Keeping up with the CSR Joneses: The impact of industry peers on focal firms' CSR performance

The study examines whether and how corporate social responsibility (CSR) decisions of industry peers influence a focal firm’s CSR performance. Grounded in institutional theory and peer effects literature, the authors posit that firms in the same industry face similar norms, pressures, and competitive conditions, leading to mimetic behavior in CSR. Using U.S.-listed firms (2000–2015) and defining industries primarily by 3-digit SIC codes, the paper investigates the magnitude of peer effects on CSR and the contexts in which such effects are stronger or weaker. The work highlights the strategic and societal importance of CSR and addresses the mixed evidence on its benefits by focusing on external, industry-level drivers of CSR adoption.

Prior research offers mixed views on CSR, ranging from it being a cost to shareholders to being a strategic asset enhancing reputation, loyalty, and performance. Stakeholder pressures from investors, consumers, media, communities, and interest groups shape CSR. A broad literature documents peer effects in corporate policies (financing, dividends, misconduct, reporting), suggesting firms mimic peers due to shared environments and competition. Institutional theory (mimetic, normative, coercive isomorphism) predicts firms emulate peers’ practices to gain legitimacy and resources. CSR may therefore diffuse via peer imitation, reinforced by executive reputation concerns. Building on these insights and prior evidence of CSR peer effects, the paper formulates the hypothesis that peer firms’ CSR positively influences focal firms’ CSR performance (H1).

Data: CSR measures are from MSCI ESG Stats (KLD); accounting data from Compustat; stock returns from CRSP; analyst coverage from IBES; institutional holdings from Thomson 13F; and industry/product market similarity from the Hoberg and Phillips Data Library (TNIC). Sample: U.S.-listed firms from 2000–2015, excluding financials (SIC 6000–6999) and utilities (SIC 4000–4949), with necessary data available. Final sample comprises 2,593 firms and 16,634 firm-year observations. Variables are winsorized at the 1st and 99th percentiles by year. CSR construction: Following Lins et al. (2017), CSR is the sum across five stakeholder-oriented categories (community, diversity, employee relations, environment, human rights) of each category’s score defined as average strengths minus average concerns (each scaled 0–1), yielding an overall CSR in [−5, 5]. For robustness, an alternative six-category measure includes product quality, and dimension-level CSR measures are also constructed. Peer CSR: Main measure is the average CSR of other firms in the same 3-digit SIC industry (excluding the focal firm). Robustness uses 2-digit SIC and TNIC-based peers (TNIC-2, TNIC-3), with peer sets defined by text-based product-description similarity above median or 75th percentile. Regression design: Main specification is a firm- and year-fixed-effects model predicting next-year CSR (CSR_{i,t+1}) from contemporaneous Peer CSR_{i,t}, controlling for: firm size (MV), leverage, cash, capital expenditures (CAPX), dividends, institutional ownership, market-to-book (MTB), return on assets (ROA), sales growth, and analyst coverage. Additional analyses assess moderating effects of: geographic proximity (log median HQ distance to peers), salient peers (SP500 indicator for any peer in S&P 500), peers’ relative size, focal firm characteristics (market share, profitability, global operations, analyst coverage), industry competition (1−HHI and TNIC similarity), and CEO reputation (tenure). Robustness checks include alternative peer/CSR measures, instrumental variables (peer firms’ idiosyncratic returns) to address endogeneity/reflection concerns, quantile regressions, and a dynamic analysis using changes in CSR and Peer CSR.

- Baseline peer effect: Peer CSR positively predicts focal firms’ next-year CSR. Main coefficient ~0.204 (t≈6.00), implying that a one standard deviation increase in peer CSR is associated with about 10.15% of a standard deviation increase in focal CSR (0.204 × 0.390 / 0.784). - Contexts strengthening peer effects (peer characteristics): Effects are stronger when peers are geographically closer (negative interaction with log distance), include salient peers (SP500 × Peer CSR positive and significant), and when peers are relatively larger than the focal firm (positive interaction with relative size). - Contexts weakening peer effects (focal firm characteristics): Effects are weaker for industry leaders (higher market share), more profitable firms (higher ROA), and multinational firms (Global × Peer CSR negative). Peer effects are stronger under greater external monitoring (more analyst coverage). - Industry competition: Peer effects are amplified in more competitive environments; Competition × Peer CSR interactions are positive and significant using both 1−HHI (3-digit SIC) and TNIC similarity measures. - CEO reputation: Using CEO tenure as a proxy, interactions with Peer CSR are insignificant, suggesting CEO reputation concerns (as proxied) do not materially moderate peer effects. - Economic consequences of following peers’ CSR increases: A Follow indicator (increase in a firm’s CSR when peers increased CSR in the prior year) is associated with lower subsequent product market share by ~0.2 percentage points and lower ROA by ~0.3 percentage points (t+1 to t+3 averages), but higher firm value (Tobin’s Q) by ~3.4%.

Findings support institutional theory and mimetic isomorphism: firms observe and emulate peers’ CSR practices, particularly under strong competitive pressures, when peers are salient, proximate, or larger. The moderating patterns suggest that firms with weaker market positions, lower profitability, or greater external monitoring are more susceptible to peer influence, while industry leaders and multinationals are less inclined to follow. Although mimicking peers’ CSR may entail short-run product market and earnings costs, it is associated with improved firm value, indicating capital market recognition or longer-horizon benefits. Robustness checks using alternative peer definitions, CSR measures, instrumental variables, quantile regressions, and dynamic specifications reinforce the main conclusions and mitigate concerns about measurement error and endogeneity.

The study documents a robust, positive peer effect in CSR: higher peer CSR leads firms to raise their own CSR, with stronger effects when peers are closer, salient (S&P 500), and larger, and weaker for industry leaders, more profitable firms, and multinationals. Product market competition intensifies imitation, while CEO tenure does not significantly moderate the effect. Following peers’ CSR boosts valuation despite modest declines in product market share and ROA, suggesting firms should avoid blindly copying peers and instead invest strategically in CSR. Policy implications include roles for industry associations to catalyze CSR diffusion and for public policies that subsidize or incentivize CSR to amplify positive spillovers.

- Potential endogeneity and reflection problems in peer effects are acknowledged; addressed via instrumental variables (peer idiosyncratic returns) but residual concerns may remain. - Measurement error in CSR and peer identification is possible; mitigated through alternative CSR constructions (including six-category and component-level measures) and multiple peer definitions (SIC and TNIC), yet construct validity limitations persist. - Sample restricted to U.S.-listed firms from 2000–2015 and excludes financials and utilities, which may limit generalizability across countries, time periods, or regulated sectors. - Some dimensions (e.g., diversity) show weaker or insignificant peer effects, indicating heterogeneity across CSR categories. - CEO reputation is proxied only by tenure, which may not capture broader reputation constructs.