Exploring the relationship between strategic attention breadth and performance: The role of corporate social responsibility

Abstract

The attention-based view contends that executives possess limited attentional capabilities that must be carefully allocated across different strategic issues. Although many scholars contend that narrow strategic attention breadth leads to better performance, others argue that broad strategic attention breadth may be more beneficial due to better opportunity scanning. We posit that the relationship between strategic attention breadth and performance will be inverted U-shaped, where strategic attention breadth is positively related to firm performance up to an optimal point, after which firms will see declining benefits due to executive cognitive overload. Furthermore, we propose that executives’ assessment of strategic opportunities will be influenced by the firm’s corporate social responsibility perspective, as the firm’s environmental and ethical commitment may mitigate executive blind spots and enhance opportunity selection. We support our hypotheses with multiple measures of firm performance and a content analysis of annual reports corresponding to a 5-year longitudinal sample of 2,245 S&P 500 firms.

Introduction

Strategic attention has attracted consistent academic interest in recent years (Cho & Hambrick, 2006; Joseph & Wilson, 2018; Ocasio, 1997) as an important determinant of firm behavior. The attention-based view (ABV) purports that firm behavior will be largely influenced by what executives pay attention to (Ocasio, 1997), and that different patterns of organizational attention influence firm outcomes differently (Ocasio & Joseph, 2018). The attention-based literature generally recognizes that performance differences between firms can be explained, at least partially, by the opportunities to which their managers pay attention and by how these managers understand different environmental signals (Rerup, 2009). Although strategic attention may influence a variety of firm outcomes, such as firm growth (Joseph & Wilson, 2018) or strategic change (Cho & Hambrick, 2006), this study is specifically interested in the effect of executive attention breadth on firm performance.

Strategic attention breadth (SAB) refers to ‘both the number of strategic issues to which executives pay attention, and the relative level of attention that executives give to each of these issues’ (Eklund & Mannor, 2021, p. 3). SAB may vary from narrow (low) to broad (high) and manifests itself on a continuum. Two conceptual perspectives provide incongruent explanations on the type of relationship to be expected between SAB and firm performance. On one hand, the ABV literature suggests that focusing on a few strategic issues leads to quicker, more effective information processing (Funk, 2014; Ocasio, 1997). On the other hand, the managerial cognition literature suggests that broader strategic attention and environmental scanning allows for the selection of better opportunities (Danneels, 2003) and a better ability to respond to environmental change (Tripsas & Gavetti, 2000). The opposition between both streams of literature has created a gap in the literature (Eklund & Mannor, 2021) regarding the benefits of a middle ground, more nuanced SAB. Indeed, executives with a middle ground SAB may be better able to balance between focusing on current strategic issues while still being cognitively flexible and searching for the most relevant information (Kiss, Libaers, Barr, Wang, & Zachary, 2020). For example, middle ground SAB may allow CEOs to prioritize key strategic issues while still remaining aware of the broader environment and being sensitive only to the most relevant outside opportunities. However, these benefits may not be experienced by CEOs with the broadest SAB, as they could experience cognitive overload likely to undermine the benefits achieved from identifying a new opportunity (Eklund & Mannor, 2021).

Much literature highlights how corporate social responsibility (CSR) is important decision-making criteria for the modern executive (Carroll, 1999; Dahlsrud, 2008; Durand & Jacqueminet, 2015; Fehre & Weber, 2019; Wang & Huang, 2018). For example, research has suggested that a firm’s own stance on CSR will impact how much attention they pay to their suppliers’ behaviors (Harness, Ranaweera, Karjaluoto, & Jayawardhena, 2018) – e.g., in the real world, a firm that values environmental sustainability would not entertain doing business with a polluting supplier. In this study, we add to extant literature on the effect of economic moderators of the relationship between SAB and performance (e.g., Eklund & Mannor, 2021) by proposing that CSR – a normative moderator – may be a key to understanding the link between SAB and performance. As such, CSR is the focal construct of this study, and is defined as a firm’s ethical and prosocial commitment to its employees, the environment, human rights, and the community at large (consistent with prior frameworks; cf. our literature review). CSR is posited to influence the dynamics between SAB and performance, as research has shown that firms applying CSR principles are likely to find new strategic opportunities (Kiessling, Isaksson, & Yasar, 2016) and to avoid certain strategic mistakes (Bansal, Jiang, & Jung, 2015). Furthermore, CSR is known to inform strategy selection when managers are overloaded with too many opportunities, as CSR guides an executive’s attention toward ethical options (Harness et al., 2018) while also eliminating opportunities that are misaligned and would have otherwise required attention from the firm (Peters, Waples, & Golden, 2014). Thus, CSR may lighten the cognitive overload associated with broader SABs.

