wPDS 3X POLICY RESEARCH WORKING PAPER 3 02 8 Does a Country Need a Promotion Agency to Attract Foreign Direct Investment? A Small Analytical Model Applied to 58 Countries Jacques Morisset The World Bank Foreign Investment Advisory Service U April 2003 POLICY RESEARCH WORKING PAPER 3028 Abstract Establishing an investment promotion agency has empirical analysis indicates that agencies devoting more become a central part of most countries' development resources on policy advocacy are more effective because strategies. Today there are more than l50 investment such activity is not only beneficial to foreign investors promotion agencies worldwide. Yet very little is known but also to domestic investors. In contrast, investment about what these agencies have been really doing, generation or targeting strategies appear expensive and notably in emerging countries, and whether they have risky, especially in countries with poor investment been effective in influencing investors' decisions. climates. Using data from a new survey on 58 countries, Finally, certain internal characteristics of the agencies Morisset shows that greater investment promotion is are associated with greater effectiveness. The agencies associated with higher cross-country foreign direct that have established reporting mechanisms to the investment (FDI) flows, on top of the influence of the country's highest policymakers (the president or prime country's investment climate and market size. minister) or to the private sector have been systematically But this result has to be qualified on several couIIts. more efficient at attracting foreign direct investment. First, the effectiveness of the agency depends on the Such institutional links are crucial because they country's environment in which it operates. An agency in contribute to strengthen the government's commitment a poor investment climate is less effective at attracting as well as reinforce the agency's credibility and visibility investment. Second, the scope of activities that an agency in the business community. undertakes influences its performance. Morisset's This paper-a product of the Foreign Investment Advisory Service-is part of a larger effort in the Bank to understand foreign direct investment flows. Copies of the paper are available free from the World Bank, 1818 H Street NW, Washington, DC 20433. Please contact Monte Feghali, room 19-110, telephone 202-473-6177, fax 202-522-3262, email address mfeghahi@ifc.org. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at jmorisset@worldbank.org. April 2003. (22 pages) The Policy Research Working Paper Senes dtsseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the finditgs out quickly, even if the presentations are less than fully polished. The papers carry the names of the auithors and shotld be cited accordingly. The findings, interpretations, and concluszons expressed in this paper are entirely those of the authors They do niot necessanly represent the view of the World Bank, its Executive Directors, or the countries they represent. Produced by the Research Advisory Staff DOES A COUNTRY NEED A PROMOTION AGENCY TO ATTRACT FOREIGN DIRECT INVESTMENT? A small analytical model applied to 58 countries by Jacques Morisset' l Lead Economist, Foreign Investment Advisory Service (FIAS). I am especially grateful to Professor L. Wells for his support at various stages of this research project. I also benefited from comments on another version of this paper from J. Battat, F. Sader, N. Roger, J. Bergsman, D. Weigel, and S. Djankov. K. Johnson, A. Minoux, N.Smith and A. Alvarez contributed to this research. 1. Introduction Establishing an investment promotion agency has become a central part of most countries' development strategies. Today, there are worldwide more than 160 national investment promotion agencies (IPAs) and over 250 sub-national ones (UNCTAD, 2001). This trend is relatively new since only a handful of these agencies existed 20 years ago. Yet, the reality is that very little is known about what these agencies are actually doing in most countries, especially in the developing world. The objective of this paper is to assess to what extent does investment promotion help explain cross-country variations in foreign direct investment (FDI) flows. We propose a simple analytical model in which the promotion agency's behavior is assumed to reflect the interest of the government (or a supervisory board) under the constraint of limited budgetary resources. It follows a certain tradition in public finance, see Heller, 1975. Our model remains simple, as it is principally use to derive an equation that can be easily estimated for our sample of countries. It should also be seen as a complement to the seminal work of Wells and Wint, 2001, who have provided the theoretical underpinnings of adopting investment promotion policies to attract FDI.2 Due in part to a lack of reliable data, no broad empirical study of investment promotion agencies and their effectiveness in attracting FDI has been done to date3. This empirical gap means that the debate on the effectiveness of IPAs is still very open. We use data 2 Wells and Wint (2001) define investment promotion as "activities that disseminate information about, or attempt to create an image of the investment site and provide investment services for the prospective investors". This definition encapsulates the two most important analytical justifications for IPAs. The first consists of its role in communicating and disseminating information. Since this latter can be considered as a public good, it is possible that the private sector behavior will not lead to the optimal social welfare. As a matter of fact, local firms may voluntarily restrict information flows to prevent the entry of new potential competitors. The second justification is that the IPA can play a role of coordinating most activities aimed at improving the business environment in the host country. This role can range from providing assistance to potential and existing investors in their daily problems to lobbying for key policy and legal reforms. 