Differences in the structure of social networks can have significant implications for people''s accumulation of human capital for themselves and their children – and, as a consequence, for growth and inequality. That is the central message of research by Tiago Cavalcanti and Chryssi Giannitsarou, published in the August 2017 issue of the Economic Journal.
For example, the study shows how inequality may persist in less cohesive and more segregated societies. But in societies with more network cohesion, social groups with different human capital will eventually end up interacting with each other, resulting in less inequality over time, though usually at the cost of lower economic growth.
The researchers'' analysis indicates that the human capital of an individual grows faster when it is relatively low compared with the average human capital of the household''s network neighbours. In other words, a low-income household located in a relatively affluent neighbourhood – being ''a small fish in a big pond'' – can make more progress than when it is located in low-income neighbourhoods.
Such results have far-reaching implications for a range of social policies, including those related to the design of school catchment areas, school vouchers and affordable housing policy.
More…
What are the effects of social structures on the dynamics of human capital acquisition, growth and inequality? How does the design of school catchment areas affect the distribution of education and its dynamics? Can such dynamics be driven by peer effects and local externalities that depend on how people interact in society? These are some of the questions addressed in the new study.
In the economic environment that the researchers analyse, the human capital of children depends on how much their parents invest in their education. But it is also dependent on a local externality, represented by the human capital of others with whom they interact.
The novelty of the researchers'' approach is to use network theory to describe social structures. This allows them to study how a variety of network structures affect investment in education and the dynamics of an economy. This is an important addition to previous research, which has typically only considered very stylised social structures, such as segregated versus integrated societies.
The approach in the new study is much more flexible and allows the researchers to analyse many different groups of households, which can be connected through different networks. Through this innovative methodology, Cavalcanti and Giannitsarou are able to reach striking results about social stratification, individual and aggregate outcomes.
First, they find that heterogeneity in human capital may imply outcomes in which inequality persists over time. Importantly, this result does not rely on credit market imperfections or discontinuous choice of human capital (for example, college versus non-college), two of the standard mechanisms proposed in previous research.
Instead, their mechanism works solely through the local externality that can be relatively positive or negative, depending on whom individuals interact with. In societies that have less cohesion and are potentially more segregated, inequality may perpetuate over time.
If, however, the network that describes the social structure has enough cohesion, social groups with different human capital will eventually end up interacting with each other, resulting in less heterogeneity in human capital and less inequality over time, but usually at the cost of lower economic growth.
How much network cohesion is required to avoid inequality depends on the incentives households have to invest privately in the education of their children. If education is provided as a public good and households have less urgency to invest individually in education, then the network externality effect for human capital accumulation becomes relatively more important and less cohesion is required to smooth out differences across households.
This is a powerful result because it links a simple, one-dimensional network statistic (network cohesion) to long-run outcomes with low income dispersion. When network cohesion is low relative to the importance of investment in education for accumulating human capital, heterogeneity of human capital of individuals and inequality persist in the long run.
The researchers also find that the human capital of an individual grows faster when it is relatively low compared with the average human capital of the household''s network neighbours – that is, whenever the household is a small fish in a big pond.
Their theory predicts that by being located in a relatively affluent neighbourhood, a low-income household can grow faster than when it is located in low-income neighbourhoods. Therefore, younger generations have more chances of climbing upwards in the distribution ranking.
Recent empirical evidence supports exactly this view: that when young children from low-income households relocate with their families to high-income areas, they are very likely to have substantially higher income when they become adults than their parents, therefore increasing their chances of upward social mobility.
''Growth and Human Capital: A Network Approach'' by Tiago Cavalcanti and Chryssi Giannitsarou is published in the August 2017 issue of the Economic Journal. The authors are at the University of Cambridge.
