

If I have one quibble with the paper, it's that it passes too quickly over the standard problems of time series analysis (what's the counterfactual growth path? which approach do you use to deal with seasonality?), and underplays the ability to learn about the through microdata - either via studies that exploit random or quasi-random timing of spending, or those that ask households whether they spent or saved the money. Here's an op-ed and an Australian paper with more details. And this is my favourite US paper combining both micro approaches. Fortunately, the micro approach backs up Chris's finding from macrodata: fiscal stimulus was a very good deal for the Australian taxpayer.
Do you like this post?