3 The only empirical examination of the impact of FDI promotion on FDI flows was conducted by Wells and Wint, 2001, based on data gathered in 1988. This study had serious shortcomings in terms of the concept used to measure the promotion 3 from a new survey that was conducted in 58 investment promotion agencies between February and May 2002.4 At the outset, it is worth underscoring that our research should be viewed as a first attempt to fill the existing empirical gap and has obvious limitations. The most important is that we have been only able to examine empirically the relationship between promotion and foreign direct investment at one point in time, specifically the year 2001. Unfortunately, data are simply not available for additional years. We believe that it nonetheless provides some answers to four sets of questions that should help IPA managers and policy makers to develop a better understanding of the conditions -both external and internal to the agency-- that influence the effectiveness of promotion: o How does the amount of spending on investment promotion affect its effectiveness? Does an agency need to exceed a minimum level to have any effect on international investors? o To what extent does the business environment or the country's characteristics affect the effectiveness of investment promotion? Does the quality of the general business environment matter? o Does the effectiveness of investment promotion vary according to the functions or activities on which it focuses? Should an IPA devote more resources toward policy advocacy or image building? o Is the effectiveness of investment promotion influenced by different agencies' characteristics such as their structure, mandate, sources of funding, and institutional relationships? The paper proceeds as follows. In Section II we begin by presenting a small analytical model through which we explain the behavior of an investment promotion agency. Section Em[ discusses the selection of variables as well as the main sources of data used for the application of the above model. Section IV presents the main empirical results 4The questionnaire was sent via email or fax to 114 agencies, both from developing and industrial countries. We received responses from 75 agencies, yielding a response rate of 66%. The rate was higher in Latin America and in Eastern and Central Europe where it reached 86% and 71%, respectively. A description of the main responses can be found in the FIAS website (www.fias.net). 4 and provides a few interpretations. Finally, Section V concludes and offers a few directions for future research. II. A Small Analytical Model To our knowledge, there has been no attempt to capture analytically the behavior of investment promotion agencies. Here we adopt a simple approach based on the assumption that governments will strive to maximize the level of FDI inflows through its promotion effort and at the same time, minimize resources allocated to this effort. The assumption that the agency attempts to maximize the level of FDI is based on the observation that for most countries the effectiveness of promotion is measured by its capacity to attract (foreign) private investment. It is true that some sophisticated IPAs try to do more: some aim at increasing the quantity as well as the quality of FDI, where quality might be measured by investments impact- in terms of job creation, exports or technology transfers. Nonetheless, all developing countries are first and foremost interested in attracting more PDI; suggesting that the IPA performance can be evaluated on the basis of this shared goal. At the same time than maximizing FDI inflows, the government will attempt to minimize the use of public resources required to finance the promotion effort. Again, this assumption might be excessive because it could be argued that promotion agencies have access to other sources of funding. Still, public funds account for the about 70% of EPA budgets worldwide -this percentage exceeds even 80% for developing countries. Alternative sources of funding, such as private or external sources, have been inconsistent or weak complements because of the public nature of investment promotion. The government, or a set of policymakers, maximizes their utility, taking into account uses of public resources required to finance the promotion effort (P) and the objective to attract FDI. In any period, we assume the utility function of the IPA defines as: 5 U = ao + a, (FDI - FDI*) - 22 (FDI-FDI*)2 _a3(P_P*)+a4 (p_p)2 (1) 2 2 with a, >- O Following Heller,1975, the selection of this functional form ensures diminishing marginal utility for any increase in FDI and increasing marginal (dis)utility for P, as these variables rise above a level determined by their target levels (defined with the symbol *). It reflects a compromise between the need for an estimable functional form with desirable utility function properties.S The target FDI*, or the desired level of FDI, is set in the context of the literature on FDI that assumes that long-term FDI flows are defined by a series of structural variables such as the level of economic development and the quality of the investment climate in each country. FDI = yC + Y2Y (2) where IC is an indicator of the quality of the investment climate and Y the level of economic development or the market size. These two variables have been retained following the recent literature on the determinants of FDI. The importance of these two factors is linked with the two main motivations for FDI: investments are especially sensitive to the investment climate, since multinational can generally choose between locations; investments aimed at the local market are most easily attracted by large 6 markets. In our attempt to select these extemal variables, we also considered others factors, in particular infrastructure and education variables, as well as regional dummies, but those added only limited and inconsistent explanatory power. 