Effects of Corporate Life Cycle on Corporate Social Responsibility: Evidence from Korea

Abstract

Few studies examine how firms make strategic decisions over time. In this study, we test whether a firm undertakes corporate social responsibility (CSR) activities as a function of its life-cycle stage. Drawing on prior CSR research that finds ethical concerns and opportunistic behavior to be two key motivations that underpin CSR activities, we hypothesize that firms in their growth stage are positively associated with CSR, while firms in stage of decline are less likely to invest in CSR. The empirical findings of our study—derived by leveraging a sample of South Korean listed firms—are consistent with these predictions. We further find that in the growth stage, group-affiliated firms are more engaged in CSR than are unaffiliated firms. Given that affiliated firms can share the resources of other group-member firms, this evidence supports the slack resource hypothesis. Overall, our results indicate that firms have different CSR strategies, depending on their life-cycle stage.

1. Introduction

Firms develop over their life cycle, but there has been little research on how firms make strategic decisions over time. Firms may not only take part in operating activities, but also participate other activities. This study attempts to identify whether corporate social responsibility (CSR) activities differ by corporate life-cycle stage. While most CSR studies examine firm-level cross-sectional differences, the current study examines historical evolution by employing life-cycle approach.

Prior research highlights two motives that underpin CSR. One comprises ethical or reputation concerns, which drive managers to become involved in CSR activities [1,2]. The other is financial performance or opportunistic behavior [3,4,5]. Leveraging corporate life-cycle models used in prior research, we expect that growth-stage firms are more likely to engage in CSR activities: firms in this stage require both a sound ethical reputation and strong financial performance, if they are to survive. They need to build trust with external stakeholders [2] and also achieve the financial success essential to firms starting new businesses. Next, we have no predictions with respect to the CSR activities of mature-stage firms, firms that are less likely to increase their financial performance, as they already generate sufficient earnings; nevertheless, they know it is essential to maintain their ethical reputation. Therefore, we make no ex ante predictions, either positive or negative, with regard to this stage. However, we do expect decline-stage firms to be less likely to engage in CSR activities; rather, they are more likely to engage in more “extreme” adjustments, such as restructuring and mergers and acquisitions. Additionally, as firms’ slack resources are important determinants of investment in CSR activities [6], decline-stage firms will be unable to invest in CSR; therefore, we expect that demand for CSR should be the lowest at this stage.

In using a sample of firms listed on the South Korean stock market in the 2013–2016 period, we derive results consistent with our predictions. Firms are more likely to engage in CSR activities when they are in a growth stage. We find an insignificant coefficient on the maturity stage; more interestingly, we find a negative correlation between the decline stage and CSR. We further document that in the growth stage, group-affiliated firms have a higher probability of engaging in CSR activities than are unaffiliated firms. Given that slack resources constitute an important factor in whether or not CSR activities are undertaken, we interpret these results thus: affiliated firms in the growth stage share the resources of other group-member firms, and this enables them to have a positive CSR relationship.

The current study contributes to the literature in the following ways. First, empirical research on CSR activities has not been forthcoming in documenting the effects of the corporate life cycle on CSR involvement. Instead, most of the related literature focuses on firm-level characteristics as determinants of CSR activities. The current study fills this gap by applying a life-cycle approach. Second, we use South Korean data and find that group affiliation is an important factor that affects a firm’s CSR strategy. As a firm’s survival and operational continuity is contingent on the satisfaction of various stakeholder groups, future research can extend our findings to other ownership structures.

The remainder of this paper is organized as follows. Section 2 presents the prior research, as well as the hypotheses of the current study. Section 3 describes this study’s research design. Section 4 reports our empirical findings, and Section 5 provides concluding remarks.

2. Prior Research and Hypotheses Development

2.1. Prior Research on Corporate Life Cycle

The corporate life cycle is differentially defined in the literature. Earlier studies suggest that this life cycle consists of five different stages (i.e., birth–growth–maturity–revival–decline) [7]. The birth phase or “stage one” [8,9] is the period in which a new firm becomes an entity. In the growth stage, a firm competes with market competitors and has experience with initial success. The third stage is the maturity phase, during which a firm’s innovative activities are in decline but its internal operations are still functioning effectively [10]. In this stage, a firm’s profitability is maximized, and it accumulates cash. In the revival stage, firms try to diversify their product offerings, so as to survive in the market and not lose market share. In some cases, there is also an emphasis on more sophisticated control and planning systems [7]. In the last stage, decline, firms (or sometimes entire markets) “dry up” and begin to collapse. Failures to diversify and to complete mergers or acquisitions can lead to business withdrawal.

