The current issue of The Public Interest includes an exchange on "Low-Wage America: Two Views" (between Robert F. Ferguson and Marvin H. Kosters). I think the writers missed the big story.
Published in July of 2003, the Census Bureau's "Dynamics of Economic Well-Being: Poverty 1996-1999" is the follow-up to the July 1998 report with the same title (but subtitled "Trap Door? Revolving Door? Or Both?"). Both Census reports highlight income-mobility and dynamics.
The recent study notes that, " ... a majority of poor individuals do not remain poor for very long periods of time ... poverty is chronic for a small but significant proportion of the poverty population."
The average poverty rate in 1999 was 12.8%; 19.5% were poor for at least 2 months but 2% were poor all 48 months.
In 1996-99, the report shows that 7.6 million people entered poverty but 14.8 million left poverty. Immigrants, legal and otherwise, who entered the U.S. in this interval are, by definition, not counted.
But, how many immigrants were there in that period? Three-million "authorized" and more "unauthorized" (census labels).
Some points emerge. 1) Comparing poverty levels but not cohorts is practically meaningless; and 2) it would be wonderful if the SIPP mobility data were comparably available for recent immigrants as well as for non-immigrants; at least, flag these folks in the survey.
The U.S. poverty story is complex but is difficult to fathom unless these two points are addressed.