-section data of 40 countries in 7 years, statistical results show that the presence of high corruption and low transparency significantly hindered the inflow of FDI to host countries.
* This study (1) contributes explicitly to our knowledge about how corruption and transparency undermine the FDI and (2) presents the practical implications and challenges facing international managers contemplating entering into foreign markets.
* Based on a cross-country data of 40 countries in 7 years, the results show that the presence of high corruption and low transparency significantly hindered the inflow of FDI to host countries.
Corruption as a cultural, political and economic phenomenon has attracted a lot of public attention around the world in recent years, while the issue of transparency has been viewed with little significance. Transparency often exists in more implicit forms within the purview of government administration, while corruption manifests itself in a variety of forms. Corruption can range from explicit conduct of paying "tips" and "speed money" to implicit ones of favor exchanges and "commissions". In fact, corruption has been so rampant that the Financial Times, in its year-end editorial on December 31, 1995, characterized 1995 as the year of corruption.
The importance of eradicating corruption to ensure a better business environment is widely recognized. In 1997, OECD countries reached a bribery ban agreement that applies to firms headquartered or located in member states (Wall Street Journal September 1997, Hershman 1998). The World Bank and International Monetary Fund threatened to suspend loans and aid to those countries that were reluctant to fight against corruption.
Corruption in general is condemned as a social evil. The phenomenon has led many scholars to search for an array of political, economic and social/cultural factors. Many studies typically attributed the occurrences of corruption to a number of factors including the degree of state intervention in an economy; the degree of governmental discretionary power and monopoly (Alam 1990); intermediation costs; and the wage level of civil servants (Mookherjee/Png 1995, Flatters/ McLeod 1995). However, the scholastic views on its economic impacts have been divergent. The majority of the studies centered on the overall economic impacts of corruption (Mauro 1995). Some argue that corruption converts perfect competition to monopoly and is detrimental to social morale and political stability. Others hold that there is a redeeming value in corruption.
In spite of the growing awareness of the issue, the affect of corruption on the flow of FDI remains primarily anecdotal. Moreover, the issue of transparency of host governments has been relatively ignored in both the public and academic sectors. For these reasons, the question as to how corruption and transparency affect FDI arises naturally. Further segmenting the issue requires consideration of impacts on inward versus outward FDI. For example, less developed countries (LDC)—particularly the emerging economies of East Asia—evidence high levels of corruption. However, until recently, Asian newly emerging economies seemed to experience growth in inflow of foreign direct investment. This leads to a multi-faceted analysis of the relationships between corruption, transparency and FDI.
This analysis uses a sample of 40 countries to establish the empirical evidence on the impacts of corruption and transparency on FDI. We do not question the destructive impacts of corruption on society as a whole, nor the fact that multinational enterprises (MNEs) in general repudiate and reject the practice of corruption. However, we take on a need for a close and systematic examination of the actual and specific effects corruption and transparency may have on FDI, accounting for other crucial factors.
This paper is organized as follows. The next section discusses the related literature. Section three presents the hypotheses to be tested. Data and test methods are elaborated in section four. The analytical results are reported in section five. The final section concludes the analysis and points to the potential research topics.
Review of Related Literature
Despite recent reports to the contrary, corruption remains an ubiquitous feature of the modern world economy. Theories that have contributed to the understanding and explanation of corruption include rent-seeking theory from economics (Besley/McLaren 1993, Murphy/Shleifer/Vishny 1993), public choice (Rose-Ackerman 1978), transaction cost (Husted 1994), institution and social cost (North 1990), property right (Davies 1977) and socio-cultural perspectives (Husted 1999). However, the opinions on how corruption affects the economic activities remain divided. Most scholarly works analyzing corruption tend to focus on its overall economic effects. The mainstream studies hold that corruption distorts the market: it gives rise to the inefficient allocation of resources which, then, hinders economic growth (Mauro 1995).
Economists view corruption as rent-seeking behavior that increases the transaction costs (Husted 1994, Besley/McLaren 1993). These costs may be spent for obtaining information about market conditions in any given foreign market; identifying appropriate partners, negotiating, drafting and enforcing contracts; financing the transaction and bearing the risks of default at subsequent stages and changing contracts as circumstances change. Other analyses showed that corruption led to income inequality (Gupta/Davoodi/Alonson-Terme 1998). Corruption also brings extra costs to host countries in the form of reduction of tax revenue and distorts the effects of industrial policy and productivity on investment (Ades/Di Tella 1999, Tanzi/Davoodi 1997).
Nonetheless, one stream of economic analysis took a functionalist view contending that corruption had positive effects on political and economic development, and found some redeeming value in corruption. Early studies by Leff (1964) and Huntington (1968) advanced the view that corruption could enhance efficiency by removing government-imposed rigidities that practically impede investment and obstruct other economic decisions favorable to economic growth. A number of later studies also shared the same view. Beck and Maher (1986) and Lien (1986) developed models showing that, in bidding competition, those who are most efficient can afford to offer the highest bribe. Therefore, bribes can promote efficiency by assigning projects to the most efficient firms. Comparing bribes and taxes, some argued that bribes could be the second-best choice. Based on a theoretical analysis of corruption in an overall competitive capitalist environment, Braguinsky concluded that "limited corruption (that is, corruption occurring up to some limit) actually is conducive to growth because it erodes monopoly and spreads innovation" (1996, p. 15). In another way, bribes supplement low wages, thus can allow the government to maintain a lower tax burden, which in turn facilitates growth (Tullock 1996).