We fill this gap in the literature by moving away from linear explanations of the SAB-performance link and by proposing that the relationship between SAB and firm performance is in fact curvilinear and moderated by CSR. First, we empirically reconcile both the ABV and managerial cognition perspectives by positing an inverted U-shaped relationship between SAB and firm performance, where narrow SAB can be as detrimental as excessively broad SAB, whereas moderate SAB is most beneficial. Second, we posit that CSR is a key moderator of the link between SAB and firm performance, since it has the potential to inform and condition what opportunities firms pay attention to (Harness et al., 2018) and is likely to increase the effectiveness of managerial cognition (Peters et al., 2014).

This study is organized as follows. First, we provide a literature review covering this study’s constructs and conceptual background. Second, we build on prior literature to introduce our two hypotheses. Third, we provide an in-depth presentation of our methodology, including an explanation of the computer-assisted-text-analysis (CATA) technique applied to assess constructs from 2,245 annual reports of S&P 500 firms. Fourth, our analysis and results are presented, several proper checks of curvilinearity are performed (Haans, Pieters, & He, 2016), and we back our findings with two relevant firm performance dependent variables – Tobin’s Q and productivity. Lastly, this investigation is concluded with a discussion of our findings and their theoretical and practical implications, their limitations, future research prescriptions, and a conclusion.

Literature review

ABV

The ABV purports that a firm’s behavior is largely dependent on what a firm’s decision-makers pay attention to (Ocasio, 1997; Ocasio & Joseph, 2005). A core assumption of ABV is that managers possess limited attentional capabilities (Simon, 1947), and therefore cannot pay attention to all information relevant to decision-making, thereby resulting in a bounded capacity to be rational. Two tenets of ABV may have a marked impact on the nature of strategic attention: first, executives select which strategic issues they attend to at any point in time, and decision making can only happen for these issues to which executives pay attention to; and second, the focus of decision-makers is largely influenced by the situations to which they are confronted (Ocasio, 1997).

The attention-based literature and the managerial cognition perspective highlight a tension between the benefits and drawbacks of narrow versus broad SAB. On one hand, the ABV purports that managers should have narrower executive attention, only focusing on a limited set of opportunities (Ocasio & Joseph, 2005). According to Ocasio (1997), successful decision-makers ‘selectively focus their attention on certain characteristics of the organization and its environment, and ignore others’ (p. 203), and firms with a focused attention to their environment are able to successfully adapt to rapidly changing environments. Extant research has begun to highlight many of the benefits of a narrow SAB. For example, scholars have suggested that executives with focused attention may be able to craft better strategies (Ocasio & Joseph, 2018), achieve superior firm growth (Joseph & Wilson, 2018), and better scan the environment for emerging competitive threats (McMullen, Shepherd, & Patzelt, 2009). On the other hand, the managerial cognition perspective assumes that managers are information workers who need to absorb, process, and disseminate all strategy-relevant information if they are to make good decisions for the firm (Walsh, 1995). They need to scan, process, and make sense of vast amounts of information for the firm as a whole (Nadkarni & Barr, 2008), which may warrant broader SAB. Here too, research has suggested benefits, as managers with broad SAB may be able to identify better opportunities (Danneels, 2003), create more innovative firms (Manral, 2011), and make better decisions on when to engage in strategic turnarounds (Lant, Milliken, & Batra, 1992).

SAB and firm performance

There is a lack of agreement in the literature on the proper level of SAB needed to achieve superior performance. On one hand, some scholars have highlighted the benefit of focusing attention on fewer strategic issues (i.e., narrow SAB) (Magretta, 2012). Indeed, prior research suggests that placing a greater emphasis on specific strategic issues may be beneficial. For example, CEOs with a strong emphasis on a single strategic issue, such as human resources, may achieve superior results (Chadwick, Super, & Kwon, 2015). Similarly, executives directing their attention toward the issue of competitiveness in the marketplace have been suggested to experience greater economic success (Connor, 2003). On the other hand, other scholars have suggested that paying attention to multiple strategic issues simultaneously (i.e., broad SAB) allows for better opportunity identification and greater responsiveness to market changes (Kaplan, 2011; Shepherd, Mcmullen, & Ocasio, 2017). For example, CEOs with broad SAB may be better able to quickly deploy and utilize firm resources toward achieving the firm’s strategic goals (Li & Yang, 2023), and be better able to launch new products successfully (Srivastava, Sahaym, & Allison, 2021).