Additionally, some empirical research relates to the corporate life cycle. For instance, Anthony and Ramesh [11] look at the various aspects of accounting information over the life cycle. They find that the growth of sales and capital expenditure is an important factor in the growth stage, and that while it shows a positive relationship with market returns, it decreases in the decline stage. Black [12] compare the usefulness of accounting earnings and cash flow in assessing the value of a firm in each life-cycle stage; the paper finds that in the growth and maturity stages, accounting earnings are more value-relevant than cash flow, while in the decline stage, cash flow is more relevant than accounting earnings. Hribar and Yehuda [13] argue that the sustainability of accounting earnings and the ratio of reported losses differ with life-cycle stages, and thus differ in terms of value relevance and cost of capital, depending on the stage. They find that the cost of capital is lowest in the maturity stage and relatively high in the growth and decline stages. Martin et al. [14] review the published life-cycle studies of bio-based products to investigate the extent to which they include important sustainability indicators; their results suggest that there is a discrepancy between the indicators that are found to be important and the indicators that are frequently included in the literature. They argue that there is a need to develop a method by which to model important sustainability indicators.

In summary, the results of the aforementioned studies indicate that the economic circumstances that a company faces in each of its life-cycle stages vary. This implies that CSR strategies, as well the motives for investing in CSR, may also systematically differ at each stage.

2.2. Prior Research on Corporate Social Responsibility Activities

Investigations of CSR activities now constitute a major stream of business research. One of the research substreams therein suggests motivations for undertaking CSR activities, with two in particular coming to the fore. First, earlier studies provide a theoretical background for CSR by relating it to business ethics. For instance, Carroll [15] suggests that socially responsible firms should obey the law, be ethical, and be compelled to report good financial performance. Second, theories also suggest that CSR is a strategy relating to opportunistic behavior, rather than to stakeholder trust [16]. Similarly, Petrenko et al. [5] show that a CEO’s personal need for attention can be a source of investing in CSR.

Academic research has also examined the demand and supply of CSR activities. For instance, McWilliams and Siegel [17] argue that consumers and stakeholders—including investors, employees, and community members—are the two major sources of CSR demand. They hypothesize (and find) that a firm’s CSR level depends on its size, diversification level, research and development, and advertising expenses. Capital, materials and services, and labor resources are considered CSR supplies. Waddock and Graves [4] find that prior financial performance correlates positively with CSR, thus supporting the assertion that the existence of slack resources is a significant supply factor in undertaking CSR activities. More recently, Lee and Kim [18] document that organizational culture may play an important role in enhancing CSR activities.

Empirical investigations across multiple research streams have find a linkage between CSR and firms’ various outcomes (e.g., a relationship between CSR and financial performance, accounting quality, and other corporate outcomes). For instance, Cochran and Wood [19] find that firms with older assets tend to have lower CSR ratings. In their study, even after controlling for firm-level assets, there is still a positive correlation between CSR and financial performance. McGuire et al. [3] use both stock market-based performance and accounting-based performance to find that a firm’s prior performance measures more closely relate to CSR than do subsequent performance measures. More recently, Lins et al. [2] document that high-CSR firms experienced higher stock returns during the 2008 financial crisis suggesting that the trust between a firm and its stakeholders pays off when a firm suffers from an external shock. Similarly, Hao et al. [20] find a negative association between CSR and stock price crash risk. They further find that the internal controls moderate the relationship between CSR and crash risk. Consistently, Chung et al. [21] find a positive relation between CSR and firm value proxied by Tobin’s Q. As CSR is a form of corporate disclosure, other studies compare CSR to quality of accounting information. Atkins [1] claims that the CSR of public firms is to report transparent financial statements; likewise, Kim et al. [22] find that CSR firms report accounting information of a higher quality than that of non-CSR firms. Specifically, they find evidence that CSR firms are less likely to engage in both accruals and real activities earnings management. Moreover, their results support the assertion that CSR activities reduce the likelihood of being the subject of U.S. Securities and Exchange Commission investigations. Chih et al. [23] examine CSR and earnings management across 46 countries and derive mixed findings. They find that earnings management type affects the relationship between CSR and earnings management. For instance, they find evidence of a negative relationship between CSR and both earnings-smoothing and loss avoidance; however, they also find that aggressive earnings management correlates positively with CSR. Researchers that have found a relationship between CSR and other corporate outcomes have tended to focus on the costs of raising capital. Dhaliwal et al. [24] find that firms with a high cost of equity capital tend to initiate CSR disclosure in order to reduce their equity-raising costs. They further document that, as a consequence, this initiation attracts institutional investors and analysts. Similarly, El Ghoul et al. [25] find that firms with better CSR scores can reduce their cost of equity capital. In particular, they examine and document investments which improve responsible employee relations, environmental policies, and product strategies related to lower cost of capital.

In summary, there is no consensus in the literature as to whether CSR has only positive effects on firms. For instance, from an agency problem perspective, previous studies suggest that managers might engage in CSR activities to accrue personal benefits, rather than at the interest of stakeholders [15,16]. Thus, while recent empirical studies consider the positive effects of CSR activities on firms, the debate on the effectiveness on CSR continues.

It is noteworthy that in the above discussion, many studies examine those corporate outcomes affected by CSR, but only a few examine the factors that affect CSR. The current study addresses this research lacuna by examining the effect of the corporate life cycle on CSR activities.