While the study of transparency influences on economic activities is almost nonexistent, empirical studies of effects of corruption on FDI are also scarce. Even the four literature sources of empirical analyses yielded from a wide search render opposite findings. In an early study of FDI location selection by American firms, Wheeler and Mody (1992) found no significant correlation between FDI size and corruption incorporated into a risk factor of host countries. They concluded that the significance of the risk factor in FDI decisions should be "discounted." This finding was confirmed by a related study later (Hines 1995). Based on the analysis of FDI by American firms and the level of corruption in host countries since the Foreign Corrupt Practice Act (FCPA), Hines' study (1995) showed an overall insignificant effect of corruption on FDI but a significantly negative impact of corruption on FDI from United States after 1977.
Contrary to these previous studies, in a more recent study of bilateral FDI from fourteen source countries to forty-five host countries, Wei (1997a) found empirical support of the negative effect of corruption on inward FDI. Using a series of dummies, the study also suggests that investors from East Asian countries were no less sensitive to corruption than investors from other source countries and that the American investors were not necessarily more risk averse to host country corruption than other OECD investors. Following the same study, Wei (1997b) examined the effect of corruption on FDI using a new concept of corruption-induced uncertainty. In this new concept, corruption, unlike tax, is nontransparent and not pronounced. Because of this nature "corruption embeds arbitrariness and creates uncertainty" in the enforcement of agreements (Wei 1997 a, p. 1). The index of corruption-induced uncertainty was constructed based on an unpublished survey of individuals by the World Economic Forum. Wei's statistical analysis based on this index confirmed his previous study suggesting a significantly negative impact of corruption on FDI.
The proceeding review of literature shows that the majority of studies suggest an inverse relationship between corruption and in-bound foreign direct investment. Following the previous studies and our reasoning about transparency as a risk factor, we postulate that both high corruption and a low level of transparency are negatively associated with FDI, though the strength of these relationships may differ. Hence, we propose in detail the following two hypotheses.
Corruption and FDI
Corruption in this study is defined as the "sales by government officials of government property for personal gain" (Shleifer/Vishny 1993, p.599). While the existence of corruption can be attributed to the quality and size of public institutions (Kaufmann/Wei 1999), the lack of market competition (Ades/Tella 1999, Laffont/Guessan 1999), and cultural and institutional traditions (Treisman 2000), the causal relationship may point in the opposite direction. When the corruption in a country is high, the national investment tends to be depressed (Mauro 1995, Keefer/Knack 1995) and the growth of GDP is likely to be slow (Brunetti/Kinsunko/Weder 1997). To investors, the effects of corruption can be dynamic, varying over time according to the way corruption is organized. Corruption in the form of a monopoly where rent-seeking bureaucrats have a somewhat transparent bribe schedule is more like a tax, except that the bribes benefit the individual's rather than the government treasury (Shleifer/Vishny 1993). However, in the form of what Shleifer/Vishny called "the industrial organization of corruption," corruption carries more uncertainty to investors over total payments, when bureaucrats can demand bribes from investors independently and when the entry is free for bureaucrats to impose additional fees. The effects of corruption may also change over time. FDI and corruption may exhibit a u-shaped relationship over time, assuming that the anti-corruption efforts are institutionalized by countries and persistent. Initially the high level of corruption of a country deters firms from investing in that country. This downward curve of inward FDI can be averted eventually when the causes of corruption become weakened or start disappearing.
The current study chooses to focus on the static effect (average level) of corruption on FDI rather than on the dynamic effects, because the deficiency in the information and data prevents us from such a test. The consequences of corruption are approached from the perspective of the potential costs and uncertainties corruption carries. Specifically corruption negatively affects investors in a number of ways. First, from the investment perspective, corruption means high transaction costs. The high transaction costs due to corruption arise when extra payments have to be made in order to facilitate the business transactions. This extra cost that is eventually reflected in product costs will reduce the investors' price competitiveness. Another less tangible cost to investing firms is the damage to a firm's reputation when the firm is exposed to a corruption scandal. The ripple effect of this image damage would be ultimately felt in the negative perception of firm's brand and erosion of its competitive advantage as a "good citizen" of the host country. Finally, corruption is a risk. A country that exhibits a high presence of corruption may foreshadow a potential social unrest and political instability since ultimately corrupted governments result in economic and social polarizations in a society. Based on the preceding reasoning, we hypothesize the relationship between FDI and corruption to be tested as follows:
Hypothesis 1. Other factors being equal, corruption is likely to lead to reduced FDI. The higher the corruption level, the lower level of FDI inflow.
Transparency and FDI
Transparency refers to how well the government communicates its policies and regulations clearly to the public (IMD). In a country where the degree of transparency is high, the government must be open to the public, and the government official must be transparent in their decisions and actions and will not withhold information that is in the public interest. Holders of public office "should give reasons for their decisions and restrict information only when the wider public interest clearly demands." (1) Lacking prior empirical evidence on the linkage between transparency and FDI, we propose that the presence of low transparency in a host country would negatively influence the inward FDI. This statement is grounded on two related rationales. First is the cost-inducing nature of low transparency. The presumption underlying the negative impacts of low transparency on inward FDI is that, ceteris paribus, foreign investors would opt for a national market in which the regulatory regime is more open. An open regulatory regime is instrumental to the inducement of inward FDI and conducive to fair competition. In the case of a foreign market characterized with an inadequate transparency, the poor transparency implies high transaction cost potential, impeding foreign investment. The reason is that the low transparency increases the bureaucratic inefficiency that can be translated into potential hidden costs in the form of indefinite delay of issuing licenses and unexpected and frequent changes of investment rules and regulations. Given this cost-inducing nature of poor transparency, the low transparency affects the expected flow of return from a particular investment and thus a number of investing firms. The second is the legal implication of poor transparency on foreign investors. To the extent that the legal environment of countries affects the external financing (La Porta et al. 1997) and that FDI represents a source of external capital to a host country, for foreign investors who are unfamiliar with the host country's legal and regulatory systems, the low degree of transparency implies inadequate legal protection. In countries where investment-related regulations and policies are murky and elusive, interpretations and implementation of these rules and policies are likely to be obscured. Therefore, foreign investors are likely to refrain from committing themselves in these markets. The perceived high risk associated with non-transparent government procedures can be heightened because the low re-deployability of assets and resources of FDI makes foreign investors more vulnerable to an unfavorable change in the country business environment.
