NotesFAQContact Us
Search Tips
ERIC Number: ED508266
Record Type: Non-Journal
Publication Date: 2009-May-22
Pages: 30
Abstractor: ERIC
Reference Count: 50
A Detailed Picture of Intergenerational Transmission of Human Capital
Nichols, Austin; Favreault, Melissa
Urban Institute (NJ1)
Using data from the Health and Retirement Study (HRS), the authors consider how parental education relates to four separate outcomes in the children's generation: education, lifetime earnings, health and (financial) wealth. They relate parents' educational ranks to children's ranks on these four outcomes. By focusing on ranks, they are able to see full distributions of outcomes and can pick up within-group differences that even a relatively disaggregated analysis, like a quintile-based transition matrix, can obscure. They particularly focus on the tails of the children's distribution--the 10th and 90th percentiles--plus the median. The "children" in this study are Americans who have recently retired or are approaching retirement age. The authors find a mixed story. For a wide swath of the middle of the parental education distribution, the distribution of children's outcomes is extremely broad: for a given level of parental education, the most successful children (that is, those at the 90th percentile) end up at the top of the overall education, wealth, health and earnings distributions, and the least successful children (that is, those at the 10th percentile) for that same level of parental education end up at the bottom. This suggests a fluid and mobile society since children in the middle do not just end up in the middle: they end up at all points in the distribution in nearly equal measures. At both tails of the parental distribution, however, they see far closer correspondence between parents' and children's outcomes. The most successful children of parents with low educational ranking (that is, the 90th percentile of the children whose parents have the least education) have only average wealth, health or education; they attain about the median outcome. This pattern seems to be more pronounced for education, health and wealth than for lifetime earnings, which is arguably a poorer measure because of data limitations, most notably top-coding in the administrative earnings history data. The top-coding is quite pronounced for part of the period. To provide a preliminary look at educational outcomes for three generations, the authors include the eldest children of the near-retiree HRS children in the first analysis. They find that persistence in the tails across three generations is higher than they would expect if they assumed that transition probabilities stayed the same across both generations. This suggests the need to caution when making inferences about how many generations it takes to overcome the effects of parental privilege or disadvantage based on measures estimated using data from just two generations. Distributions of Children's Characteristics is appended. (Contains 17 figures, 6 tables and 23 footnotes.)
Urban Institute. 2100 M Street NW, Washington, DC 20037. Tel: 202-261-5687; Fax: 202-467-5775; Web site:
Publication Type: Reports - Research
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: Urban Institute