The Effects of Ex-slave Repatriation on Indigenous and Immigrant Networks in Liberia

 

Research Question

How did the early social networks amongst early Liberian settlers affect long-term development outcomes?

 

Background

After slave trade was abolished in the early 1800s, many slaves brought to the United States from the colonies were repatriated in West Africa. Over a period of 50 years, Liberia transformed from a slave colony, used primarily to export slaves to the new world, into an ex-slave colony. The initial settlers were mostly freed slaves who assumed important roles as emissaries from the US, commissioned by the American Colonization Society. Americo-Liberians, as they came to be known, distinguished themselves from the indigenous populations by creating institutions modeled on those they had become accustomed to in the US. Soon after the declaration of independence in 1847, the Republic of Liberia became the first independent African state with a new “elite” or “ruling” class of individuals that divided the society into various factions. Identity became a focal issue in the division of power between different factions: Americo-Liberians, though few in number, came to dominate politically and financially. These divisions in society are often cited as the main reason for Liberia’s two civil wars.

Objective

Using social network analysis of census data collected prior to Liberia’s independence, I hope to achieve three things:

a)      Examine how the network evolved over time as new settlers were repatriated into the colony from the US;

b)      Analyze the determinants underlying the determinants of factions over time and some of the basic elements that distinguished these factions;

c)      Hypothesize and test whether there are any links to the dynamics of the social networks and the long-term development outcomes of different areas in Liberia.

 

Methodology

The 1843 census of Liberia has basic socio-demographic data along with a module that records whom each person in the census knew. I intend to create two separate datasets: a one-node matrix with (directional) relationships between the full sample of individuals in the census and an attributes file with individual node characteristics. I will then examine the natural sub-groups (factions) that form and generate group level measures on centrality and closeness.

 

Hypothesis

Long-run outcomes are poorer in places where there are factions forming in the social network.

 

Potential limitations

The endogeneity of the social network makes it an empirical challenge to identify the causal impacts of factions on long-term development. Nevertheless, an analysis of the evolution of the network over time and (potentially across space) may make provide useful insight into the complex hierarchies that have formed within networks and between groups.