Interestingly, the O'Neil data don't add much information to the official stats. Graphing the two over time with each measured on different Y axes, the O'Neil measure moves in the same general direction as the official measure, while the latter has has more cyclical movement (rising with unemployment). With the O'Neil measure as the dependent variable, the intercept in a simple regression is about 7.82 (so his measure averages about 7.82 points higher than the official measure), while the coefficient on the official stat is 0.995 (with a t-stat equaling a respectable 6.354), meaning that the two measures move in step on average. (BTW, I've never been impressed by the _level_ or _rate_ of poverty. What matters is the directions of movement over time.)
On 10/8/07, Doug Henwood <[EMAIL PROTECTED]> wrote: > A guy named Jack O'Neil, a Census Bureau analyst, used to compute > such a relative defintion of poverty on his own. He's either retired > or died; the last estimate I have from his series was for 1998. > Here's what I have from his series (the second column is the relative > poverty rate, i.e., less than half the median; the third is the ratio > of that rate to the official rate). The Luxembourg Income Study uses > a relative defintion of poverty - they don't do it annually, but at > frequent enough intervals to give you the trend. For that, see: > > <http://www.lisproject.org/keyfigures/full_kf.xls>. > > Relative poverty rate and ratio to official poverty rate > source: Jack O'Neil, Census Bureau (unpublished data) > > 1969 18.0% 148.8% > 1970 18.1% 143.7% > 1971 18.3% 146.4% > 1972 18.9% 158.8% > 1973 18.9% 170.3% > 1974 18.7% 167.0% > 1975 19.6% 159.3% > 1976 19.3% 163.6% > 1977 19.8% 170.7% > 1978 19.6% 171.9% > 1979 20.1% 171.8% > 1980 20.3% 156.2% > 1981 20.9% 149.3% > 1982 21.4% 142.7% > 1983 22.0% 144.7% > 1984 21.8% 151.4% > 1985 21.7% 155.0% > 1986 21.8% 160.3% > 1987 22.1% 164.9% > 1988 22.0% 169.2% > 1989 22.1% 172.7% > 1990 21.8% 161.5% > 1991 22.3% 157.0% > 1992 22.8% 154.1% > 1993 22.8% 151.0% > 1994 22.6% 155.9% > 1995 22.2% 160.9% > 1996 22.3% 162.8% > 1997 22.3% 167.7% > 1998 22.3% 175.5% > -- Jim Devine / "The truth is at once less sinister and more dangerous." -- Naomi Klein.
