The Northern Ireland Longitudinal Study (NILS) as a resource for investigating neighbourhood peer effects: A case study using natural experiment(s) in fertility and labour market participation

Our neighbours can potentially influence our behaviour. For instance, poor health behaviours amongst neighbours may normalise and reinforce our poor health behaviours. This is an example of a peer effect. Imitative behaviour can cause small initial changes in individual behaviour to spread amongst their social networks and result in a social multiplier effect. Understanding the size and mechanisms behind the social multiplier effect allows for more effective health interventions. It also helps us understand why persistent health inequalities exist across different neighbourhoods and social groups. In an ideal experiment, we would randomly allocate people into treatment and control groups and change the behaviours of persons A in the treatment group (directly or through incentives). Then we would observe the effects of changes in person A's behaviour on their neighbour person B. This ideal experiment is practically and (possibly) ethically unfeasible. Furthermore, data on a large enough sample of people and their neighbours is very expensive to collect. Instead of a normal experiment, we can use natural experiments which change person A's behaviour. In this paper, we use a well-known natural experiment that affects women's fertility and, indirectly, their labour market participation. Furthermore, data on people and their neighbours are available from the Northern Ireland Longitudinal Study (NILS). The Northern Ireland Longitudinal Study (NILS) is a longitudinal study consisting of 28% of the NI census and roughly 50% of all households. The NILS is linked to a database of all addressable properties in NI which include the coordinates of residences. In theory, NILS contains a large sample of households in NI and their close neighbours that can be used for studying peer effects. This project is a proof of concept for studying peer effects using NILS. If any random intervention exists (e.g. natural, quasi- or actual randomised trial) then NILS can always be used to study peer effects (amongst neighbours). This is significant because peer effects are notoriously hard to study due data limitations and a large number of credible natural experiments exist in health research. This project's contribution is the discovery that NILS almost uniquely placed as resource for studying peer effects in the UK.