A Facility Location Model for
Multinational Enterprises Considering
Corporate Social Responsibility Effect
Lingzi Tian S3293289
Email: ayakotian@gmail.com
First Supervisor: dr. X. (Stuart, Xiang) Zhu Second Supervisor: C. (Chengyong) Xiao
Co-assessor: dr. J. (Jasper) Veldman
MSc Supply Chain Management &
MSc Technology Operations Management
Abstract
In this study, a model is built and examined to help multinational enterprises with new overseas subsidiary location decisions by considering the effect of Corporate Social Responsibility. The study is twofold in originality. First, we improve one economy geography model to serve the need of supply chain and technology operations management. We achieve the goal by adding in a new influencing factor which reflects the uncertainty effect of Corporate Social Responsibility on the model and applying a Conditional Value-at-Risk tool. Second, we examine the new model and prove the importance of the added factor and how it influences the outcome of the model by using empirical data and simulation study. Managerially, this study suggests multinational enterprises to choose overseas subsidiaries require less input in Corporate Social Responsibility during their production process there.
1. Introduction
Corporate Social Responsibility (CSR) has recently been faced with a paradox. On the one hand, more multinational enterprises (MNEs) have the intention to build more positive images on CSR since an increase in the level of CSR can improve the performance and profitability of MNEs (Husted & Allen, 2006). On the other hand, they cannot completely get rid of corporate social irresponsible activities (CSiR) if those CSiR are included in their vital production process (Strike et al., 2006).
This paper aims at building up such a model that can quantify the influence of an outstanding related CSR factor on the performance of subsidiaries. This model can then be used by MNEs as a guide on location selection. The biggest challenges in building this model will be finding a proper way of involving this CSR factor that is found to be relevant and representative in business decisions in current literature and clarifying its relationship with other classical relevant factors. These challenges will be dealt with in the later parts of this paper. Additionally, this model also provides insights on improving the strategic level and multidimensionality of current subsidiary location decisions. Finally, this enhances the importance of subsidiaries for MNEs since CSR is generally regarded as a strategic management factor (Rugman et al., 2001).
Based on the gap in the subsidiary location model building, the research question of this paper will be:
What would be a good subsidiary location choice for an MNE when taking the influence of CSR into consideration?
limitation statement. In general, this paper contributes to finding a numerical reference to the location decision of MNE subsidiaries considering CSR.
2. Theoretical Background
According to the research aim of this paper, three theoretical aspects will be combined to form a new mathematical model on the optimal subsidiary location choice for a company considering the impact of CSR. The theoretical background will then be arranged as follows. First, theories on the classical location models will be reviewed and summarized to provide a basic skeleton of the model. Second, conditional value-at-risk models and related literature will be reviewed to add up for the classical location model from the perspective of uncertainty contained in the model. Finally, literature regarding CSR will be reviewed to clarify the exact CSR aspect that will be used in this model. All of them contribute to a complete, reasonable and comprehensive literature background of the target model. The underlying logic is as follows. First, there are multiple existing models that contribute to the location decision from different aspects and one proper base model needs to be chosen and reviewed through literature. After that, when the new factor (i.e. CSR) is added into one chosen existing model on subsidiary location, theoretical support on what to be included and how to properly include it should be determined from a qualitative review. Finally, since the proposed new model is predictive in decision making, the risk of uncertainty regarding real-life management needs to be considered, which makes the Value-at-Risk model review necessary.
2.1 Location Model
companies. There have been different types of location models with different targets since the development of the famous p-median model of Daskin in 1997 (Drezner et al., 2001; Revelle et al., 2008; Sridharan, 1995). Later, these models become more realistic by relaxing the boundaries and providing insights from different aspects (Hinojosa et al., 2000). Besides, facility location choices have gradually become strategic by being built no longer alone but in the scope of a whole supply chain both horizontally and vertically (Altiparmak, 2006).