2.3. Hypotheses Development

Two key factors lead to systematic differences in CSR activities among the various life-cycle stages. The first are those ethical or reputation concerns that drive managers to become involved in CSR activities [1]; the other is financial performance or opportunistic behavior [3,4]. The former relates to ethical expectations (i.e., CSR represents an effort wherein firms must expend resources to stakeholders and society); the latter relates more to opportunistic behavior, and how managers who engage in CSR activities do so for their own incentive purposes. Firms have different strategies, structures, and decision-making processes at each stage of the corporate life cycle. Thus, managers can determine their optimal CSR level by undertaking a cost–benefit analysis, based on these two key factors that depend on their firm’s life-cycle stage.

Growth-stage firms require both an ethical reputation and financial performance. Growing firms need to build trust with external stakeholders. In any case, a firm’s survival and operational continuity depend on the satisfaction of stakeholder groups (e.g., investors, internal employees, consumers, and governmental regulators). For instance, Lins et al. [2] show that a trust between a firm and its stakeholders is built through investments in CSR. Thus, by undertaking CSR activities, a start-up firm can enhance its reputation, and this may in turn enhance its business efficiency. However, there will still be the opposite logic that a firm in the early stage has much more pressing activities than CSR. For instance, Al-Hadi et al. [26] find that CSR reduces financial distress and this relation is more pronounced for firms in mature stage therein firms at the other stages may be more concerned about CSR investment. Nevertheless—and quite obviously—financial success is essential to firms that are starting a new business. As the organizational goal of publicly held firms is to maximize profits, CSR is viewed as an investment mechanism [17]. By investing in CSR, firms can surpass their performance benchmark [4]. In particular, Cochran and Wood [19] find that firms’ young assets tend to have higher CSR ratings, and this implies that growing firms are expected to undertake CSR activities. Thus, we propose our first hypothesis as follows:

Hypothesis 1.

Firms in the growth stage are positively related to CSR activities.

Next, it is unclear whether mature-stage firms invest more or less in CSR, relative to firms in other stages. On one hand, mature-stage firms have sufficient internal resources to conduct CSR activities. As resource availability often dictates the costs of engaging in CSR, firms at this stage might engage in CSR more so than other firms in other life-cycle stages. On the other hand, mature-stage firms are less likely to increase financial performance opportunistically, given that they already generate sufficient earnings. Moreover, these firms are now in the “right” direction in terms of their operating strategy, and this disincentivizes CSR engagement. Nevertheless, even for firms at this stage, maintaining ethical reputation is still on the table. Therefore, we make no ex ante predictions with respect to CSR activities among mature-stage firms: whether firms at this stage conduct more CSR activities than those at other stages remains an empirical question to be answered. Our second hypothesis is as follows:

Hypothesis 2.

Firms in the mature stage are not related to CSR activities.

Decline-stage firms tend to be engaged in more “extreme” adjustments—such as restructuring and mergers and acquisitions—rather than CSR activities, which are indirect means of survival. Additionally, when selling a firm in decline, company-to-company activities will be more vigorous than company-to-stakeholder activities. As firm-level slack resources are important determinants of investment in CSR activities [6], firms in decline are not as able to invest in CSR. Under agency theory, firms in earlier life-cycle stages have more growth opportunities, and those growth options diminish over time [27]. Thus, at some point in the decline stage, when growth-oriented resources have been exhausted, firms begin to acquire growing companies, or they are merged into another companies. Indeed, at this stage, conducting CSR activities is very much secondary to survival; we therefore expect the demand for CSR activities to be the lowest at this stage. We present third hypothesis as follows:

Hypothesis 3.

Firms in the decline stage are negatively related to CSR activities.

In summary, the overall demand for CSR activities declines as a company moves from the growth stage to the decline stage. We therefore propose these hypotheses.

3. Research Design

3.1. Variables

3.1.1. Proxy of Corporate Life Cycle

Studies in the literature differentially define corporate life-cycle stages. Earlier studies of corporate life cycle assume there to be five stages, comprising birth, growth, maturity, revival, and decline [7]. Black [12] uses four classifications, which are start-up, growth, mature, and decline. Anthony and Ramesh’s [11] scheme includes five stages, which are growth, growth/mature, mature, mature/stagnant, and stagnant. Roughly in line with these classifications, the current study uses five categories, with the second and fourth categories denoting intermediate stages. Thus, we refer to the growth, intermediate 1, maturity, intermediate 2, and decline-stage classification.

The current study measures the corporate life cycle by using five indicators—namely, sales growth rate, growth rate of tangible and intangible assets, market-to-book ratio, employee growth rate, and retained earnings ratio. The following are the previous studies that use the above five indicators, respectively. First, sales growth rate and growth rate of tangible and intangible assets are used by Anthony and Ramesh [11]. Second, market-to-book ratio is introduced by Bens et al. [28]. Third, employee growth rate is from Hribar and Yehuda [13]. Finally, retained earnings ratio is from DeAngelo et al. [29]. Growth-stage firms generally have higher sales growth rates, higher market capitalization, and higher investment in both tangible and intangible assets. Additionally, growth-stage firms tend to increase employment to grow their business. Although performance from investment is generally not revealed for earlier-stage firms, such dividends are expected to be low, given these firms’ low profit surplus levels. These characteristics change throughout the mature and decline stages.