The role of CSR

CSR has been a part of corporate life and a crucial part of executive decision making for well over 60 years, and the concern for CSR from both academics and practitioners has grown consistently (Wang, Tong, Takeuchi, & George, 2016). A recent impactful example of CSR becoming a top corporate imperative is perhaps Blackrock’s 2020 ‘Letter to CEOs’ which emphasized the need for environmental, social, and stewardship policies as a possible criteria for institutional investment – which is seen by many as a ‘tectonic shift’ (Strampelli, 2020) for how business is to be conducted. Consequently, CSR today is a prominent part of the executive decision-making process (Carroll, 1999; Dahlsrud, 2008; Durand & Jacqueminet, 2015; Fehre & Weber, 2019; Wang & Huang, 2018), and recent data shows that 98% of S&P 500 firms voluntarily participate in CSR reporting (G&A Survey, 2023).

CSR may play a crucial role in executive attention given its importance in today’s corporate environment. Indeed, CEOs that apply CSR principles to the strategic issues they pay attention to may not only get the favors of the world’s largest institutional investor (Strampelli, 2020), but may also achieve greater firm performance (e.g., Okafor, Adeleye, & Adusei, 2021; c.f., Wang et al., 2016 for a review). Defining CSR early in this investigation is important, as the construct is socially constructed, and may mean different things to different people and organizations (Dahlsrud, 2008). We adopt the World Business Council’s definition of CSR, which is widely accepted and reads as follows: ‘CSR is the continuing commitment by business to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as of the local community and society at large’ (Pencle & Mălăescu, 2016; p. 112). This definition of CSR is multidimensional, and encompasses the distinct areas of employee, human rights, social/community, and environmental CSR – four underlying components present in most popular CSR frameworks (e.g., GRI, KLD, IIRC) and on which prior research has found agreement and consensus (Pencle & Mălăescu, 2016).

A key research question in this investigation concerns whether CSR can act as a leading normative moderator of the relationship between SAB and firm performance. Indeed, extant literature has suggested that a relationship exists between a firm’s CSR stance and the strategy it chooses to pursue (Galbreath, Lucianetti, Tisch, & Thomas, 2022), and research has suggested that firms applying CSR principles may be more likely to select certain strategic opportunities (Kiessling et al., 2016) and to reject other opportunities, thereby avoiding certain strategic mistakes (Bansal et al., 2015). Indeed, CSR may be intimately linked to the size of the opportunity set considered by executives, as CSR could serve as a guide to select certain more ethical options (Harness et al., 2018) while also eliminating opportunities that are misaligned (Peters et al., 2014). Thus, CSR may lighten the cognitive overload associated with broad SAB, as it constrains CEOs to only consider a reduced set of opportunities that match their CSR stance.

Hypotheses

SAB and firm performance

Extant literature suggests that when it comes to SAB, too little can be as detrimental as too much. On one hand, the managerial cognition literature (Kaplan, 2011) suggests that narrow SAB is undesirable and leads to lower firm performance, since narrow SAB may lead firms to miss opportunities in their market due to executive blind spots (Danneels, 2003) and to an inability to properly respond to changing environments (Tripsas & Gavetti, 2000). Conversely, the managerial cognition perspective views broad SAB as helpful, since it allow for better opportunity scanning (Danneels, 2003), better new product launches (Srivastava et al., 2021), better conviction in identified opportunities (Shepherd et al., 2017), better sensemaking of the new information (Nadkarni & Barr, 2008), and better resource utilization (Li & Yang, 2023). Yet, the cognition perspective is in disagreement with the ABV, which views broad SAB as undesirable and potentially leading to lower firm performance. Indeed, per the ABV, broad SAB can lead executives to become overloaded with too many opportunities and with more information than they can effectively process (Funk, 2014), as well as slower, less accurate information processing (Ocasio, 1997). Contrarily, studies suggest that focusing on just a few strategic issues (i.e., narrow SAB), such as human resources (Chadwick et al., 2015) or product competitiveness (Connor, 2003), may lead to greater firm performance. The ABV thus favors the view that narrow SAB is more beneficial to firm performance.

These contradictory views may hint to trade-offs lying at the core of the relationship between SAB and firm performance. Indeed, the aforementioned separate research streams (ABV and managerial cognition) suggest that both too narrow and too broad SABs may be detrimental to performance, which implies that there may be an optimal level where SAB is most beneficial, and declining benefits once that optimal level is exceeded. Thus, for this hypothesis, we posit that the nature of the SAB–firm performance relationship will be curvilinear and taking the form of an inverted U, where a middle point SAB is most associated with firm performance. Indeed, a mid-point moderate SAB may be most beneficial for firms, as the most detrimental drawbacks of either a narrow SAB (i.e., missed opportunities, executive blind spots) or broad SAB (i.e., information overload, slow decisions) are avoided. Based on the present argument, we propose that there is an inverted U-shaped relationship between SAB and firm performance. Formally:

Hypothesis 1: There is an inverted U-shaped relationship between SAB and firm performance.

