This report is built on the approach in Grini and Johnsen (2021), which studied earnings levels and attributed connected with low pay for all employments, including apprentices. In this publication we focus on workers who are connected to the labour force only as employees, meaning that those who combine work with education (which includes apprentices), or participation in some welfare program, are disregarded.
Amongst jobs in the lower end of the earnings distribution, workers are typically employed in a small number of occupations and industries, like accommodation and food services, wholesale and retail, and health and social work. Jobs in these industries are predominantly in Service and sales, and often have low educational requirements, are part-time, and more frequently are temporary positions as opposed to a contract with indefinite duration.
Tips, as a component in earnings, and membership in a labour union have not been presented in statistics on earnings before. In this /publication we find that tips are concentrated in low paying jobs, and especially in accommodation and food services. The prevalence of union membership, as reported in the Tax Returns, shows that membership decreases in the lower pay brackets compared with membership rates in the middle of the earnings distribution.
The OECD publishes statistics on earnings for full-time employees, among many statistics, one is a measure of low pay frequency. Low pay frequency measures the frequency of jobs with earnings below two thirds of median earnings (OECD, 2023). Statistics Norway uses a calculation of full-time equivalents, so part-time jobs can compare with full-time jobs when it comes to earnings. In this /publication we utilize OECD’s measure, but we do not limit our study to just full-time work. The result is that our study gives a more comprehensive overview of how jobs in the low end of the earnings distribution are composed compared to jobs above the two thirds threshold.
Finally, we use carefully selected statistical models that provide results in accordance with the descriptive findings that industry and several other variables correlate with jobs being situated in, either at a point in time or over several years, the low end of the earnings spectrum. We use statistical models in two ways. The first is concentrated on the description of the effect different variables have on the movement in the earnings distribution from 2017 to 2022. We also introduce a model that tries to address which variables are most correlated with staying in low paying jobs. In this last case we work towards the study of the probability of staying in such jobs.