A Multiple Criteria Decision Analysis for the FDI in Latin- American Countries

Proceedings of the 2009 Industrial Engineering Research Conference A Multiple Criteria Decision Analysis for the FDI in Latin- American Countries Levis R. Cabrera, Germán E. Giraldo Department of Industrial Engineering University of Puerto Rico, Mayagüez, Puerto Rico 00680, USA Abstract The interest of Foreign Direct Investors in Latin American countries is growing rapidly; in the period of 1992-2006 the Foreign Direct Investment (FDI) has tripled, from US$27,500 to US$72,440 millions. Investors have the challenge of deciding which country is better for their investments. The problem of selecting a country to invest in is a common topic in the current economic environment. In this study, an analytical hierarchical process (AHP) is used to help investors make a rational decision. The results show consistence with the ranking obtained by the World Economic Forum, where Chile appears with an advantage of 34% as best option. Keywords Foreign Direct Investment, Analytical Hierarchical Process, Latin America, Country Selection 1. Introduction The Foreign Direct Investment (FDI) on Latin American Countries has increased in the recent years. The Economic commission for Latin America and Caribbean Countries (CEPAL) has informed that in the period from 1992-2006 the amount of Foreign Direct Investment (FDI) has tripled, from US$27,500 to US$72,440 millions [17] This economic interest in the development on emerging economies and of several of our Latin American countries in particular has triggered a new challenge for Company executives, entrepreneurs and investors. Facing the problem of deciding what country is better to put their investments. It is important to recognize that each country is different and each one has different factors FDI attractive. This is an important problem because FDI is deemed an engine with respect to growth. This decision has several aspects to be analyzed in order to take a correct decision based on different important factors. Traditionally it would be desirable to have a criterion of decision based on experience, costs, and benefits, but in this case it is not possible to measure the factors in those terms. These factors are called [2] nonmonetary attributes. A more formal approach provides the decision maker with a rational method to evaluate the alternatives with respect to a multiple nonmonetary attributes and assist them in their country selection problem. The Analytical Hierarchical (AHP) Process developed by Saaty [11] involves a multiattribute, multiperiod and multiperson problem hierarchy for the selection of the most desirable alternative. The method constructs a hierarchy based on pairwise comparisons of the elements. In this case, these are factors that attract the investment (attributes) and the selection of the country (Alternatives) and create a scale of the strength with the elements that dominates another. The result of the process is a scale of alternatives that allows the decision maker to justify the selection of the country to the stockholders or owners in a rational fashion, based on the strength of the alternative. Unfortunately for foreign investors interested in Latin America, the countries attributes are not sufficiently organized in order to help investors to make decisions. The main objective of this paper is to collect each country attribute and develop a scale of Latin American countries as investment alternatives using the AHP methodology, so that it helps investors to resolve conflicting interests and goals to make a decision. Other objectives that are involved in this research are: the identification of the most attractive countries, the identification of important attributes common to all and the gathering of data for diverse sources to fit the AHP model. For engineering, it is useful to develop new models that explain and help analyze critical decisions that are involved in important economical factors. Given the resources that will be invested in a host country and the risk involved for the stockholders, the analysis of factors, alternatives and their relation; it is an important area of interest for research and development which is subsequently important to engineering and science. 137

