Saving lives with smarter decisions: Operations Research in animal shelters

Joining me today is Katalia Chen, recipient of INFORMS Doing Good with Good O.R/ Award, which recognizes outstanding work using operations research to create meaningful social impact.

Katalia’s research tackles a question that affects hundreds of thousands of animals every year: how can we help animal shelters make better decisions when resources like kennel space, staff time, and foster homes are limited?

By combining operations research, optimization, survival analysis, and causal inference, her work demonstrates that smarter allocation of these scarce resources can dramatically increase the number of lives saved – without building new facilities or increasing capacity.

We find fostering actually improves the adoption probability, or probability of positive outcome, by about 25%. The reason why we started this research is we found by running a simple regression, we know the fostering effect is actually not scientific enough. There exists endogeneity – which is a very typical word in empirical research – meaning there is selection bias for foster dogs. For example, some people might favor cute dogs and they might want to foster dogs that are already highly likely to get adopted. On the other hand, the shelter managers might put dogs that need the help most – for example medical dogs – in the fostering pool. So that made the selection bias very obvious. That's why we want to find instrumental variables that could solve the endogeneity problem. And we did multiple robustness check to make sure that our conclusion is robust. And this 25% makes sense, we did get that conclusion.

Interviewed this episode:

Related Episodes