Scottish Longitudinal Study
Development & Support Unit
Occupational Mobility and Neighbourhood Effects in Scotland
Evan Williams (University of Glasgow)
Nick Bailey (University of Glasgow)
5 May 2015
The notion of a neighbourhood effect alludes to the claim that the neighbourhood in which somebody lives has an independent impact on their life chances over and above their individual characteristics . There is a large body of observational studies that investigate the existence of neighbourhood effects on a variety of individual outcomes, but the majority suffer from endogeneity . The central challenge is to distinguish between causal neighbourhood effects and self-selection into neighbourhoods, since both processes might explain an observed correlation between neighbourhood context and individual outcomes .
With this issue in mind, the SLS has been used in two studies to investigate the relationship between neighbourhood context and employment outcomes [4, 5]. By disaggregating their regressions for homeowners and social renters, these studies find significant neighbourhood effects for homeowners but not for social renters. Social renters in Scotland in the early 1990s were, in general, randomly allocated by housing officers, and so the observed difference between social renters and homeowners is arguably explainable by selection effects for the latter group.
There is little neighbourhood effects research, however, that explicitly investigates the link between neighbourhood context and occupational outcomes. Existing research that does explore the link between the two, furthermore, does not attend to the issue of causal neighbourhood mechanisms or selection effects . In contrast, van Ham and Manley  use the SLS to consider neighbourhood effects on occupational mobility. Disaggregating their regressions by tenure, they argue that selection effects explain the finding that the correlation between characteristics of the residential environment and occupational mobility is strongest for homeowners.
The proposed research contributes to this literature by testing the finding that selection effects – as opposed to neighbourhood effects – explains the difference between social renters and homeowners in Scotland. Using a distinct measure of occupational status and a revised modelling strategy, the study will compare results from the 1991-2001 period with those from the 2001-2011 period. The random allocation of social renters that arguably occurred in the early 1990s was soon replaced by the use of points-based allocation mechanisms, in addition to the choice-based agenda that emerged towards the end of the decade . This suggests that there should exist an observable distinction between the two time periods, which will allow the proposed research to better inform the neighbourhood effects/selection effects debate.
The study aims to investigate the impact of neighbourhood characteristics on individuals’ occupational outcomes. Specifically, the project will ask:
(i) To what extent do 1991 neighbourhood characteristics influence the probability that those who are low-skilled in 1991 are high-skilled in 2001?
(ii) To what extent do 1991 neighbourhood characteristics influence the probability that those who are high-skilled in 1991 are low-skilled in 2001?
Regressions (i) and (ii) will be disaggregated according to certain individual and household characteristics, such as age or housing tenure, in order to consider how neighbourhood effects vary by such factors. A second set of questions will ask:
(iii) To what extent do 2001 neighbourhood characteristics influence the probability that those who are low-skilled in 2001 are high-skilled in 2011?
(iv) To what extent do 2001 neighbourhood characteristics influence the probability that those who are high-skilled in 2001 are low-skilled in 2011?
Regressions (iii) and (iv) will also be disaggregated by certain individual and household characteristics. The outcomes from regressions (i)-(ii) and (iii)-(iv) will be compared, and the results used to inform discussion around the identification of causal neighbourhood effects. This will be particularly useful for those regressions disaggregated by housing tenure, which have the potential to provide useful information on the relative importance of neighbourhood versus selection effects.
1. van Ham, M., Manley, D., Simpson, L., Bailey, N. and Maclennan, D. (2012). Neighbourhood Effects Research: New Perspectives. In M. van Ham, D. Manley, L. Simpson, N. Bailey, & D. Maclennan (eds.), Neighbourhood effect research: New perspectives. Dordrecht: Springer
2. Manley, D., van Ham, M. and Doherty, J. (2012). Social mixing as a cure for negative neighbourhood effects: evidence-based policy or urban myth. In: G. Bridge, T. Butler and L. Lees (eds.), Mixed communities: gentrification by stealth? Bristol: Policy Press
3. Durlauf, S. (2004). Neighborhood effects. In: J. V. Henderson and J. F. Thisse (eds.), Handbook of regional and urban economics, vol. 4, cities and geography. Amsterdam: Elsevier
4. van Ham, M. and Manley, D. (2010). The effect of neighbourhood housing tenure mix on labour market outcomes: a longitudinal investigation of neighbourhood effects. Journal of Economic Geography, 10(2): 257-282
5. Manley, D., & van Ham, M. (2012). Neighbourhood effects, housing tenure and individual employment outcomes. In: M. van Ham, D. Manley, L. Simpson, N. Bailey, & D. Maclennan (eds.), Neighbourhood effect research: New perspectives. Dordrecht: Springer
6. Musterd, S., Ostendorf, W. and De Vos, S. (2003). Neighbourhood effects and social mobility: a longitudinal analysis. Housing studies, 18(6): 877-892
7. van Ham, M. and Manley, D. (2014). Occupational Mobility and Living in Deprived Neighbourhoods: Housing Tenure Differences in ‘Neighbourhood Effects’. Applied Spatial Analysis and Policy: 1-16
8. Pawson, H. and Kintrea, K. (2002). Part of the problem or part of the solution? Social housing allocation policies and social exclusion in Britain. Journal of Social Policy, 31(4): 643-67