2. Multicriteria FDI Analysis Methods Using Latin American Countries Relevant studies in which FDI in Latin American countries are compared versus a set of important factors are analyzed. To the best of our knowledge there is no a study performed with AHP model for Latin America, but some authors have made specific approaches on Latin countries using other techniques. The following is a summary of these studies. Treviño and Mixon uses cross-country differences in macroeconomic and institutional environments to explain the factors that affect the FDI inflows by Multinational Enterprises (MNE) to seven Latin American countries [5]. Argentina, Brazil, Chile, Colombia, Mexico, Peru and Venezuela where analyzed for the period of 1988-1999, while Frenkel et. al. [6], examine the determinants of FDI flows to emerging economies by analyzing a compiled data set of bilateral FDI flows the period of 1992-2000. Five countries of investment were selected: France, Germany, Japan, UK and USA, and 21 countries from Asia, Latin America, and Central and Eastern European countries for FDI were employed, were five are Latin American: Mexico, Chile, Brazil, Argentina and Venezuela. These authors employed regression analysis technique. Galan and González [4] proposed three main hypothesis for potential growth, manufacturing should be close to the customer, the cultural affinity. [3], presents the most close work to our study. They used a Multiple Criteria Decision Analysis (MCDA) based on the use of fifteen factors and different index (rank) for each one of them. A hierarchical process was made in order to compare pairs of alternatives. The result of the study shows a rank of countries for selection in a utility percentage. The study is different in the way that the analysis was made. Our analysis use the AHP developed by Saaty [12] and the present work is based on the work of Nijkamp and van Delft [9]. Table 1 summarizes the studies of multicriteria analysis on Latin American countries. None was made on the basis of the AHP process. Table 1.Studies of multicriteria analysis on Latin American countries Year Author Method of analysis Countries Attributes 2004 Treviño and Mixon Regression 7 6 2004 Frenkel et al. Regression 5 6 2006 Galan and Benito ANOVA 11* - 2006 Beim and Lévesque MCDA 3 15 2008 Trevino et al. Regression 16* 4 *The name of the country are Unknown 3. Selection of Alternatives (Countries) and Attributes (Factors) To determine main alternatives and factors two studies are used. Fist, Estrada [7] presented an analysis of the most attractive countries for investment during the World Economic Forum on Latin American. The study takes in to account eight important factors for evaluation of alternatives: Economic stability, business law, risk conditions, ease access to information, financial security, FDI and government stability as principal factors. They determine that Chile, Brazil, Colombia and Peru lead the region over the level of attraction offered by the climate for private investment. Second, World Bank [17] in the annual report of 2007 uses principally the FDI to classify the most attractive countries. The countries with mayor FDI in % of the total investment are Mexico (26.14%), Brazil (25.93%), Chile (11.12%), Colombia (8.69%) and Peru (4.78). These countries add a sum 76.66% of the total FDI. Then, the alternatives for this study are: Brazil, Chile, Colombia, Mexico and Peru. The analysis identifies the important factors that impact the FDI as follows: time to start a business, tax rates, borrow rate, labor cost (salaries), inflation and population growth, given by the research in many countries and in the official data that they collect. The selected factors for the present study are: financial security, growth rate, purchase power parity, business law, lack of crime, labor regulations, risk conditions, bureaucracy, lack of corruption, government stability, time to start a business, tax rates, borrow rate, labor cost (salaries), inflation and population growth. These factors are the most important factors that affect the FDI in a Latin American country, then this are the factors that are used in the model. 3.1 Attributes ranking Beim and Lévesque [3] and World Bank [17] evaluate ranking of the alternatives based on economic, political and cultural criteria. Table 2 summarizes the ranks and the indexes used are recommended by the Foreign Investment Advisory Service of the World Bank Group (World Bank and Foreign Investment Advisory Service) and other important sources. In order to take a final decision we made calculations for the ranking index (shown in Table 3) 138

