From DECKERS@ACFcluster.NYU.EDU Mon Aug 1 16:30:10 1994 Return-Path: Received: from AXP3.ACF.NYU.EDU by nber.harvard.edu.nber.harvard.edu (4.1/SMI-4.1) id AA03883; Mon, 1 Aug 94 16:30:07 EDT Received: from ACFcluster.NYU.EDU by ACFcluster.NYU.EDU (PMDF V4.3-8 #5342) id <01HFEGFHXR1SPZ8L8G@ACFcluster.NYU.EDU>; Mon, 1 Aug 1994 16:26:51 EDT Date: Mon, 01 Aug 1994 16:26:51 -0400 (EDT) From: DECKERS@ACFcluster.NYU.EDU Subject: mare-winship To: feenberg@nber.harvard.edu Message-Id: <01HFEGFHXVRMPZ8L8G@ACFcluster.NYU.EDU> X-Vms-To: IN%"feenberg@nber.harvard.edu" Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Content-Transfer-Encoding: 7BIT Status: R Dan: I don't completely understand the story myself. 1966 is a weird year in a lot of ways. For the rest (1964, 65 and 67), I was tempted to say that answer is that the 1964-67 cps only includes people aged 15 and over, so the total should sum to less than the total population. This reasoning seems to imply that 1965 is okay (132 million shoulds about right for total pop. if you exclude "children"). But 1964 and 1967 are still a little too low for this to be the whole explanation. I am not very familiar with using the cps weights in general. I know in the 1980s, I use something called a "march supplemental weight" which doesn't really exist for the 1960s. Did you try other weights than the "famwght" one? Aren't there separate person and family weights? Good luck. Let me know if you solve the mystery. Sandy