Abstract
Indonesian laws mandate that every employer should provide health insurance and work accident insurance to their employees. Nevertheless, there is a significant gap in the coverage of employer-sponsored insurance among Indonesian workers. This study examines the coverage of employer-sponsored insurance and work accident insurance and analyses the characteristics of the uninsured working population in Indonesia. We analysed nationally representative cross-sectional data from the National Labour Force Survey (NLFS) conducted between 2018 and 2022. The primary dependent variable was the provision of health insurance and work accident insurance by employers. The independent variables included having any physical disabilities, number of working hours, duration of employment, labour union membership, earning at least the provincial minimum wage, having a written contract and working in high risk jobs. Logistic regression was employed using the R statistical software. The findings indicate that coverage of employer-sponsored health insurance is low in Indonesia-ranging from 36.1% in 2018 to 38.4% in 2022. Workers with a written contract, earning at least the provincial minimum wage, were members of a labour union, employed for at least 5 years and working more than 40 hours a week were more likely to be insured. By contrast, workers who had physical disabilities or were employed in high-risk jobs were less likely to be insured. Our study concludes that having a written employment contract is the single most influential factor that explains the provision of employer-sponsored health insurance in Indonesia. The country's labour laws should therefore formalize the provision of written employment contracts for all workers regardless of the type and nature of work. The existing laws on health insurance and work accident insurance should be enforced to ensure that employers meet their constitutionally mandated obligation of providing these types of insurance to their workers, particularly those engaged in high risk jobs.