To serve the need of choosing a proper subsidiary for an MNE considering CSR, the model raised by Kheder and Zugravu in 2012 is chosen. Compared to other facility location models that are very detailed within one geographic zone, this model focuses on the profit a subsidiary expects to get considering macro-economic drivers (i.e., capital and labor) which directly show the overall productivity and market potential. Therefore, the model can reflect the characteristics of an economy. This model also provides a background that sketches out both the main tangible and intangible factors. Meanwhile, literature show that the environmental regulation, which is highly related to CSR for MNEs (Russo & Perrini, 2010), can be included in this model. Besides, there are unstudied residuals that play significant roles in the base model, which provides possibility on further complication (Kheder & Zugravu, 2012). Considering the target model in this paper, this residual part will be extended to be the uncertain effects that the chosen CSR factor can have on location decisions. By taking this into account, the base model will be clearer in showing how the CSR factor other than the classical capital and labor is affecting the facility location decision. However, the challenge is to predict the pattern of the location decision making when taking the CSR uncertainty into consideration, which complicates the model extension process and will be dealt with in the later parts of the thesis.
2.2 Conditional Value-at-risk Model
Conditional Value-at-risk (CVaR). Being more comprehensive, CVaR considers both reward and risk and has better computational characteristics as well as stochasticity (Chen et al., 2009), which makes it more widely applied in studies of operations management to look for a more realistic model building process. Based on the studies towards the classical Newsvendor model using CVaR, it’s found that business decisions would be influenced by the uncertainty level of the producer faces at market (Ahmed et al., 2007). Meanwhile, it’s found that there are relatively universal robust solutions for optimization problems in operations management when considering risk (Ben-Tal et al., 2013), which further strengthens the necessity for this paper to take risk into account since the generality of the factor can be ensured.
When considering facility location, it has long been assumed that some universally accepted index generated from tangible world trading data can represent the market potential of an economy (Head & Mayer, 2006), which therefore ignores or leaves alone the uncertainty led by intangible issues about the market potential which could influence the location choice. However, it is found that consumer’s awareness of CSR and their level of recognition and investment on products from environmental-friendly producers in local markets can change over time and have an influence on the market share of producers (Lii &Lee, 2012). Meanwhile, the market potential can also be influenced by the local market competitiveness, technology patent, policy and other tangible and intangible characterized factors which are all related to CSR (Rugman et al., 2011). What worth mentioning is that this influence is easy to be found existed but hard to measure because of the contingency occurred from the subjective change of consumers’ view towards CSR (Grappi et al., 2013). This adds to the value of putting CVaR of CSR’s influence on market potential in this model to elaborate it since CVaR method helps to measure this kind of uncertainty that has not yet been considered in the existing studies on the subsidiary location models.
2.3 Social Corporate Responsibility
make sense in Operations Management process. Although firms of different sizes are critically different in behaviors and decision-making process (Welsh, 1981) and this paper focuses on the location decision of MNEs (implicitly medium or large size), it’s found that some CSR factors that characterize them are the same (Russo & Perrini, 2010). Common factors include environment regulation, market competition and business culture etc. (Anderson et al., 2009). Among these, environment regulation gets stressed in the literature (Cruz & Wakolbinger, 2008) and will be taken as the main influential testing factor in the model. According to the base model of profitability of MNES that is used in this thesis (Kheder & Zugravu, 2012), the environment regulation will be considered together with capital and labor. However, the market competition overlaps with the market potential factor to a considerable level (Rugman et al, 2011) and has already been measured in Kheder and Zugrazu’s model, which makes it unnecessary to be focused again. On the other hand, the business culture is hard to quantify and stays constant for all subsidiary locations regarding one focal MNE. Therefore, the specific CSR factor included in this proposed model will be the environmental regulation while the other two will be left alone. The following part of this section will include more related literature review to provide a closer look.
Environment Factor
provides a more precise and comprehensive look at the benefit an MNE can get from building a potential new subsidiary based on the current literature.
