This paper develops an empirical model of dynamic interactions within an employment network with two sectors, a loosely organized nonprofit sector, and a for-profit one. Entry by agents into the network, their turnover within it, along with their quitting from the network, endogenously determines the creation, growth, decline and disappearance of the establishments that employ them. We apply the model to the real estate market in a large North American metropolitan area and provide novel evidence on the network patterns and their role in determining agents productivity, turnover and organizational structure. Networking naturally arises in our framework because agents must team together in order to produce. Repeated interactions increase human capital by facilitating future production. Consistent with agents' preference for complementary and internal links, we find that agents whose skills are better matched to the rest of the office are more productive and stable. Given their flexible compensation structure, franchised brokerage offices exhibit greater allocative efficiency in agent composition than non-franchised ones. We then provide conditions that nonparametrically identify the distribution of unobserved variables, in order to estimate the role that networks play in individual career decisions, the evolving distribution of establishment size and the relative importance of the two sectors.
Please use the following link to set up a time to meet with Lu.
Lunch will be provided in room 2-90 ten minutes prior to the seminar.
For more information, please visit the Stern CREFR website.
For questions, please contact Annelise Bell