ABSTRACT
The purpose of this study is to investigate the moderating effect of slack resources (namely, profitability and cash holdings) on the relationship between corporate social responsibility disclosure (CSRD) and firm market value. The authors test the hypothesis through performing panel data analysis for a sample of 95 non-financial Jordanian firms listed on the Amman Stock Exchange from the period 2011 to 2016. Content analysis was employed to evaluate the level of CSRD based on an index of 42 items and four themes. While the findings show no significant relationship between CSRD and firm market value, higher slack resources were found to positively moderate this relationship. The outcomes therefore suggest that slack resources accentuate the effect of CSRD on firm market value. The use of Jordanian firms is distinctive and provides valuable insight into contexts in which the market tends to place less weight on a firm's CSR activities.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Correction Statement
This article has been corrected with minor changes. These changes do not impact the academic content of the article.
Notes
1 We take the natural logarithm of the market to book ratio to approximate its normal distribution and thus consistency with the basic assumptions of a linear regression model (i.e., linearize the relatiosnhip) (Wooldridge Citation2020).
2 We identified one firm as being influential and excluded it from the analysis. The firm had extreme values of the PRESS statistic, and likelihood distance. Excluding it reduced the sample to 95 firms (570 observations). The inclusion of the firm in the further analysis does not change the results presented in the study.
3 Under ANTE(1), the residual within firms is assumed to be heterogeneous across time (Heck, Thomas, and Tabata Citation2013). e ANTE(1) assumes that the correlation between any two observations depends on that between any two adjacent observations of repeated measurements (Lit tell et al. Citation2006). Heteroscedasticity was checked by the test proposed by White (Citation1980). The null hypothesis of constant variance is rejected, χ2 (167) = 261.34, p < .001; the residuals do exhibit heteroskedastic variance. Moreover, the Wooldridge (Citation2020) test for serial correlation shows that the residuals exhibit serial dependency. The resulting test statistics are F(1, 95) = 53.405, p < .001. These were dealt with using the ANTE (1) covariance structure, which allows the residual to be both heterogeneous and serially correlated.
4 This was tested using the Hausman test. The resulting Chi-squared statistic (χ² = 5.22) allows us to accept the null hypothesis that both fixed and random effects models provided efficient and constant estimates of the model's parameters (Wooldridge Citation2020). In the manuscript, we present the results of a fixed effects model () which are identical to those obtained when using a random effects model.