5 See Heller, 1975, or Mosley et al., 1987, for a similar utility function. 6 These two motivations are sometimes referred as vertical and horizontal FDI. For a good summary on these views, see S. L. Brainard, 1997. 6 We assume that the level of promotion has no effect on the desired level of FDI in long- run following the arguments developed by Wells and Wint, 2001, that justify the use of investment promotion principally on a temporary basis. Promotion agencies are viewed as vehicles for addressing coordination and information issues that tend to disappear over time. Since promotion does not influence FDI flows in the long run, we can write that the desired level of public resources devoted to promotion is, equal to 0. P* = 0 (3) Next, we capture the eventual temporary influence of investment promotion on FDI flows by: FDI = AP (4) with A Ž0 Maximizing equation (1) with respect to FDI and P and subject to constraint (4) yield tc the following equation: FDI = /o + A3P +821C +/33Y with fi0 = - ; =-; = 2 = r, 3 = Y2 a2 a2 From equation (5) we define PI as a measure of IPA effectiveness. Since our model will be estimated in logarithmic form, the value of PI describes the elasticity of the investment promotion effort and FDI inflows. The main feature of equation (5) is that the IPA effectiveness is measured by taking into account the influence of external variables and IPA characteristics. Such influence has to be considered because an increase in FDI flows can occur independently of a greater promotion effort, for example through an improvement in the macro-economic stability, and a consequent positive relationship observed between the promotion effort and FDI may only be-spurious. 7 The basic model remains can be extended in three directions. First, the coefficient fi, is likely to vary depending on the country's policy environment even though it is not clear a priori in which direction. For example, one might suspect that investment promotion is more effective in a good than in a poor policy environment because it is easier to convince potential investors to come to an attractive country. However, in such a context, it could also be argued that the agency is redundant. Most investors are well aware of opportunities in their field or industries, and do not really need to contact (or to be contacted by) an investment promotion agency. To account for the influence of the environment on IPA effectiveness, equation (5) can be modified by associating multiplicative dummy variables to the promotion effort. These dummies will be defined to test whether threshold values in the external variables, say the quality of the investment climate influences the effectiveness of the promotion agencies. This approach lies on the belief that above or below specific values the role and performance of the promotion agency differ. Alternatively, we can explore the possibility of a linear relationship between the effectiveness of promotion and the country's environment. Following Burnside and Dollar, 2000, this can be done by adding an interaction termn between the promotion effort and the investment climate into equation (5): FDI = 3o +/AP+/2EV+/3 (PxIC) (6) where IC is defined as the log of the investment climate indicator used. Taking the derivative of the FDI inflows with respect to the promotion effort, we can identify the relationship between the IPA effectiveness and the quality of the investment aFDI climate, = i + J,63IC. The IPA effectiveness is defined as a linear function of the quality of the investment as long as the coefficient ,83 is significantly different from 0. The second extension is to assess whether IPA effectiveness is influenced by which functions does the agency carry on. To capture these functions, we simply rewrite our 8 basic equation (5) by separating the promotion effort (P) in the several categories defined by Wells and Wint, 2001; that are image-building, investor servicing, investment generation, and policy advocacy. These activities have been extensively detailed by these two authors and will not be explained in this paper (see Annex A for a brief summary). Here it suffices to indicate that they are generally well known by managers of investment promotion agencies and experts in that field. A third and final extension is to incorporate a number of internal characteristics of the promotion agencies as explanatory variables into equation (5). In doing so, we follow again Wells and Wint, 2001. In particular, these authors have argued that such effectiveness rnight be strengthened when the agency benefits from the involvement of the private sector and a focus mandate on investment promotion. III. Definition of Variables and Sources of Data This section discusses, first, the selection of variables used for the application of the model developed above. Then, we briefly describe some of the methodological choices that we had to face. We used as a first indicator total gross FDI inflows as defined by UNCTAD, which offers a more complete country's coverage than the IMF or the World Bank's database. This indicator presents the double advantage to be relatively homogenous across countries and relatively well used in the economic literature, therefore facilitating comparisons. It may be however biased for our purpose of measuring the IPA effectiveness since many agencies have reported that they do not have the main responsibility for privatization and for the mining sector --two important