Abstract
The risk of unemployment varies between individuals, between occupations and industries, and between places. The Sample of Anonymised Records drawn from the 1991 Census is used to model geographical, demographic, and socioeconomic variations in male unemployment by means of multilevel logit models. The underlying structure of the problem is such that cells in a multiway cross-tabulation of individual characteristics (level 1) are nested within places (level 2). Geographical variations in male unemployment are found even after allowing for age, marital status, ethnicity, higher education qualifications, social class, and industry.
Original language | English |
---|---|
Pages (from-to) | 611-628 |
Number of pages | 18 |
Journal | Environment and Planning A |
Volume | 29 |
Issue number | 4 |
DOIs | |
Publication status | Published - 1997 |