3. Methodology
This paper aims at finding out the optimal solution for the subsidiary location of an MNE with consideration of CSR. Considering the mainstream research methods that are used in related studies and the basic quantitative need of the research question, a mathematical model will be built to analyze the issue. As is stated above, the conditional Value-at-Risk model and classical Location models will serve as the base with CSR factors added in. To build this model, we refer to two previously developed models, which are the profitability calculation model for an economy developed by Kheder and Zugravu in 2012 and the market share calculation formula developed by Becchetti et al. in 2013. The details of the model will be introduced in the following.
3.1 The basic model
Preliminaries and assumptions
quantified. This paper aims at completing this step. Finally, the share of MNE’s investment on the local environmental regulation during their production process will be a response to the emphasis local customers have on the CSR level of the MNE (Dawkins and Lewis, 2003).
Based on these assumptions, Kheder and Zugravu built a profitability model in 2012 (referred to as K&Z model in the following part of the paper). This model improves the classical economy geography model of Head and Mayer in 2004 (referred to as H&M model in the following part of the paper). K&Z model adds the factor of environmental regulation into the marginal cost section of H&M model to measure its influence on profitability together with the classical capital and labor factors. However, K&Z model does not consider the influence of environmental regulation on the market potential for a newly entered MNE, which has been proved to have connections (Luo and Bhattacharya, 2006). This connection is argued to be reflected through the market share a company can get. In this sense, the market share calculation method considering socially responsible choices can be added into K&Z model to provide more insights (Becchetti et al., 2013).
regulation influences.
The following model is built in several steps. First, the K&Z model serves as the skeleton of an MNE’s profitability in one potential subsidiary with uncertainty. Second, market share with uncertain CSR influence level is divided by market share with base CSR influence level to generate an influence index with uncertainty. Then, the index is applied to the K&Z model on the market potential part to reflect its uncertainty and influence on profitability thereafter. Next, the CVaR tool is applied by considering the difference between H&M model (i.e., profitability considering the basic profitability without any effect from CSR) and the new method of profitability calculation built in this paper. Finally, all potential subsidiaries are calculated in the model and the one with the highest profitability is chosen as the optimal new overseas subsidiary for the MNE.
Decision variables
t 䇅 ⺁ 䇅 䁓 香 䁓 䇅 䁓 ݁
Parameters
I set of subsidiary countries, indexed by i
financial estimation of the profit generated by the subsidiaries without considering CSR factors
E expectation operator
price of the MNE for final goods at subsidiary i price of the competitors for final goods at subsidiary i
σ elasticity of substitution between production varieties of the subsidiary subjective consumers’ preference impact of CSR factors on consumer utility in monetary terms in subsidiary i ( < less than proportional
proportional > more than proportional)
w cost of labor cost of capital
cost of environment regulation (represents the CSR factor) α share of labor input in the firm’s production process β share of capital input in the firm’s production process
θ share of environment regulation input in the firm’s production process A total factor productivity (TFP)
sum of external factors that affect marginal cost and their applicability degree of risk aversion of the company in profit expectation (regarded neutral in the empirical test period)
⺁⺁ conditional value at risk regarding firm profit with certain degree of uncertainty
φ influential index for the subsidiary with uncertain level of influence of CSR in the local market (see Becchetti et al, 2013 for details)
M Krugman Market Potential (See Head and Mayer, 2004 for details)
Note that: + + θ for any θ ( ) for any ⺁䇅 ݑ > ݄ 香⺁݁ 䁓 䁓 t ⺁ ⺁tt 香 香 香⺁݁ 䁓 香䁓 ݄ 香⺁݁ 䁓 䁓 t ⺁ ⺁tt 香 香 香⺁݁ 䁓 ⺁⺁݁݁ ⺁䁓 ⺁ ⺁݊ t݁ 香 ݁ ݊ ݁ 䁓t 䁓 䁓t ⺁ t⺁ 䁓 䁓香 䁓 䇅 䁓香 ݁ h 䇅香 t䁓 ⺁ 䇅 䁓香
⺁⺁ is the conditional value at risk for the profitability that a firm could get from the choice of a subsidiary i.
is the profitability of a firm located in another country i and has trade with other countries.
