ECON 361: Income Distributions and Problems of Inequality David Rosé Queen s University March 12, 2018 1/33
Last class... Social Assistance in Ontario (Adams, Chow, and Ros, 2018) Started Inequality and Health (Case and Deaton, 2017) Today... Research Proposal Wrap up Ineq. and Health Income Mobility (Corak, 2013) Education Mobility (Ayedmir, Chen, and Corak, 2013) 2/33
Definitions 1. Social mobility: movement of individuals, families or HH between social classes (i.e. income quanties). 2. Equality of opportunity: the notion that everyone should have access to the same opportunities, regardless of social status. 3. Intergenerational mobility: social mobility between generations. 4. Upward mobility: likelihood an individual will move from a lower quantile to a higher one. 5. Downward mobility: likelihood an individual will move from a higher quantile to a lower one. 3/33
Equality of Opportunity Project Raj Chetty, Nathaniel Hendren, Emmanuel Saez, and Nicholas Turner http://www.nytimes.com/2013/07/22/business/ in-climbing-income-ladder-location-matters.html?&_r=0 4/33
I m a dreamer. I have to dream and reach for the stars, and if I miss a star then I grab a handful of clouds. Mike Tyson 5/33
The American Dream Being free to say or do what you want. Being free to accomplish almost anything you want with hard work. Being able to succeed regardless of the economic circumstances in which you were born. 6/33
Inequality in a Democracy: Why do lower earners tolerate inequality in democratic countries? Why don t lower earners just vote for redistributive policies all the time? Candidate Explanation: The Prospect of Upward Mobility Hypothesis: Encapsulated by the words of Senator Marco Rubio... There are two kinds of people in this room. Those who have made it and those who are about to make it! [... amiright, amiright!?] 7/33
Corak (2013) Recent evidence suggests that high income inequality is resulting in family income playing a bigger role in determining outcomes. We observe low levels of intergenerational mobility in more unequal countries like Italy, the U.S., the U.K., etc. Research Question: To what degree does income inequality limit economic mobility for the next generation of young adults? 8/33
The Great Gatsby Curve: graph that captures the positive correlation between income inequality (Gini coefficient) and the intergenerational earnings elasticity. Intergenerational Earnings Elasticity: the elasticity between paternal earnings and a son s adult earnings. Derived using the following regression: ln Y i,t = α + β ln Y i,t 1 + ɛ i Y i,t is the paternal permanent earnings and Y i,t 1 is the son s permanent earnings β is the intergenerational earnings elasticity 9/33
10/33
11/33
12/33
What do the graphs tell us? Figure 1: Intergenerational elasticity in the U.S. is about 0.5, approximately double the Canadian measure. Figure 2: Sons born to TD fathers are more likely to end up in TD than any other decile. Figure 3: Sons born to BD fathers are more likely to end up in BD than any other decile. Trends are much more dramatic for U.S. than Canada = more intergenerational mobility in Canada. Middle deciles don t look that too bad in terms of intergenerational mobility. Overall intergenerational mobility tends to be driven by what happens at the tails (BD and TD). 13/33
... thinking about elasticities Do we want an intergenerational earnings elasticity of zero? A popular idea: we need to distinguish between circumstances and choices (Roemer, 2012). Some circumstantial factors that may affect opportunity: Race Parents (occupation, education) Region of birth Disabilities Inequality of opportunity is the link between income inequality and social mobility. High inequality = unequal distribution of opportunities = persistence of parental SES. 14/33
The Great Gatsby Curve is not a causal relationship. But, we might want to go beyond the correlation does not imply causality story. The process of development from childhood to adulthood may be a recursive process: 1. Early nutrition (even in utero) affects early cognitive and social development. 2. These building blocks are reinforced by neighbourhood and pre-school quality which influences elementary school outcomes. 3. Success in elementary school feeds into high school success... and so on... Also some evidence that family connections and resources affect schooling and job outcomes. 15/33
Returns to Schooling and Social Mobility In general, countries with higher returns to college have lower intergenerational mobility. Some context, average U.S. college graduate earns 70% more than a HS graduate. In Canada the premium is approximately 30%. Mazumder(2012) finds that the father-son earnings elasticity and the return to education (per year of schooling) follow very similar trends. 16/33
17/33
Human Capital Investment and Social Mobility Families with more human capital (schooling, work experience, connections ) tend to invest more in their children. Monetary investments - e.g. cool daycare, sports clubs, tutoring Non-monetary investments - e.g. reading to children, helping with homework, etc. 18/33
Monetary investments 19/33
Credit constraints are more likely to bind for lower income families: Unavailability of financial aid can limit college choice. Even college application fees can deter applying to prestigious schools (which usually offer excellent financial aid). http://www.nytimes.com/2014/06/10/upshot/ a-case-study-in-lifting-college-attendance.html 20/33
Policy Discussion: Canadians also have a Tysonesque dream... Similarly to Americans, we: Value equality of opportunity. Recognize the importance of individual responsibility. Are averse to equality of outcomes as a desirable goal. Unlike Americans: Canadians have a much more positive view about the role of public policy and trust in government intervention. 21/33
Access to Schooling: In 2008, the U.S. spent about $15,000 per student on schooling (more than almost all other OECD countries). U.S. spends less (on average) on students from disadvantaged backgrounds than on other students. and High quality teachers are seldom found in disadvantaged schools (school funding largely comes from local property taxes). 22/33
Let s look at spending data ourselves (share of GDP): 23/33
Back to the paper... Bradbury et al. (2012) study vocabulary development and behavioral problems of 4-5 year old kids. Australia, Canada, the U.K., and the U.S. Find inequalities according to family income and mother s education in all 4 countries. Inequalities greatest in the U.S., most muted in Canada. This suggests an important role for public provision of services (health care, education, etc.) that weaken the link between childhood development and parental income/education. 24/33
Aydemir, Chen, and Corak (2013) Motivation: There is evidence that successive immigrant cohorts are doing relatively worse in the labour market than the Canadian born population. Are the children of immigrants also having increasing difficulty integrating? 25/33
Research Questions: 1. What is the degree of intergenerational education mobility in Canada? 2. Has the strength of the tie between the education of immigrant parents and their children changed over time? 26/33
Results preview: 1. There is a weak link between parents and children s years of schooling. 2. This link is even weaker for children of immigrants vs. children of Canadian born parents. 3. 2 nd generation immigrants and their parents are more educated on average than 3 rd -plus generation immigrants and their parents. And MORE! (They study differences across origin-regions, across time, etc.) 27/33
Data: 2001 Canadian long form census... 2016 results are coming soon! Can identify the country of origin of an individual s parents but not parents schooling (large sample) 2002 Ethnic Diversity Survey From the EDS,can identify the country of origin and schooling of individual s parents (small sample) 28/33
Econometric framework: Y i,t = α t + βy i,t 1 + ɛ i α t represents the average years of education for cohort t. Y i,t and Y i,t 1 are the child and parent s years of education, respectively. β captures the degree of intergenerational mobility of education. Note: β is not an elasticity (this is not a log-log regression, it s a linear specification). 29/33
30/33
31/33
32/33
33/33