channels of FDI worldwide. For this reason, using the UNCTAD data, we adjusted gross FDI flows by the amount of Merger & Acquisitions (M&A) in each country. The idea is to capture more accurately the FDI flows that are under the direct control of most IPAs. 9 The promotion effort (P) is measured by the annual budget of each agency converted into US dollars. This conversion assumes that each dollar spent by the agencies has the same value across countries, which might introduce inconsistencies for cross-country comparisons. While the costs of certain promotional activities are independent of the IPA location (e.g., the cost of an advertisement in an international newspaper would in principle be the same for Ethiopia and Germany), some expenses vary greatly across countries. For these reasons, we also measure the promotion effort by the number of professional staff dedicated to investment promotion in each agency. This proxy does not vary with the exchange rate and accounts for different purchasing value across countries. However, it makes the strong assumption that labor (rather than money) determines the level of promotion effort in each country. Both variables -the IPA budgets and staff- have been extracted from the survey conducted by the Foreign Investment Advisory Service of the World Bank in early 2002. As explained earlier, the promotion effort has been separated into four categories so that we can capture better what IPA do. We follow Wells and Wint, 2001, and distinguish the following activities: investment generation (IG), policy advocacy (PA), investor servicing (IS), and image-building (IB). The budgetary resources allocated to each of these functions have also been identified using the FIAS survey. The internal characteristics of the agency have been captured through the FIAS survey, including a) age of the agency, b) legal status (founded by law or decree), c) institutional affiliation and linkages with government (public, semi-public, autonomous, or private body), d) linkages with the private sector (financial contribution, frequency of meetings and inputs, degree of private sector representation on board), e) reporting arrangements (board, government body, prime minister or president), f) overseas offices, g) number of mandates on top of foreign investment promotion (e.g., export promotion, privatization programs,) and f) staff's characteristics and salary policy. We tested several indicators of the quality of the investment climate such as the ones developed by the World Economic Forum, Harvard, the Heritage Foundation, 10 Intemational Country Risk, etc. After numerous tests, we selected the Heritage Index as it had added the highest explanatory power to our model.7 Similarly, we used three as alternative indicators for the local market size: GDP, population, and income per capita. All these variables appear highly correlated for our sample of countries, and the level of income generated the most robust results per capita, hence we retained this variable. We explore the relationship between the investment promotion effort and FDI inflows in 58 countries during 2001.8 Our sample is diversified in terms of regions, level of income per capita, as well as the magnitude of investment promotion efforts. To keep some homogeneity, we include only countries that have reported national investment promotion agencies. We exclude countries, such as the US, China, Brazil, and India, that have only regional agencies since it was difficult to assess their contribution to FDI flows at the national level. This omission may be important because these countries account for a substantial share of FDI worldwide. While the data obtained from the FIAS survey are unique, they are only available for the year 2001. For this reason, our approach can only be cross-sectional. By doing this, we do not necessarily assume that promotion has an immediate effect on overall FDI flows because we use a cross-country rather than time-series analysis. Our results would be significantly biased only if we had depicted large variations in the promotion effort over the last few years. Fortunately, it appears that variations in the promotion effort across 7 Indeed, as recently confirmed by Batra and Moody, 2002, variations in cross-country FDI flows are well explained by these factors over the past decade. The Heritage Foundation Index captures 50 independent variables divided into 10 broad factors of economic freedom related to internal and external macroeconomic conditions, economic openness, and political and institutional health. The higher the score on a given factor, the greater the level of government interference in the economy and the less economic freedom a country enjoys. The index assigns scores inversely to the state of the investment climate; hence we expect a negative sign on the coefficient of the estimation. 8 'he list of countries is the following: Armenia, Bangladesh, Belgium, Benin, Bolivia, Bosnia And Herzegovina, Bulgaria, Cape Verde, Chile, Colombia, Costa Rica, Cote D'Ivoire, Czech Republic, Denmark, Dominican Republic, Ecuador, El Salvador, Estonia, Ethiopia, Finland, Georgia, Greece, Guyana, Honduras, Hong Kong,, Indonesia, Ireland, Israel, Jamaica, Jordan, Kenya, Korea, Latvia, Lesotho, Malawi, Maldives, Mali, Mauritius, Morocco, Mozambique, Nepal, Pakistan, Panama, Paraguay, Peru, Portugal, Senegal, Slovakia, Slovenia, South Africa, Sri Lanka, Saint Lucia, Thailand, Trinidad and Tobago, Tunisia, Uganda, Vanuatu, Venezuela. countries reported for the year 2001 are not significantly different than those that existed in the late 1990s.9 We use the Ordinary Least Squares (OLS) technique to estimate our basic FDI equation. In doing so, we assume that the level of promotion effort is weakly exogenous in the sense that the promotion effort is not contemporaneously