Macroeconomists had been kicking around at the idea of a very long propagation mechanism through the 60’s and 70’s – the techniques were hard, and not many believed the results.
The big paper that got everyone’s attention was Nelson and Plosser’s piece in the Journal of Monetary Economics in 1982. Nelson was, and still is, at the University of Washington. Plosser was at the University of Rochester, but he’s now President of the Federal Reserve Bank of Philadelphia, and this year he’s a voting member of the Federal Open Market Committee (the Federal Reserve System committee that decides monetary policy).
They tested a variety of macroeconomic data for unit roots, and concluded that most of the series had them. A unit root means that the sum of the coefficients on the lags in an autoregression is one (i.e., a unit), and it implies that impulses to that series have permanent effects. The spreadsheet work I did in class last Wednesday is an extension of what they did.
There’s an interview with Plosser, and what he thinks about macroeconomic policy in the February 12th issue of The Wall Street Journal.