φ is an index of the percentage increased on Krugman Market Potential by considering production variety of the subsidiary under increasing returns to scale in an environment of monopolistic competition in country i influenced by the CSR factor (i.e. environmental regulation) with an uncertainty level X which is positive and follows a Uniform Distribution. There are logical restrictions put on φ and (see Becchetti et al, 2013 for details)
+
Explanation of formulas:
: This formula is derived from the H&M model (Head and Mayer, 2004) which means a basic profitability expected by a potential subsidiary location. This formula only considers Capital and Labor without CSR involved. The other factors including TFP, market potential and elasticity of substitution. This formula contains no uncertainty and can be calculated for each potential subsidiary based on and in the given data for test. The reasons for keeping the ratio of t unchanged is that the relationship of capital and labor should remain unchanged either CSR is in or out.
index reflects to what extent the uncertainty increases or decreases the market potential in the profitability formula.
：This formula is generated based on the formula of K&Z model (Kheder and Zugravu, 2012). However, it adds the uncertainty index (i.e. φ ) to the K&Z model by multiplying the index to the Krugman Market Potential while other parts remain the same. Through this multiplication, the revised market potential part of the
formula contains the effect of uncertainty in CSR.
⺁⺁ : This formula is modified based on the general definition of CVaR (Rockafellar and Uryasev, 2000) regarding a function (i.e., in this study), which is ⺁⺁ max⺁ + min } This definition asks to choose a maximized out of its set of possible values considering its uncertainty. Therefore, the general definition is modified to form the current ⺁⺁ because there is no set of baseline profit (i.e. ) for one potential subsidiary. The uncertainty in this study hides in the profitability function considering environmental regulation (i.e.
).
This model aims at measuring the profitability an MNE can obtain from opening a new subsidiary overseas more comprehensively and realistically with a combination of three developed and proved methods. This strongly eliminates the possibility of failure. However, it’s realized that the model testing process is still a must and the following section will show the results and use of the model.
4. Model Testing and Results
to the K&Z model which sets as the base of the new model building? Third, how sensitive the new model is towards the change of the important factors serving as the aim of the study? Finally, is the new model capable of performing well under various sources of data and assumptions?
To answer these questions in good logic, the experiments will be designed as follows. First, the new model contains a decision-making nature. This makes interesting to dig out the relationship between the two crucial CSR related parameters (i.e. θ and t) for the optimal choices with largest CVaR profitability (i.e. ⺁⺁ ). Therefore, the pattern of these parameters in the chosen subsidiaries of the new model will be discovered through numerical experiments. Second, since this new model is built on the basic K&Z model with more complexity and comprehensiveness, an experiment that can make comparison between them towards their characteristics, effectiveness and insights should be conducted. The comparison will target at calculated profitability and important parameters of the optimal chosen subsidiaries. With them, the advantages and possible disadvantages of the new model can be driven. Third, a sensitivity test will be done to have a check on three CSR related important parameters and analysis will be done based on its results. Finally, another check of robustness will take place to ensure the universality of the model when alternatives of data are possible.
4.1 Empirical Data and Estimation Method
More than 1000 sets of results are generated during the process for analysis.
