Extra homework and lessons do not raise test scores and hurt high schoolers’ behavior

Using US time diaries, economists found the last weekly hour of homework, reading or lessons has no effect on test scores, while in high school each extra hour lowers non-cognitive skills by 0.039 standard deviations.

Share this:

For the typical American child, one more hour a week of homework, reading or lessons does nothing measurable for test scores. In high school it does harm, lowering behavior and emotional skills by about 0.039 standard deviations for each extra weekly hour.

That is the conclusion of economists Carolina Caetano and Gregorio Caetano of the University of Georgia and Eric Nielsen of the Federal Reserve Board. Their paper was published online in Economics of Education Review on January 4, 2024, in the journal’s February 2024 issue.

Every hour of enrichment replaces something

The authors start from a simple point. A child’s day has 24 hours, so an extra hour of homework or tutoring is an hour taken from sleep, play or time with friends. The question is not whether enrichment helps in isolation, but whether the last hour is worth what it displaces.

They used time diaries from the Child Development Supplement of the Panel Study of Income Dynamics, from the 1997, 2002 and 2007 waves. Each diary records a full weekday and weekend day for a child, from pre-kindergarten through 12th grade.

Children averaged about 45 minutes a day of enrichment. Homework made up two-thirds of it, reading a book 14% and before- or after-school programs 7%. Lessons such as piano or soccer were a small share.

Cognitive skills came from reading and math achievement tests. Non-cognitive skills came from parents’ answers to 36 questions about behavior, covering both acting out and anxiety, depression and self-esteem.

Using the kids who do none at all

Children who do more enrichment differ in many hidden ways from those who do less, so simple comparisons mislead. Indeed, raw correlations showed enrichment linked to better skills of both kinds.

About 30% of the children did no enrichment at all. The authors argue that among these children, the hidden factors still vary while the amount of enrichment does not. That lets them measure how the hidden factors alone affect skills and strip that out of the estimates.

After the correction, the positive link with test scores vanished. Across all ages, the effect of an extra weekly hour on cognitive skills was about -0.004 standard deviations, statistically indistinguishable from zero. The effect on non-cognitive skills was -0.019 standard deviations.

High school is where the cost shows up

Split by grade, test score effects were near zero at every stage. The harm to non-cognitive skills was concentrated in grades 9 to 12, at -0.039 standard deviations per extra weekly hour. That is when enrichment shifts toward solitary homework and away from reading and social programs, and when some teens are pushing for selective college admission.

For scale, the authors note that a one standard deviation better teacher, on non-cognitive value-added, raises these skills by about 0.08, roughly twice their estimate. They argue a few extra hours a week could add up to meaningful losses in later-life outcomes such as wages.

Results held with a broader definition of enrichment that included sports, arts and volunteering. When enrichment was limited to homework alone, the non-cognitive harm was even larger and more significant.

Study Details:

  • Title: Are children spending too much time on enrichment activities?
  • Authors: Carolina Caetano, Gregorio Caetano, Eric Nielsen
  • Journal: Economics of Education Review
  • Publication Date: January 4, 2024
  • DOI: 10.1016/j.econedurev.2023.102503