using the Ordinal Scaling or method of paired comparisons to Canada et.al. [2] where pairwise comparisons were made by putting a letter "P" in the row factor that is preferred than the column factor. Table 4 summarizes the results obtained which turned out to be consistent with the literature. Table 2. Ranks used and weighted index Attribute Description Source Rank Index % * perceptions of the quality of governance of a large number of Government stability enterprises, citizens, non governmental World Bank. (2008) 0-12 10.66% organisations, commercial risk rating agencies and think tanks Business law laws in country that protects the investment Frazer Institute. (2008) 0-10 9.93% Lack of crime law and order Frazer Institute. (2008) 0-10 9.56% likelihood that changes in the business environment Risk conditions adversely affects operating profits or the value of assets The PRS GROUP. (2007) 0-100% 9.56% in a specific country Lack of corruption Corruption Perception Index which Transparency reflects the degree to which corruption International. (2008) is perceived 0-6 9.56% time in days when you decide to start the operation of a Time to start a business new business until its runinng World Bank. (2008) days 7.72% Financial security the legal right given to a creditor The PRS GROUP. (2007) 0-10 6.25% Purchase power parity Tax rates Borrow rate Growth rate per capita Gross National Product adjusted by purchasing power. company taxes that are paid as a result of operations rate that investors paid to acces a credit Increase or decrease of the Gross Domestic Product in per cent World Bank (2007) Bank of different countries Bank of different countries Labor cost (salaries) the minimum salarie per country World Bank. (2008) Labor regulations employment contracts, conditions of employment, job security, collective bargaining, worker participation in management, and collective disputes Money (dollars) 5.88% 0-100% 5.88% 0-100% 3.31% World Bank (2007) 0-100% 2.94% Money (dollars) 2.94% World Bank. (2008) 0-100% 2.57% FDI Foreign Direct Investment CEPAL (2007) Money 2.57% Inflation is a rise in the general level of prices over time World Bank. (2008) 0-100% 1.47% Population growth is the change in population over time, and can be quantified as the change in the number of individuals in a population using "per unit time" for measurement * Percentage of contribution (importance) of each attribute Table 3. Ordinal Scaling World Bank. (2008) 0-100% 0.38% A B C D E F G H I J K L M N O P Q Total % * G Government Stability = P P P P P P P P P = P P P = P 14.5 10.66% A Business Law = = P P P = P P P P P = P P = P 13.5 9.93% L Lack of Corruption P = P P P = P P P P = P P = P 13.0 9.56% M Lack of Crime = P = P P P P P P P = P P = P 13.0 9.56% P Risks Conditions = P = P P P = P P P P = = P P = 13.0 9.56% C Bureaucracy = P = P P P P P P = = P P = = 12.0 8.82% Q Time to star business P = = P P P P P P P P = 10.5 7.72% D Financial Security P = P P = P P P P = 8.5 6.25% H Tax Rates P = P P P P P P = 8.0 5.88% O Purchasing Power Parity P P P P = P = P P 8.0 5.88% B Borrow Rates = = = P = = P 4.5 3.31% F GDP growth rate = P = P P 4.0 2.94% J Labor cost = = P = P = 4.0 2.94% E Foreign Direct Investment = P = P = 3.5 2.57% K Labor Regulations = P = = P 3.5 2.57% I Inflation = = P 2.0 1.47% N Population growth = 0.5 0.37% Ranking: G>A>L=M=P>C>Q>D>H=O>B>F=J>E=K>I>N * Percentage of contribution (importance) of each attribute 139

4. The Analytical Hierarchical Process The Analytical Hierarchical Process developed [11] is a hierarchy based on pairwise comparisons of two types of elements: attributes and alternatives. The principal objective of the process is to rank the alternatives using weights on the attributes and sub-attributes that indicate the strength of each. This is a technique that has been improved over time. The AHP was improved principally by Saaty [8,10-15]. The mayor contributions of the author are in the fields of Network, dependence and feedback Systems. Table 4 below presents the research of the authors for improvement of the technique through the years, but the present study use principally the AHP process that was explained by Canada et. al. [2] follows: Construction of a decision hierarchy by identifying the elements and decision alternatives, determination of the relative importance of attributes and sub-attributes (if any), determination of the relative standing (weight) of each alternative with respect to each next-higher level of attribute or sub-attribute, determination of the indicator(s) of consistency in making pairwise comparisons and determination of the overall priority weight (score) of each alternative. Country Selection Business Law Active Law Bureaucra cy Financial Security FDI GDP grow rate Goverment stab. Tax Rates Inflation Labor Cost Labor Regulati. Lack Corrupti. Lack of Crime Population growth PPP Risks Time to start Chile Brazil Mexico Colombia Peru Figure 1. Model for the AHP hierarchy process Table 4. Evolution of the technique Author Year Title Saaty, T.L. 1980 Interaction and Impacts in Hierarchical Systems Aczel, J. and Saaty 1983 Procedures for synthesising ratio judgments Aczel, J. and Alsina 1986 On synthesis of judgments Buede, D. and Maxwell, D.T. 1995 Rank disagreement: a comparison of multi-criteria methodologies Saaty, T.L. and Vargas, L.G 2003 The Possibility of