To make good use of the data, the parameters will be changed in a proposed order to generate a pool of results to test he model. The test will be as the following. First, there are several parameters in the model that are set to be constant as its mean value. Logically, it is because that they are less related to the study focus of this paper, which is the CSR’s effect on subsidiary location choice. There are also other reasons for doing so. First, the total factor productivity (A ) will be regarded as constant because the change of it has been insignificant during the period of data collection (2% on average). Meanwhile, elasticity between substitution of varieties ( ) reflects how an MNE can improve the production performance by adjusting its input on capital and labor but this has been restricted by the base model author in the basic assumption ( >1) and its range is therefore vey much limited to a Cobb-Douglas function (for more detail please refer to Brainard, 1993). The risk-averse level of the decision maker (i.e., the MNE) is regarded as neutral which is 0.5 numerically. Moreover, when coming to test the φ function which represents the uncertain influence of the CSR environmental regulation factor, the involved final price is settled as the selling prices of one 330ML can cola from Coca-Cola ( ) and Pepsi ( ) company in 2005 respectively. The reasons are as following. On the one hand, they are two of the largest MNEs in the world which both seek to expand their production map internationally and they mostly form a duopoly in markets which makes it reasonably to conduct the φ formula based on its original assumptions (Becchetti et al., 2014). On the other hand, they lie in the food and beverage industry which suits other empirical data collected from the basic profitability formula (Kheder and Zugravu, 2012).
parameters are grouped together and the the CVaR values of them are calculated using the new model. Then one set of parameters out of the ten that generates the highest profitability (i.e., largest CVaR) will be picked out. By doing so repeatedly, 1000 chosen optimal parameter sets are collected. Among them, sets with all cost of labor, capital and environmental regulation that are at their boundaries (i.e. either minimum or maximum) will be kicked out because it is unrealistic to have a potential geographical zone with all extreme cost parameters. Then, with the set of CVaR profitability, the profitability calculated through K&Z model and all parameters that are used during the empirical test, the comparison will be made. The results are shown as below.
4.2 Simulation Analysis Results
an average θ of 0.351. The percentage of change is 6.84% thereafter.On the other hand, the downside shift of t for chosen and unchosen subsidiaries is not significant with a mean value of 53.002 and 53.047 respectively, changing only 0.08%. Therefore, it comes to the first conclusion that optimal subsidiary locations tend to require less share of input regarding environmental regulation during production process. However, the trend of cost led by environmental regulation in the subsidiary remains insignificant. .
Figure 1
in this figure lead to the second conclusion that the profitability of a potential overseas subsidiary location will be lower when environmental regulation influences both market potential and the marginal cost of a certain subsidiary as assumed in the new model. This can be explained by the fact that the uncertainty of environmental regulation’s effect on the market potential, which can negatively influence it, lowers the expectation of profitability that market potential can contribute to the MNE if a new subsidiary is built there. In other words, unknown future adds risk to the final decision. Therefore, the new model will provide more conservative decisions. To be more precise, the profitability of all chosen optimal subsidiaries in the simulation calculated by applying both models are shown in Figure 3.
Figure 2
Figure 3
requirements in different potential subsidiaries (i.e. t and θ), the new model suggests MNEs to choose subsidiaries that require higher share of environmental regulation input during production process and has lower local cost of environmental regulation, comparing to what the K&Z model is likely to suggest.
Figure 4
Figure 5
4.3 Sensitivity Test towards Important CSR Related Factors
checked for their sensitivity regarding the maximized CVaR of profitability.
The same method is used to test the sensitivity for three parameters. First, 1000 sets of variables indicating potential subsidiaries with different characteristics are generated to decide a base optimal CVaR profitability. Then, with all other variables unchanged, the 1000 sets are put into calculation again with a given set of , θ and M respectively. These three variables are changed by 10 percent both positively and negatively. Finally, a new set of maximized CVaR of profitability comes out regarding the changes of the three variables and the percentage of its change is displayed below. By taking the average of the change rate of all the 1000 sets of results, the following Figure 7 shows the sensitivity of , θ and M .
Figure 7
low with a percentage of only 1.2% and 2.6% respectively at smallest. Therefore, the focus of the paper, which is the influence of CSR on the profitability of subsidiaries can be proved to be of good importance. First, this sensitivity test proves that the consideration of the uncertain nature of market potential is valuable because θ has a significant influence on the profitability and M itself is not a sensitive parameter for profitability. This helps with describing the source of uncertainty, which is the uncertainty index containing θ , more precisely to get a more reliable result from the model. Second, between the two concerned variables in the previous analysis (i.e., θ and ), profitability is not sensitive to . On the one hand, it implies the may not play an important role in the profitability of a subsidiary location alone. On the other hand, the K&Z model proves the joint influence of θ ݁香 , which suggests that the joint influence might mainly come from θ , as is shown in the previous conclusions in this paper.
