Presentation on theme: "Economic transitions amongst the over-50s Lorenzo Cappellari Richard Dorsett Getinet Haile."— Presentation transcript:
Economic transitions amongst the over-50s Lorenzo Cappellari Richard Dorsett Getinet Haile
Format Policy background and research questions Analytical approach Summary of the data used in the analysis A description of the over-50s, 1993-2003 Results of modelling transitions Results of modelling time until change Summary
Policy background and research questions Pensions Green Paper states commitment to raising employment among older people Research questions: –Who are the over-50s? –How should over-50s employment be characterised? –How much movement between states is there? –How quickly does this movement happen? –What factors are associated with movement?
Analytical approach Focus on those aged 50 – SPA Use longitudinal data to examine changes Two approaches: –Markov models examine transitions between two points in time (state dependence) –Duration models examine the time taken until a change from initial state (duration dependence)
Data – longitudinal LFS 5 quarter LLFS observes households 5 times (quarterly) over a year Data are weighted – nationally representative Data cover 1993-2003 Resulting dataset has 26,000 individuals
The changing size of the older working age population
Summary of characteristics Mostly white (97, 97%: men, women) Mostly partnered (82, 77) Mostly without dependent kids (86, 90) Mostly own/mortgage accom (78, 79) Long-term health probs common (45, 41) About a quarter on benefit (23, 24) Men better qualified: no quals (26, 39); NVQ3+ (51, 27); apprenticeship (32, 5)
Examining transitions between states with Markov models
Markov models transition probabilities: estimating probability of being in a particular state conditional on the state occupied at an earlier time. control for differences in characteristics (but do not consider duration issues) allow assessment of state dependence. Policy relevance: –causal effects of states – via eg signalling, human capital changes, loss of motivation – may be more important than personal characteristics –so all individuals may become trapped in, for example, unemployment once they experience the state.
Examining timing of transitions with duration analysis
Duration analysis Models the hazard: the probability of a transition at some point for those who had not experienced a transition before that point Allows assessment of duration dependence. Policy relevance: the longer individuals are in an adverse labour market state the harder it may become to help them.
Age: older men & women more likely to exit employment Qualification: no significant marginal effect Accommodation: men and women with a mortgage less likely to exit employment Having a partner: women: less likely to exit. Non-working partner more likely to exit (for both men and women) Occupation: no significant marginal effect Industry: men in all industries but real estate & other more likely to exit Hours of work: men with <30 hrs/wk more likely to exit; women <17 hrs more likely to exit Region: Men in all but York & H., Wales, Scotland and N. Ireland less likely to exit; Women in York & H., EM, rest of SE, WM, & Wales less likely to exit. Employment exits: marginal effects
Age : older men & women less likely to exit Disability : temp and perm more likely to exit Mortgage & rented : more likely to exit Dependent children : more likely to exit Partner : non-working partners less likely to exit Region : NI less likely to exit Marginal effects of covariates
Summary Evidence of state dependence and duration dependence. Former shows potential risk for all individuals of being trapped in inactivity. Duration dependence shows importance of escaping this trap relatively early. Policy faces trade-off between intervening soon and risking deadweight or intervening later and risking individuals becoming hard- to-help. However, inactivity spells are mostly long (average about 7 years for men and 10-14 for women). For those in employment, need to help individuals remain in work for long enough to overcome early shake-out (most exits are concentrated in first 3 years). Part-time work as bridge to full-time work. Scope for policy to support bridging role of part-time work.