Group Choice: Pairwise Comparisons and Merging Functions Saaty, T.L. and Ozdemir, M. 2005 The Encyclicon Saaty, T.L. 2005 Theory and Applications of the Analytic Network Process Saaty, T.L. 2006 Fundamentals of the Analytic Hierarchy Process Saaty, T.L. and Tran, L.T. 2007 On the invalidity of fuzzifying numerical judgments in the Analytic Hierarchy Process Saaty, T.L. and Peniwati, K 2008 Group Decision Making 4.1 Assumptions There are many types of investors depending on different factors such as nationality, language, culture, distance to the host country and many others. The assumptions are that the analysts are consistent in their decisions and the evaluation is subjective. Other investors could give different values according to their perception, analysis or preferences and change the ranking. We have to make generalizations of the data that serves as reference for other person based on average or common factors. Ranks for different factors differ among themselves. We have changed the ranks of the factors that appear to be common to all ranks, so that they can be compared between each other; the assumption is that all factors are moving in the same range for easy comparisons. An implicit assumption that affects factor rank is that the factors can be affected by each other; this is that the factors cannot be completely independent of each other. For example labor cost is affected by inflation, or the lack of crime affects the risk conditions of the country, but for the study purpose of this study we assume that are independent of each other. 5. Results The final results obtained by the AHP process are given in form of a ranking of evaluation of the alternatives. This is a model that gives to the investor a useful tool for making a decision about the country for investment; the model 140

developed is useful for any kind of entrepreneur, investor or company executive, but the aim of the ranking is not to take an optimal decision, they try only to give a rational base of reference for the decision, that can change with the perception of the analyst, the country or with some factors that are changing constantly with time (political, cultural or social phenomenon). Consistency ratio is the value that expresses the internal consistency of the judgments that have been entered in the model [2]. All results in the model are consistent with CR with a value of 0.09738 which is acceptable (a value < 0.1 are acceptable [11]). Table 5 presents the ranking of the attributes. The results show that the principal factor in the selection are Government Stability with a 14.4% followed by Business law, Risk conditions and Lack of Crime with values between 11.3 and 11.8% as the four most important factors. The other weights are showing below. Figure 2 shows the ranking of countries. The first option is Chile with a 34.37% of importance; this is a logical result because Chile is the strongest economy right now in Latin America given its last reforms to the economic and political system. Chile is the most attractive country by the World Forum Economics. The second place is for Mexico with a 21.36%, who receives mayor part of the FDI compared to other Latin Countries; Mexico has a competitive advance given its important relation with the US market. The third place is for Brazil with a 16.97%, the fourth for Peru with a 14.16% and the fifth for Colombia with a 13.14% of the importance in the ranking. The difference between the last three countries is in the order of a 1.02% and 2.8%. There is a close decision, a little change on the variables can make that the order of alternatives change in these results, the analyst has to consider that limitation for further analysis. Table 5. Weightings for attributes Attribute Weight Attribute Weight Government Stability 0.144 Purchasing Power Parity 0.039 Business Law 0.118 Active Rate 0.025 Risks Conditions 0.115 GDP growth rate 0.023 Lack of Crime 0.113 Labor cost 0.019 Lack of Corruption 0.102 Labor Regulations 0.016 Bureaucracy 0.080 Foreign Direct Investment 0.016 Time to star business 0.069 Inflation 0.013 Financial Security 0.054 Population growth 0.009 Tax Rates 0.044 Ranking of Alternatives Colombia Counrties Peru Brasil Mexico Chile 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00% 35.00% Percent (%) Chile Mexico Brasil Peru Colombia Ranking of Alternatives 34.37% 21.36% 16.97% 14.16% 13.14% Figure 2. Ranking of Alternatives 6. Conclusions We developed a model for the Analytical Hierarchical Analysis of the factors that influence the selection of a host country. The model was shown to be consistent with the data collected and preferences evaluated. A ranking of Alternatives for investment and factors was presented as a tool that helps Company executives, entrepreneurs and investors to make a more rational choice. Other authors establish that a minimum market size below a country can be eliminated [3]. We propose a modification of the factor size of the country because the factor growth is more important than size, which is not a significant argument to eliminate a country and can be the case for other criteria. 141