4.4 Robustness Check towards Parameters with Restrictions
Because that the model testing process is empirical and based on data generated from descriptive confidential data, some of the parameters are set with restrictions. This may have some negative influence on the universality of the results and conclusions generated from the model testing process. Therefore, several robustness checks are made in this part regarding these parameters in the new model. What worth mentioning is that most of the robustness checks of the data with restrictions have been finished in the original K&Z model thoroughly by changing the data pool in terms of the geographical zone, development level, focal industries and measurement method of different index involved of the potential subsidiaries (see Kheder and Zugravu, 2012 for details). Based on this and the fact that the descriptive data used in this thesis are the same as theirs, only the following restrictive parameters will be checked for their robustness to generate model results.
Table 1
Table 2
Regarding the risk preference level, 9 more models are simulated based on different levels of risk preferences and the same checking principle is applied. The results in Table 2 share the same testing logic with Table 1 and shows that in most cases (9 out of 10) the new model has stable outcome when the risk preference level of MNE ( ) changes. However, it also implies that under certain extreme occasions (i.e. when the decision maker is extremely risk averse and has an or less) the subsidiary location choice can be changed. Nevertheless, this is not regarded as an outstanding problem for the stability of the model because the probability is low.
regulation level itself (i.e. t) turns out to be insignificant individually. This indicates that a potential subsidiary with lower requirement of CSR production input will be more attractive for the MNEs. This finding is also supported by current literature as mentioned in the theoretical background.
On the other hand, if the new model is not analyzed alone but compared to the current K&Z model, there are some interesting findings as well. First, the K&Z model considers input factor and local environmental regulation factor together in the profitability calculation formula (i.e., θ ݁香 ) and proves the significance level of it. According to the testing results in this paper, the individual influence of is not working well and the main influence comes from θ . Besides, it is interesting to find that the new model has a lower profitability in general when considering the value of uncertainty. This supports the important hypothesis in this paper of the uncertain nature of CSR’s effect on subsidiary locations because the uncertainty index that contains the input share of CSR lowers the influence power of market potential which positively adds to profitability. This leads to the downside shift of profitability thereafter. Also, comparing with the K&Z model’s requirement, the new model requires an optimal subsidiary to have relatively higher CSR input share during the production process and lower local environmental regulation cost. The reason is also that K&Z model only measures the joint effect of the two aspects of CSR (i.e. θ ݁香 ) on marginal costs and puts less stress on their independent influence.
5. Conclusions and Limitations
In this study, the effect of one of the most outstanding CSR factors, Environmental Regulation, on the location decision model of overseas subsidiaries for MNEs has been modeled and tested. Several important conclusions can be drawn with practical implications in managerial field.
expectations about the setup of new overseas subsidiaries, MNEs will need to be more careful about their input distribution strategies and cost that might occur due to local environmental regulations. Second, the new model developed in this study suggests different optimal subsidiary location choices when comparing to previously used models to a considerable level. This is important managerially since no matter what the prediction process is, the decision makers only want a more reliable result anyway. By considering CSR’s effect more thoroughly, this new method for location decision is believed by the author to provide more insights. Namely, MNEs can try to find potential subsidiaries that allow the company to invest less in CSR if no other restrictions or comparable models are applied.
Although almost all the involved parameters are tested to stabilize the model successfully, there are limitations found in this study which may serve as drawbacks and need further improvement. On the one hand, the assumption for a duopolistic overseas market is applied for the consistency and convenience of model testing but it is not regarded as universal in real life problems. More types of markets regarding competition should be studied in further work. On the other hand, environmental regulation is taken as the representative of CSR’s effect to contribute to market potential for consistency but this is not complete. There might exist other CSR factors that has considerable influence and requires further study. Last of all, as is mentioned in the base model used in this study by Kheder and Zugravu in 2012, the dataset has its limitation by not considering the time effect.
