NERO Conference Paris, 22 June 2015 INCOME INEQUALITY SOCIAL MOBILITY AND ECONOMIC GROWTH* Federico Cingano, ELS Employment Analysis Division *Source: OECD (2015) In It Together: Why less inequality benefits all OECD Publishing, Paris.
OECD work on income distribution Long-standing interest in data collection and analysis Early works mid 1970s; regular data collection mid 1990s Documented the (increasing) patterns of income inequality across MCs in many publications Latest one warns: The gap between rich and poor is at its highest level since 30 years Shares of bottom, middle and top incomes in total income, OECD average 1985 to 2011/12, 1985 = 1 Traditional definition Palma definition 1.20 bottom 10% middle 60% top 10% 1.20 bottom 40% 50% to 90% top 10% 1.10 1.10 1.00 1.00 0.90 0.90 0.80 0.80 1985 1990 1995 2000 2005 2010 1985 1990 1995 2000 2005 2010 Source: OECD Income Distribution Database (www.oecd.org/social/inequality.htm),. Note: Income refers to cash disposable income adjusted for household size. 2/15
... contributed to a growing public discussion Rising concerns about the potential impact of these trends on our societies Is increasing inequality compatible with social cohesion and political stability? and economies: Does it imply lower social mobility? Does it affect economic growth (recovery)? To a large extent, the answers depend on the impact of inequality on (human and physical capital) investment, on entrepreneurship and innovation etc. Incentives vs Opportunities 3/15
Inequality & mobility: what do we know? Social mobility varies substantially across OECD countries Children education (or income) depends on parents education (or income) intergenerational persistence in outcomes Intergenerational earnings mobility More mobiilty 0.85 0.80 0.75 0.70 0.65 0.60 0.55 0.50 ESP NOR CAN KOR AUS SWE DEU FRA DNK FIN JPN NZL CHE USA ITA GBR Note: The charts plots estimates of social mobility (= 1 - the intergenerational earnings elasticity). A higher number means more mobility. A value of 0.85 (=1-0.15) in DNK means that if an individual earns 10,000 less income than average, the children will earn 1,500 less than average (as opposed to 5,000 less than average in the UK). Source: OECD calculations from different sources 4/15
Inequality & mobility: what do we know? and it is negatively correlated with inequality (the GCC ) Intergenerational persistence in outcomes increases with outcome (income) disparities Intergenerational earnings mobility More mobility 0.85 0.80 0.75 0.70 0.65 0.60 DNK SWE FIN AUS NOR CHE ESP CAN JPN KOR 0.55 FRA USA ITA 0.50 0.20 0.25 0.30 0.35 0.40 GBR Gini coefficient, working-age population Note: estimates of social mobility (1 - the intergenerational earnings elasticity) against (Gini) inequality. Both indicators increase with inequality and mobility. Source: OECD calculations from different sources DEU NZL 5/15
Inequality & mobility: what do we know? Does this imply that increasing inequality would lower mobility? Difficult to argue from cross country correlations inequality might correlate with the quality of the educational system, or with other policies and institutions that affect outcomes Aggregate evidence (the GCC) is silent on the underlying mechanisms is it due in particular to underinvestment by the poor? 6/15
Recent OECD evidence on inequality & (educational) mobility Look at the consequences of inequality on attainments by individuals with different parent educational background (PEB) HH i,t,c = β 1 PEB i,t,c + β 2 PEB i,t,c IIII t,c + μ c + μ t +ϵ i,t,c HC β1 captures average educational outcomes by PEB (the intercepts). HC PEB=HIGH β2 measures how such averages (mobility) vary with inequality (the slopes) A measure of educational mobility PIAAC Data HC PEB=LOW Inequality Estimates exploit within country variation (µc accounts for fixed country characteristics) HC: human capital (individual i, in country c, age cohort t) PEB: index for parents education being low, medium or high Ineq: inequality in c, measured when i was 14 yrs old 7/15
The role of inequality and family background for formal education (i) Inequality lowers the probability of Tertiary education, but only among individuals with low parental education Probability of tertiary education by parental educational background (PEB) and inequality Low PEB Medium PEB High PEB 0.50 0.45 Probability of tertiary education 0.40 0.35 0.30 0.25 0.20 0.15 0.10 18 20 22 24 26 28 30 32 34 36 38 Inequality (Gini coefficient) Interquartile range A 6 Gini pts. increase in inequality (the US-Canada differential in 2010) is associated to 4 ppts. lower probability of tertiary education of Low PEB individuals Note: Low PEB: neither parent has attained upper secondary education; Medium PEB: at least one parent has attained secondary and postsecondary, non-tertiary education; High PEB: at least one parent has attained tertiary education. The bars indicate 95% confidence intervals. Source: OECD (2015) In It Together: Why less inequality benefits all OECD Publishing, Paris. http://dx.doi.org/10.1787/9789264235120-en 8/15
