Welfare Regimes and Poverty Dynamics: The Duration and Recurrence of Poverty Spells in Europe

1. Welfare Regimes and Poverty Dynamics: The Duration and Recurrence of Poverty Spells in Europe Didier Fouarge & Richard Layte Presented by Anna Manzoni Socrates Program 2008, Komotini 4-14 March

2. Motivation Dynamic perspective on poverty: • Mobility • Persistence • do we see the same or different people in poverty? • Are poverty spells short/long in duration? Do people remain in poverty? • Is poverty recurrent? ‘Time’ has to be part of the definition of poverty use of longitudinal data to understand the processes behind cross-sectional statistics, the events leading individuals to and from poverty and the impact of this povert on their living standards

3. Previous findings • There is a great deal of turnover in the stock of people living in poverty • The majority of poverty spells are rather short in duration • However, many of those who have left poverty return relatively quickly • A substantial minority experience persistent poverty

4. This paper • How well the different welfare states of Europe perform in terms of preventing recurrent and persistent income poverty • What household and individual characteristics influence poverty duration

5. Esping-Andersen (1990) “Three worlds of welfare capitalism”

6. Research questions • To what extent does the level of recurrent and persistent poverty vary across countries? • To what extent is this a function of the distribution of different socio-economic variables in the country rather than the welfare regime type? • Are transition rates from poverty and average durations of poverty in different countries related to the welfare regime? • To what extent is the probability of leaving poverty influenced by particular socio-economic predictors (employment status, human capital, hh context) rather than welfare regime type?

7. Hypotheses • HP 1: Social democratic countries have lower rates of both recurrent and persistent poverty compared to corporatist • HP 2: Disadvantaged groups and those who are unable to participate in the LM will have a higher risk of recurrent and persistent poverty in all other regimes compared to the social democratic • HP 3: Social democratic welfare regime countries have higher initial exit rates but lower levels of incentives will lead to sharply falling exit rates from poverty as duration increases.

8. Summary indicator of longitudinal poverty:Poverty profiles (Muffels, 1999) • never poor • single-year or transient poor • recurrent poor (poor more than once but never longer than two consecutive years) • persistent poor(at least three consecutive years in poverty) Data: 5 waves of ECHP, 1994–8 Poverty line = 60% median standardised hh income

10. From our data • 1/3 of the EU population is found to experience poverty at least once in the 1994-98 period • 1/3 of the ever poor are only poor for a single year • 12% of the EU population is found to be persistently poor • Lowest poverty incidence: The Netherlands • Highest poverty incidence and persistence: Greece and Portugal •  social democratic countries display lower rates of poverty - corporatist countries - Higher, more recurrent and persistent poverty: residual and liberal welfare regimes

11. Determinants of poverty profiles • Ordering among the profile categories • Predicting factors: • HH formation events • LM events • Control vars: • Personal and HH characteristics • Socio-economic characteristics Model 1: LM and HH status vars Model 2: add country dummies Model 3: add regime type dummies

13. Model fit (pseudo-R2) Country model: 0.152 … Regime model: 0.150 …

16. Duration of spells • Model exits from poverty (discrete-time hazard model) • Account for left and right censoring • Account for multiple spells • Correct for unobserved heterogeneity

17. Hazard rates by country • Exit probabilities fall quickly between the 1st and the 2nd year • Across regimes: • The overall spread of rates in the 1st year is quite small • Social democratic countries have the highest initial exit rates • The Southern and liberal countries tend to have the lowest rates in the 1st year but over e 3-year period any resemblance of a regime order is lost •  there is very little variation across countries in exit rates over a 3-year period, but this overall similarity stems from different patterns of exit between countries in different regimes

18. Weibull Discrete-Time Hazard Model • To estimate the probability of making a transition from poverty and its dependence on time • The hazard of exit from poverty decreases with duration: each additional period in poverty decreases the probability that the person will leave the state • Covariates: - female head - old head - low educated head slow down exit from poverty - each additional child - unemployment/non participation • Across countries: no significant difference. Despite the differences in poverty persistence among welfare regimes, different patterns of exit from poverty lead, over the medium term, toroughly similar rates of exit from poverty for countries with different welfare regimes

19. Couples with children have a greater probability of experiencing poverty • Positive and significant effect for the # of children • Tentative explanation: - young children coming into the HH induce an additional financial burden that is generally less than compensated by child benefits; -children leaving the HH are generally older and have their own market income, which may have negative consequences on the HH’s income position

20. Conclusions • Poverty profiles = useful summary measure of poverty • Consider household labour supply in poverty studies • Country and welfare regime effects • Persistent and recurrent poverty highest in liberal and southern regimes • Vulnerable groups less at risk of poverty in social democratic countries. Significant higher risk in liberal countries • Despite their dissimilar patterns of poverty duration, European welfare states display rather similar probabilities of exit from poverty, once controlling for duration, though these similar rates may be the result of very different processes with social democratic and corporatist countries having high exit rates that decrease quickly whereas liberal and Southern European countries have moderately high rates that remain more constant over time