Exposures, mechanisms and outcomes: which is which?

Authors: Gerry McCartney, Heather Brown, Jayne Galinsky, Michaela Mazzei, Jennifer McLean, Neil Craig

In research, an ‘exposure’ is typically understood as something that influences outcomes — but in practice, when does it remain an exposure, and when does it become a ‘mechanism’ or even an outcome itself? Welcome to the dilemmas we face as researchers from the CoWBELLS project, seeking to understand the impacts of Community Wealth Building (CWB) - a local economic development strategy which seeks to share economic wealth more equitably – as it is introduced in Scotland. 

CWB involves actions by local 'anchor institutions' (such as local authorities, the NHS, universities and colleges) and others across five areas: employment; procurement; plural ownership of the economy; better use of land and assets, and of financial resources. It is primarily an economic strategy orientated to address issues such as poverty, unemployment, inequality and economic growth through increasing the distribution, circulation, and democratic ownership of wealth within local economies. However, given the importance of the economy for health, and the potential for CWB to reshape economies, it is of huge interest to those of us concerned about public health. 

CWB is an intervention where researchers do not control its implementation, and which is complex and interacts with multiple other systems. To evaluate (so-called) ‘natural experiments’ like this, an early step is to develop a theory of how it will work based on evidence and the insights of people working in the area. This means detailing the likely ways in which policy actions will impact via various mechanisms on a series of intermediate outcomes, which ultimately change health. In the case of CWB, this might be changes in procurement that encourage more local  businesses to supply local hospitals, leading to more employment in the local economy, leading to improvements in health. Although complex, many aspects of the theory we have developed are familiar to public health researchers. We have long theorised and researched how changes in incomes, employment, working conditions and other aspects of the economy impact on health. Yet we have found ourselves in much more tricky territory in developing our theory at the 'left side' of our diagrams: the bit detailing the CWB policy.

Why has this been so problematic? After all, CWB is (reasonably) clearly defined, and the links to the economy are transparent...

Put simply, how do we know and how can we tell where, when, and to what extent, CWB has been implemented across different areas of Scotland? Is it when organisations (often local authorities) publish a CWB plan or strategy? Or might some organisations have implemented something very like CWB, but without using that term? Or does it begin when new approaches to local procurement and employment are introduced? Or is it when there are changes in how financial resources flow and how people work? Or is the real test more tangible – improvements in jobs or green space, increases in the amount of community ownership of land and assets, more community groups, higher incomes, or improved working conditions in the communities exposed? 

This is not a trivial issue. Defining what CWB is, and what it means to have 'implemented' CWB, is critical to any evaluation. To understand the impacts on the economy and health, we need to decide and be able to distinguish which populations are 'exposed' to CWB and which are not. This is where it gets tricky. If we define 'exposure' to CWB in terms of economic outcomes, we risk confusing the exposure and the outcome (i.e. cause and effect). However, if we take a lighter-touch definition – such as the publication of a CWB plan, strategy, or appointment of a CWB 'officer', we may end up classifying communities or including populations in our 'exposed' group who have not actually experienced any change in policies or practices ‘on the ground’ as a result of CWB, biasing the evaluation. 

So what should we do? There is no simple answer, but fortunately it is a problem we anticipated.

To distinguish between populations that have and have not been exposed to CWB over time, we plan to interview and survey key informants across Scotland. This will help us build a clearer picture of when, and to what extent, CWB has been implemented in each local authority area, which we can then compare to any differences in economic or health outcomes between areas. Using this mixed-methods approach we aim to disentangle rhetorical commitment to CWB (or explicit non-implementation) from actual changes in policy and practice. In doing so, we can keep our measures of CWB exposure distinct from measures of economic change - as an important outcome itself, but also as the mechanisms or key pathways linking CWB to health outcomes. Even so, there is no perfect solution. The risks of misunderstanding where and when different populations have been exposed to CWB will remain a challenge for the evaluation.

You can find more about the study, who is involved, and the outputs so far, on the CoWBELLS website: www.cowbellsstudy.com