The flattening effect of CSR

CSR may moderate the relationship between SAB and firm performance, as an executive’s own stance on CSR may serve as a crucial lens for directing attention (Fehre & Weber, 2019), may be closely linked to which strategy is chosen (Galbreath et al., 2022), and may help the firm achieve superior firm performance (Okafor et al., 2021; Wang et al., 2016). In Hypothesis 1, we posited that a middle ground SAB would have the greatest association with firm performance, with both the narrow and broad SAB ends of the posited inverted U-shaped curve showing a lower association to firm performance. This second hypothesis, Hypothesis 2, is concerned with the steepness of the inverted U-shaped relationship between SAB and firm performance as CSR is introduced as a moderator. Indeed, CSR may flatten or attenuate the steepness of both ends of the curve, as extant research highlights many beneficial effects of CSR in opportunity selection processes (Kiessling et al., 2016) and opportunity rejection processes (Bansal et al., 2015), thereby guiding the firm’s attention towards the best aligned opportunities (Harness et al., 2018) and making misaligned opportunities easier to see (Peters et al., 2014).

On the ‘narrow’ end of the curve, executives with narrow SABs coupled with high CSR inclinations may achieve greater performance than executives with low CSR stances. Indeed, although the drawbacks of narrow SAB include executive blind spots and the risk of missing opportunities, the presence of a high CSR emphasis is likely to guide the executive’s scarce attention only toward the most appropriate choices, since CSR restricts the size of the opportunity space an executive can consider (Harness et al., 2018) and maximizes the chances of choosing strategically aligned opportunities (Kiessling et al., 2016). In other words, the executive blind spots associated with having narrow SAB may constitute only a mild drawback for the CSR-inclined firm, as many of the missed opportunities could not have been considered by the high CSR executive to begin with.

Similarly, the benefits of high CSR inclination should be visible on the ‘broad’ end of the curve, as executives with broad SABs and high CSR stances may also reach greater performance than executives with low CSR stances. While broad SAB may lead executives to become overloaded with too many opportunities and with more information than they can effectively and accurately process (Funk, 2014; Ocasio, 1997), simultaneous high CSR may be able to lighten the cognitive overload associated broad SAB. Indeed, CSR may lead managers to evaluate opportunities differently, and CSR selection criteria – such those related to the firm’s responsibilities to employees, society, and the environment (Harness et al., 2018) – may lead to the elimination of many opportunities which would have otherwise been evaluated and would have required attention from the firm (Peters et al., 2014). Thus, CSR may lead to more effective information processing despite a firm’s broad SAB and may lighten executive cognitive overload. In summary, synthetizing the present section, we posit a flattening of both the ‘narrow’ and ‘broad’ ends of the previously intimated inverted U-shaped relationship curve between SAB and firm performance as a result of CSR, as follows:

Hypothesis 2: The inverted U-shaped relationship between SAB and firm performance is significantly flattened by CSR.

Method

Sample

We use a 5-year longitudinal sample composed of all firms listed on the S&P 500 index during the 2010, 2011, 2012, 2013, and 2014 fiscal years. We chose the S&P 500 as a sample target due to its representativeness of large firms present in the US economy and its potential for more generalizable results. Our dependent and control variables were calculated from accounting, stock price, and employee count data retrieved from COMPUSTAT and CRSP, and our independent and moderating variables were assessed through a text analysis of 10-K annual reports collected from directly from the SEC database (the CATA methodology is described later in this section). We adopted a listwise deletion method to treat incomplete observations, and firms were only included in the sample if they remained continuously listed on the S&P 500 over the course of 5 years. Per the presented selection criteria, our final sample consists of 449 firms followed over a 5-year period, for a total of 2,245 firm-year observations.

Firm performance: Tobin’s Q and productivity

Firm performance is measured with two dependent variables: Tobin’s Q – a summary variable of performance – and productivity – a human capital efficiency measure. We use Tobin’s Q as a performance measure as it provides a useful evaluation of firm market value relative to replacement cost, and is widely used in strategy research (Carpenter, 2002). Consistent with prior operationalization (e.g., Brown & Caylor, 2006), we calculated Tobin’s Q as the market value of all assets owned by the firm, divided by the book value of such assets. Tobin’s Q values typically reflect both short and long-term performance expectations (Allen, 1993), and higher values (e.g., greater than 1) typically suggest that firms are performing well (Copeland & Weston, 1988). Productivity, a widely used measure of organizational performance (Richard, Devinney, Yip, & Johnson, 2009), was chosen as our second dependent variable, and is defined as revenue divided by the total number of employees, in line with prior operationalization (Huselid, 1995). For enhanced interpretation, we use a scaled variable (divided by 1,000). Productivity is a useful indicator of performance for this study, as it is a human capital efficiency measure that foretells long-term success (Guthrie, 2001).

