The nature of employment and excess mortality in Glasgow and Scotland

Abstract

Background At 77 years for males and 81 years for females, Scotland has the lowest life expectancy in Western Europe. The ‘excess mortality’ (higher mortality over and above that which is explained by higher levels of deprivation) exists in all geographical areas and at all levels of deprivation, but is most pronounced in the most deprived areas, particularly in Glasgow and the west of Scotland. Recent work has focused on comparing the cities of Glasgow, Liverpool and Manchester, where deprivation levels have been identified as being similar, to try and better understand the causes of this excess mortality. Despite these similarities in socioeconomic deprivation, Glasgow has consistently shown excess mortality compared to its English neighbours. A number of hypotheses have been proposed to help explain this excess mortality, ranging from differences in poor health behaviours (such as alcohol consumption, smoking, low physical activity and poor diet) to lower social capital and even climate. Although many of these factors have been empirically explored, much of the excess mortality remains unexplained. Differences in historical and contemporary employment conditions (for example, types of occupations, occupational health and safety, occupational precariousness, trade union membership, unemployment etc.) remain as possible, yet under-researched, explanations for this excess mortality. Aims and objectives The overall aim of this project is to assess whether—and to what extent— aspects of the labour market and employment differ between Scottish and English populations, specifically between those living in Glasgow compared to residents of Liverpool and Manchester. Objectives include: • Develop plausible hypotheses about how employment may contribute to Glasgow and Scotland’s excess mortality • Bring together knowledge on employment and work characteristics to understand which aspects have differed over the last 60 years between the three cities of Glasgow, Liverpool and Manchester • Assess the feasibility of conducting quantitative analyses to test the hypotheses linking employment and health inequalities Method A. Plausible Theories Linking Employment & Excess Mortality We developed hypotheses as to how the nature of employment may contribute to Scotland and Glasgow’s excess mortality based on reviewing relevant articles and reports and distilling the key possible employment characteristics important for health. B. Literature Review A rapid review of the literature on differences in employment conditions and peoples’ experiences of the labour market in Glasgow, Liverpool and Manchester was carried out. We used four approaches to identify relevant articles. First, an electronic search was conducted, described in detail below. Second, the reference sections of articles deemed suitable for review were scrutinised. Third, each team member checked his or her own files for relevant material. Finally, experts in the field were contacted. This multidimensional approach was used to help identify articles not readily identified through traditional database searching alone. Search terms included a number of words/phrases relevant to the theories identified in section A and relevant place terms for Glasgow/Clydeside, Liverpool/Merseyside and Manchester. After applying relevant inclusion and exclusion to filter the number of studies down to those most relevant to the research question, we then extracted information on study participants (location, sample size, age range, sex, study design), employment characteristics, data type (e.g. quantitative versus qualitative) and main results from all relevant fulltexts. C. Identify and Appraise Datasets for Further Analyses We aimed to identify and appraise existing datasets that could be used to explore similarities and differences in factors related to the labour market between the three cities/regions. The same employment search terms identified for the literature review were used to search the UK Data Service to aid identification of datasets that may allow further analysis of Scotland and Glasgow’s excess mortality by employment characteristics. In addition, the research team used their knowledge and checked their own files for relevant datasets. Information on study type (sample, time dimensions and method of data collection), locations, year(s) covered, demographics (population type, representativeness, ages, sexes included and sample size), relevant employment data and relevant health data (morbidity, symptoms, general health and wellbeing, medications, biomarkers and administratively linked health records) were extracted from the Data Service online data catalogues, data documentation and through searching the data files. Datasets were then assessed on their strengths and limitations according to nine criteria across study quality and relevance to the research topic. Summary of main or key results A. Plausible Theories Linking Employment & Excess Mortality Seven characteristics were identified: overall employment levels; job types; precarious employment; pay; physical work environment; psychosocial work environment; and employee representation. B. Literature Review The literature review focused on identifying literature that had included any of these characteristics and the three cities. Fifty-one records were identified, with the focus for most being on employment status, job type and pay, with only 11 looking at other employment characteristics. Overall, there was little evidence that Glasgow suffered worse employment features compared to Liverpool and Manchester across multiple datasets and evidence sources for these employment characteristics, although there were notable gaps for precariousness, the physical and psychosocial work environments and employee representation. Across employment levels, job types and pay, data was typically limited to the last 20-30 years with single measurements multiple years apart, but even where data stretched back further, the similarities across the three cities/regions were notable. There remains a lack of research investigating precariousness, the physical and psychosocial work environments and employee representation as possible links between employment and excess mortality in Glasgow. C. Identify and Appraise Datasets for Further Analyses There were a number of datasets identified that could be available to investigate this issue further (28 identified here), although none focus on the three cities specifically, or even Liverpool and Manchester alone (there is one Glasgowbased study) to allow cross-cohort comparisons. However, many datasets do have UK data with local authority region identifiers, but have not been used to compare regions within the literature identified in this report given the need for specific requests for regional data and decreasing sample sizes at the lower city level. All the datasets contained data on employment status and/or job types, but there was much more limited and less consistent data for the other measures, as seen with the literature review. Some basic data descriptives were extracted from freely available data looking at employment status (time-series aggregate-level data) and pay (cross-sectional health survey data, 2003), but again this data is limited in being relatively modern and/or only available at one time-point. However, as seen with the literature review, few differences between the three cities were identified for employment levels, job types and pay. Conclusions This comprehensive and novel review has highlighted that while there is currently insufficient evidence to link employment as a major explanatory factor for the excess in mortality in Glasgow, there are gaps in the research that could be explored further. Previous research and hypotheses have perhaps failed to fully acknowledge the complex and nuanced nature of the term ‘employment’, meaning that important features such as precariousness and the psychosocial work environment have been neglected in terms of data collection and research. However, emerging data in the UK does hold the potential to explore this topic in more detail, but it may require many years to fully link employment exposures to health outcomes.