6.1 Practical Contributions of the Technique We first establish two important matters. First, we see in the actual economic environment a remarkable interest of investors to move their investment to more advantageous (profitable) economies. We call that movement as capital flow and it is strongly correlated with the FDI. Second, the use of AHP is helpful to analyze complex problems with multiple attributes to aid decision making. The problem of country selection involves many factors in which the investor (as a decision-maker) have to choose an alternative that represents the interest of their stockholders. The contributions of this research are basically the use of the technique in a multiattribute decision problem that are practically unexplored under AHP process, and that is concerned to an actual economic problem. The process gives the investor a different point of view of this matter for making decision in a real world situation. 6.2 Limitations of the Technique One limitation of this technique is that it is based on personal judgment (subjective), so if two people are evaluating the same problem they could reach two different solutions. This limits the application up to the evaluator. The AHP's ranking of attributes is subjective and therefore depends on the evaluations that have been made previously. Another limitation is the confidence of the sources used to gather data. Not always is possible to access information sources (garbage in, garbage out). Finally, the calculations are made based on mathematical operations with matrixes, but there is no statistical theory that allows us to endorse the results obtained under this technique. Acknowledgements We would like to thank Alexandra Medina-Borja, Ph.D. for her support in the realization of this research as well as her commitment to us as her students from the graduate class of ININ 6030 and for the subsequent revisions. We also extend our thanks to Ernesto Aponte. References 1. Bank, World. (2007). World Investment Report 2007: Transnational Corporations, Extractive Industries and Development. New York and Geneve: United Nations. 2. Canada Jhon R., William G. Sullivan, Dennos J. Kulonda and Jhon A. White. (2005) Capital Investment Analysis for engineering and management. 3rd Ed. Prentice-Hall. P 545. 3. Gina Beim, Moren Lévesque (2006). Country Selection for New Business Venturing: A Multiple Criteria Decision Analysis. Long Range Planning, Volume 39, Issue 3, Pages 265-293 4. José I. Galán, Javier González-Benito (2006). Distinctive determinant factors of Spanish foreign direct investment in Latin America. Journal of World Business, Volume 41, Issue 2, Pages 171-189 5. Len J. Treviño, Franklin G. Mixon Jr. (2004). Strategic factors affecting foreign direct investment decisions by multi-national enterprises in Latin America. Journal of World Business, Volume 39, Issue 3, Pages 233-243. 6. Michael Frenkel, Katja Funke, Georg Stadtmann (2004). A panel analysis of bilateral FDI flows to emerging economies. Economic Systems, Volume 28, Issue 3, Pages 281-300. 7. Mia, I., Estrada, J., & Geiger, T. (2007). Comparative analysis of the attractiveness for private investment in national infrastructure. Santiago: World Economic Forum. (in spanish) 8. Saaty, T.L. (2006) Fundamentals of the Analytic Hierarchy Process, RWS Publications, Pittsburgh, PA. 9. Nijkamp and A. van Delft (2002), Multi-Criteria Analysis and Regional Decision Making, from the series Studies in Applied Regional Science vol. 8. 10. Saaty, T.L. (2005) Theory and Applications of the Analytic Network Process, RWS Publications, Pittsburgh, PA. 11. Saaty, T.L. (1980) Interaction and Impacts in Hierarchical Systems, Academic Press, Inc., pp.29 77. 12. Saaty, T.L. (1980), The Analytic Hierarchy Process. New Cork. McGraw-Hill Book Company. 13. Saaty, T.L. and Ozdemir, M. (2005) The Encyclicon, RWS Publications, Pittsburgh, PA. 14. Saaty, T.L. and Peniwati, K. (2008) Group Decision Making, RWS Publications, Pittsburgh, PA. 15. Saaty, T.L. and Tran, L.T. (2007) On the invalidity of fuzzifying numerical judgments in the Analytic Hierarchy Process, Mathematical and Computer Modeling. 16. Saaty, T.L. and Vargas, L.G. (2003) The Possibility of Group Choice: Pairwise Comparisons and Merging Functions. 17. World Bank (2008). Foreign Investment Advisory Service. - World Bank, 2008. http://www.fias.net/. Retrieved May 01, 2008. 142