Our method for calculating firm-specific life-cycle stage is as follows [11,12]. First, we select the median value for each indicator by using the data from the four previous years, including the current fiscal year (i.e., five years in total). Because the data may contain outliers, we use a firm’s median values over the five previous years to reduce volatility [11]. Thus, the current study’s data sample includes firms for which there are essential data over five consecutive years. Each firm-year’s values are sorted into quintiles and assigned between one and five points, based on size. Because we use the five aforementioned indicators, the total score for a given firm is between five and 25 points. These total individual-level scores are sorted by year and then classified into five quintiles. While different researchers differentially define life-cycle stages, the consensus is that the life cycle flows from growth to maturity and decline [12]; thus, we assume that the first, third, and fifth quintiles are the growth, maturity, and decline stages. As mentioned, we consider the remaining second and fourth quintiles intermediate stages.

A description using specific variables is as follows. For example, if we use the sales growth rate as an indicator for 2013, the sales growth rate of a firm needs to satisfy the five consecutive years (i.e., 2009–2013). From the five sales growth rates from 2009 to 2013, we derived the median value. All sample firms’ median values are divided by size into five quintiles, and a score of 1 to 5 is assigned. Since there are five indicators (including the sales growth rate), the sum of the indicators will range from five to 25 points. Therefore, the firms in 2013 are divided into five quintiles based on their total score, and the companies belonging to the first, third, and fifth quintiles are defined as being in the growth, mature, and decline stage, respectively. The remaining second and fourth stages, as discussed, are named intermediate 1 and intermediate 2.

3.1.2. Proxy of Corporate Social Responsibility Activities

In South Korea, the Korea Economic Justice Institute (KEJI) provides the “Best corporate citizen index,” which in prior research has been used as a CSR index value [30]. In 1991, the KEJI started to give “Economic Justice Corporation” awards to firms who have actively participated in CSR activities. Additionally, since early 2000s, it has also reported detailed CSR index values, and its evaluation method has been continuously improved. For instance, by applying more evaluation indicators after 2010, the KEJI changed its total CSR score more fine-grained, changing it from 75 to 100. For its most recently improved index scheme, the CSR scores of the top 200 listed firms are available; for this reason, the sample period of the current study starts in 2013.

The KEJI calculates its CSR index by undertaking the following three-step procedure.

(1)

Calculate the actual value of a given indicator;

(2)

Convert actual values to a 100-point scale, based on the rating formula (interpolation method);

(3)

Calculate a final score weighted by indicator.

Regarding the first step of the procedure, six categories are used as indicators—namely, soundness (25 points), fairness (20 points), social contribution (15 points), consumer protection (15 points), environmental management (10 points), and employee satisfaction (15 points). See Yu and Lee’s [23] Appendix A for the detailed indicators.

In the second step, the interpolation method by which to convert actual values to rating values is as follows.

Rating value = min value + [(max value − min value) * (actual value − actual min value)]/(actual max value − actual min value)

In the third step of the procedure, the scaled values derived via the above equation are weighted by each indicator. The CSR scores used in our study are final scores derived by applying this three-step procedure.

3.2. Research Model

To examine the association between corporate life cycle and CSR activities, we run the following regression model:

CSR_i,t = α₀ + α₁Growth_i,t + α₂Maturity_i,t + α₃Decline_i,t + α₄Size_i,t + α₅Leverage_i,t + α₆ROA_i,t + α₇Accruals_i,t + α₈Big_i,t + Year FE + Industry FE + ε_i,t,

(1)

where CSR is a CSR score from the KEJI database; Growth, Maturity, and Decline are corporate life-cycle stage dummy values equal to 1 if a firm belongs to the growth, maturity, or decline stage, respectively; Size is the natural logarithm of total assets; Leverage is calculated as total liabilities divided by total assets; ROA is return on assets, calculated as net income divided by total assets; Accruals is calculated as net income minus operating cash flows, divided by total assets; Big equals 1 for having a “Big 4” auditor, and 0 otherwise; and Year FE and Industry FE are year and industry fixed-effect dummies, respectively.

The positively (negatively) significant coefficients of each life-cycle stage indicate that firms belonging to that stage are more (less) likely to engage in CSR activities. As predicted in our hypotheses, we expect positive α₁, insignificant α₂, and negative α₃. To derive robust results, we first present the results of including each stage, one by one; from there, we show in one regression the results of including all three stages.

