Contextual consolidation and intergroup inequality
Author(s): Martin Arvidsson
Thursday 15 | 13:20-13:40
Room: TP52
Session: Frontiers in segregation research
A large body of research in the social sciences demonstrates how social contacts across various social domains—households, schools, workplaces, neighbourhoods—significantly influence individuals in numerous ways, including their health, educational attainment, and productivity, to name a few. Because individuals are non-randomly distributed across social contexts, these influence effects can have important consequences for intergroup inequality: High segregation levels mean that some individuals receive many positive exposures, while others receive few.
Although many important contributions have been made within this broad area of research, a pervasive limitation is its treatment of social contexts as isolated entities. Illustrative of this is that each domain has its own, separate literature, focused on its causes (e.g., residential segregation, workplace segregation) and consequences (e.g., neighbourhood effects, workplace premiums) of segregation. However, we still know rather little about how these different domains and social contexts integrate into a larger whole.
This domain-centric perspective has two key limiting effects on our understanding of how segregation impacts intergroup inequality. First, it overlooks that an individual may be differentially segregated across various domains. Second, it ignores that individuals, by being situated in multiple domains simultaneously (for example, working in workplace A and living in household X), create linkages between these domains (e.g., WPA-HHX) that allow for spillover effects. Such linkages and indirect exposures are overlooked by viewing social contexts as isolated.
To address these limitations and understand the joint exposure inequality resulting from individuals being situated in multiple social contexts, I draw upon the concept of consolidation proposed by Peter Blau and colleagues; extending it to apply to social contexts rather than individual attributes. More specifically, I argue that, in order to understand how segregation affects the inequality of exposures between different groups, we must consider how consolidated the exposures from the different domains are. For instance, knowing all my colleagues are vaccinated against COVID-19, what can we infer about my sibling’ vaccination status? Under high (low) consolidation, the answer is a lot (not very much).
Using a combination of computational and empirical analyses, I demonstrate the key role that contextual consolidation (CC) plays in shaping intergroup inequality. With simulations, I first show that—counterintuitively—exposure inequality can shift substantially without any change in segregation levels within individual domains, merely due to variation in CC. Second, while exposure inequality reduces monotonically with decreasing CC, these simulations also show that when ties within contexts form based on homophily, too low a CC results in crosscontextual ties that are not sufficiently reinforced to sustain the diffusion of behaviours. This suggests that, for behavioural adoptions as opposed to mere exposures, a moderate CC level may more effectively reduce inequality than a very low level. Lastly, using Swedish administrative register data, I demonstrate how CC constitutes a key mechanism for accounting of how exposure inequality has changed in Sweden in the past two decades.