
Impact of Factory Inspections on Understanding of and Compliance with the Factories Ordinance in Sri Lanka
By: T. H. K. Liyanage and A. M. P. Arandara
Abstract
In Sri Lanka, occupational safety and health legislation is implemented by the factories ordinance (FO), which is enforced by factory inspecting engineers through factory inspection and partly by awareness programs. Awareness programs are not frequently conducted, and sometimes do not cover every type of industry. Therefore, the main way to enforce FO is through the inspection of factories. The objective of inspections is to make them more compliant with the provisions of the FO, which may result in making workplaces safer and reducing work-related accidents. However, although previous studies have shown that the size of the firm, safety culture, and financial capability affect safety performance in factories, there are no studies on safety and health inspections and how they affect the understanding of occupational safety, health legislation, and compliance with legislation. Therefore, this study aims to evaluate the impact of workplace inspections on the understanding of and compliance with FO.
This study uses primary data collected through a questionnaire during the inspection of 102 factories as the unit of analysis. Among the 102 factories in the study, 50 factories had been inspected (treatment) previously, and the remaining 52 had not been inspected (control) during the last five years. The outcome variables are understanding of and compliance with the FO. Details such as the registration of the factory under the Board of Investment, the size of the factory, the nature of the industry, and the gender, age, and work experience of the respondents were also collected for use as covariates. Because the factories for inspection were not chosen randomly, a selection bias was present. To address this issue, the study utilizes the propensity score matching technique to estimate data that closely match factories from both the treatment and control groups based on similar characteristics. The estimation results reveal that inspections have a positive effect on the understanding of FO and compliance with it in factories where officers in charge of inspected factories have 2.406 higher scores (out of eight) compared to officers in the noninspected factories. Similarly, the inspected factories had approximately 1.872 (out of 9) higher compliance scores than non-inspected factories. This is based on the results provided by one of the matching methods, the nearest neighbour, and these results are consistent with the radius matching, kernel matching, and coarsened exact matching results that were conducted as a robustness check.
This study uses primary data collected through a questionnaire during the inspection of 102 factories as the unit of analysis. Among the 102 factories in the study, 50 factories had been inspected (treatment) previously, and the remaining 52 had not been inspected (control) during the last five years. The outcome variables are understanding of and compliance with the FO. Details such as the registration of the factory under the Board of Investment, the size of the factory, the nature of the industry, and the gender, age, and work experience of the respondents were also collected for use as covariates. Because the factories for inspection were not chosen randomly, a selection bias was present. To address this issue, the study utilizes the propensity score matching technique to estimate data that closely match factories from both the treatment and control groups based on similar characteristics. The estimation results reveal that inspections have a positive effect on the understanding of FO and compliance with it in factories where officers in charge of inspected factories have 2.406 higher scores (out of eight) compared to officers in the noninspected factories. Similarly, the inspected factories had approximately 1.872 (out of 9) higher compliance scores than non-inspected factories. This is based on the results provided by one of the matching methods, the nearest neighbour, and these results are consistent with the radius matching, kernel matching, and coarsened exact matching results that were conducted as a robustness check.
DOI: https://doi.org/10.4038/sljda.v7i1.7154 | Journal eISSN: 1391-8214
Language: English
Page range: 115 - 156
Published on: Jul 29, 2025
Published by: Sri Lanka Institute of Development Administration
In partnership with: Paradigm Publishing Services
Keywords:
© 2025 T. H. K. Liyanage, A. M. P. Arandara, published by Sri Lanka Institute of Development Administration
This work is licensed under the Creative Commons Attribution 4.0 License.