Our conjecture on the negative effect of transparency is further justified by examining several distinct attributes of transparency. First, low transparency implies high uncertainty. When host government apparatus follow "hidden agenda," "internal documents" and the vagueness of the policy and regulations, the uncertainty of running business in uncharted water increases. This enhanced uncertainty in the business environments would compound complexity of managing the FDI in a foreign land. Second, the potential cost due to low transparency tends to be intangible. While a cost caused by corruption to a foreign investor can be approximated by using publicly released information, the cost of low transparency is hard to predict and measure and it may lead to system failure as the recent Asian financial crisis manifests. The third is the untraceable nature of transparency. Severity of corruption in a host country is more or less a known fact. Information on corruption is often exposed in the public and the public at large plays important roles in denouncing corruption and pressuring for punishment. Hence, corruption may be less a threat to investors since the likelihood to be detected and penalized is high (Rahman 1986). In contrast, transparency tends to be invisible and its impacts on business often fail to draw public attention but may force foreign investors to shirk away from a market. Following the above reasoning, we propose the following hypothesis to be tested.
Hypothesis 2. The presence of low transparency in a host country is likely to hinder the inflow of FDI to that country. The lower the transparency, the lower the level of inward FDI
The Method and Data
Based on the above reasoning, the relationship between FDI, corruption and transparency can be modeled by simply positing that the level of FDI inflow into a country is minimized when high corruption and low transparency are present in that country. However, FDI decisions cannot be made in isolation. On the contrary, previous studies of factors influencing FDI flow showed that decisions to undertake foreign investments were often made based on comprehensive assessments of overall business environment (Green/Cunningham 1975, Kobrin 1976, Pugel 1981, Hollander 1984, Schneider/Frey 1985, Klein/Rosengren 1990, Brewer 1991, Dunning/Narala 1994, Dunning 1996). Hence, we also incorporate a number of business environment variables in our model to control their effects. The importance of host country environment is obvious since unfavorable conditions in the host country would not make it an attractive market for foreign investors. We ran three statistical tests of the hypotheses using the following models, with the subscripts i referring to country:
(1) [FDI.sub.i,t-1] = [[beta].sub.i] - [[beta].sub.1] [CORRUP.sub.i] + [[beta].sub.2][EROI.sub.i] + [[beta].sub.3][STABLE.sub.i] - [[beta].sub.4][INFLATE.sub.i] - [[beta].sub.5][KCOST.sub.i] - [[beta].sub.6][INTERVEN.sub.i] + [[beta].sub.7][EXPGROW.sub.i] + [alpha]
(2) [FDI.sub.i,t-1] = [[beta].sub.i] - [[beta].sub.1] [TRANSP.sub.i] + [[beta].sub.2][EROI.sub.i] + [[beta].sub.3][STABLE.sub.i] - [[beta].sub.4][INFLATE.sub.i] - [[beta].sub.5][KCOST.sub.i] - [[beta].sub.6][INTERVEN.sub.i] + [[beta].sub.7][EXPGROW.sub.i] + [alpha]
(3) [FDI.sub.i,t-1] = [[beta].sub.1] - [[beta].sub.1] [CORRUP.sub.i] + [[beta].sub.2][TRANSP.sub.i] + [[beta].sub.4][EROI.sub.i] - [[beta].sub.4][STABL.sub.i] - [[beta].sub.5][INFLATE.sub.i] - [[beta].sub.6][KCOST.sub.i] + [[beta].sub.7][INTERVEN.sub.i] + [[beta].sub.8][CORRUP.sub.i] x [TRANS.sub.i] + [alpha]
In all the models, FDI is a one-year lagged dependent variable measured as the percentage of total GDP. CORRUP and TRANSP are measures of corruption and transparency respectively. Model (1) and (2) are aimed to estimate the separate effects of CORRUP and TRANSP for the reason discussed in the previous section. Because low TRANSP often creates opportunities for bribed bureaucrats to seek personal gains, it is highly likely that low TRANSP gives rise to more CORRUP. In view of the high correlation of TRANSP and CORRUP, we include an interactive factor of CORRUP and TRANSP in model (3) to examine its compounding effects. In addition, we also control a number of variables in the tests. The definitions of these control variables are explained in the following section.
We also apply model (3) to three different country groups (Asia, OECD and emerging markets). We select these country groups because they are at the different level of economic development and the perceived levels of corruption and transparency in these groups also differs. The country group tests serve two purposes. First, the tests allow us to compare whether the impacts of corruption and transparency on FDI vary across country groups. If the effects are insignificant across country groups, it would lend more validity to our hypotheses. Second, these tests also serve as sensitivity tests for the model. As the model is applied to different country groups, fixed effects can be estimated to isolate the potential estimation bias due to distinctive characteristics associated with a specific country group. Thus, the consistent results of a model that takes into account country differences would add more predictive power to the model.
The unbiasedness of ordinary least square model (OLS) "between" (i.e. cross-country) and "random effects" estimators, requires that the independent variables not be correlated with the country-specific effects (often reflecting omitted variables) counted in the error term. The typical econometric solution to this issue is to first establish the data differentials and then estimate a relationship by using instrumental variables. However, we do not apply this procedure for two reasons (2). First, our measure of corruption exhibits very little variation over time. Thus, the information content of the data corresponds in large part to the cross-country variation in the data. First differencing estimation will therefore suffer from low power. Second, the lack of variation in corruption and transparency used in the test indicates that corruption and transparency may occur in a moving average process with long lags. When an appropriate lag is unknown and an inappropriate lag-structure is specified, the first differencing method may produce inconsistent estimates of the long-run relationships between variables. Hence, estimation based on cross-country means with the raw data covering a reasonably long time-span would provide a consistent estimate of the long-run relationships among variables (Pesaran/Smith 1995).
