(1/16) Q: How should we seek to improve child outcomes? More cash for low-income parents, or more direct investments in children's skills and health?

A: Best evidence suggests direct investments much more promising.

cc #EconTwitter

(2/16) Big NYT article today promotes parental income approach as a "child allowance" – e.g. $3k/child in cash for parents.

IMO the article is based on a very misleading and outdated lit review.

https://www.nytimes.com/2020/08/22/sunday-review/coronavirus-poverty-child-allowance.html?smid=tw-share

cc @JasonDeParle

(3/16) In my work (Hilger 2016), I find very small impacts of parental job loss on child college enrollment+quality, even for low-income kids. Financial aid appears ~100 times more effective than parental income per dollar at improving college outcomes. https://drive.google.com/file/d/0B8J_qdFYwNJ6Qmo4bkEzMEZieDA/view

(4/16) Why such small impacts? Likely because parents don't spend extra money on helping kids attend college.

Best guess in data: only $2-5 out of every additional $100 of income spent on greater college assistance.

(5/16) Bulman et al 2018 replicate before/after college design with lottery winnings, even more compelling. Same result: tiny impact of parental wealth on child college outcomes.

This is not promising for a child allowance!

https://people.ucsc.edu/~gbulman/lottery_2020.pdf

cc @sarena_goodman

(6/16) Cesarini et al 2016 find similar results in Sweden, also using lottery winnings—small impacts of large parental wealth shocks on children's outcomes.

https://academic.oup.com/qje/article/131/2/687/2606947

cc @eriklindqvist1

(7/16) Many who support parental income approach cite Akee et al 2010. Why do they find such radically different impacts than these other highly credible studies?

https://www.aeaweb.org/articles?id=10.1257/app.2.1.86

cc @indigenalysis

(8/16) One possibility—intervention also transformed local labor market through new casinos. Some new jobs required high school degrees --> altered incentives to graduate.

(Other thoughts Randy @indigenalysis? This has always puzzled me.)

(9/16) Hoynes et al 2016, 2020 find impacts of SNAP on long-term child outcomes, but only for kids under age 5.

(Duncan et al 2011 also discuss larger impacts of parental income for very young children).

cc @HilaryHoynes @dwschanz @martha_j_bailey @maya_rossin @reed_walker

(10/16) But SNAP is not cash: families spend much larger share of SNAP $$ on food (Hastings et al 2018, Bruich 2014). If nutrition more important for young kids than other parental expenditures, SNAP impacts >> child allowance.

https://www.aeaweb.org/articles?id=10.1257/aer.20170866

cc @JHastings_Econ

(11/16) EITC improves child outcomes (Dahl et al 2012, Hoynes et al 2015), but impacts small per dollar compared to impacts of e.g. WIC, which provides more direct investments in child health (Kreider et al 2018, Rossin-Slater 2013, Jackson 2015).

(12/16) Historical lit mixed. Big lottery wins had no impacts on children in 19th century (Bleakley et al 2016), but early 20th century cash transfers did (Aizer et al 2016). IMO these (great) studies both have pretty big limitations.

cc @JosephPFerrie @LlerasMuney @ShariJEli

(13/16) Lareau's (2003) up-close observations of families led her to conclude that parental income likely played only a small direct role in divergent child outcomes; deep parenting practices more important.

Aside: Lareau deserves a Nobel Prize in Economics.

cc @AnnetteLareau

(14/16) Bergman et al's (2019) beautiful work on CMTO in Seattle also shows how cash alone does little to help parents improve opportunities for kids; deeper support needed to navigate complex bureaucracy/information/etc.

https://opportunityinsights.org/wp-content/uploads/2019/08/cmto_paper.pdf

cc @peterbergman_ @OppInsights

(15/16) Hendren et al (2020) review many studies that credibly estimate impacts of interventions helping kids; efficacy much greater for many direct investments than policies that raise parental income.

https://scholar.harvard.edu/files/hendren/files/welfare_vnber.pdf

cc @nhendren82 @bsprungkeyser

(16/16) Child allowance estimated cost of $100b/year.

With direct investments, that $ could instead help to fill the most glaring chasms in US child development ecosystem: early childhood, summer breaks, college.

Thank you for reading, all feedback/comments welcome.