Using Big Data to Make Policy Decisions

Congress and the Biden Administration are trying to figure out who should get a new round of stimulus. Many Republicans, and some Democracts, have expressed reservations about sending checks to everyone and want to target more of those in need. Very good idea. Except how to go about doing it?

Three economists, Raj Chetty, John Friedman and Michael Stepner, have a suggestion: use big data. Using data from financial service companies, they were able to track the spending behvior of high and low income households following the reciept of their January 2021 stimulus checks. 

What the economists found was that spending activity diverged dramatically before and after the stimulus by economic strata. The chart below illustrates the effect (click to enlarge): 

The authors calcuate that in the first month after the initial $600 stimulus checks were received, households with annual income above $78,000 spent only $45 of the $600; while housesholds making less than $78,000 spent $235 of the $600. 

Why? Because, as Chetty, etal show, most high-income workers got their jobs back fairly quickly after the economic shutdown when corporations shifted work online. And they mainly used the government aid to pad their savings. By contrast unemployment for low-wage workers continues to persist and resulted in lower spending prior to getting the stimulus checks,

Chetty, etal estimate that high-income households will spend only $105 of the additional $1,400 of government aid they receive. That means $200 billion of additional government expediture will lead to only $15 billion of additional spending. Not much of a money multiplier. Targeting additional aid to only low-income households, who need it, will be far more effective. Hope government considers Chetty, Friedman and Stepner's analysis.