Coautores: Kevin Donovan (Yale) y Terence Johns on (Notre Dame)

Abstract:

"An increasingly utilized class of gen
eral equilibrium models includes inter-firm knowledge spillovers through di
ffusion. Standard methods to calibrate critical diffusion parameters requir
e making assumptions about the economic environment, then using the resulti
ng structure to map these parameters onto more easily observed empirical mo
ments. Within this class of models, we prove that randomly varying interact
ions uniquely identifies a small set of parameters characterizing the diffu
sion process independent of the remaining economic environment. We provide
an application of our results in Kenya, where we conduct a randomized contr
olled trial matching firms from the left tail of the profit distribution to
those from the right. Despite matching the quick fade out of the small-sca
le experimental treatment effect, the model simultaneously implies a large
general equilibrium diffusion externality. Key is that critical parameters
push the partial and general equilibrium magnitudes in different directions
. This matters: if a policy-maker selected economies in which to implement
optimal policy based solely on the magnitude of their experimental impact,
she would in fact minimize the possible welfare gains. Thus, the ability to
properly estimate such parameters is critical not only for measuring the e
quilibrium importance of diffusion but alsofor the interpretation and extra
polation of smaller-scale empirical studies."

DTSTAMP:20191023T024118Z DTSTART:20191024T173000Z DTEND:20191024T184500Z SEQUENCE:0 TRANSP:OPAQUE END:VEVENT END:VCALENDAR