Measuring dynamic inefficiency in the presence of corporate social responsibility and input indivisibilities

Magdalena Kapelko*, Alfons Oude Lansink, Spiro E. Stefanou

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

10 Citations (Scopus)

Abstract

This article proposes a model for evaluating inefficiency accounting for firms’ corporate social responsibility engagement as part of their broader output production activities. The model combines the production of marketable outputs, socially responsible outputs, and undesirable outputs into overall measures of firm performance using data envelopment approaches. Methodologically, the article builds on the dynamic by-production model, which accounts for adjustment costs related with investments, allowing for non-convexities of the production set and input indivisibility, as well as firm corporate social responsibility activities. This study compares dynamic technical inefficiency scores for each input, output, and investment, estimated assuming the presence of input indivisibility (non-convexity) and its absence (convexity). The empirical application focuses on European firms in three industries (offering capital, consumption, and other goods) for the period 2010–2017. The results show significant differences between inefficiencies with and without the convexity assumption and find evidence for non-convexity of firms’ production set and inputs’ indivisibility. Overall, among all outputs, the results reveal the highest inefficiency in the production of socially responsible outputs.

Original languageEnglish
Article number114849
JournalExpert Systems with Applications
Volume176
DOIs
Publication statusPublished - 15 Aug 2021

Keywords

  • Corporate social responsibility
  • Data envelopment analysis
  • Non-convexity
  • Undesirable outputs

Fingerprint

Dive into the research topics of 'Measuring dynamic inefficiency in the presence of corporate social responsibility and input indivisibilities'. Together they form a unique fingerprint.

Cite this