To preclude the omitted variable problem, we include several control variables. First, we control for firm size, as larger firms have economies of scale that allow them to save costs by spreading CSR costs across multiple divisions; small firms do not have this capability [17]. Additionally, large firms generally have slack resources, and this makes them more likely to engage in CSR [6]. We include the financial leverage ratio to control for the impact of financial conditions on CSR decisions. Prior research finds there to be a positive correlation between CSR and financial performance [3,4], and so we control for ROA. To mitigate the effect of accounting quality on CSR activities, we include total accruals [1,22]. The “Big 4” indicator is included to control for the auditor–CSR relationship [31]. The Year FE and Industry FE variables are year fixed dummies and industry fixed dummies, respectively. Throughout this paper, the reported t-values are corrected for firm-level clustering [32].

To investigate increasing or decreasing trends in CSR activities as a function of life-cycle stage, we run the following model.

CSR_i,t = α₀ + α₁LifeCycle_i,t + Controls + Year FE + Industry FE + ε_i,t

(2)

Here, LifeCycle is a variable that takes a value of 1–5, depending on whether the firm is in the growth, intermediate 1, maturity, intermediate 2, or decline stage, respectively; hence, a larger value indicates that that a firm is closer to being in the decline stage. In line with the predictions inherent in our hypotheses, we expect a negative coefficient on the LifeCycle variable. The same control variables in Equation (1) are also included in Equation (2).

3.3. The Sample

The sample data pertain to the 2013–2016 firm years. As mentioned, our sample period starts in 2013 because the KEJI score had been newly updated in that year. We use those observations that satisfy the following criteria:

(1)

Firms do not operate in a financial industry;

(2)

Listed firms have made available the financial data needed to calculate the variables used;

(3)

Firms’ group affiliations are identified.

We drew financial and group-affiliation identifier data from the Total Solution (TS) 2000 database, provided by the Korea Listed Companies Association. The sampling procedure is shown in Panel A of

Table 1. Sample selection process and distribution.

Panel A: Sample selection
Initial sample					800
Less:
(Financial industry or delisted)					(9)
(Missing financial data to calculate life cycle and control variables)					(126)
Final sample					665
Panel B. Sample distribution by year
Year	2013	2014	2015	2016	Total
Observations	156	161	175	173	665

This table shows sample selection process and yearly distribution.

. The initial sample starts with 800 firm-years, as the CSR scores of 200 listed firms are disclosed every year and we use a four-year sample period. We then delete firms that operate in a financial industry or which have been delisted. We also exclude observations for which the financial data needed to calculate the variables used herein are unavailable. Following this procedure, the final sample comprised 665 firm-years. Panel B of Table 1 presents the yearly distributions: as one can see, the number of firms by year is generally stable. Although not tabulated, 156, 100, 51, and 39 firms appear once, twice, three times, and four times in our sample period, respectively. With respect to the number of firms by life cycle stages, 92, 129, 207, 114, and 123 firm-years belong to the growth to decline stages, respectively.

4. Empirical Findings

4.1. Variable Description

Table 2 reports descriptive statistics and Pearson correlation values. Panel A of Table 2 presents the descriptive statistics of the main variables. All continuous variables are winsorized at the top and bottom 1 percent, in order to preclude outlier problems. The mean value of CSR—which represents the KEJI’s CSR score—is 63.64. In our sample, the 1st (99th) percentile’s score is 59.29 (69.42), and 14 percent, 31 percent, and 18 percent of our sample firms are classified as being in the growth, maturity, and decline stage, respectively. The mean value of LifeCycle is 3.07, which indicates that if we are to assign a score of 1 to 5 to the growth to decline stages, respectively, the average would be 3.07. The mean value of Size is 19.79, and firms have 36 percent of their liabilities scaled by their total assets (Leverage). ROA and Accruals have mean values of 0.05 and −0.02, respectively. “Big 4” auditors are contracted with 66 percent of our sample firms, and about one-half of the sample firms (49 percent) are group-affiliated (Business Group).

Panel B of Table 2 presents mean values of variables by each stage. The mean values of CSR get smaller as stages move from growth to decline. ROA also declines, however, there are no monotonic trends in other variables.

Panel C of Table 2 presents Pearson correlation values. The correlation matrix indicates that CSR correlates positively with Growth, has no significant correlation with Maturity, and correlates negatively with Decline; all of these findings align with our hypotheses. Additionally, each of Size, ROA, and Big positively correlates with CSR, and this is consistent with the results of previous studies. Finally, Leverage and CSR negatively correlate with each other.

Table 3. Determinant of Corporate Life Cycle and comparison by corporate social responsibility (CSR) score.