The strategy to cope with this issue of estimation bias adopted in this paper is to examine three separate estimates: full model tests with and without lag (3) and a censored model (fixed effects). The model with lag helps to find time series related effects and the censored model intends to show how significant the independent variables are on the dependent variable after the samples are selected discretionally. The appropriateness of the estimate will hold if the results are consistent for all these tests. Otherwise the misspecification of the model is suspected.
Obviously, using annual aggregates for any one source-country does not provide a sufficient number of observations for a robust test. Therefore, pooled cross-country time series data between 1991-1997 were constructed for 40 countries where corruption indices are available from the annual survey of Transparency International. This panel data yielded 280 observations. The data were drawn from several sources, as outlined below. The detailed variable definitions and sources are given in Appendix 1.
FDI/GDP data are from annual International Financial Statistics by International Monetary Fund (IMF).
The data employed to measure CORRUP are from annual indexes published by Transparency International (TI). TI is a non-profit research organization housed at Goettingen University in Germany (4). Corruption Perception Index representing a `poll of polls', constructed by a team of researchers at Goettingen University. While several other corruption indices are available and no one index is perfect, the TI index that standardizes and combines a number of measures from various sources into a single composite index strengthens its validity and reliability (Lancaster/Montinola, forthcoming). Despite its limitations (Husted 1999), a corruption index from TI has been used in a number of academic studies (Wei 1997a, Heidenheimer 1996, Lambsdorff 1997, Husted 1999, Treisman 2000). The original corruption scores from TI range from 1 to 10, the higher the score the less corrupted. To facilitate the interpretation, we recalculated the corruption score by subtracting the original scores from 10. Consequently the recalculated high score refers to a high presence of corruption. The data for research variable TRANSP are from The World Competitiveness Yearbook (formerly titled as the World Competitiveness Report) of the Institute for Management Development (IMD). IMD is located in Lausanne, Switzerland. The information in the yearbook measures and compares the extent to which a country provides an environment that fosters the domestic and global competitiveness of companies within its borders. IMD conducts a survey of a panel of more than 3,500 executives. IMD gathers both qualitative and quantitative information on 139 hard criteria. The measure of TRANSP is the ranking of 40 countries with 1 being the most transparent and 40 as the lease transparent.
Since firms make decisions about foreign investment based on expected returns and the effects of corruption and transparency are not necessarily exogenous parameters that one can vary in a FDI model without considering other key economic factors, we've incorporated into the statistical analysis a number of control variables suggested in prior literature, such as financial, socio-political and infrastructure factors. Sources of these data are given in Appendix 1. The expected return on investment (EROI) is a key determinant of FDI. A country with a perceived high return is likely to draw more FDI than a country with low EROI. Lacking data on expected return on investment does not permit us to use an exact measure of EROI. Alternatively we calculated an approximate EROI by dividing investment income by total investment. The data for investment income and total investment are drawn from Balance of Payments published by the International Monetary Fund (IMF). Inflation rate (INFLATE) and cost of capital (KCOST) are also included since high levels of both may erode foreign investors' competitive positions and earning potentials (Grosse/Trevino 1996). The indicator of KCOST is IMD's ranking index of countries in terms of the cost of capital. The high ranking index refers to high capital costs. To control for the effect of external trade of a country on its inward FDI, we used the export growth rate of each country in the model (EXPGROW). Social and political stability is an important factor in a FDI decision making (Kobrin 1980, Tallman 1988). Social stability index (STABLE) is used to approximate this effect. We also include two variables to count the impacts of local government intervention (INTERVENE).
Description of the Data
Table 1 presents the descriptive statistics of mean, standard deviation and correlation of variables in the analysis. CORRUP and TRANSP are highly associated, indicating an interactive effect that was estimated in our analysis. Although a number of variables such as INTERVENE, STABLE and KCOST are correlated, the inclusion of these variables in the model gives us some confidence that their coefficients will not necessarily be capturing effects of other explanatory variables. Furthermore, given the nature of time series data, multicollinerity among variables is expected.
We examined the severity of multicollinearity using the diagnostic statistics of variance inflation factor (VIF) (Neter/Wasserman/Kutner 1983, Gujarati 1995). Except variable STABLE that has a VIF exceeding 10, the VIF tests of other variables indicated no undue influences of multicollinerity on the least squares estimates.
Figures 1 and 2 provide scatter plots of FDI versus CORRUP and TRANSP for the 40 countries. Figure 1 exhibits a weak correlation between CORRUP and FDI. However, this speculated relationship can be misleading because of the obvious extreme cases such as Singapore and Ireland and because of partial association without controlling for other factors. For Figure 2, there is a somewhat clear pattern indicating strong associations between FDI and TRANSP. However, again the exceptions of Singapore and Ireland are noticeable.
[FIGURES 1-2 OMITTED]
Results of Multiple Regression Analysis
Table 2 presents the statistical estimates of model parameters. The results for the general models include all 40 countries with FDI as a lagged dependent variable. Apparently the results from the three models confirm our hypotheses that high corruption and low transparency led to the reduced FDI in host countries.
* The corruption only model (column 1) shows that CORRUP was only marginally associated with the FDI (p < 0.10). However the direction of influence supported our hypothesis number one with regard to its negative impacts on FDI.
* As a contrast to CORRUP, TRANSP (column 2) turns out to be more negatively correlated with FDI judging by its coefficient (-0.477; p 0.01). Again this result strongly confirmed our hypothesis 2 that the low transparency of a host country tends to increase business uncertainty resulting in less inward FDI.
