Social networks play an important role in the labor market. Various surveys document that 30-60% of jobs are found through friends or relatives. To better understand how networks operate in the labor market, I examine how networks that were formed involuntarily as a result of the American Civil War and the First World War draft affect the postwar labor market outcomes of veterans in 1880, 1900, and 1930. My study uses two data sets. The first, contains new data on 1,295 drafted American Infantrymen who served together overseas during World War I, and was formed by matching military service records, prewar draft records, and postwar information from the 1930 Census, as well as information on up to sixty of a veteran's nearest neighbors in 1930. The second, collected by Fogel et al. (2000), matches 35,570 Civil War veterans to postwar censuses. I exploit the time-series feature of the Union army sample and eliminate all unobserved individual and group-level fixed effects. For both samples, the military unit's overall unemployment rate has a negative and statistically significant effect on a veteran's own likelihood of employment. The findings are consistent with a model in which information about job vacancies is communicated through the network. Both samples are fairly representative of the white working-age male population, therefore contributing to the external validity of the results. I introduce a new framework which allows one to further decompose the social effect into its two components, the endogenous ("the effect of others' outcomes"), and the contextual ("the effect of others' characteristics"). I show that the two effects are separately identified, provided that some people belong to more than one group. I apply the framework using two types of reference groups for each veteran, those who had served in his unit and his neighbors. I find the endogenous effect to be much stronger than the contextual effect, indicating the presence of a large social multiplier: a change in an individual's employment propagates through the network and affects the employment of others. The framework is also applicable in other settings, since in many cases individuals are potentially affected by multiple types of reference groups

Last modified

• 05/28/2018

Creator

• Ron Laschever

DOI

• https://doi.org/10.21985/N26H5F

Subject

• Economics

Keyword

• Employment
• First World War Veterans
• Identification
• Social Interactions
• Civil War Veterans
• Multiple Reference Groups

Date created

• 2007-05-10

Resource type

• Dissertation

Rights statement

• In Copyright