Is the social origin pay gap bigger than we thought? Identifying and acknowledging workers with undefined social origins in survey data

This Bulletin summaries the findings from: Vallely, M., Findlay, J. and Hermannsson, K. “Is the social origin pay gap bigger than we thought? Identifying and acknowledging workers with undefined social origins in survey data”. Research in Social Stratification and Mobility. Available at: https://doi.org/10.1016/j.rssm.2024.100952

Introduction

Surveys on labour market outcomes now ask more detailed questions about people’s social origin, allowing researchers to study how someone’s social origin affects their earnings. The data show that even after taking people’s education and occupational status into account, earnings are influenced by the occupational status of the previous generation of their family. Social origin pay gaps of this sort have been observed in different datasets across several high-income countries leading some to refer to a ‘class ceiling’. The present research investigates whether previous empirical studies have underestimated the social origin pay gap by omitting respondents whose social origins could not be defined. Specifically, it focuses on individuals who were not assigned a social origin in the data because they were not living with their family, nobody was earning in the household, or the main wage earner’s occupation could not be identified. 

Data and methods 

We use the UK Labour Force Survey (LFS) between 2014, when questions about social origin were first included, through to 2021. To assign social origin, respondents were first asked if they were living with their family at age 14. If so, they were then asked to identify a main wage earner in the household. If respondents identified a main wage earner, they were then asked about that person’s occupation. Thus, the resulting dataset has no indicator of social origin for individuals who were not living with their family when they were 14, those who could not identify a main wage earner and those for whom the main wage earner’s occupation could not be matched to the Standard Occupational Classification (SOC). We investigated the characteristics and earnings of these individuals with missing social origin information.

Results

People with undefined social origins in the data (comprising 10.5% of the working age population) were disproportionately likely to be social disadvantaged, as indicated by their educational attainment, housing tenure, employment status and earnings. Estimating class pay gaps with people with undefined social origins omitted from the calculation results in significant underestimation. We find that individuals who were not living with their family at age 14 experience a pay gap of 11.4%, compared to individuals from higher managerial and professional origins. For those whose parent(s) occupation could not be matched to a SOC code, the estimate is 7.9% and for those who could not identify a main wage earner it is 7.4%; all larger than any other group with defined social origins. Of those with defined social origins, the largest pay gap was 6.4% for individual’s whose parent (main earner in the household) worked as a process, plant or machine operative.

Conclusions

Our results highlight how respondents from non-traditional backgrounds are not captured by the indicators introduced in the LFS, with consequences for our understanding of pay gaps. Individuals with undefined social origins have several characteristics which indicate disadvantage in education, housing and employment. Those who do not fit the occupational classification experience the largest pay gap. They share similar material outcomes to those from working class origins but are demographically different and do not fit well into an occupation-based social class schema.