Organisational culture is assumed to be a key factor in large-scale and avoidable institutional failures (e.g. accidents, corruption). Whilst models such as “ethical culture” and “safety culture” have been used to explain such failures, minimal research has investigated their ability to do so, and a single and unified model of the role of culture in institutional failures is lacking. To address this, we systematically identified case study articles investigating the relationship between culture and institutional failures relating to ethics and risk management (n = 74). A content analysis of the cultural factors leading to failures found 23 common factors and a common sequential pattern. First, culture is described as causing practices that develop into institutional failure (e.g. poor prioritisation, ineffective management, inadequate training). Second, and usually sequentially related to causal culture, culture is also used to describe the problems of correction: how people, in most cases, had the opportunity to correct a problem and avert failure, but did not take appropriate action (e.g. listening and responding to employee concerns). It was established that most of the cultural factors identified in the case studies were consistent with survey-based models of safety culture and ethical culture. Failures of safety and ethics also largely involve the same causal and corrective factors of culture, although some aspects of culture more frequently precede certain outcome types (e.g. management not listening to warnings more commonly precedes a loss of human life). We propose that the distinction between causal and corrective culture can form the basis of a unified (combining both ethical and safety culture literatures) and generalisable model of organisational failure.

Investigations of institutional failure in healthcare typically use staff narratives to identify the cultural factors contributing to the incident. But, to what extent can staff, who are embedded in the culture and who were part of the failing, reflect on and report on the culture? We investigate this by comparing 40 witness statements from staff and 53 witness statements from patients and relatives collected by a public inquiry into a major UK healthcare failure (Clostridium difficile outbreak). Through quantitative text analysis, we found that, while staff and external stakeholders both recognised problems in care, they diverged on the factors considered paramount. Staff emphasised underlying factors such as under‐resourcing and training (causal culture), while patients and relatives emphasised corrective behaviours such as communication for identifying and taking precautions against the spread of C. difficile (corrective culture). The results indicate that patients and relatives may be able to report on cultural factors that staff do not report or are unaware of, thus allowing a more complete analysis. Even in light of an institutional failure, staff may have incomplete accounts of the contributing cultural factors, with implications for learning and postincident improvement.