affected by FDI flows. Although we cannot, in the absence of time-series analysis, prove that causality runs from promotion to investment, it seems unlikely that increases in foreign investment will typically cause an increase in the promotion effort. Recent experience suggests that IPA budgets are determlined and approved within a business plan that usually covers a 3-5 year period. They are rarely revised as a result of FDI flows, at least not in the short term. Moreover, since the main source of IPA funding is the government, some degree of inertia is expected to be associated with public finance decisions.10 lWv. Empiricall ResuRts Our main empirical results are summarized in the Table 1. Overall, the results appear satisfactory, as the explanatory power of the regressions is relatively high, in spite of the limited number of explanatory variables. By assuming that causality runs from promotion effort to FDI flows, we find that investment promotion is positively associated with cross-country variations in FDI flows -the elasticity coefficient is equal to 0.25. More importantly, this finding holds when we account for the influence of the investment climate, as measured by the Heritage Index, and the level of development in each country. It also continues to be robust using 9 For fuller details on this point, see Morisset and Johnson, 2003. 10 Of course, abrupt changes in IPA budget are always possible due to budgetary crises in government or the termination of external assistance. However, these events occurred in a very few countries of our sample, hence we believe that they will not significantly affect our empirical results that capture trends across a sample of 58 countries. 12 alternative definitions of FDI inflows. Still, our simple and restricted correlation between the IPA budget and FDI must be interpreted with caution. The positive and significant correlation does not necessarily imply that promotion always positively affects FDI, but it makes it more difficult to argue that promotion is in fact bad for attracting FDI. The positive correlation, in other words, restricts the range of possibilities. Table 1: Estimated Elasticity Coefficients a/,b/, c/ (t-statistics in Parenthesis) FDI FDI (ADJUSTED FOR M&A) IPA Budget .25 .31 (2.41) (3.33) IPA Staff .26 .20 (1.56) (1.14) investment climate-Heritage -1.62 -1.64 -1.78 -1.90 (4.09) (-3.80) (4.15) (4.12) GDP per capita .57 .73 .42 .49 (2.48) (3.10) (1.67) (2.11) Observations 58 49 51 43 AdjR2 .682 .683 .696 .681 Notes: a/ All variables are in logs except for the Heritage Index. for which an improvement captured by a decline in the indicator. b/ We omitted to report the constant term. c/ Numerous other external variables were tested in the regressions including GDP, regional dummies, other investment climate indicators. In general, the results were robust independently of the inclusion or exclusion of these variables. This positive association is found when the promotion effort is measured by the IPA budget. When effort is measured by the level of human resources, the relationship is also positive, but not statistically significant. It is possible that the close relationship of FDI to expenditures means that promotion activities require less labor than money. Anecdotal evidence collected at the country level reveals that fixed labor costs usually account for less than one-third of an agency total budget --the bulk of expenses is associated with buildings, promotion materials, advertisement, and promotional trips. 13 One concern is that the IPA effectiveness is affected by the size of their budgets, which vary considerably across countries. 1 ' Using a kernel or neighborhood function'2 we find that up to a minimum budget of about US$64,000 IPAs seem to have no (or even negative) impact on FDI. Between a budget of US$64,000 and US$2 million -IPAs are effective, and the elasticity is about 0.5. Between US$2 and US$11 million, the estimated elasticity reaches its maximum value. On average, we find that a 10 % increase in the IPA budget is associated with a 7.5% increase in FDI flows, which is three times higher than the elasticity estimated for the entire sample. Finally, above a budget of about US$11 rmillion, promotion expenditures are ineffective. Although these results should be interpreted with caution, they suggest that small agencies are not really effective at attracting FDI. Presumably, the existence of a threshold reflects the significant fixed costs associated with numerous activities, such as advertising and image building.13 Still, agencies that are too big have their own problems. There is a maximum level of resources beyond which there are decreasing economies of scale -even though this limit appears out of reach of many agencies. The existence of decreasing marginal returns can also be explained in the light of the analytical arguments used to justify the creation of IPAs. Beyond a limit, it is unlikely that the agency can contribute more to resolving the information and coordination issues that had justified its creation. It is worth underscoring that we have not been able to depict any significant patterns associated with the size of IPAs budgets. For example, we did not find that IPA budgets were significantly correlated to country's size or levels of development -at for our sample of countries. Granted, IPAs located in industrial countries have generally higher budgets than those located in developing countries, but this positive correlation does not hold for other categories such as low income and middle income countries. 12 This method is adaptive and dynamic in the sense that the slope parameter changes along the regression curve depending on the location of the varnables in the sample. More specifically, we minimize the weighted sum-of-squared errors between the actual and the fitted by allowing the program to compute a regression at every point in our data set. A kernel fit is superior to the regression analysis because the latter assumes constancy of the slope parameters. 