When weak sanctioning systems work: Evidence from auto insurance industry fraud investigations

Danielle E. Warren, Maurice E. Schweitzer

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


To deter auto insurance fraud, insurance companies and law enforcement agencies investigate and prosecute suspicious claims. We describe this sanctioning system and perceptions of this system by integrating unique datasets: insurance company records, interviews with insurance fraud investigators, state law enforcement data (CA, NY), and surveys of automotive insurance customers. We identify organizational constraints, such as public relations concerns, that limit the effectiveness of the formal sanctioning system (fewer than 1% of claims that are flagged as suspicious are ever prosecuted for fraud). We also identify psychological factors that deter consumers from committing fraud; consumers over-estimate the probability of detection, over-estimate the consequences of prosecution, are sensitive to social sanctions (e.g., negative publicity), and anticipate high emotional costs, such as shame and embarrassment, that make the prospect of committing fraud highly aversive. That is, psychological factors substantially deter fraud even though the economic sanctions are weak. Our findings integrate scholarship on sanctioning systems (Tenbrunsel & Messick, 1991) and highlight the role of organizational constraints and psychological factors in deterring fraud.

Original languageEnglish (US)
Pages (from-to)68-83
Number of pages16
JournalOrganizational Behavior and Human Decision Processes
StatePublished - Sep 2021

All Science Journal Classification (ASJC) codes

  • Applied Psychology
  • Organizational Behavior and Human Resource Management


  • Deception detection
  • Emotion
  • Fraud
  • Risk perception
  • Sanctioning systems


Dive into the research topics of 'When weak sanctioning systems work: Evidence from auto insurance industry fraud investigations'. Together they form a unique fingerprint.

Cite this