Measuring SAB and CSR

This study’s independent and moderating variables (SAB and CSR, respectively) were assessed from 10-K annual reports through CATA methodology. CATA is a method of text analysis that allows researchers to assess constructs directly from business texts, such as annual reports (Jancenelle, Wei, & Ang, 2020) or IPO regulatory filings (Pencle & Mălăescu, 2016). CATA methodology is considered to possess strong empirical validity for strategic management research (Short & Palmer, 2008) due to its advantages over human coders. Indeed, CATA offers the unique feature of being able to analyze hundreds of thousands of words within minutes (Duriau, Reger, & Pfarrer, 2007) and with near-perfect reliability (as similarly coded CATA software will yield the same results; Neuendorf, 2016). Another distinctive feature of CATA for secondary research is its ability to generate formative independent variables (in contrast with reflective variables; e.g., a survey scale) which represent the ‘manifest characteristics of messages’ (Neuendorf, 2016, p. 32). As such, CATA measures yield scores representative of written words that are objectively visible and countable.

First, we assess SAB from annual reports. SAB refers to the range of issues to which executives pay attention, and the construct was measured through a Herfindahl computation of 13 strategic issue dictionaries. Both the Herfindahl computation and the 13 strategic issue dictionaries were proposed, developed, and validated by Eklund and Mannor (2021) through inductive and deductive best practices in CATA measure development (Short, Broberg, Cogliser, & Brigham, 2010). The 13 strategic issue dictionaries (with example sample words in parenthesis) are as follows: alliance partner strategies (JV, partner); customer-orientated strategies (customer, salesforce); external stakeholder management strategies (crisis, communication); financial and risk management strategies (debt, budget); internal organizational orientated strategies (culture, talent); low cost and efficiency strategies (lean, low cost); mergers, acquisitions, and firm scope strategies (M&A, takeover); new market entry strategies (emerging, international); product marketing strategies (brand building, advertise); resource and capability development strategies (capabilities, knowledge); social strategies (CSR, green); product innovation strategies (revolutionize, innovate); and business model innovation strategies (platform, blue ocean). The word lists for all dictionaries are presented in Eklund and Mannor (2021) and were designed for use with LIWC software – which we used to analyze our sample’s annual reports for all 13 word-count dictionaries.

We now turn our attention to the SAB measure computation. We begin by calculating the proportion of words associated with each strategy issue k. There are 300 words in total, spread over 13 separate strategy issue dictionaries, and equation 1 is first used to estimate p_i,k,t.

(1)\begin{equation}{p_{i,k,t}} = \,\frac{{300}}{{13\,x\,{n_k}}}\,x\,\frac{{{c_{i,k,t}}}}{{{q_{i,t}}}}\end{equation}

For equation 1, c_i,k,t corresponds to the word count for firm i on strategy issue k at time t in each annual report, q_i,t equals to the total number of words in the annual report, and n_k is the total number of words included in the dictionary corresponding to strategy issue k. We now turn our attention to computing the strategic breadth variable, operationalized with equation 2.

(2)\begin{equation}strategic\,breadt{h_{i,t}} = 1 - \frac{{\mathop \sum \nolimits_{k = 1}^{13} {{\left( {{p_{i,k,t}}} \right)}^2}}}{{{{\left( {\mathop \sum \nolimits_{k = 1}^{13} {p_{i,k,t}}} \right)}^2}}}\end{equation}

Equation 2 shows a usual Herfindahl measure derived from the proportion calculation presented in equation 1. We note that the Herfindahl calculation is subtracted from 1 so that scores closer to 1 indicate a greater breadth of strategic attention. The full development process and a more in-depth explanation of the measure is provided by its authors in Eklund and Mannor (2021).

Second, we measure CSR from annual reports through a previously developed and validated CATA measure of widely agreed-upon CSR dimensions (Pencle & Mălăescu, 2016). The measure was designed to be compatible with other research, and is composed of four dictionaries intended to represent CSR dimensions in line with the World Business Council’s CSR definition. The dimension dictionaries and selected sample words are as follows: environmental CSR (clean energy, global warming, sustainable); human rights CSR (fairness, minorities, diversity); socal/community CSR (charitable giving, community impact, job creation); and employee CSR (participatory, employee equity, employee well-being). All dictionaries were summed to create a total CSR score for each observation, and, similarly to strategic intention breadth, we operationalized the CSR measure through LIWC software.