13 This is easy to understand: an advertising campaign in an international newspaper can reach several thousands of US$ dollars; promotional trips and participation to fairs can be equally expensive; as are the needs of potential investors when they visit the country. 14 Our empirical analysis clearly confirms that the quality of the investment climate and the level of development have a significant effect on IPA performance. 14 The better the investment climate, the greater IPA effectiveness. Similarly, the higher the level of development; the more effective an IPA is. These linear relationships between the IPA effectiveness and these external variables are captured in the two following equations (all variables are expressed in logs): dFDI/dPE = 0.552 - 0.289 IC (4.59) (-4.31) dFDI/dPE = -0.110 +0.042 Y (-0.56) (2.79) More concretely, by applying these relationships to our sample of countries, they indicate that the IPA effectiveness ranges from 0.16 for the agency located in the country with the worst investment climate to above 0.35 for the country with the best investment climate. Similar large differences in the impact of the promotion effort are also depicted in function of the country's income per capita.. We now turn to the question whether promotion effectiveness is influenced not only by the amount of resources spent by the agency but also by how these resources are allocated across its functions or activities. We estimated equation (5), except that we broke IPA spending into the four components identified earlier to detect their influence on FDI inflows. Each function was estimated separately to avoid multi-co linearity problems between these variables. The estimated elasticity coefficients are presented in Table 2. The estimated results reveal that policy advocacy appears to have the strongest impact on F'DI inflows, followed by image-building, investor servicing, and investment generation. However, there does not appear to be a significant difference between image-building and investment servicing, both of which appear equally effective and not far behind 4 We were unable to identify threshold values using multiplicative dummues. 15 policy advocacy. 1516 Our finding that policy advocacy seems to be the most effective IPA function is consistent with our previous result that IPA effectiveness is positively correlated with the quality of the investment climate. Most IPAs are in a strategic position to carry out policy advocacy activities because of their interface between the private and the public sector. By contrast, the weak elasticity coefficient associated with the investment generation function can be explained by the high cost and high degree of expertise required to effectively carry out targeted programs. An agency needs specialists, by sectors or enterprises, and usually time and attention to convince a particular n investor to locate in its country. Most practitioners would agree that for each US$1 spent, investment generation has the lowest return, and especially in our sample that includes a majority of countries with relatively poor investment climates. Table 2: Elasticity Coeffcient by IPA Functions FUNCTION/ ELASTICITY COEFFICIENTS .30 Policy Advocacy (2.14) Image Building .25 (2.32) Investor Servicing/Facilitation .24 (2.17) Investment Generation .18* (1.79) *All coefficients are statistically significant at 5%, except the one associated with investment generation, which is at the 10% level. 15 The three coefficients associated to these variables are not significantly different from each other at the 5% level, but only at the 10%. 16 These estimated results represent an average for all IPAs, they do not account for the charactenstics of each country in which the IPA is located. As we did earlier for the overall IPA budget, we explore whether the effects associated with each function vary depending on the country's environment but, in that case, we were unable to find any statistical variations in our sampleWe used the same approach than for the overall EPA budget, but apply it to each EPA function. Unfortunately, these series of tests were non-conclusive in depicting any significant cross-country differences in the estimated impact of each EPA function. 16 The conceptual framework developed by Wells and Wint, 1990 and revised in 2001, suggests that an IPA's effectiveness is influenced by its institutional structure and reporting mechanisms. Although we tested several internal characteristics using both dummy variables and interactive tenns associated with the IPA budget variable, only three of them seem to have significant effects on the effectiveness of IPAs across countries (Table 3)17. First, FDI flows are significantly lower in countries where the IPA is part of a ministry in contrast to be an autonomous body or a joint private-public institution. Second, the effectiveness of IPAs is enhanced when the agency reports to a supervisory board that includes representatives of the private sector ---the higher the number of private members, the greater IPA effectiveness18. Finally, IPA effectiveness is enhanced when it reports directly to a country's president or a prime minister. Table 5: The influence of IPA characteristics on FDI Inflows (T-statistics in parenthesis) Dependant Variable/Explanatory Variable FDI FDI FDI Public agency*dummy -.79 (-2.43) Number of private representatives in the Board *Budget c/ .052 (2.72) Prime Minister/President Dummy .53 (1.80) IPA Budget .21 .22 .25 (1.95) (2.01) (2.29) Observations 56 36 55 AdjR2 0.71 0.72 0.68 Notes: a/We omit to report the results associated with the constant term, GNI per capita, the Heritage Index that wei e also included in the above regressions. b/All variables are in logs, except for Heritage Index and the dummy variables. c/ The number of representatives starts from 1 to the maximum registered in our survey. 