The Illusion of Spells: Making Sense of Poverty Dynamics

Jan O. Jonsson, Swedish Institute for Social Research
Carina Mood, Institute for Futures Studies

Dynamic poverty research mainly uses hazard regressions to study exit rates of inflow cohorts. Though sound in itself, this approach has contributed to a biased view of poverty as primarily a brief and transient experience. For Sweden, we find that half of the poor exit within one year, but only 9 percent of all years spent in poverty are contributed by persons that have only one calendar year of poverty. We discuss which relevant questions about durations that the existing research does and does not answer, and so reconnect to the studies that sparked the dynamic line of poverty research. Using Swedish population data spanning two extreme business cycles (1991-2007, N(persons*years)=102,754,809) enables us to avoid common problems of attrition and censoring and makes it possible to provide precise and reliable calculations of completed durations and their change over time.