Control variables and statistical analysis

Several control variables are necessary in our study. First, we control for size each year through the log of market capitalization (calculated at the end of each fiscal year). Second, due to the nature of our performance dependent variables, we view it necessary to control for research and development spending, cost of goods sold, and long-term debt. Third, industry controls were also included, using the Global Industry Classification Standard. We used the energy industry (0-industry) as our unestimated base industry, and coded the other 8 industries in the following way: Materials (1), Industrials (2), Consumer Discretionary (3), Consumer staples (4), Healthcare (5), Financials (6), Information technology (7), and Telecommunication services (8). Due to the longitudinal nature of our sample, we cannot use OLS regression (as panel data is likely to show autocorrelation, which would be an OLS assumption violation; Lattin, Carroll, & Green, 2003). Thus, we turned our attention to panel methodologies, and elected to use fixed-effects regression as the appropriate tool of analysis for our study since the Hausman specification test (Allison, 2005; Hausman, 1978) rejected the use of random effects. Our models and hypotheses were tested with SAS 9.4.

Results

We present this study’s means, standard deviations, minimum and maximum values, as well as a correlation matrix in Table 1. As shown, all correlations are well below 0.70 (the typical maximum level allowed for regression techniques; Lattin et al., 2003) and multicollinearity can be ruled out. Of note in Table 1 is also the fact that Tobin’s Q and Productivity have a near 0 correlation, thereby indicating that both dependent variables of performance add value to this study, as they measure different conceptualizations of performance.

Table 1. Descriptive statistics and correlation matrix

Italicized = p < .05.

We present our models and hypotheses for the Tobin’s Q variable in Table 2, and for the productivity variable in Table 3. We begin by turning our attention to Hypothesis 1. Our first hypothesis predicted an inverted U-shaped relationship between SAB and firm performance. For both performance dependent variables, the equations to be tested are written as follows: ${{\text{Y}}_{{\text{Tobin's Q}}}}{\text{ = }}{{{\beta }}_{\text{0}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{1}}}{{\text{X}}_{{\text{SAB}}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{2}}}{{\text{X}}_{{\text{SAB}}}}^{\text{2}}\;{\text{and }}{{\text{Y}}_{{\text{Productivity}}}}{\text{ = }}{{{\beta }}_{\text{0}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{1}}}{{\text{X}}_{{\text{SAB}}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{2}}}{{\text{X}}_{{\text{SAB}}}}^{\text{2}}$ (we did not add control variables to the equations for illustrative simplicity). We follow the prescriptions of Haans et al. (2016) to test for curvilinear relationships, and acknowledge that two conditions must be met to establish their presence: first, β₂ must be significant and of the expected sign (i.e., negative for inverted U-shapes); and second, the slopes at each end of the curve must steep enough for a turning point to be identified. The first condition is met both for the Tobin’s Q and productivity dependent variables, as each corresponding β₂ predictor is negative and significant (see Model 3 in both Tables 2 and 3). The second condition requires us to calculate a turning point for each slope, with the following calculation: −β_1/2β₂ (Haans et al., 2016). For Tobin’s Q, the turning point is located at 0.83, that is, well within the range of X_SAB (with a min of 0.65 and a max 0.90; cf. Table 1), and for productivity, we find that the turning point is located at 0.77, also well within the range of X_SAB. Hence, the curve appears to be sufficiently steep in both cases, and the second condition to establish curvilinearity is met for both dependent variables, thereby giving full support for Hypothesis 1. A final plot¹ with the Hypothesis 1 relationship for Tobin’s Q is presented in Fig. 1a, and for productivity in Fig. 1b.

Figure 1. (a) The relationshipbetween SAB and Tobin’s Q, (b) The relationship between SAB and Productivity.

Table 2. Fixed-effects regression for Tobin’s Q

† p < .10; *p < .05; **p < .01; ***p < .0001.

Note: Year and company fixed-effects are included in the models.

Table 3. Fixed-effects regression for productivity

† p < .10; *p < .05; **p < .01; ***p < .0001.

Note: Year and company fixed-effects are included in the models.