Variable	Growth	Intermediate1	Maturity	Intermediate 2	Decline
Life cycle determinants
1. Sales growth rate	0.16	0.09	0.04	0.02	−0.02
2. Tangible, intangible assets growth rate	0.12	0.08	0.04	0.01	−0.02
3. Market-to-book ratio	1.35	0.79	0.68	0.52	0.37
4. Employee growth rate	0.08	0.05	0.02	0.01	−0.02
5. Retained earnings ratio	13.29	13.57	10.20	17.09	16.09
Overall CSR score	64.34	64.02	63.64	63.51	62.85

reports the determinant of each life-cycle stage. To define life cycle stages, the median values for each indicator, by using the data from the five previous years, are selected by yearly basis. The numbers shown in Table 3 are the mean values of these selected median values. By definition, the sales growth rate, tangible/intangible assets growth rate, the market-to-book ratio, and the employee growth rate are all highest in the growth stage; conversely, they are lowest in the decline stage. However, the retained earnings ratio tends to increase as firms approach the decline stage. The retained earnings ratio should be the highest in the decline stage and lowest in the growth stage. However, by merging life-cycle data and CSR data, the remaining sample shows slightly different results. Although the retained earnings ratio in the raw life-cycle data monotonically increases at increasingly advanced stages, the values in our final sample differ from those in the raw data. Specifically, in Table 3, the retained earnings ratio is highest in the intermediate 2 stage and lowest in the maturity stage.

More interestingly, the overall CSR score is highest in the growth stage and lowest in the decline stage. If we plot these numbers on a dotted line graph (Figure 1), we find that CSR scores decrease gradually from the growth stage to the decline stage. In summary, the results presented in Table 3 and Figure 1 support our hypotheses. Having found a nearly monotonic decline in CSR activities as a function of the life-cycle stage, we next test this association by applying multi-regressions.

4.2. Test Results of Hypotheses

Table 4. Regression results of testing hypotheses.

Dependent Variable = CSR
Independent Variables	(1) Growth	(2) Maturity	(3) Decline	(4) G-M-D	(5) LifeCycle
	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)
Intercept	58.807 *** (40.28)	58.347 *** (39.09)	58.889 *** (40.11)	59.133 *** (41.11)	59.764 *** (40.29)
Growth	0.507 ** (2.38)			0.429 * (1.88)
Maturity		0.124 (0.81)		0.043 (0.25)
Decline			−0.752 *** (−4.60)	−0.684 *** (−3.77)
LifeCycle					−0.248 *** (−4.13)
Size	0.271 *** (3.62)	0.280 *** (3.67)	0.268 *** (3.58)	0.264 *** (3.59)	0.262 *** (3.50)
Leverage	−0.496 (−1.16)	−0.470 (−1.11)	−0.758 * (−1.78)	−0.765 * (−1.78)	−0.703 (−1.64)
ROA	5.271 *** (2.78)	6.407 *** (3.35)	5.060 *** (2.61)	4.314 ** (2.21)	4.060 ** (2.07)
Accruals	−0.249 (−0.20)	−0.770 (−0.60)	−0.273 (−0.57)	−0.378 (−0.30)	−0.217 (−0.17)
Big	0.074 (0.44)	0.077 (0.46)	0.085 (0.52)	0.084 (0.52)	0.065 (0.40)
Year	Included	Included	Included	Included	Included
Industry	Included	Included	Included	Included	Included
Adj. R-sq.	0.4091	0.4036	0.4184	0.4205	0.4210
N. of obs.	665	665	665	665	665

T-statistics reported in parentheses are firm-level clustered. *, **, and *** denote significance at the 0.1, 0.05, and 0.01 levels (two-tailed), respectively.

presents the main results vis-à-vis our hypotheses. The dependent variable is CSR, and the first column shows the result of using Growth as the variable of interest. The coefficient of Growth is 0.507 and statistically significant at the 5-percent level (t-value = 2.38); this is a strong supportive evidence of Hypothesis 1, which predicts that growth-stage firms are more likely to engage in CSR activities. Firm size and performance (Size and ROA) correlate positively with CSR, which indicates that slack resources are important determinants of CSR. In the second column, we use Maturity as an indicator of a firm’s life-cycle stage. The coefficient of Maturity is insignificant, and this expectation is inherent in Hypothesis 2. In the third column, the coefficient of Decline is negative and significant at the 1-percent level (coefficient = −0.752, t-value = −4.60); this is a consistent supportive evidence of Hypothesis 3, which predicts that decline-stage firms are less likely to engage in CSR activities. We include all three life-cycle variables in the fourth column, and derive similar results. In the fifth column, we include the LifeCycle variable, which takes a larger value as firms more closely approach the decline stage. The negative and significant coefficient (coefficient = −0.248, t-value = −4.13) indicate that a firm that is more closely approaching the decline stage is less likely to undertake CSR activities. Overall, the results presented in Table 4 strongly support our hypotheses. The results have important implications to both manager and stakeholder as they indicate that the ability and the tendency of investing in CSR vary across life cycle stages.

4.3. Additional Tests

4.3.1. The Effect of Business Group on the Relation between Life Cycle and CSR

By demonstrating that prior financial performance correlates positively with CSR, Waddock and Graves [4] suggest that holding slack resources is a significant supply factor in undertaking CSR activities. As we find a positive relationship between the growth stage and CSR activities, it is unclear how firms in this stage could invest in CSR, given that their resources are quite restricted relative to those of firms in the other stages. Here, we use group affiliation as a moderator that facilitates CSR activities. In South Korea, there are considerable numbers of group affiliated firms. For instance, 2083 firms are affiliated to 60 business groups in year 2018 ([33]). Chang and Hong [34] show that by sharing intangible and financial resources, Korean group-affiliated firms can maintain their financial performance. Thus, although many growth-stage firms have insufficient resources, they can share and use the resources of other group-member firms, thus allowing them to invest in CSR activities. Consistently, Lee [33] document that group affiliated firms are more likely to engage in CSR. To adopt these prior findings into our tests, we divide our sample into group-affiliated and unaffiliated firms.

