國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Although, the Brookings Institute is a policy-focused organization, their findings have lent a new perspective to our understanding of the current theoretical paradigm of local FDI attraction. Their findings, do not establish a new phenomenon, but rather, are merely an initial look into what has already been taking place in local governments.
Local, state and county level policy makers have been allocating funds to diplomatic initiatives in the hopes to build relationships for the purpose of economic exchange (specifically FDI). These “Relationship building” activities, however, are not characterized by direct economic impact, things like; cultural and artistic exchange, incubation of study abroad programs, support for ethnic organizations and cultural events, and sister city exchanges.
The budgetary allocations set aside to these local diplomatic agencies are supplemental to those already set aside for economic development, establishing competitive FDI attraction incentives, labor and workforce development, the building and maintenance of infrastructure, constructing a competitive investment promotion strategy and other proven avenues to international economic development. This indicates a previously established recognition of “relationship building” as an influential determinant of FDI attraction.
2.2 Framework: Dunning’s OLI Eclectic Theory
The framework for our study will follow Dunning’s OLI model as it refers to (L) location specific advantages. Dunning’s model categorizes the (L) variables into four respective areas which explain a firm’s choice through the utility of profit maximization.
In essence, a firm will seek discounted profits through a range of considerations related to
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
the location choice of an FDI project. Dunning separates location choice by a firm into the following areas:
1) Resource Seeking: Resource seeking behavior refers to MNEs making location choices based on the availability of; low rent or land cost; workforce availability (either skilled or unskilled depending on the industry); and the local cost of raw materials and other factors of production.
2) Market Seeking: Market seeking behavior refers to MNE’s making location choices based on; local market size or growth; preferences of the consumer for individual products; market structure; distance from the host country or region; and access to regional or global markets.
3) Efficiency Seeking: Efficiency seeking behavior refers to MNE’s making location choices based on; cost of resources and capabilities (i.e. electric power, land cost, rent, etc.); transportation and communication costs; industrial agglomeration factors (i.e. relative existence of R&D, level of employment in a certain industry, etc.); and whether the local government entity is a cosignatory of regional trade agreements.
4) Strategic Asset Seeking: Strategic asset seeking behavior refers to MNE’s making location choices based on; technical and managerial assets; infrastructure (i.e.
roads, ports, telecommunications, etc.); entrepreneurial or educational know-how;
regulations on industrial competition.17
17 Daria Zvirgzde & Daniel Schiller & Javier Revilla Diez, 2013. "Location choices of multinational companies in Ukraine," ERSA conference papers ersa13p219, European Regional Science Association.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
2.2.1 Variables: Reformatting the OLI and IPA Theories
Our research program is meant to incorporate local diplomacy and “relationship building” into the leading theories on FDI. Our study will make use of those variables deemed significant through the OLI and IPA theories in order to juxtapose our new variables, associated with the practice of relationship building. As we are narrowing the scope to account only for local government initiatives, those variables under the control of the national government, will be excluded or reformatted in such a way that they are made to account for the difference in domestic FDI dispersion. Below is a graphic of the variables to be included in this study.
Figure 1: OLI and IPA Frameworks Accounting for “Relationship Building” Variables
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
In order to evaluate the relevance of each independent variable we will need to separate our study into two sections, quantitative and qualitative. This will allow us to first determine any possible relevance of a “relationship” variable before conducting in depth qualitative analysis. Ideally, the quantitative section will guide the research scope while a follow up with qualitative investigation could solidify theoretical relevance. For example, if “spending on diplomatic efforts” appears to be significant in the quantitative results, we will investigate further with local diplomats to determine specific policy objectives or areas of priority. Finally we would follow up with a survey of foreign companies in that city, to ask specifically about that policy. For this purpose, the quantitative section will precede the qualitative analysis, and both will remain separate with independent conclusions. In chapter 5, we will combine lessons from both testing methods to evaluate the position of a “relationship” variable in future theory.
2.2.2 Quantitative Testing: Location Choice Trends in Taiwan
First we will carry out a conditional logistic regression as popularized in FDI literature by McFadden (1974).18 This model prioritizes the relative profit maximization value sought by a firm, which is provided to the firm by choosing either location-j or location-k. Below is the function for profit by firm-i which is derived by investing in location-j. The value-z indicates the specific qualities of a particular city (in our case;
Taipei, New Taipei City, Taichung, or Kaohsiung).
(1)
18 McFadden, D. L. Conditional logit analysis of qualitative choice behavior. In
Zarembka, P., editor, Frontiers in econometrics, pages 105{142. Academic Press, 1974.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
However, if the firm decides to invest in location-k. The function changes to read as the following;
(2)
Given the above functions for firm choice in locations-j and -k, then the following function assigns a Y=1 value for the firms choice to invest in location-k instead of
location-j;
(3)
In the conditional logistic model for determining which characteristics-z of each location are relevant, we code the dependent variable (investor choice) to be a case value of “1” for the location that was chosen for an investment, while coding the control value of “0” for all other locations in the group where the firm did not choose to invest. Given that our Y-i variable is random and indicates investor choice, then we can assume that the following function allows us to assess the maximum likelihood of investor choice based on a set of independent, location specific variables;19
(4)
Since this project will not be analyzing firm specific variables, we will not make use of dummy variables to account for qualities of the investing MNE, but rather will conduct an exploratory, macro level data analysis, in order to determine the potential
19 Deichmann, Karidis, and Sayek. (2003). “Foreign Direct Investment in Turkey:
regional determinants”. Applied Economics. 2003, 35, 1767-1778.
