Kieran Marray

Background

Phd candidate at Department of Spatial Economics, Vrije Universiteit Amsterdam and the Tinbergen Institute. Supervised by Michael Konig, and Ozan Candogan (University of Chicago).

Fellow in POPulation-scale social NETwork analysis project at the Institute for Advanced Studies, University of Amsterdam.

Interests

Econometrics of networks, economics of networks, machine learning for economics.

Selected research

Estimating spillovers from sampled connections (working paper on ArXiv )

Keywords: Networks, Sampling, Peer Effects

Summary: Show when sampling networks biases spillover estimates from linear and nonlinear regression models. Propose unbiased estimators based on aggregate network statistics.

Optimal disease mitigation with endogenous network response (with Michael Konig, Ozan Candogan, and Frank Takes).

Keywords: Networks, Epidemics, Rewiring

Summary: Introduce model of spread of disease on network where susceptibles adjust connections in response to spread of disease, and derive tractable optimal policy problem with many groups. Estimate for Covid-19 in Netherlands using new population-level social network data.

Teaching

Econometrics 1, Tinbergen Institute (TA)

Phd level introductory econometrics course. Covers OLS, maximum likelihood, GMM, testing, discrete and multinomial choice models.

Lecture notes I wrote on R for students.

Urban economic challenges and policies, VU Economics (TA)

Masters course covering applied and structural econometrics for policy evaluation on urban economics.

Course website with interactive notebooks.

Software

PowerLawSamplers.jl (repo)

Random graphs with power law degree distributions and samplers for discrete power-law distributions

GraphicalLassos.jl (in progress)

Pure julia implementation of graphical lasso and path graphical lasso algorithms