* Column 3 reports the results of the model tests for both CORRUP and TRANSP, and their interaction effects. Interestingly TRANSP showed no significant impact on FDI (p > 0.10). The sign is correct indicating the negative relationship. However, both CORRUP and the interactive effect of CORRUP x TRANSP exhibited significant and negative influences on FDI. This result may suggest that corruption together with low transparency would exert more negative impacts on FDI than that of low transparency alone.
To test the sensitivity and stability of the model and to examine whether the influences of corruption and transparency vary across different country groups, regression was also run for Asian, OECD and emerging economies (fixed effects). The estimated fixed effects are presented in columns 1, 2 and 3 of Table 3 respectively. As indicated, even after accounting for the trend effects, the explanatory powers of all the tests are high accounting for 76, 65 and 66 percent of the variances in the level of FDI respectively. The signs and the significances of the coefficients of CORR UP and TRANSP were consistent in all country groups. These results further supported our hypotheses that the presence of high corruption and low transparency are likely to hinder the inflow of FDI even after controlling the positive and significant effects of EROI. Except the OECD model, the interactive effects of CORRUP and TRANSP also show the significant reduction of FDI inflow due to the simultaneous presence of two factors.
With regard to control variables, for all the six models the variable EROI consistently showed significant and positive impacts on the inflow of FDI in each of the four models. The high magnitudes of influences as indicated by the coefficients strongly suggest that EROI is a far more important determinant in FDI and confirm the existing financial theory that the foremost consideration of investors is the expected return on their investment when making investment decisions in an international environment (Jorion/Khoury 1996). The next important factor is INTERVENE. Except the case of OECD, government intervention showed a significant inverse relationship with FDI, suggesting that the high government intervention perceived by investors as a hurdle to investment and repatriation of income led to decreased FDI. The result for OECD, however, is no surprise since the economies in the majority of OECD countries are relatively subject to less government intervention.
The STABLE factor showed consistently as an influencing variable explaining the different level of FDI. The more stable a nation, the more likely FDI inflow. This factor is not significant for emerging economies though. Nevertheless, it carries the correct sign. KCOST also turned out to be a key determinant of FDI. With the exception of the Asian countries, the high cost of borrowing (KCOST) in local financial markets may negate the likelihood of inducing FDI.
In general, EXPGROW of host countries were significant in attracting FDI. This factor is also marginally significant for OECD countries and emerging markets (respective of p < 0.010). For the Asian countries it is not significant, but the positive sign of the coefficient points to the correct direction. This weak relationship may be explained by the fact that foreign investments in Asian economies were mainly to access the low-cost labor, the abundant resources and the local markets.
The comparison of the three country group models revealed that EROI, KCOST, STABLE, and INTERVEN exerted similar influences on the Asian and emerging markets, as judged by the direction and significance levels of coefficients. For OECD countries, social stability and high expected returns seem to explain the high level of FDI. But for the Asian and emerging markets groups, in addition to the high returns, the economic growth and inflation rates influence the FDI flow.
This study sets out to test two political and cultural factors on inward FDI. It argues that corruption and transparency, though different, produce similar impacts on FDI given the presence of similar expected returns on investment. Two hypotheses were postulated. The statements that the impact of high corruption and low transparency on FDI is likely to negatively influence the inward FDI were then tested. These statistical tests were based on panel data of 40 source countries and confirmed these hypotheses. This is robust in three aspects.
* First, unlike previous empirical studies that control a few variables, we embraced factors of economic, social and financial factors that have been explained or proved in finance and economic theories, since the existing literature in international business illustrate that more factors play significant roles in the process of international investment decision making than in the domestic investment (e.g., Rugman/Hodgetts 1995).
* Second, the sensitivity tests performed in three country group models further strengthen the rigor of the analyses. The tests show that the results are stable and consistent.
* Third, the sample size is large and covers a longer time frame than previous studies. Therefore, we are confident to conclude that our models are adequate to describe the actual situation.
Our analysis made two contributions to the study of FDI. First, this paper--unlike the existing literature that basically examines the relationship between corruption and domestic economic and social activities--studied a different dimension by focusing specifically on the impact of corruption on FDI. Despite the increasing public concerns over the negative impacts of corruption on firms doing business internationally, few systematic studies have been undertaken to provide empirical evidence. This study represents one of a few such efforts and its finding sheds valuable lights on how corruption as a social phenomenon influences the inward FDI.
The second contribution is that the results of this study add another dimension to the existing knowledge concerning FDI and its influencing factors by providing further empirical evidence on the relationship between corruption and FDI. The extant literatures of international management by and large ignored the effect of corruption and, at best, subsumed corruption into the political risk. However, our study treats corruption explicitly as an independent factor influencing the inflow of FDI. By incorporating this factor, our study enriched our understanding of the influences of external factors on FDI decisions.
The third contribution is the introduction of transparency as an important factor in the FDI model. Lacking prior studies on the relationship between transparency and FDI, our findings on the negative impact of low transparency on FDI may be the first of its kind. Based on the empirical estimates, we can infer from the existing finance and economics theories that low transparency, as a systematic risk, is hard to diversify or minimize and its effect on foreign investment is considerably negative. In more transparent countries, investors are more able to appraise the government policy and regulations and, hence, are in a better position to make rational and informed investment decisions. On the contrary, in less transparent countries, investors not only are susceptible to systematic risk, which they cannot diversify in the market, but also may be subject to some other unforeseeable risks, such as expropriation and discriminatory policy and regulations.