‧
degree of foreign ownership. Additionally, within our quantitative methodology we take on a level of limitation when using the conditional-logit model for only 4 choices for investment in Taiwan. This concept, known as the Independence of Irrelevant Alternatives (IIA), refers to the absence of additional choices that may have or may not have been considered by decision makers (i.e. Tokyo, Bangkok, Guangzhou or other industrialized cities in the region), and is commonly accounted for by distance and relative difference between choice options. However, in our study the nearness and similarity of choice in Taiwan provides additional obstacles. Our study can account for this limitation by assuming that those investors included in this study were already planning on operating in Taiwan and were therefore choosing between suitable locations on the island. Further, the primary test variable of relationship building, is only better tested by comparing local government variables once a company has chosen to invest in a country or region.20Our selection of indicators for each independent variable deemed relevant by the literature were chosen on the basis of access to information and the metric’s ability to be representative of the location specific quality. Each case provides us with a glimpse at the relative capacity for profit maximization in the four cities chosen for our study, Taipei, New Taipei City, Taichung and Kaohsiung.
20 Mukim and Nunnenkamp. (2010). “The Location Choices of Foreign Investors: A District-level Analysis in India”. Kiel Working Papers. Kiel Institute for the World Economy. No. 1628. June 2010.
Arrow, Kenneth Joseph (1963). Social Choice and Individual Values (2nd ed.). Wiley.
‧
Table 2 gives an overview of the data collected for each location specific (or city) variable and also includes a proxy variable for the IPA method as well as our own contribution, in the form of relationship building and diplomatic efforts. 21
Table 2: Location Specific Variables
Variable Explanation Exp. Sign Min. Max. Mean
Investment Dependent variable where denoting choice to
invest 0 1 .25
GOVQ Air quality as a proxy for good governance,
total (suspended particulates) − 38 172 78.13
MKTS Average family income (100,000s NTD) + 10.32 16.98 12.89
INFRA Govt. spending on transportation (billions NTD)
+ .75 35.5 11.34
AVGC Average family expenditure on rent and
utilities (1,000s NTD) − 14.38 30.59 20.32
AGGL Number of workers employed in the relevant
industry (1,000s of workers) + 0 197.01 37.95
IPASp Govt. spending on investment promotion + 1.79 25.38 6.04
DIPSp Govt. spending on diplomacy (millions NTD) + 10.33 25.72 17.01
SISCi Investor headquartered in a sister city + 0 1 .03
SISCo Number of sister cities in the relevant source country
+ 0 17 5.19
21 Joyez, Charlie. (2015). “Location choices and foreign direct investment motives of heterogeneous firms.” LEDa - DIAL, Paris-Dauphine University.
‧
2.2.3 Qualitative Testing: Surveying Taiwan’s MNE Managers
The qualitative chapter will be broken into two sections. First, local government diplomats will be interviewed in order to determine policy commonality and difference among cities in in Taiwan. A semi-structured interview format will be used in person in either Chinese or English. Interviews will not be recorded and will not exceed 2 hours in length. As the purpose for the interviews is exploratory, topics may vary depending on the mood and direction of conversation, while maintaining a common vein which can properly measure one city to the next. Interviews are expected to uncover variables which can be tested to indicate success of common and specific practices alike. Therefore, the organization of interview content will fall into two categories; “general” and “specific”.
After the content of the interviews has been analyzed, our study will use that content to inform targeted survey questions, designed to gauge recognition of policies by investing managers. We will use an established 7-point, Likert scale survey to measure attitudes of relative importance from investing companies. The precedent for using such a method can be found in similar research projects concerning FDI determinants; especially Luo (2001, 2007) and Liu (2009).22 Combining interviews with surveys in this way affords our research an informed perspective when formulating our survey. As such, we will be able to see, precisely, the attitudes of companies towards municipal diplomatic policy. In other words, was the decision to invest by the FOE influenced by the specific
22 Liu, Yulong. (2009). “Factors Determining Locations Choice of Foreign Direct Investment in China: A Perspective from an Inland Province”, Massey University, New Zealand.
Luo, Y. (2001). Determinants of Entry in an Emerging Economy: A Multilevel Approach. Journal of Management Studies, 38, pp. 443-472.
Luo, Y. (2007). Are joint venture partners more opportunistic in a more volatile environment? Strategic Management Journal, 28(1), pp. 39-61.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
policies uncovered in the interviews? Did the work of local diplomats in Taiwan make any difference?