We now turn our attention to testing Hypothesis 2. This study’s second hypothesis was concerned with testing the moderating (flattening) effect of CSR on the inverted U-shaped relationship identified between SAB and firm performance in Hypothesis 1. The moderation of a curvilinear relationship is computed as follows (Haans et al., 2016), for both Y_{Tobin’s Q} and Y_Productivity, that is: ${{{\beta }}_{\text{0}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{1}}}{{\text{X}}_{{\text{SAB}}}}{ }{{\text{ + }}_{ }}{{{\beta }}_{\text{2}}}{{\text{X}}_{{\text{SAB}}}}^{\text{2}}{\text{ + }}{{{\beta }}_{\text{3}}}{{\text{X}}_{{\text{SAB}}}}{{\text{Z}}_{{\text{CSR}}}}{\text{ + }}{{{\beta }}_{\text{4}}}{{\text{X}}_{{\text{SAB}}}}^{\text{2}}{{\text{Z}}_{{\text{CSR}}}}{\text{ + }}{{{\beta }}_{\text{5}}}{{\text{Z}}_{{\text{CSR}}}}$. Hypothesis 2 requires us to test for the flattening of the inverted U-shape, and this is done by assessing β₄’s direction and significance (Haans et al., 2016). More specifically, for a flattening, we would expect a positive and significant β₄X_SAB²Z_CSR coefficient. As shown in Model 4 for both dependent variables (cf. Tables 2 and 3), β₄ is significant and positive for both the Tobin’s Q and productivity dependent variables, thereby providing full support for Hypothesis 2. The flattening of the inverted U-shaped relationship caused by CSR is plotted for Tobin’s Q in Fig. 2a, and for productivity in Fig. 2b.

Figure 2. (a) The effect of CSR on the SAB-Tobin’s Q relationship, (b) The effect of CSR on the SAB-Productivity relationship.

Robustness checks

To further support this investigation’s main curvilinear relationship, we performed additional checks. The inverted U-shaped relationships found between strategic breadth and both dependent variables (Tobin’s Q and productivity) were also tested for S-shaped relationships. To do so, we ran additional models with the addition of a cubic term (SAB³), as is usually recommended (Haans et al., 2016). The cubic term added to the models was not found to be significant for either dependent variable, and did not improve model fit, thereby strengthening this study’s support for the main quadratic relationship.

Discussion

Our results were in line with expectations for both hypotheses. The directional consistency of our findings across two measures of firm performance – Tobin’s Q and productivity – heighten the relevancy and generalizability of our study. Inverted U-shaped relationships between SAB and firm performance were found for both the Tobin’s Q and productivity variables. The turning point for productivity (cf. Figure 1b) does appear to occur at narrower SAB levels than for Tobin’s Q (cf. Figure 1a), but this may due to the fact that Tobin’s Q is partially market-based, and it may take some time for the market to react to a firm’s SAB position. However, both relationships are markedly curvilinear, thereby suggesting that moderate SAB is ideal across both types of performance. This finding is consistent with our hypothesizing for Hypothesis 1, where we posited that middle ground SAB would be most beneficial for firms, as the most detrimental drawbacks of either narrow SAB (i.e., missed opportunities, executive blind spots) or broad SAB (i.e., information overload, slow decisions) are avoided.

Regarding CSR, it appears that a high CSR inclination coupled with either narrow or broad SAB is beneficial to performance, as a flattening of the inverted U-shaped relationship between SAB and firm performance is observed for both performance dependent variables when CSR is introduced (cf. both Fig. 2a and 2b for a figure; Fig. 2b even suggests that the flattening leads to a minor shape-flip). The takeaway regarding CSR is that all levels of SAB benefit from a high CSR positioning, with the benefits of CSR being most pronounced for firms with broad SAB. This finding is in line with Hypothesis 2, as we posited that CSR would reduce executive cognitive overload due to elimination of many opportunities which would have otherwise needed to be evaluated and would have required attention from the firm (Peters et al., 2014). In short, CSR serves as a guide to executive attention by constraining top managers to only consider opportunities that match their CSR stance.

Theoretical implications

Our findings suggested that SAB is most beneficial to firm performance up to a middle point after which broad SAB yields diminishing returns due to cognitive overload and less effective information processing (e.g., Funk, 2014). Narrow SAB was also associated with lower performance than middle ground SAB, thereby highlighting the nonlinear nature of the relationship between SAB and performance. This finding adds nuance to one of ABV’s core tenets regarding the limited attentional capabilities of managers (Simon, 1947) and the benefits of focusing on just a few strategic issues (Magretta, 2012). Indeed, our findings suggest that focus is valuable, but that too much focus, or focusing on too few strategic issues, will be detrimental to firm performance. Conversely, our findings also suggested lower performance for the broadest end of SAB, in line with ABV’s arguments on the needs for managers to ‘selectively focus their attention on certain characteristics of the organization and its environment, and ignore others’ (Ocasio, 1997, p. 203). Thus, a key theoretical implication of our result is that there is indeed a ‘lower bound’ conditioning the effectiveness of focused attention, and that the narrowest forms of SAB may not be desirable.

