1. 4th Russia-India-China Conference, New Dehli, November 20-21 Entry to and Exit from Poverty in Russia: Evidence from Longitudinal Data Irina Denisova New Economic School, Moscow

2. 4th Russia-India-China Conference, New Dehli, November 20-21 Motivation Long-term sustainability requires social stability and could be undermined by high poverty levels. Poverty in Russia declined during the last five years from 29% of population in 2000 to 15% in 2006. Still more than twenty five million people have incomes that are lower than subsistence level The economic crisis is likely to increase poverty. The most vulnerable groups? Significant body of literature on poverty in Russia –literature analyzes primarily stocks of poverty –only some indirect insights about flows in and out of poverty –an understanding of how entry to poverty and exit from poverty are shaped is lacking For the field: initial income distribution in transition is very condensed. The role of educational and health endowments could be convincingly highlighted (less correlation with household wealth endowment)

3. 4th Russia-India-China Conference, New Dehli, November 20-21 Literature Poverty incidence is sizeable but poverty gap is small for most families –Limited opportunities to smooth consumption Most poverty is transitory. Observable determinants of transitory and chronic poverty are the same The vulnerable: –Large families –Families with children (particularly single parent) –Rural households –Families with unemployed family heads or wage arrears The Buffers –Pensions –Education

4. 4th Russia-India-China Conference, New Dehli, November 20-21 Research questions What are the determinants of entry to and exit from poverty (duration of poverty and non- poverty spells)? Are the two processes symmetric or there are important asymmetries? What are the time-related properties of the flows? Are there changes brought to life by the economic growth that followed the economic decline?

5. 4th Russia-India-China Conference, New Dehli, November 20-21 Data Russian Longitudinal Monitoring Survey Waves 5 – 13 (1994 to 2004) –Nationally representative, about 5,000 households in each round –Very complete income and expenditure information –Follows same household structure –Panel attrition is a serious issue

6. 4th Russia-India-China Conference, New Dehli, November 20-21 Attrition of Households from Round 5 (from B.F.Mills (2007)) Percent of NumberInitial Households in Round 54,293 Households in Rounds 5 - 63,48681 Households in Rounds 5 - 72,59560 Households in Rounds 5 - 82,15250 Households in Rounds 5 - 91,77341 Households in Rounds 5 - 101,57237 Households in Rounds 5 - 111,43433 Households in Rounds 5 -121,28830

7. 4th Russia-India-China Conference, New Dehli, November 20-21 Comparing Panel Sample with Period Samples (Mills, 2007)

8. 4th Russia-India-China Conference, New Dehli, November 20-21 Study measures Absolute poverty concept Total household income is main measure of household economic well-being Region specific poverty lines –Local diets –Regional prices –Estimated household equivalence scales

9. 4th Russia-India-China Conference, New Dehli, November 20-21 Poverty rates (RLMS definition)

10. 4th Russia-India-China Conference, New Dehli, November 20-21 Methodology Survival analysis –Eliminates estimation bias due to non-normality of time to event and right-censoring –Allows using an unbalanced panel Proportional hazard model Cox model specification (non-parametrical baseline) Cluster on id (to take into account multiple failures) Try strata on settlement type and period (allows baseline hazard to vary across the groups)

11. 4th Russia-India-China Conference, New Dehli, November 20-21 Methodology Explanatory variables (X) include –demographic characteristics of a household; –labor market attachment of adult family members; –eligibility for public transfers; –characteristics of human capital of a household; –settlement type and economic region; –time period Changes in variables in X are considered

12. 4th Russia-India-China Conference, New Dehli, November 20-21 Panel statistics

13. 4th Russia-India-China Conference, New Dehli, November 20-21 Gaps in data record: an example 1, 3 and 5 – underestimate poverty spells 4, 6 and 8 – underestimate non-poverty spells 2 and 7 - overestimate poverty and non-poverty spells

14. 4th Russia-India-China Conference, New Dehli, November 20-21 Poverty and non-poverty incidence and duration

15. 4th Russia-India-China Conference, New Dehli, November 20-21 Survival function of staying in poverty

16. 4th Russia-India-China Conference, New Dehli, November 20-21 Survival function of staying out-of- poverty

17. 4th Russia-India-China Conference, New Dehli, November 20-21

18. Survival of staying in poverty, by settlement type

19. 4th Russia-India-China Conference, New Dehli, November 20-21 Survival of staying in non-poverty, by settlement type

20. 4th Russia-India-China Conference, New Dehli, November 20-21 Survival of staying in poverty, by period

21. 4th Russia-India-China Conference, New Dehli, November 20-21 Survival of staying in non-poverty, by period

22. 4th Russia-India-China Conference, New Dehli, November 20-21

24. Results Symmetries –Presence of children weakens hh position –Households headed by retired males and females tend to be in a stronger position than headed by an adult man –Adult female-headed households are indistinguishable from adult male-headed households –High share of adults with university degree is an effective buffer against poverty: it reduces entry to poverty and increases exit from it. –Families that live in urban areas have lower chances to get into poverty and higher chances to get out of it.

25. 4th Russia-India-China Conference, New Dehli, November 20-21 Results Asymmetries –larger families are doing better when getting out of poverty and average when slipping into poverty –higher share of unemployed reduces hazard from poverty but does not affect the rate of entry to poverty –higher share of employed in the public sector increases entry to poverty but is insignificant for poverty exit rate –involvement in subsistence farming is a sign of lower exit rate and has no influence on entry rate –bad health increases chances to enter poverty but does not affect chances to escape it –economic growth lowers chances to slip into poverty but also reduces hazards from poverty.

26. 4th Russia-India-China Conference, New Dehli, November 20-21 Conclusions and policy implications Find significant asymmetries between determinants of entry to and exit from poverty –A better design of policies to fight poverty should distinguish between measures to prevent from slipping into poverty, and measures to get out of poverty for those who are poor. –Households in poverty in the period of economic upturn are to be paid special attention to Confirm the role of educational and health endowments in preventing poverty