There’s a complex and dynamic world of data, incentives and decisions that lies behind many of our daily activities, from each web purchase you make to how you interact with various government policies.
That world is where Alexandre Belloni operates.
Belloni is a professor of decision sciences at Duke’s Fuqua School of Business. He’s interested in how you respond to incentives, treatments and offers. More precisely, on measuring the causal impacts of interventions — like ballot design on election outcomes — and on designing systems to account for participants’ incentives, like procurement auctions with budget constraints.
You may not know what made you buy a product at a particular price; but he probably knows how likely it was for you to do it.
Belloni studied electrical engineering in his native Brazil before earning a master’s degree in mathematical economics there as well. He then came to the United States and earned a doctorate in operations research from MIT. He joined Duke in 2007 and has carved out a niche as an expert in economics, operations research and statistics.
For one project, Belloni is studying how to estimate the effect of interventions/treatments in the presence of self-selection and interference across participants. With interference between individuals, even the widely accepted randomized A/B testing would be biased. In another project Belloni considers designing an auction in situations where we can monitor and reward participants that bid accurate values leading to better allocations.
In 2019 he became an Amazon Scholar, a program that recruits academics to turn their scholarship into practice at the online marketplace and help Amazon solve difficult technical problems.
One problem he has tackled: the need for Amazon to provide enough – but not excessive – storage space for goods sold by third-party sellers on the platform.