The findings of this study pose practical and challenging tasks for international business practitioners. The negative effects of corruption on FDI imply that firms contemplating to move into international markets need to carefully examine the degrees of corruption and transparency in the target countries. Even if the return on investment is expected to be high for a given target country, firms have to face the ethical and legal dilemma and need to make trade-offs. Though firms may not be able to fight effectively corruptions in a country as individuals, contingency plans and strategies aimed at coping with corrupt bureaucracy and obscure rules must be formulated based on the sound judgment of the country-specific corruption practices. However, firms need to be aware of the difficulties in making such a judgment given that the level of corruption of a country is deeply rooted in the culture of that country (Husted 1999) and cultural values are difficult to change (LaPalombara 1994).
The negative impacts of poor transparency on FDI imply that firms, before making foreign investment decisions, should conduct careful evaluation of the government apparatus and operations of a host government since the ambiguous interpretations of policies and regulations and the practice of double-standards through internal documents could have more direct and immediate impacts on any FDI projects. Further, our study refutes the argument that FDI decision making should be different in countries where cultural-orientations manifested in the different views and practices of corruption and transparency. While it is true that the views on and practices of corruption may vary across countries due to the different culture and traditions, the investment decisions by MNEs should account for these differences and incorporate measures to counteract the negative impacts on the business.
The finding of this study also suggests that the attention should be also paid to the impact of transparency by firms as well as policy-makers. Up to now, the effect of transparency of government policies and regulations has largely gone unnoticed by firms and general public, while corruption has remained the target of the public crusade and governments have taken various legal measures to fight the corruption practices. Though low transparency and corruption are highly correlated and lead to depressed FDI, they carry different policy implications in respect to the improvement of FDI environment. Our findings suggest that for countries to attract inward of FDI, policy-makers should fight on two fronts. The practical implications are that the public as well as governments should also devote efforts to the enhancement of transparency of the regulatory regime, in addition to fighting corruptions using various legal means.
There are two limitations to our study. First, our data are at the national level and corruption and transparency data compiled by IMD and TI are perceptual and subjective. The test results may be more robust if we could use industrial or even company level data. However, the unavailability of the data prevents us from performing analyses at industrial and firm levels. Future research can extend the current analysis by overcoming this limitation. Hence, we suggest that future research efforts should be made to create and collect objective information on both measures in the future.
In the same vein, the classifications of various forms of corruption and transparency are needed in the future research to capture the differential impacts on FDI. For instance, the corruption and transparency measures can be classified by the severity and scope, and by industrial sectors. Should this classification scheme be developed in conjunction with refined data collection, more dynamic linkages between FDI and corruption and transparency may be captured.
Another weakness is that our study ignored event information. Event information on specific announcements about corruption and transparency and the investment-related announcements at company and country levels may help us to capture and estimate better the relationship between corruption and transparency. Again the unavailability of concurrent data for all the countries does not allow us to analyze the impact of such events.
Appendix 1. Definition of the Variables FDI/GDP Data are from International Financial Statistics of IMF. CORRUP Corruption index from the annual report by Transparency International. The index value of one means the least corrupted. TRANSP Transparency index from the World Competitiveness Yearbook published by IMD, Lausanne, Switzerland. EROI Value of total investment income divided by total investment. Data are from the Balance of Payment published by IMF. EXGROWTH Export growth rate reported in the World Development Report by the World Bank. INFLATE Data are collected from International Financial Statistics of IMF. KCOST Country Ranking Index for the cost of capital by IMD. STABL Index from the World Competitiveness Yearbook of IMD. The index value of one denotes the highest stability. INTERVEN Index about the degree of government intervention in business activities, composed by IMD. Table 1. Descriptive Statistics: Mean, Standard Deviation and Correlations Variable Mean Std. 1 2 3 4 Dev. 1 FDI 6395 10923 1.00 2 CORRUP 6.078 2.461 -0.085 1.00 3 TRANSP 23.92 13.14 -0.042 0.791 1.00 4 EXOGROW 9.09 3.563 0.060 0.218 -0.012 1.00 5 EROI 0.1919 5.638 0.633 -0.034 -0.312 0.036 6 INFLATE 30.45 179.6 -0.055 0.045 0.000 -0.031 7 KCOST 23.22 13.81 -0.288 0.200 0.053 0.019 8 STABLE 23.02 12.89 0.042 -0.485 0.188 -0.002 9 INTERVEN 2.49 13.32 -0.080 -0.564 0.104 -0.335 Variable 5 6 7 8 9 1 FDI 2 CORRUP 3 TRANSP 4 EXOGROW 5 EROI 1.00 6 INFLATE 0.006 1.00 7 KCOST 0.048 -0.043 1.00 8 STABLE 0.002 -0.063 0.032 1.00 9 INTERVEN -0.042 -0.035 -0.120 0.009 1.00 Table 2. Regression Results of FDI/GD[P.sub.t-1] on CORRUPT and TRANSP (1) (2) (3) Research Variables CORRUP -0.144 * -0.363 *** (-1.62) (-3.61) TRANSP -4.77 *** -0.027 (-5.99) (-1.56) Control Variable EROI 0.618 *** 0.755 *** 0.802 *** (10.734) (13.09) (13.19) EXPGROW 0.108 ** 0.102 ** 0.107 ** (2.31) (.496) (1.707) INFLATE -0.04 -0.44 -1.77 *** (-.093) (-.999) (-1.32) KCOST -2.13 *** -0.180 ** -0.152 (-2.61) (-2.38) (-2.33) STABLE 0.191 ** 0.300 *** 0.318 *** (2.61) (3.59) (3.61) INTERVENE -.300 *** -0.341 *** -0.345 *** (-3.63) (-4.62) (-4.46) CORRUP x TRANSP -0.350 *** 0.447 (-2.897) Adjusted R2 0.447 0.509 0.521 F 27.72 *** 35.40 *** 29.87 *** Standardization coefficients with T statistics in parentheses. * p = < 0.10; ** p = < 0.05; *** p = <0.01 Table 3. Regression Results of FDI/GD[P.sub.t-1] on CORRUPT and TRANSP (a) Asian OECD Emerging Model Model Mkt Model (1) (2) (3) Research Variables CORRUP 1.337 *** -0.300 * -0.699 ** (-6.024) (-1.894) (-2.247) TRANSP -1.215 *** -1.046 *** -0.732 ** (-4.569) (-3.879) (-2.28) Control Variables EROI 0.329 *** 0.844 *** 0.454 *** (2.252) (11.11) (5.561) EXPGROW 0.035 0.083 * 0.119 * (-0.469) (1.707) (1.788 INFLATE -0.196 * 0.76 -0.113 * (-1.824) (1.277) (-1.643) KCOST -1.407 *** -0.152 -0.571 *** (-10.58) (-1.566) (-5.543) STABLE 0.680 *** 0.424 *** 0.073 (3.957) (5.249) (.702) INTERVENE -1.206 *** -0.047 -0.327 *** (-9.03) (-.527) (-3.691) CORRUP x TRANSP -1.011 *** -0.26 -651 *** (-3.988) (-1.372) (-1.728) adjusted [R.sup.2] 0.766 0.653 0.659 F 20.59 *** 29.06 *** 17.06 *** (a) Standard coefficient with t-statistics in parentheses. * p < 0.10; ** p < 0.05; *** p < 0.01.