The role of inequality and family background for formal education (ii) while increasing their probability of (at most) lower secondary education Probability of lower secondary education by parental educational background (PEB) and inequality Low PEB Medium PEB High PEB 0.50 Probability of lower secondary education 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 18 20 22 24 26 28 30 32 34 36 38 Inequality (Gini coefficient) A 6 Gini pts. increase in inequality is associated to a ~5 ppts higher probability that Low PEB individuals attain at most lower secondary education Note: Low PEB: neither parent has attained upper secondary education; Medium PEB: at least one parent has attained secondary and postsecondary, non-tertiary education; High PEB: at least one parent has attained tertiary education. The bars indicate 95% confidence intervals. 9/15
The role of inequality and family background for skill proficiency (i) Inequality lowers (literacy and numeracy) skills, but only among individuals with low parental education Average PIAAC numeracy score by parental educational background (PEB) and inequality Low PEB Medium PEB High PEB 300 290 Numeracy score 280 270 260 250 240 18 20 22 24 26 28 30 32 34 36 38 Inequality (Gini coefficient) A 6 Gini pts. increase in inequality is associated to lower Numeracy score (by ~6 pts) by low PEB individuals (35% of the baseline differential relative to Med PEB) Note: Low PEB: neither parent has attained upper secondary education; Medium PEB: at least one parent has attained secondary and postsecondary, non-tertiary education; High PEB: at least one parent has attained tertiary education. The bars indicate 95% confidence intervals. 10/15
The role of inequality and family background for skill proficiency (ii) Inequality lowers skill proficiency of low PEB individuals, even conditioning on the level of formal education Average PIAAC numeracy score conditional on education by parental educational background (PEB) and inequality Low PEB Medium PEB High PEB 300 290 Numeracy score 280 270 260 250 240 18 20 22 24 26 28 30 32 34 36 38 Inequality (Gini coefficient) A 6 Gini pts. increase in inequality is associated to lower Numeracy score (by ~5 pts) by low PEB individuals Note: Low PEB: neither parent has attained upper secondary education; Medium PEB: at least one parent has attained secondary and postsecondary, non-tertiary education; High PEB: at least one parent has attained tertiary education. The bars indicate 95% confidence intervals. 11/15
The role of inequality and family background for labour market outcomes Inequality increases the probability that low PEB individuals are not employed over their working life Fraction of working life spent out of employment by parental educational background (PEB) and inequality Low PEB Medium PEB High PEB 0.2 Probability of not being employed 0.15 0.1 0.05 18 20 22 24 26 28 30 32 34 36 38 Inequality (Gini coefficient) A 6 Gini pts. increase in inequality is associated to an increase in this probability by 3 ppts Note: Low PEB: neither parent has attained upper secondary education; Medium PEB: at least one parent has attained secondary and postsecondary, non-tertiary education; High PEB: at least one parent has attained tertiary education. The bars indicate 95% confidence intervals. 12/15
Inequality & growth: what do we know? Huge literature started in 1990s. Mostly focused on reducedform growth regressions Largely inconclusive: Early (cross-country) works: mostly negative estimates Later (panel) analyses: often positive (or non-significant) estimates Possible explanations: Data quality and their coverage Estimation approaches and inequality indicators Recent research based on OECD data shows that when income inequality rises economic growth falls 13/15
Recent OECD evidence on Inequality & growth Looking across OECD countries (1970-2010) the effect is significantly negative Increasing income inequality by 1 Gini pt. lowers the growth rate of GDP per capita by ~0.12 pp per year, with a cumulative loss of ~3% after 25 years. Result is driven by disparities at the bottom of the distribution The negative effect is not just for the poorest income decile but involves the lower middle classes (the bottom 40%). Top inequality is less, if any, relevant for growth 14/15
Conclusions In general, policies that help to limit or reverse inequality may not only make societies less unfair, but also wealthier. Which does not imply that any redistributive policy would work. (from this and other works) measures that might help create greater equality of opportunities & growth in the long run include: Promoting access to education, particularly for students from disadvantaged backgrounds. Improving job-related training and education, and access to formal education over the working life Facilitating access to jobs (and career prospects) for under-represented groups (youth, women) 15/15