6. Acknowledgement
Appendix: Data Summary
Table 3. Descriptive statistics for variables used in simulations.
Reference: confidential data provided by Dr. Natalia Zugrazu used in her research in 2012.
Table 4. Data sources
Table 5. Data for Figure 2
Table 7. Data for Figure 4
References:
1. Ahmed, S., Çakmak, U., & Shapiro, A. (2007). Coherent risk measures in inventory problems. European Journal of Operational Research, 182(1), 226-238.
2. Altiparmak, F., Gen, M., Lin, L., & Paksoy, T. (2006). A genetic algorithm approach for multi-objective optimization of supply chain networks. Computers & industrial
engineering, 51(1), 196-215.
3. Andersen, M., & Skjoett-Larsen, T. (2009). Corporate social responsibility in global supply chains. Supply chain management: an international journal, 14(2), 75-86.Husted, B. W., & Allen, D. B. (2006). Corporate social responsibility in the multinational enterprise: Strategic and institutional approaches. Journal of international business studies, 37(6), 838-849. 4. Battaglia, M., Bianchi, L., Frey, M., & Iraldo, F. (2010). An innovative model to promote
CSR among SMEs operating in industrial clusters: evidence from an EU project. Corporate social responsibility and environmental management, 17(3), 133-141. 5. Becchetti, L., Palestini, A., Solferino, N., & Tessitore, M. E. (2014). The socially
responsible choice in a duopolistic market: A dynamic model of “ethical product” differentiation. Economic Modelling, 43, 114-123.
6. Ben-Tal, A., Den Hertog, D., De Waegenaere, A., Melenberg, B., & Rennen, G. (2013). Robust solutions of optimization problems affected by uncertain probabilities. Management Science, 59(2), 341-357.
7. Brainard, S. L. (1993). A simple theory of multinational corporations and trade with a
trade-off between proximity and concentration (No. w4269). National Bureau of Economic
Research.
8. Carlton, D. W. (1983). The location and employment choices of new firms: An econometric model with discrete and continuous endogenous variables. The Review of Economics and
Statistics, 440-449.
9. Chen, Y., Xu, M., & Zhang, Z. G. (2009). A risk-averse newsvendor model under the CVaR criterion. Operations research, 57(4), 1040-1044.
production economics, 116(1), 61-74.
11. Dam, L., & Scholtens, B. (2008). Environmental regulation and MNEs location: Does CSR matter?. Ecological Economics, 67(1), 55-65.
12. Daskin, M. (1997). Network and discrete location: models, algorithms and applications. Journal of the Operational Research Society, 48(7), 763-764.
13. Dawkins, J., & Lewis, S. (2003). CSR in stakeholde expectations: And their implication for company strategy. Journal of Business Ethics, 44(2-3), 185-193.
14. Drezner, Z., & Hamacher, H. W. (Eds.). (2001). Facility location: applications and theory. Springer Science & Business Media.
15. Grappi, S., Romani, S., & Bagozzi, R. P. (2013). Consumer response to corporate irresponsible behavior: Moral emotions and virtues. Journal of business research, 66(10), 1814-1821.
16. Head, K., & Mayer, T. (2004). Market potential and the location of Japanese investment in the European Union. Review of Economics and Statistics, 86(4), 959-972.
17. Head, K., & Mayer, T. (2006). Regional wage and employment responses to market potential in the EU. Regional Science and Urban Economics, 36(5), 573-594.
18. Hinojosa, Y., Puerto, J., & Fernández, F. R. (2000). A multiperiod two-echelon multicommodity capacitated plant location problem. European Journal of Operational
Research, 123(2), 271-291.
19. Jorion, P. (1997). Value at risk: the new benchmark for controlling market risk. Irwin Professional Pub.
20. Kheder, S. B., & Zugravu, N. (2012). Environmental regulation and French firms location abroad: An economic geography model in an international comparative study. Ecological
Economics, 77, 48-61.
