Too much of economics is pure theory with no data. Take a look at James Hamilton’s commentary at Econobrowser on a paper by Chevalier and Kashyap. The authors actually used grocery store price data. What variability is hidden behind national economic data?