Labor Networks

The problem of employment is, in a sense, a search problem. Workers are looking for places to work, and employers are looking for competent employees that can perform productive work for them. Implicit in this search is the need for matching the skills and interests of the prospective worker, and the technical needs and interests of the prospective employer. So how can we model this?

Together with my colleagues Omar Guerrero and Rob Axtell, I have developed a mathematical framework that addresses this problem in a new way. We begin with the concept of the Labor Flow Network which is a network that capture the likely job transitions that individuals can perform between any pair of firms in an economy. This network captures the affinity that relates different firms with each other in the sense that individuals are likely to be useful employees to firms that are connected. In this framework, job mobility is modelled as a set of walks that workers perform inside the network. Labor flow networks are constructed from microdata that is specific down individual forms and workers, and therefore provides a tremendously realistic picture of employment in an economy.

I use methods of stochastic processes on graphs to solve the equations of the model, and obtain very interesting results. Among those, I am able to reproduce the firm-size distribution of companies, something that has, up to now, been explained with other models that mostly lack microscopic level data. The economic models that explain the important parameters of the stochastic process also provide explanations for the origin of the so-called employer-size premium, and offer a new way to look at unemployment. From the model, we have also built an agent-based simulation platform to help us visualize concrete situations, and explore the consequences of particular policy and economic shocks.

Our framework differs from the broad picture of currently used models, because those utilize partial or complete aggregation for available workforce and vacancies, and therefore, loose information. The different, high resolution nature of our approach establishes a new way to develop microfoundations (code word for a micro to macro mechanism that explains a process) of job search and matching process that is specific down to the level of firms and individuals. This level of detail provides our approach with the potential to change the way in which economic shocks and policy incentives relevant to employment are understood.

Select publications:

1. Firm-to-firm labor flows and the aggregate matching function: A network-based test using employer–employee matched records

2. The Network Picture of Labor Flow

3. The Network Composition of Aggregate Unemployment