Our paper further contributes to literature on CSR and ABV by highlighting the value of CSR in moderating the drawbacks associated with both narrow and broad SABs. Indeed, CSR was found to flatten the inverted U-curve between SAB and firm performance, thereby reducing the drawbacks associated with having either too narrow or too broad SABs. In other words, CSR makes selecting the right opportunities easier (when executives have narrow SABs), and also makes eliminating the wrong opportunities easier (when executives have broad SABs), as CSR serves as a guide for directing attention toward strategic options which align with the firm’s CSR stance. A possible theoretical implication for ABV is that executives with higher CSR stances may have an ability to focus on and analyze a wider range of strategic options, as the nature of CSR itself means that executives must analyze the business world through a lens which includes not only the business itself, but also its employees, the communities served, the environment, and society at large (Pencle & Mălăescu, 2016). CSR could therefore create a habit of holistic thinking which facilitates the use of executive attentional resources.

Implications for practice

This research suggests that focusing executive attention on too few strategic opportunities can be as bad as focusing on too many, and executives with a balanced focus should be rewarded with the highest levels of firm performance. Thus, executives should be careful about not concentrating on too few strategic opportunities, which could lead to missing important strategic options; but should also not spread their attention too thin across too many opportunities, which could lead to cognitive overload and a dilution of their attention. In practice, executives can use various techniques to open their minds to a balanced set of opportunities, such as training in strategic surveillance, external environment scanning, and fostering a firm environment where diverse ideas and feedback flow freely from all levels of the organization. Furthermore, a firm’s board may want to set up proper incentives and align executive compensation with a balanced and holistic set of organizational objectives.

Furthermore, CSR was found to attenuate the performance drawbacks associated with either too narrow or too broad SABs. Thus, the implementation of CSR should be seen as a positive towards achieving greater firm performance regardless of an executive’s SAB. In practice, executives may want to analyze the firm’s opportunities and strategic options with a CSR lens, as their own stance on CSR may facilitate opportunity analysis, selection, and elimination. Executives’ knowledge of CSR should be fostered and encouraged, as executive CSR lenses should lighten the cognitive overload associated with opportunity assessment. Corporate boards can foster CSR by providing training and guidance to the firm’s executives, but also by aligning executive compensation and providing other incentives toward the achievement of CSR goals.

Limitations

This research comes with a few limitations. First, our study may not be generalizable to firms of a smaller size, whose scope of operations may be markedly narrower than large S&P 500 firms. Second, although CATA methodology has many benefits, such as the ability to analyze text quickly, consistently, objectively, and reliably (Duriau et al., 2007; Neuendorf, 2016), it also is unable to provide researchers with the nuance of qualitative content analysis or case study methodologies. Finally, while there are benefits to using two distinct performance dependent variable measures for our study, these measures are still narrower than multidimensional measures of organizational effectiveness, such as survey-based measures (Richard et al., 2009).

Future research

First, future research may want to consider the nature of the relationship between SAB and other important firm outcomes, such as innovation, growth, diversification, or new market entries. Uncovering conditions under which SAB has a linear versus curvilinear effect on strategic outcomes would help us understand the boundaries to the applicability of both the ABV and managerial cognition perspectives. Second, extant research should seek to replicate and extend this study in other contexts – such as small and medium enterprises, or family businesses – as these contexts may bring unique challenges to how managers direct their strategic attention (e.g., Fehre & Weber, 2019).

Lastly, future research should investigate whether the effect of CSR on the SAB – performance relationship holds true in different market environments, in different industries, and for firms of different types and/or sizes. For example, CSR could have a limited moderating effect in a green industry such as sustainable energy but may have a great moderating effect in the fossil fuel sector. Additionally, different forms and intensities of CSR may impact performance differently, and firms may need to display a minimum level of concern for CSR for them to start witnessing performance benefits. This is especially true as some firms may engage in inauthentic forms of CSR such as greenwashing (e.g., Wu, Zhang, & Xie, 2020), that are unlikely to be beneficial.

Conclusion

This study drew on observations from the ABV and managerial cognition literature to develop hypotheses on the nature of the relationship between SAB and firm performance. We highlighted a core tension in the literature, and the trade-off that underlies SAB positioning: on one hand, narrow SAB may lead to missed opportunities (Danneels, 2003), while on the other hand, broad SAB may lead to cognitive overload, slow decision making, and ineffective information processing (Funk, 2014; Ocasio, 1997). Consequently, we posited an inverted U-shaped relationship between SAB and firm performance. This study also contended that attention to CSR would improve information processing and strategy selection, thereby flattening this study’s focal relationship. We tested our hypotheses with a 5-year longitudinal sample of S&P 500 firms, and a CATA analysis of 2,245 annual reports. Our hypotheses were supported with two performance dependent variables – Tobin’s Q and productivity – and the inverted U-shaped nature of the relationship between SAB and firm performance was confirmed using best practices in curvilinear relationship assessment (Haans et al., 2016). Our discussion highlights this study’s theoretical and practical implications, its limitations, and future research directions.