Estimation of the static corporate sustainability interactions

ABSTRACT

The empirical literature considers firm-specific aspects affecting corporate sustainability decisions but generally omits the influence of the competition. We advocate that sustainability actions of a company impact its marketplace and vice versa. Therefore, the sustainability return of the single firm is a function of the other firms’ sustainability decisions. We approach sustainability decisions as strategic decisions and evaluate the effect of competition and spillovers in a static market entry game. We estimate the parameters of the discrete choice model using the social performance ratings from MSCI KLD 400 Social Index as proxy for sustainability decisions and financial information from Wharton Research Data Services’ COMPUSTAT dataset. When strategic interaction is not accounted for, we find that an increase in the number of competitors increases the likelihood of sustainability investments, seemingly shows the spillover effect dominates the competition. When we apply the multi-stage approach, which incorporates competitive interaction, we provide empirical evidence that the effect of competition on the likelihood of entry into the sustainability market dominates the effect of spillover. We find that strategic motives, typically ignored in the empirical literature, appear to be an important factor in understanding sustainability-related decisions.

KEYWORDS:

• Sustainability
• empirical study
• strategy
• applied statistics
• probabilistic models

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 https://www.mckinsey.com/business-functions/operations/our-insights/operations-driven-sustainability?cid=other-eml-alt-mip-mck&hlkid=6e9f6a62040549009363f75a09314d72&hctky=9256414&hdpid=3ff4b99a-bac0-4f33-a0e5-68b08783214c# accessed on 8 September 2020.

2 https://www.msci.com/documents/10199/904492e6-527e-4d64-9904-c710bf1533c6.

3 http://www.csrhub.com/.

4 www.globalreporting.org.

5 http://www.sustainability-indices.com/.

6 https://wrds-www.wharton.upenn.edu/documents/1353/MSCI_ESG_KLD_STATS_1991-2015_Data_Set_Methodology.pdf.

7 For instance, an estimation that does not take this endogeneity into account, will likely obtain a positive coefficient for the cumulative effect. That if interpreted as the effect of competition will lead to incorrect conclusions about the effect of competition.

8 Corporate inertia is a term used to describe established companies’ lag in adapting business models, operating conditions, and making strategic decisions which can be beneficial in the long run.

9 More recent papers such as Gallant, Han, and Khawaja (2018) and Tan (2019) estimate a dynamic game of entry decisions of moderate number of entrants with spillover effects in Pharma industry. Similar dynamic setting can be applied to strategic interactions in sustainability market. However, the computational burden and data requirements caused by the large number of entrants in the dynamic version of the sustainability market entry game is a challenge that needs to be solved. Thus, this remains for future research.

10 In this model a company is considered as an entrant into the sustainability market if $w_i>0$. The model can be extended to companies, which have taken substantial sustainability initiatives to enter the sustainability market. Then a company will be considered as an entrant if the sustainability investments $w_i$ exceed a threshold value.

11 The net benefit of player i depending on each possible action taken by the competitors is multiplied by its probability of occurring, and the resulting products are summed to produce the expected value. Thus, the expected value of the random variable net benefit ${\mathrm{\Pi }}_i\left(x_i,\mathit{S};\vartheta \right)$ can be calculated.

12 While S denotes the vector of state variables in the first stage, ${\mathit{S}}^{\mathbf{\prime }}$ denotes the vector of state variables S with the inclusion of a market-specific component in the second stage.

13 Entry into the sustainability market is not an entry game in the classical sense, and hence staying in the no sustainability state does not necessarily lead to a zero payoff. However, as explained in Section 3.3. in detail we won’t be able to identify the net benefit from investing into sustainability and staying in the no sustainability state separately. We can identify the difference between the net benefit of investing in sustainability as opposed to not investing. Thus, we assume that the net benefit of not investing in sustainability is equal to zero.

14 The limiting distributions for the minimum or the maximum of a very large collection of random observations from the same arbitrary distribution can only be described by generalized extreme value distributions models – specifically, the Gumbel, Fréchet, and Weibull distributions also known as type I, II and III extreme value distributions.

15 The difference of two type-I extreme value distributed variables follows a logistic distribution, of which the logit function is the quantile function.

16 We thank an anonymous referee for pointing out this. In the earlier version of the paper, probabilities were calculated using just the estimates from the first stage.

17 https://www.msci.com/resources/factsheets/index_fact_sheet/msci-kld-400-social-index.pdf.

18 The rating methodology of MSCI KLD 400 Social Index has changed over the years. The methodology is consistent over the selected time period.

19 For detailed information about the construction of data set and variables (see Soytaş and Uşar 2017).

20 The first stage of the IV estimates indicates a significant association between the number of competitors and the market size variables. The corresponding F-statistics is significantly high. Also, the Wald test of exogeneity employed produces 4.95 for the chi-squared (1) with the corresponding p-value of 0.026.

21 In Stata IV-probit is implemented where the variable number_of_competitors is instrumented with the log of market revenue.

22 In general, this is not required for the model identification but incorporating an extra variable into the estimation, that supplies independent variation for each company will make the identification easier. Otherwise the model is identified depending on a functional form.

Additional information

Notes on contributors

Mehmet Ali Soytaş

Dr Mehmet Ali Soytas (Ph.D., Economics, University of Pittsburgh) is an Associate Professor in the Accounting and Finance Department at King Fahd University of Petroleum and Minerals Business School (KBS). His research covers topics in labour economics, family economics, income inequality, intergenerational mobility, corporate sustainability and strategic behaviour. Dr Soytas’s work has been published in the Quantitative Economics, Energy Economics, International Journal of Production Economics and Federal Reserve Bank of St. Louis Review, among others.

Damla Durak Uşar

Dr Damla Durak Usar (Ph.D., Operations Management, Ozyegin University) is an Assistant Professor of Operations Research at Faculty of Engineering, Turkish- German University. Her research interests are sustainable operations, supply chain efficiency, corporate sustainability and strategic behaviour. Dr Durak Usar’s work has been published in the International Journal of Production Economics.

Meltem Denizel

Dr Meltem Denizel (Ph.D., Decision and Information Sciences, Warrington School of Business, University of Florida) is an Associate Professor of Supply Chain Management at Ivy College of Business, Iowa State University. Her research interests are sustainable operations, supply chain efficiency, mathematical modelling and production lot sizing. Dr Denizel’ s research is published in, for example, Operations Research, Journal of Business Logistics, Production and Operations Management, Transportation Journal, Decision Sciences, IIE Transactions, and International Journal of Production Economics.