Table 5. The impact of group-affiliation on the relation between life cycle and CSR.

Panel A. Group-Affiliated Firms
	Dependent variable = CSR
Independent Variables	(1) Growth	(2) Maturity	(3) Decline
	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)
Intercept	59.119 *** (28.60)	58.291 *** (26.82)	58.822 *** (27.41)
Growth	0.922 *** (3.37)
Maturity		0.107 (0.40)
Decline			−0.699 ** (−2.31)
Size	0.313 *** (3.17)	0.312 *** (3.01)	0.295 *** (2.89)
Leverage	−0.717 (−0.93)	−0.573 (−0.75)	−0.726 (−0.95)
ROA	1.310 (0.41)	3.946 (1.23)	2.778 (0.85)
Accruals	0.559 (0.24)	−0.998 (−0.42)	−1.323 (−0.55)
Big	−0.021 (−0.09)	−0.005 (−0.02)	0.021 (0.08)
Year	Included	Included	Included
Industry	Included	Included	Included
Adj. R-sq.	0.4535	0.4324	0.4424
N. of obs.	329	329	329
Panel B. Unaffiliated firms
	Dependent variable = CSR
Independent Variables	(1) Growth	(2) Maturity	(3) Decline
	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)
Intercept	62.711 *** (21.62)	62.297 *** (21.56)	62.925 *** (22.55)
Growth	−0.117 (−0.26)
Maturity		0.268 (1.28)
Decline			−0.902 *** (−4.39)
Size	0.251 * (1.75)	0.267 * (1.88)	0.248 * (1.80)
Leverage	−0.464 (−0.78)	−0.460 (−0.79)	−0.805 (−1.41)
ROA	9.395 *** (3.27)	8.311 *** (3.12)	7.222 *** (2.80)
Accruals	−0.089 (−0.05)	−0.208 (−0.13)	−0.072 (−0.05)
Big	0.123 (0.52)	0.114 (0.49)	0.121 (0.54)
Year	Included	Included	Included
Industry	Included	Included	Included
Adj. R-sq.	0.4102	0.4137	0.4356
N. of obs.	336	336	336

T-statistics reported in parentheses are firm-level clustered. *, **, and *** denote significance at the 0.1, 0.05, and 0.01 levels (two-tailed), respectively.

shows the results. The observations in Panel A are group-affiliated firms. In the first column, there is a positive and significant coefficient of Growth (coefficient = 0.922, t-value = 3.37). However, in Panel B—which uses unaffiliated firms as a sample—Growth is insignificant. Thus, in comparing the two panels, we find that in the growth stage, only affiliated firms engage in CSR activities. Next, the coefficients of ROA are significant only in unaffiliated firms. Collectively, this evidences are consistent with slack resource theory. For both subsamples in the other columns, the coefficients of Maturity are insignificant and the coefficients of Decline are significantly negative; these results are consistent with our prior findings. Thus, only growth-stage firms are affected by group affiliation.

4.3.2. Further Analysis of Using Sub-CSR Scores

Next, we additionally test the association between life cycle and individual CSR activities. There are six categories of CSR activities on KEJI’s evaluation procedures. Those are soundness, fairness, social contribution, consumer protection, environmental management, and employee satisfaction. As we document that firms are investing less in CSR as their life cycle become closer to the decline stage, we further examine this relation by using sub-CSR scores.

Table 6. Analysis of using sub-CSR scores.

Panel A. Soundness, Fairness, and Social contribution
	Dependent variable = CSR
Independent Variables	(1) Soundness	(2) Fairness	(3) Social contribution
	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)
Intercept	11.202 *** (10.28)	23.463 *** (20.67)	4.188 *** (4.09)
LifeCycle	−0.181 *** (−3.39)	0.019(0.55)	−0.161 *** (−3.89)
Size	0.307 *** (5.83)	−0.294 *** (−5.54)	0.125 ** (2.58)
Leverage	−1.491 *** (−3.84)	0.119 (0.42)	0.790 ** (2.48)
ROA	1.520 (1.04)	−0.511 (−0.41)	5.267 *** (4.13)
Accruals	−2.512 *** (−2.62)	0.055 (0.07)	2.254 ** (2.51)
Big	0.218(1.60)	−0.033 (−0.38)	0.099 (0.89)
Year	Included	Included	Included
Industry	Included	Included	Included
Adj. R-sq.	0.4671	0.4107	0.2573
N. of obs.	665	665	665
Panel B. Consumer protection, Environmental management, and Employee satisfaction
	Dependent variable = CSR
Independent Variables	(1) Consumer protection	(2) Environmental management	(3) Employee satisfaction
	Coefficient (T-stat)	Coefficient (T-stat)	Coefficient (T-stat)
Intercept	9.536 *** (14.65)	1.756 *** (3.42)	9.806 *** (12.24)
LifeCycle	0.011 (0.56)	0.019 (0.93)	0.053 (1.56)
Size	−0.058 * (−1.94)	0.235 ** (9.74)	−0.062 * (−1.69)
Leverage	−0.058 (−0.40)	0.131 (0.85)	−0.147 (−0.52)
ROA	−1.630 ** (−2.36)	−0.121 (−0.21)	−0.443 (−0.44)
Accruals	0.043 (0.10)	0.108 (0.26)	−0.098 (−0.13)
Big	−0.043 (−1.05)	−0.016 (−0.27)	−0.169 * (−1.72)
Year	Included	Included	Included
Industry	Included	Included	Included
Adj. R-sq.	0.3552	0.4104	0.1766
N. of obs.	665	665	665

