Repository logo

An Analysis of the causes of construction accidents in South Africa: a case study approach.

Thumbnail Image



Journal Title

Journal ISSN

Volume Title



It is well-known that construction represents a challenging regime in which to manage health and safety exacerbated by the enormous diversity in terms of the size and range of its activities (HSE, 2001). Health and safety improvements in the workplace are developed and shaped because of the knowledge gained and assumptions made from accidents that have occurred (Gibb, Lingard, Behm and Cooke, 2014). Understanding what causes accidents is important as it helps distinguish between factors that require attention and remedial action and factors that are unimportant and can be ignored (Swuste, 2008). There are several theories that exist created by scholars in attempts to provide understanding of the causation of accidents on construction sites. Some of these theories are discussed in this study, namely:  Accident Proneness Theory;  Goals-Freedom-Alertness Theory;  Adjustment-Stress Theory;  Distractions Theory;  Chain of events (Domino and Updated Domino theory);  Multiple Causation Model;  Reason’s Framework for Accident Causation;  Constraint-Response Theory;  Human Error Theories; and  Systemic accident model Despite these theories, accidents have continued unabated. Typically, these theories have focused on the construction worker as being the primary cause of accidents – a basic tenet of the behavioural safety approach espoused by Krause and Hidley (1990) and others. According to Whittington et al. (1992), emphasis on individual failures resulted in a reliance on short-term solutions rather than uncovering more fundamental management or organisational problems. Generally, the proposed remedy targeted a specific event or operative, such that no effort was made to uncover the underlying cause of the accident. The HSE (2001) observed that changes at the direct level alone would not deliver the degree of change being sought, nor would the resultant improvement be sustained. This study utilised a qualitative research approach and a combination of descriptive and analytical research methods namely both questionnaires and case studies to analyse the problem statement. The sample design used in this study is based on quota sampling as a sample of any 30 lost time accidents investigation reports were required from the large participating construction company for selection and analysis for this study. A further sample of 10 Contractors and Health and Safety Officers were surveyed to test if the findings of the case study were in line with what these professionals actually experience on site. A sample of 30 accident investigation reports were analysed by categorising every cause identified in these reports according to the relevant accident causation theories to determine which theories most prevalently or most frequently would have identified the causes as shown in the construction company’s accident investigation reports. The frequency of each accident causation theory was analysed to determine if the identified causes of the accidents were focused on the actions and failures of workers or management and if the remedial actions taken were correct and were able to prevent the accident from reoccurring according to the theories they were classified under. The causes of the 30 construction accidents as stated in the respective accident investigation reports were classified into three categories namely: Direct Causes, Contributory Causes and Root Causes to determine if the root causes were in fact identified The findings of the case study indicated that the construction accident investigation reports are flawed as they typically fault the negligence of workers as being the root cause of construction accidents. The majority of the construction professionals surveyed were of the opinion that accident investigation, although effective, can be improved. However, the case study findings indicate that accident investigation processes and methods are ineffective as 83% of accident investigation reports incorrectly identified direct and contributory causes which the suggested remedial action was not based on root causes.


Masters Degree. University of KwaZulu-Natal, Durban.