21. Lam, M. L. L. (2007). A study of the transfer of corporate social responsibility from well-established foreign multinational enterprises to Chinese subsidiaries. International Corporate Responsibility Series, 3, 343-363.
ethics, 105(1), 69-81.
23. Louveaux, F. V. (1986). Discrete stochastic location models. Annals of Operations
research, 6(2), 21-34.
24. Luo, X., & Bhattacharya, C. B. (2006). Corporate social responsibility, customer satisfaction, and market value. Journal of marketing, 70(4), 1-18.
25. McWilliams, A., Siegel, D. S., & Wright, P. M. (2006). Corporate social responsibility: Strategic implications. Journal of management studies, 43(1), 1-18.
26. Melo, M. T., Nickel, S., & Saldanha-Da-Gama, F. (2009). Facility location and supply chain management–A review. European journal of operational research, 196(2), 401-412. 27. Özsomer, A., & Gençtürk, E. (2003). A resource-based model of market learning in the
subsidiary: The capabilities of exploration and exploitation. Journal of International
Marketing, 11(3), 1-29.
28. Park, H. J., Mitsuhashi, H., Fey, C. F., & Björkman, I. (2003). The effect of human resource management practices on Japanese MNC subsidiary performance: a partial mediating model. The International Journal of Human Resource Management, 14(8), 1391-1406. 29. Perrini, F., Russo, A., & Tencati, A. (2007). CSR strategies of SMEs and large firms.
Evidence from Italy. Journal of business ethics, 74(3), 285-300.
30. Piercy, N. F., & Lane, N. (2009). Corporate social responsibility: impacts on strategic marketing and customer value. The Marketing Review, 9(4), 335-360.
31. Revelle, C. S., Eiselt, H. A., & Daskin, M. S. (2008). A bibliography for some fundamental problem categories in discrete location science. European Journal of Operational
Research, 184(3), 817-848.
32. Rockafellar, R. T., & Uryasev, S. (2000). Optimization of conditional value-at-risk. Journal
of risk, 2, 21-42.
33. Rugman, A. M., & Verbeke, A. (2001). Subsidiary ‐ specific advantages in multinational enterprises. Strategic Management Journal, 22(3), 237-250.
Management studies, 48(2), 253-277.
35. Russo, A., & Perrini, F. (2010). Investigating stakeholder theory and social capital: CSR in large firms and SMEs. Journal of Business ethics, 91(2), 207-221.
36. Ryan, S. P. (2012). The costs of environmental regulation in a concentrated industry. Econometrica, 80(3), 1019-1061.
37. Snyder, L. V., Daskin, M. S., & Teo, C. P. (2007). The stochastic location model with risk pooling. European Journal of Operational Research, 179(3), 1221-1238.
38. Sridharan, R. (1995). The capacitated plant location problem. European Journal of
Operational Research, 87(2), 203-213.
39. Strike, V. M., Gao, J., & Bansal, P. (2006). Being good while being bad: Social responsibility and the international diversification of US firms. Journal of International
Business Studies, 37(6), 850-862.
40. Surroca, J., Tribó, J. A., & Zahra, S. A. (2013). Stakeholder pressure on MNEs and the transfer of socially irresponsible practices to subsidiaries. Academy of Management
Journal, 56(2), 549-572.
41. Vahlne, J. E., & Johanson, J. (2017). The internationalization process of the firm—a model of knowledge development and increasing foreign market commitments. In International
Business (pp. 145-154). Routledge.
42. Weichenrieder, A. J. (2009). Profit shifting in the EU: Evidence from Germany. International tax and public finance, 16(3), 281-297.
43. Welsh, J. A. (1981). A small business is not a little big business. Harvard business review, July-August.
44. Wu, J., Wang, S., Chao, X., Ng, C. T., & Cheng, T. C. E. (2010). Impact of risk aversion on optimal decisions in supply contracts. International Journal of Production
Economics, 128(2), 569-576.
