To Conform or Not to Conform? The Role of Social Status and Firm Corporate Social Responsibility

Abstract

Whether firms in transition economies undertake corporate social responsibility (CSR) is an important research topic in business ethics. Applying the middle-status conformity perspective, this study uses listed companies in the transition economy of China from 2010 to 2020 to assess the influence of social status on CSR conformity. The empirical findings revealed an inverted U-shaped relationship between social status and CSR conformity. That is, firms with low- or high-level status were less inclined to adopt CSR practices than the firms with a more middling status. Moreover, performance expectation gaps strengthened, while managerial ability flattened, the aforementioned inverted U-shaped relationship. This study sheds new light on the complicated motives for firms in transition economies to adopt CSR practices and further substantiates the boundary conditions of the curvilinear relationship between social status and CSR conformity.

Appendices

Appendix 1: The Measurement of Managerial Ability

Strategies to measure managerial ability are an important topic in economics, finance, accounting, and management research. Multiple characteristics of managers (ability, reputation, or style) may affect firm performance. Previous studies have generally suggested that indicators such as firm size, executive compensation, executive tenure, and media coverage are important when measuring managerial ability. However, Demerjian et al. (2012) argued that most of these indicators are external factors beyond the control of managers.

In an effort to provide accurate measurement standards for managers’ abilities, Demerjian et al. (2012) proposed a different measure of managerial ability, which is based on the efficiency of managers in transforming firm inputs into firm outputs. Firm inputs include management and sales expenses, fixed assets, operating leases, research and development (R&D) expenditures, and intangible assets. The basic logic underlying Demerjian et al.’s (2012) suggested measurement strategy is that managers with strong ability can better understand technology and industry trends, accurately predict market demand, and invest in projects with higher returns, and that managers who are more competent at a given level of resources are expected to achieve better returns, or invest a minimum amount of resources, at a given level of returns.

Appendix 2: A Brief Introduction of Data Envelopment Analysis

Data envelopment analysis (DEA) is a new field at the intersection of operations research, management science, and mathematical economics. DEA uses mathematical programming models to evaluate the relative effectiveness (called DEA effectiveness) among “departments” or “units” (also called decision-making units, DMU) with multiple inputs and outputs. After determining whether each DMU is DEA effective, one can then judge whether the DMU is located on the “frontier” of the production possible set. The production frontier is a generalization of the production function to multiple outputs in economics. The structure of the production frontier can be determined by using the DEA method and model.

A great deal of information with profound meaning and background can be obtained by evaluating DMU efficiency with the DEA method. In measuring managerial ability, Demerjian et al. (2012) used DEA to create an initial measure of the relative efficiency of firms in their industries and formed an effective boundary by measuring the number and combination of resources used by firms in each industry to generate revenue. The score for a firm at the frontier is 1. The lower the firm’s score, the farther away it is from the frontier, and the lower its efficiency is. The calculation procedure is as follows:

$$\frac{{\sum_{(i = 1)}^s {u_i y_{ik} } }}{{\sum_{(j = 1)}^m {v_j x_{jk} } }}k = 1, \ldots \,\; \ldots \; \ldots ,\,n$$

(6)

obey:

$$\frac{{\sum_{(i = 1)}^s {u_i y_{ik} } }}{{\sum_{(j = 1)}^m {v_j x_{jk} } }} \le 1\;k = 1, \ldots \;\, \ldots \;\, \ldots ,n$$

(7)

$$v_{1} ,v_{2} , \ldots \;\; \ldots ,v_{\text{m}} \ge {0;}$$

(8)

$$u_{1} ,u_{2} , \ldots \;\; \ldots ,u_{\text{s}} \ge {0}$$

(9)

DEA measures the efficiency of a single unit (here, enterprise k) relative to a set of comparable enterprises. The objective function measures the efficiency of weighted outputs by weighted inputs, with s outputs and m inputs, represented by i and j, respectively. The number of outputs i and inputs j of enterprise k are ${y}_{ik}$ and ${x}_{jk}$, respectively. The optimizer maximizes equation (6) by selecting the weight of each output (${u}_{i})$ and input (${v}_{j})$. The weight vectors on the output (u) and input (v) are called implicit weights. Efficiency depends on the level of weighted output and the level of weighted input. For a fixed level of input, the enterprise with the highest output level has the highest efficiency (or for a fixed level of output, the lowest input level). The DEA calculates a unique set of implicit weights for each enterprise k.

The first constraint (7) scales the implicit weight so that the efficiency value of the most efficient enterprise is 1. The best weight for each enterprise k is tested to determine the best weight for all other firms comparable to enterprise k (1, … … n; ≠ k). The efficiency of each comparable enterprise is calculated based on the implied weight calculated for enterprise k in equation (6), thereby determining the relative efficiency. Constraints (8)and (9) require the implicit weights to be non-negative.

In this study, the total efficiency of enterprises is estimated by one output and seven inputs, all of which come from public financial reports. Total revenue (“sales”) is the output, because the main goal of an enterprise is to achieve sales. The inputs consist of seven items: main business costs, management expenses and sales expenses, net fixed assets, net operating lease, net R&D investment, goodwill, and other intangible assets.

According to Fama and French (1997), the DEA efficiency by industry is estimated to increase the likelihood that peer firms have similar business models and cost structures within the estimations. The resulting score ranges from 0 to 1, with 1 being the optimal output for a given mix of inputs.

Using DEA instead of traditional ratio analysis has two major advantages. First, DEA allows the weight of each input to change. Second, DEA compares each enterprise in an industry with the most efficient enterprise, whereas traditional efficiency analysis compares each enterprise with the average or median company.