Why birthplace matters so much: Sorting, learning and geography

Clément Bosquet, Henry Overman

29 April 2016



The possibility that place of residence has an influence on life chances has been a long standing concern in debates about inequality and public policy. In particular, a large and diverse literature has considered the effects of living in a deprived neighbourhood, looking for effects on education, crime, health, and labour market outcomes (e.g. Oreopoulos 2003, Durlauf 2004).

More recently, a small but growing body of literature has focused on the role of ‘initial conditions’ in determining labour market outcomes (Aslung and Rooth 2007). This new line of research doesn’t just look for an effect of current place of residence, or of conditions during childhood, but instead reaches further back in to the past to consider the effect of place and time of birth.

New evidence

In a recent paper (Bosquet and Overman 2016), we consider a particular aspect of this question by looking at whether birthplace plays a role in determining future earnings.

We focus particularly on birthplace size to try to answer whether being born in a bigger city improves earning potential.

For those of us who study the economics of cities, this is an interesting question because we have good evidence that people who live in bigger cities earn more than similar people living in smaller cities. This ‘urban wage premium’ is explained by what urban economists call ‘agglomeration economies’ whereby density makes firms and workers more productive because of labour market pooling, input sharing, and knowledge spillovers. These ideas originate with Marshall (1920) but still underpin lively academic and policy debates today.

Using the British Household Panel Survey (BHPS), we find an elasticity of wages with respect to birthplace size of 4.6%.

What this means is that on average, an individual born in London in the 1970s will earn around 7% more than an individual of the same age and gender born in Manchester; who in turn will earn 5.5% more than an individual born in Cardiff.

What could explain these effects of birthplace size on future earnings?

One possibility is that individual characteristics vary with birthplace size because of the location decisions of different types of parents and the intergenerational transmission of characteristics (‘sorting’). Indeed, the ‘urban wage premium’ literature highlights that much of the wage gap between urban and rural areas and between large and small cities is due precisely to this kind of sorting – specifically, the concentration of more productive workers in bigger cities.

A second possibility is that birthplace size somehow affects the accumulation of human capital – for example because the quality of schools varies with city size (‘learning’).

A third possibility is that birthplace influences future location decisions and through this future labour market opportunities (‘geography’). Indeed, in the extreme case of no mobility, birthplace size directly determines future labour market size and it makes little sense to try to distinguish between the effect of birthplace and current location. We consider all three of these possibilities in our research.

Our findings suggest that inter-generational transmission (sorting) and the effect of birthplace on current location (geography) both play a role in explaining the effect of birthplace. 

Effects via learning depend on when we think such learning takes place. We don’t find evidence of an effect during childhood, but there may be an effect though accumulated experience later in life (which we might think of as adult rather than childhood learning).

We find strong evidence of parental sorting in the BHPS data: 79% of people born to parents in ‘professional’ occupations are born in a city, compared with only 72% of those with unskilled parents. For London alone, the figures are 12% and 6%, respectively. The differences are, if anything, more pronounced when it comes to city size: on average, birthplace city size is 50% larger for individuals with ‘professional’ parents compared to individuals with ‘unskilled’ parents.

Parental sorting is also an important consideration when it comes to the link between birthplace size and educational outcomes. People born in big cities undertake more years of education, but this link to city size is also explained by parental sorting. Differences in education play no additional role in explaining the effect of birthplace size once we allow for the fact that parental characteristics vary with city size.

Finally, we show that birthplace city size also has an effect because it determines current location. This matters because, as we discussed above, we have good evidence of a link between wages and the size of the city in which an individual is currently working. This link from birthplace to current city size isn’t simply driven by people who don’t move. For those that work somewhere other than where they were born, current city size is positively correlated with birthplace size (consistent with a number of anecdotal observations about the differences between small-town and big-city mentalities). Interestingly, if we assume that accumulating experience in big cities also has a wage payoff, as proposed by de la Roca and Puga (2015), then this reintroduces a role for learning (at least in adulthood, rather than childhood).

Taken together, our results highlight the importance of intergenerational sorting in helping explain the persistence of spatial disparities. Low lifetime mobility reinforces the link between the location decisions of generations. We provide descriptive evidence on lifetime immobility that suggests this is an important consideration in the UK – in our data around 44% of individuals only ever work while living in the same area as they were born. In addition to immobility, as discussed above, even for those that do move there is a positive correlation between birthplace size and size of place of residence. Our findings also highlight that persistence extends across generations – nearly 54% of individuals have the same place of birth as their mother suggesting an intergenerational transmission of birthplace that is even larger than the lifetime immobility rates for work-related reasons.

What are the policy implications?

It is important to note that:

  • Nothing we document in our research downplays the importance of trying to address the intergenerational transmission of education, health, marital status, wealth, income and other labour market outcomes such as employment and occupation;
  • Many of these effects are large and, as we have demonstrated in other research (Gibbons, Overman and Pelkonen 2014), individual factors are much more important than geographical ones in determining earnings; and
  • Our research does suggest there may be an additional role for policies that try to increase geographical mobility, especially of deprived households – at least in terms of improving labour market outcomes.

Of course, in Britain (as elsewhere) the resulting improvement in nominal wages can be easily offset by more expensive housing and other disamenities of big city living. Our research strengthens the arguments for trying to deal with these – the higher mobility they allow is one way to overcome the labour market disadvantages that come from being born far from the bright lights of the big city.


Aslund, O, and D-O Rooth (2007), “Do when and where matter? Initial labour market conditions and immigrant earnings.” The Economic Journal 117(518(03)), 422-448

Bosquet, C and H G Overman (2016), “Why does birthplace matter so much? Sorting, learning and geography.” CEPR Discussion Paper 11085

Chetty, R, N Hendren and L F Katz. (2016), “The effects of exposure to better neighborhoods on children: New evidence from the Moving to Opportunity experiment.” American Economic Review 106(4), 855-902

De la Roca, J and D Puga (2015), “Learning by working in big cities”, forthcoming Review of Economic Studies

Durlauf, S N (2004), “Neighbourhood Effects.” in J V Henderson and J-F Thisse , Handbook of Urban and Regional Economics, Vol. 4, Amsterdam: Elsevier Science, pp. 2173-2242

Gibbons, S, H G Overman, and P Pelkonen (2014), “Area disparities in Britain: Understanding the contribution of people vs. place through variance decompositions.” Oxford Bulletin of Economics and Statistics 76(5), 745-763

Marshall, A (1920), Principles of economics, London: MacMillan.

Oreopoulos, P (2003), “The long run consequences of growing up in a poor neighbourhood.” Quarterly Journal of Economics 118(4), 1533-1575.



Topics:  Labour markets Poverty and income inequality

Tags:  Inequality, productivity, unemployment

Assistant Professor of Economics, University of Cergy-Pontoise (ThEMA)

Professor of Economic Geography, London School of Economics