17 The methodology was twofold. First, we used dummy variables and, second, an interactive term multiplied to the IPA budget. The first approach captures the possible effect of the EPA characteristics on the FDI flows, independently of the promotion effort. The second approach assumes a linear relationship between the EPA budget and the EPA characteristics. These two approaches have been detailed in the precedent technical appendix. 18 An agency with I private representative would report an elasticity coefficient equals to 0.22, while an agency with 8 private sector representatives in its Board will see its effectiveness increase to about 0.32. 17 These findings confirm that the most effective IPAs benefit from visibility and from participation by the private sector through their boards or through institutional relations. These characteristics reinforce the leverage of the agency and its role. They also indicate that including the private sector contributes to broaden the platform and help to achieve a consensus in the agency's effort to market the country abroad. Yet, they should not be used to overestimate the role of the investment agency. The reporting mechanism to the president or the prime minister may reflect the overall commitment of the government toward reforms. Within that context, the positive correlations reported above, while indicative, would capture this global trend in the government's effort rather than the agency's own performance. Finally, it is worth reporting that some IPA characteristics do not seem to influence the agency performance. The agency's mandate, staff qualification, and the number of overseas agencies have no significant effect on FDI flows. It is possible that these characteristics do not matter. Another possibility is that our survey has too little variations in these factors to identify their eventual impact on the effectiveness of IPAs. 'V. Concllsion and Policy Recomximendations The story that emerges from our investigation is that the IPAs are effective in influencing decision to invest. Although our empirical exercise suffers from obvious limitations, their role seems significant, on top of the influence of key factors such as the quality of the investment climate and the country's market size. The finding that promotion is positively associated with FDI inflows across countries has to be qualified because it is closely link to the environment in which the agency operates. Investment promotion is more effective in a country with a good investment climate and a relative high level of development. In fact, our results suggest that promotion can be even counterproductive in a country that offers a poor investment climate. It seems more difficult to convince an investor to come back if he was disillusioned during his first visit 18 to a country. The disappointed investor is also likely to be vocal about his disenchantment and, so, discourage other potential investors. The functions carried on by the promotion agency can influence its effectiveness. Our empirical results reveal that policy advocacy is the most effective function for attracting a dollar of investment, followed by image-building, and investor servicing. Investment generation appears to be the least efficient or cost-effective, partly because it is expensive and partly because it is often not adapted to the reality of our sample of countries that have relative poor investment climates. While the optimal level of budget that should be allocation in each function depends on the country, our results suggest that-most agencies would gain by devoting more attention to policy advocacy. Political visibility and participation of the private sector appear to be two elements that contribute to the success of IPAs. Political visibility is best attained when the agency is linked directly to the highest government officials (e.g. the president or the prime minister). Private sector involvement can be secured through participation in the board thalt supervises the agency. Our paper should be viewed as a first attempt to examine empirically the effectiveness of investment promotion agencies across countries because it contains obvious limitations. Below are proposed three directions for future research. The first direction would consist of completing the data collected in our survey, not only by including more countries but also by extending the coverage period over time. Future researchers, depending on the data availability, should carefully explore the time lag between overall investment promotion and the decision of foreign investors to invest in the country. This time lag can vary depending on the country's conditions and the functions carried out by the agency. The second direction should be to consider a more complex and realistic objective function for the IPAs. Our approach subscribes to the view that IPA's main objective is to attract more (foreign) investors. Such an assumption is certainly correct but simplistic 19 because some agencies assign more emphasis on the quality than the quantity of FDI. This is certainly true for sophisticated IPAs that focus on job creation, exports or technology transfers associated to FDI flows. Along the same lines, one may want to explore whether the IPA effectiveness varies depending on the kind of FDI (e.g., sectors, reinvested earnings vs. mergers, etc.). The third and final direction would be to proceed with individual case studies. Our study was aimed at capturing trend across a relatively large set of countries at the expense of details. A complementary approach would be to focus on a few LPAs, with a special attention on their internal characteristics and external environment. Evaluations of their particular promotional techniques (such as image building, investment service, investment generation, and policy advocacy) and of different structures (such as government and quasi-government) should add considerably to the evidence on the influence of promotion on investment flows and to an understanding of the kinds of investors likely to be influenced by promotion efforts. 