This case study illustrates how research can explore potential structural drivers of unequal experiences/outcomes when are not readily available in secondary data sources. 

This was a particular challenge for public health expert Ben Barr and his colleagues in their research on the health impact of austerity policies (i.e. cuts in public expenditure) implemented in England following the 2008 financial collapse. One of their studies sought to determine whether English regions that experienced the worst economic hardship had the worst health outcomes, in terms of increases in suicides (Barr et al., 2012).

To explore this question Barr and colleagues utilised data on temporal changes in unemployment between 2000 and 2010 as an indicator of the impact of austerity. The hypothesis they tested was whether areas with greater increase in unemployment also had greater increases in suicides. If this hypothesis was confirmed, it would be evidence that austerity was associated with increased mortality and challenge employer-friendly policies that weaken employment security.

However, the team faced significant methodological issues including 1) annual statistical fluctuations in suicide data owing to the small numbers involved and variations in the practice of coroners; 2) inaccurate/inconsistent unemployment data.  In response, Barr and his colleagues included deaths from undetermined injuries in the data from suicide to provide “a more consistent measure of all deaths that are likely to be suicides”.  They also used data on the number of people claiming unemployment benefits within each region, provided by the Office for National Statistics. They recognised that this may understate true unemployment in a period of austerity but judged it to be “the most precise and consistent measure that is officially recorded in all regions.”

Furthermore, instead of aggregated data commonly used in previous research in this area, which they argue lacks statistical power to identify underlying structural factors, Barr et al. used regional data variations. These they suggested had “the benefit of more consistent surveillance systems [and ensured that] national level factors, such as legislation changing access to the means of suicide, are kept constant”. This approach also allowed them to illuminate subregional variations in unemployment and suicide rates, previously understudied. However, they acknowledged that there could still be groups in local populations that are most vulnerable to the effects of unemployment that they did not have data on, such as people with low levels of education or with pre-existing mental health problems.