Microeconometrics & Public Policy seminar series | Can staying in school longer protect disadvantaged young people from homelessness? by Julie Moschion
Invites you to a
Microeconometrics & Public Policy seminar presented by
(University of Melbourne)
Can staying in school longer protect disadvantaged young people from homelessness?
Friday 5 August
12.00pm – 1.00pm
Via Zoom: Meeting Link
Abstract: This paper investigates whether staying in school longer decreases the likelihood of youth homelessness for disadvantaged Australians. Using a unique dataset of disadvantaged Australians who provide retrospective information on their education and housing histories, we find that disadvantaged youth with one less year of education are more likely to experience homelessness by 20 years old by 3.8pp and those dropping out before completing year 12 have a higher likelihood by 6.8pp. Because respondents with more education exhibit other characteristics that may lower their likelihood of youth homelessness, our OLS estimates control for a detailed set of childhood, family and individual characteristics to minimise biases due to unobserved confounding factors that co-determine homelessness and education outcomes. In addition, we complement the OLS strategy with an IV strategy based on respondents’ year of birth and two significant reforms that lengthened the duration of education: the increase of compulsory schooling from 14 to 15 years old in the mid-1960s and the introduction of a prep year in the 1970s. In our sample, these reforms translate to an increase in the average number of years spent in school from 9.5 years for respondents born before either reform, to 10.4 years for respondents born after the extension of compulsory schooling, and then to 10.8 for those born after the introduction of prep. The IV results confirm that staying longer in school decreases the likelihood of youth homelessness. Interestingly, both the OLS and the IV results show that staying in school longer has no effect on homelessness before the age of 14 (ie before respondents leave school), suggesting that reverse causality and (time-invariant) unobserved heterogeneity are not affecting our OLS estimates.
For further information contact: Microeconometrics & Public Policy Seminar Coordinator Dr Rebecca McKibbin (firstname.lastname@example.org)