How online public attention shapes corporate resilience

Technical Summary: How Online Public Attention Shapes Corporate Resilience

Problem Statement
In the context of the digital era and the deep penetration of social media, online public attention has emerged as a critical external factor influencing corporate behavior. While existing literature extensively documents the linear impacts of online public opinion on short-term corporate activities—such as innovation, investment preferences, and financial reporting—there is a significant theoretical gap regarding its role in shaping corporate resilience. Corporate resilience, defined as the dynamic capability to absorb shocks, adapt to changes, and achieve sustainable growth during crises, relies on both internal resource reserves and the ability to navigate external uncertainties. This study addresses the question of how online public attention affects this long-term dynamic capability, specifically investigating whether the relationship is linear or non-linear and identifying the underlying mechanisms and boundary conditions.

Methodology
The study employs an empirical analysis using a sample of China's A-share listed non-financial enterprises from 2008 to 2022.

• Data Sources: Listed company data are drawn from the CSMAR database, while online public attention data are sourced from the China Research Data Service Platform (CNRDS), specifically utilizing the "Guba" stock discussion forum.
• Variable Measurement:
  • Corporate Resilience (Dependent Variable): Measured following Ortiz-De-Mandojana and Bansal (2016) and Bai (2009). It is constructed by estimating the deviation of firm operating performance (revenue growth) caused by external shocks, capturing the magnitude and direction of performance deviation.
  • Online Public Attention (Independent Variable): Quantified as the natural logarithm of one plus the total occurrences of a company's name or stock code in investor posts on its dedicated Guba sub-forum within a given year.
  • Mediating Variables:
    • Information Disclosure Quality: Measured using the KV index (Kim and Verrecchia), which quantifies the covariation between stock trading volume and return rates. A higher KV index indicates lower disclosure quality.
    • Resource Allocation Efficiency: Evaluated using the DEA-Malmquist index model, considering inputs (labor, materials, capital) and outputs (economic and social contributions).
• Econometric Strategy:
  • Baseline Model: A regression model including firm and year fixed effects, with control variables for firm size, age, leverage, ownership concentration, board size, and CEO duality.
  • Non-linearity Test: The squared term of online public attention is introduced to test for a U-shaped relationship. The U-test (U-shaped test) is employed to confirm the existence and location of the inflection point.
  • Endogeneity Control: To address potential endogeneity, the study utilizes an Instrumental Variable (IV) approach (using the mean public attention of other firms in the same province) and the Heckman two-stage selection model to correct for sample selection bias.
  • Mechanism Analysis: A two-step mediation effect approach is used to verify the roles of information disclosure quality and resource allocation efficiency.
  • Heterogeneity Analysis: Sub-sample regressions and cross-group coefficient tests (using the suest model) examine variations based on public opinion sentiment (positive vs. negative), retail investor ownership, industry listing density, and managerial decision-making horizons.

Key Contributions
The paper makes three primary contributions to the literature:

1. Non-linear Relationship: It challenges conventional linear assumptions by uncovering a U-shaped relationship between online public attention and corporate resilience. Resilience initially declines as attention intensifies (due to market pressure) but eventually increases after crossing a specific threshold (due to supervisory governance effects).
2. Mechanism Identification: It elucidates the causal pathways, demonstrating that online public attention influences resilience through two mediating channels: information disclosure quality and resource allocation efficiency.
3. Contextual Heterogeneity: It identifies specific boundary conditions that strengthen or weaken the impact of public attention, including the sentiment of the opinion, the proportion of retail investors, industry listing density, and the time horizon of management decision-making.

Results

• U-Shaped Effect: The benchmark regression and U-test confirm a significant U-shaped relationship. The inflection point is calculated at an online public attention level of approximately 9.918. Below this threshold, attention suppresses resilience; above it, attention enhances resilience.
• Mechanisms:
  • Information Disclosure: High attention drives firms to improve disclosure quality (lowering the KV index), which reduces information asymmetry and stabilizes investor expectations.
  • Resource Allocation: High attention compels firms to optimize resource allocation efficiency, shifting focus from short-term myopic investments to strategic, long-term capabilities.
• Heterogeneity Findings:
  • Sentiment: The U-shaped effect is more pronounced under positive public opinion compared to negative sentiment.
  • Ownership: Firms with high retail investor ownership exhibit a stronger supervisory effect from public attention.
  • Industry Context: The effect is stronger in industries with a high proportion of listed companies, where information contagion and competitive linkages are more intense.
  • Management Horizon: Firms with longer managerial decision-making horizons show a more pronounced U-shaped response, suggesting they are better equipped to leverage high-intensity attention for strategic restructuring.

Significance and Claims
The paper claims to provide theoretical support for understanding how enterprises can balance public opinion management with long-term development in the digital age. By identifying the "U-shaped" threshold, the study suggests that low-intensity attention acts as a market pressure that induces myopic behavior, while high-intensity attention functions as a supervisory governance tool that forces substantive organizational learning and capability restructuring.

Practically, the findings offer insights for:

• Enterprises: To establish dynamic public opinion management mechanisms that avoid short-termism traps when attention is low and leverage high attention for crisis monitoring and strategic adjustment.
• Policymakers: To guide public opinion governance by recognizing that moderate to high levels of public scrutiny can drive corporate transparency and efficient resource allocation, thereby enhancing systemic economic resilience.

The authors maintain modesty regarding their claims, acknowledging limitations such as the need for more granular micro-level analysis (e.g., business divisions) and the potential for future research to refine the measurement of public attention by incorporating sentiment analysis and network diffusion characteristics.