(1) We thank the anonymous reviewers for the comments and suggestions in the earlier version of this paper.
(2) We ran the tests with first difference, second difference and third difference. The results do not significantly differ from the one used in the analysis. Thus, these results are not reported here.
(3) The dependent variable is lagged in one year in the full model to account for the possible autocorrelation problem that may cause the dependent variable simply follow a trend and not related to other explanatory variables in the model.
(4) Corruption measure is also available from World Competitiveness Report. It has high correlation (0.83) with that of Transparency International.
Ades, A./Di Tella, R. D., Rents, Competition and Corruption, American Economic Review, September 1999, pp. 982-993.
Alam, M. S., Some Economic Costs of Corruption in LDCs, Journal of Development Studies, 27, 1990, pp. 85-97.
Beck, P. J./Maher, M. W., A Comparison of Bribery and Bidding in Thin Markets, Economic Letters, 20, 1, 1986, pp. 1-5.
Besley, T./McLaren, J., Taxes and Bribery: The Role of Wage Incentives, Economic Journal, 103, 1993, pp. 119-141.
Braguinsky, S., Corruption and Schumpeterian Growth in Different Economic Environments, Contemporary Economic Policy, XIV, July 1996, pp. 14-25.
Brewer, T., Foreign Direct Investment in Developing Countries: Patterns, Policies and Prospects, Washington, DC: World Bank PRE Working Paper, 712, June 1991.
Brunetti, A./Kinsunko, G./Weder, B., Credibility of Rules and Economic Growth--Evidence from a World Private Sector Survey, Background Paper for the World Development Report, Washington: The World Bank 1997.
Davies, D. G., Property Rights and Economic Efficiency: The Australian Airlines Revisited, Journal of Law and Economics, 20, 1977, pp. 223-226.
Dunning, J. H./Narula, R., Transpacific Foreign Direct Investment and the Investment Development Path: the Record Assessed, University of South Carolina Essays in International Business, 10, May 1994, pp. 1-69.
Dunning, J.H., Explaining Foreign Direct Investment in Japan: Some Theoretical Insights, in Yoshitomi M./Graham, E.M. (eds.), Foreign Direct Investment in Japan, Cheltenham, UK: Edward Elgar Publishing Ltd. 1996, pp. 8-63.
Flatters, F./McLeod, W. B., Administrative Corruption and Taxation, International Tax and Public Finance, 1995, pp. 397-417.
Green, R. T./Cunningham, W. H., The Determinants of U. S. Foreign Investment: An Empirical Examination, Management International Review, 15, 1975, pp. 113-120.
Grosse, R./Trevino, L. J., Foreign Direct Investment in the United States: an Analysis by Country of Origin, Journal of International Business Studies, 27,1, 1996, pp. 139-155.
Gujarati, D. N., Basic Econometrics, McGraw-Hill International Edition 1995.
Gupta, S. H./Davoodi, H./Alonso-Terme, R., Does Corruption Affect Income Inequality and Poverty?, International Monetary Fund Working Paper, no. 98/76, May 1998.
Heidenheimer, A. J., The Topography of Corruption: Explorations in a Comparative Perspective, International Social Science Journal, 158, 3, 1996, pp. 337-347.
Hershman, M. J., Bribery under pressure, Journal of Commerce and Commercial, 416, April 1998, p. 7.
Hines, J. R. Jr., Forbidden Payments: Foreign Bribery and American Business After1977, NBER Working Paper, no. 5266, Cambridge, MA: National Bureau of Economic Research 1995.
Hollander, A., Foreign Location Decision by US Transnational Firms: An Empirical Study, Managerial and Decision Economics, 5,1, March 1984, pp. 7-18.
Huntington, S. P., Political Order in Changing Societies, New Haven, CT: Yale University Press 1968.
Husted, B. W., Honor Among Thieves: a Transaction-cost Interpretation of Corruption in the Third World, Business Ethics Quarterly, 4,1, 1994, pp. 17-27.
Husted, B. W., Wealth, Culture, and Corruption, Journal of International Business Studies, 30, 2, 1999, pp. 339-360.
Jorion, P./ Khoury, S. J., Financial Risk Management: Domestic and International Dimensions, Oxford, UK: Blackwell Ltd. 1996.
Kaufmann, D./Wei, S.J., Does Grease Money' Speed up the Wheels of Commerce? National Bureau of Economic Research, Working Paper 7093, Cambridge, MA 1997.
Keefer, P./Knack, St., Institutions and Economic Performance: Cross-country Tests Using Alternative Institutional Measures, Economics and Politics, VII, November 1995, pp. 207-227.