T-statistics reported in parentheses are firm-level clustered. *, **, and *** denote significance at the 0.1, 0.05, and 0.01 levels (two-tailed), respectively.

shows the test results. Among these six categories, the coefficients of LifeCycle are negative and significant when the dependent variables are Soundness and Social contribution. Soundness represents the soundness of governance, investment (i.e., R&D expenditure), and financing. Also, Social contribution’s sub-indices are the equality in employment contracts (i.e., equality for disable or female employees), social contribution activities (i.e., donations), and transparent tax payments. The results of Table 6 suggest that a firm getting closer to a decline stage is investing less in these activities.

4.3.3. Other Sensitivity Tests

Lastly, we examine several sensitivity tests to ensure the validity of our research models. Throughout the tests, a negative and significant coefficient of LifeCycle confirms our hypotheses. First, we conduct regression analysis by industry. Prior research finds that high-tech industries are more CSR-oriented, to run their business efficiently through improved corporate image [35]. Thus, we conduct regression analysis using subsamples partitioned by high-tech and other industries. Followings are classified as high-tech industries: manufacture of pharmaceuticals, medicinal, chemical, and botanical products; research and development; manufacture of electronic components, and computer visual, sounding, and communication equipment; computer programming, consultancy, and related activities; and manufacture of electrical equipment. The untabulated results show that the coefficients of LifeCycle are negative and significant in both subsamples indicating that operating in certain industries are not the major factor that affect CSR activities in our sample. Second, we retest our regression by each year. The untabulated results show that the coefficients of LifeCycle are negative and significant in yearly partitioned subsamples except for year 2013. Third, we divide sample by median firm size. The results show that the coefficients of LifeCycle are also negative and significant in both large and small firm subsamples. Finally, we use the natural logarithm of CSR variables to reduce variations across firms. The coefficient of LifeCycle is still negative and significant at l% level indicating that our results are robust to the different definition of dependent variable. Overall, we find that our results are not tentative to these sensitivity tests.

5. Conclusions

The two key factors that lead to systematic differences in CSR activities are ethical concerns [1,2] and opportunistic behavior [3,4]. By linking these motives to the life cycle stages, we test whether CSR activity engagement is a function of a firm’s life-cycle stage. Drawing on prior CSR research, we hypothesize that growth-stage firms correlate positively with CSR, while decline-stage firms are less likely to engage in CSR. Our empirical findings are consistent with our hypotheses. Further, we find that in the growth stage, group-affiliated firms are more likely to engage in CSR than are unaffiliated firms. Overall, our results indicate that firms have different CSR strategies, depending on their life-cycle stage.

Our findings provide insights into the direction of possible related future research. While most of the previous studies focus on the firm-level determinants of CSR, we find that firms engage differentially in CSR, depending on their life-cycle stage. Thus, future studies could investigate more deeply other firm-level nonoperating activities by applying this life-cycle approach. Moreover, our results indicate a nearly monotonic decline in CSR activities with reference to the corporate life cycle. This strong evidence provides motivation to investigate case studies of specific firms and test their life-cycle stage and CSR strategies. We document a general trend in CSR activities by life-cycle stage, and one could further examine, in detail, the CSR activities in which a firm might engage. Our results also have important implications to both manager and stakeholder as they indicate that the ability and the tendency of investing in CSR vary across life cycle stages. There is also possible anomaly that even firms that operate in industries that need CSR do not invest in these activities when they move close to the decline stage. Finally, we use data pertaining to South Korean firms, whose ownership structure is unique compared to that of firms in other countries. In South Korea, a controlling shareholder can dominate (or dictate) the general decision-making of all affiliated firms, and so the CSR strategies of such firms might be affected by this unique structure. For this reason, future research could investigate the effect of corporate governance on CSR activities by employing Korean business-group data, or data from companies in other countries.

Our study also has several limitations. First, while the descriptive statistics show that the indicators used in the paper are reasonable proxies for identifying each life cycle stage, they still could proxy for other economic circumstances. Second, the sample is based on South Korean listed firms thereby generalization of the current results must be tentative.