{"id":10608,"date":"2015-02-24T05:00:17","date_gmt":"2015-02-24T10:00:17","guid":{"rendered":"http:\/\/marcfbellemare.com\/wordpress\/?p=10608"},"modified":"2015-02-23T12:23:55","modified_gmt":"2015-02-23T17:23:55","slug":"lagged-explanatory-variables-and-the-estimation-of-causal-effects","status":"publish","type":"post","link":"https:\/\/marcfbellemare.com\/wordpress\/10608","title":{"rendered":"Lagged Explanatory Variables and the Estimation of Causal Effects"},"content":{"rendered":"

…without careful arguments on substantive grounds, lagged explanatory variables should never be used for identification
\npurposes.<\/p><\/blockquote>\n

This is from a new working paper<\/a> of mine, coauthored with Cornell’s\u00a0Tom Pepinsky<\/a>, an associate professor of government,\u00a0and\u00a0Taka Masaki<\/a>, a PhD student in government, on “lag identification.” This is a very common strategy to avoid problems of endogeneity or reverse causality, especially in political science, although table 1\u00a0in our paper\u00a0shows that this is not exactly uncommon in economics.<\/p>\n

We expose the assumption that underlies this strategy, which we term “no dynamics\u00a0among unobservables.” We also argue that this assumption is almost impossible to defend on substantive grounds, because it requires knowledge about the time series properties of a variable which is unobserved.<\/p>\n

We are\u00a0fairly certain that the germ of the idea for our paper was this post<\/a> by Phil Arena<\/a>.<\/p>\n

\n

Indeed. Prevalent in poli sci. RT @TomPepinsky<\/a>: Where is my "lagged IVs because endogeneity"? http:\/\/t.co\/H9ofBmkhS0<\/a> cc @filarena<\/a><\/p>\n

— Marc F. Bellemare (@marcfbellemare) February 13, 2014<\/a><\/p><\/blockquote>\n