Klein, M. W./Rosengren, E. S., Determinants of Foreign Direct Investment in the United States, Worchester, and Boston: MA mimeo 1990.
Kobrin, S. J., The Environmental Determinants of Foreign Direct Investment: An Exposit Empirical Analysis, Journal of International Business Studies, VI, 1976, pp. 29-42.
Kobrin, S. J., The Assessment and Evaluation of Noneconomic Environments by American Firms, Journal of International Business Studies, 9, 1980, pp. 32-47.
Laffont, J./Guessan, T. N., Competition and Corruption in an Agency Relationship, Journal of Development Economics, 60, 1999, pp. 271-295.
La Porta, R./Lopez-De-Silanes, F./Shleifer, A./Vishny, R., Legal Determinants of External Finance, The Journal of Finance, LII, 3, 1997, pp. 1131-1150.
Lambsdorff, J. G., An Empirical Investigation of Bribery in International Trade, Paper presented at the Workshop on Corruption and Development, Institute of Development Studies, Sussex, UK May 6-7, 1997.
Lancaster, T. D./Montinola, G. R., Toward a Methodology for the Comparative Study of Political Corruption. Corruption and Reform, Forthcoming.
LaPalombara, J., Structural and Institutional Aspects of Corruption, Social Research, 61, 2, 1994, pp. 325-350.
Leff, N., Economic Development Through Bureaucratic Corruption, American Behavioral Science, 1964, pp. 8-14.
Lien, D. H., A Note on Competitive Bribery Games, Economic Letters, 22, 4, 1986, pp. 337-341.
Mauro, P., Corruption and Growth, Quarterly Journal of Economics, 110, August 1995, pp. 681-712.
Mody, A./Srinivasan, T. G., Trends and Determinants of Foreign Direct Investment: Analysis of US Investment Abroad, Washington, DC: The World Bank Industry Development Division, mimeo 1991.
Mookherjee, D/Png, I. P. L., Corruptible Law Enforcers: How Should They be Compensated?, Economic Journal, 105, 1995, pp. 145-149.
Murphy, K. M./Shleifer, A./Vishny, R. W., Why Is Rent-Seeking so Costly to Growth? American Economic Review Papers and Proceedings, LXXXIII, 1993, pp.409-414.
Neter, J./Wasserman, W./Kutner, M. H., Applied Linear Regression Models, Homewood, IL: Richard D. Irwin, Inc. 1983.
North, D. C., Institutions, Institutional Change, and Economic Performance, Cambridge: Cambridge University Press 1990.
Pesaran, M. H./Smith, R., Estimating Long-run Relationships from Dynamic Heterogeneous Panels, Journal of Econometrics, 68, 1995, pp. 79-113.
Pugel, T. A., The Determinants of Foreign Direct Investment: An Analysis of US Manufacturing Industries, Managerial and Decision Economics, 2, 1981, pp. 220-228.
Rahman, A. T. R., Legal and Administrative Measures Against Bureaucratic Corruption in Carino, ed. in Asia, Bureaucratic Corruption in Asia: Cause, Consequences and Controls, Phillippines, Quezon City: NMC Press 1986.
Rose-Ackerman, S., Corruption: A Study in Political Economy, New York, NY: Academic Press 1978.
Rugman, A. M./Hodgetts, R. M., International Business: A Strategic Management Approach, New York: McGraw-Hill 1995.
Samuelson, P. A./Nordhaus, W. D., Microeconomics, 16th edition, New York: McGraw-Hill Book Company 1998.
Schneider, F./Frey, B. S., Economic and Political Determinants of Foreign Direct Investment, World Development, 13, 2, 1985, pp. 161-75.
Shleifer, A./Vishny, R. W., Corruption, Quarterly Journal of Economics, 1993, pp. 599-617.
Tallman, S. B., Home Country Political Risk and Foreign Direct Investment in the United States, Journal of International Business Studies, 1988, pp. 219-34.
Tanzi, V./Davoodi, H., Corruption, Public Investment, and Growth, IMF Working Paper, 97/139, Washington: International Monetary Fund 1997.
Treisman, D., The Causes of Corruption: A Cross-National Study, Journal of Public Economics, 76, 2000, pp. 399-457.
Tullock, G., Corruption Theory and Practice. Contemporary Economic Policy, 14, July 1996, pp. 6-13
Wall Street Journal, September 23, 1997, p. A16.
Wei, S.J., How Taxing is Corruption on International Investors?, NBER Working Paper, No. 6030, Cambridge, MA: National Bureau of Economic Research 1997a.
Wei, S. J., Why Is Corruption So Much More Taxing Than Tax? Arbitrariness Kills, NBER Working Paper, no. 6255. Cambridge, MA: National Bureau of Economic Research 1997b.
Wheeler, D./Mody, A., International Investment Location Decisions: The Case of U. S. Firms, Journal of International Economics, 33, 1992, pp. 57-76.
John Hongxin Zhao, Associate Professor of International Business, The Boeing Institute of International Business, John Cook School of Business, St. Louis University, St. Louis, MO, USA.
Seung H. Kim, Professor of International Business, Director of the Boeing Institute of International Business, John Cook School of Business, St. Louis University, St. Louis, MO, USA.
Jianjun Du, doctoral candidate at John Cook School of Business.
Manuscript received October 2000, revised April 2001, revised May 2001.…
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Article title: The Impact of Corruption and Transparency on Foreign Direct Investment: An Empirical Analysis (1). Contributors: Zhao, John Hongxin - Author, Kim, Seung H. - Author, Du, Jianjun - Author. Journal title: Management International Review. Volume: 43. Issue: 1 Publication date: January 2003. Page number: 41+. © 1999 Gabler Verlag. COPYRIGHT 2003 Gale Group.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.