20 Bibliography G. Batra and Moody, 2002, "Investment climate Indicators", mimeo, FIAS S. L. Brainard, 1997, "An Empirical Assessment of the Proximity-Concentration Trade- off Between Multinational Sales and Trade", The American Economic Review, Vol. 87, No. 4., September, pp. 520-544. Craig Bumside and David Dollar, 2000, "Aid, Policies, and Growth", The American Economic Review (U.S.); Vol. 90, No. 4:847-68, September. J. Morisset and K. Johnson, 2003, "The Effectiveness of Promotion Agencies at Attracting FDI", mimeo, FIAS. P. Mosley, J. Hudson, and S. Horrell, 1987, "AID, The Public Sector, and the Market in LDCs", Economic Journal, pp. 616. P.S. Heller. 1975," A Model of Public Fiscal Behavior in Developing Countries: Aid, Investment, and Taxation", American Economic Review, June. UNCTAD, 2001, "The World of Investment Promotion at a Glance: A Survey of Investment Promotion Practices", United Nations Conference on Trade and Development, United Nations Advisory Studies number 17, UNCTAD/ITE/IPC/3. Louis Wells and Alvint Wint, 2001, " Marketing a Country, Revisited", FIAS Occasional Paper, number 13. 21 ANNEX The Functions of an Investment Promotion Agency While there are alternative ways to separate the different functions carried on by an investment promotion agency, we have retained the grouping proposed by Wells and Wint in their seminal paper published in 1990 and revised in 2001. Based on their observation of a large of agencies, both in the developing and industrial world, these authors have proposed the following categories that are briefly described below. Image -Building is the function of creating the perception of a country as an attractive site for international investment. Activities commonly associated with image-building include focused advertising, public relations events, the generation of favorable news stories by cultivating journalists, and so on. InvestorFacilitation and Investors Servicing refers to the range of services provided in a host country that can assist an investor in analyzing investment decisions, establishing a business, and maintaining it in good standing. Activities in this area include information provision, "one-stop shop" service aimed at expediting approval process, and various assistance in obtaining sites, utilities, etc. Investment Generation entails targeting specific sectors and companies with a view to creating investment leads. Activities include identification of potential sectors and investors, direct mailing, telephone campaigns, investor forums and seminars and individual presentations to targeted investors. Investment generation activities can be done both at home and overseas. Policy Advocacy consists of the activities through which the agency supports initiatives to improve the quality of the investment climate and identifies the views of the private sector on that matter. Activities include surveys of the private sector, participation in task forces, policy and legal proposals, and lobbying. 22 Policy Research Working Paper Series Contact Title Author Date for paper WPS3011 Renegotiation of Concession J. Luis Guasch April 2003 J Troncoso Contracts in Latin America Jean-Jacques Laffont 37826 Stephane Straub WPS3012 Just-in-Case Inventories A Cross- J Luis Guasch April 2003 J Troncoso Country Analysis Joseph Kogan 37826 WPS3013 Land Sales and Rental Markets in Klaus Deininger Apnl 2003 M Fernandez Transition Evidence from Rural Songqing Jin 33766 Vietnam WPS3014 Evaluation of Financial Liberalization. Xavier Gin6 April 2003 K. Labrie A General Equilibrium Model with Robert M. Townsend 31001 Constrained Occupation Choice WPS3015 Off and Running? Technology, Trade, Carolina Sanchez-Paramo April 2003 H Sladovich and the Rising Demand for Skilled Norbert Schady 37698 Workers in Latin America WPS3016 Partisan Politics and Stuti Khemani Apnl2003 H Sladovich Intergovernmental Transfers in India 37698 WPS3017 Why is Unemployment so High Jan Rutkowski April 2003 J Rutkowski in Bulgaria? 84569 WPS30)18 Integrating the Least Developed Paul Brenton April 2003 L Tabada Countries into the World Trading 36896 System The Current Impact of EU Preferences under Everything but Arms WPS3019 Foreign Aid, Conditionality, and Ghost Thilak Ranaweera April 2003 P Rathan Raj of the Financing Gap A Forgotten 33705 Aspect of the Aid Debate WPS3020 Consumption, Health, Gender, Anne Case April 2003 D Ballantyne and Poverty Angus Deaton 87198 WPS3021 Institutional Arrangements for Public Elizabeth Currie April 2003 M Rosenquist Debt Management Jean-Jacques Dethier 82602 Eriko Togo WPS3022 Sending Farmers Back to School. Gershon Feder April 2003 P Kokila The Impact of Farmer Field Schools Rinku Murgai 33716 in Indonesia Jaime B Quizon WPS3023 Productivity Growth and Product Douglas M Addison Apnl 2003 D Addison Variety Gains from Imitation and 31188 Education WPS3024 R&D and Development Daniel Lederman April 2003 P Soto William F Maloney 37892 WPS3025 Trade Structure and Growth Daniel Lederman April 2003 P Soto William F Maloney 37892 WPS3026 Strategic Approaches to Science Robert Watson April 2003 R Watson and Technology in Development Michael Crawford 36965 Sara Farley Policy Research Working Paper Series Contact Title Author Date for paper WPS3027 Financial Intermediation and Growth: Genevieve Boyreau- April 2003 P. Sintim-Aboagye Chinese Style Debray 38526