Education can affect a child’s lifetime earning power—and that, of course, costs money. So when a family has multiple children and financial limitations, parents often face a decision about how to invest their educational funds.
While they might not articulate their decision as an economist might, they have several options, economically-speaking. They may seek to maximize their investment in their children, by spending money in a way that will return the most in terms of future income. Alternatively, they could spend an equal amount on each child, or they could spend more on the child that’s further behind to help that child earn the same amount as their sibling one day.
Which option do families most frequently choose? Research by the University of Delaware’s James Berry, Chicago Booth’s Rebecca Dizon-Ross, and University of Colorado Denver’s Maulik Jagnani finds that parents demonstrate a strong interest in treating their children equally.
The researchers worked with 300 families in rural Malawi who had at least two children in grades five to seven, telling them that two of their children would take a math test and would receive monetary awards on the basis of their scores. But parents were also given 10 lottery tickets and told one of those tickets, randomly selected, would provide an hour of tutoring for one child. The parents had to allocate the tickets between their children.
In general, parents who care only about maximizing returns should invest more in whichever child, whether high or low ability, they think will produce the greatest return on investment, the researchers note.
By contrast, parents who care most about their children earning equally should initially allocate resources toward the lower-ability child, the idea being that this child can catch up with the higher-ability child in earning potential.
And those concerned with investment equality should allocate resources—including money and time—equally to each child, regardless of their abilities and the outcome. In the experiment, any result besides an “all or nothing” distribution demonstrates concern about investment equality, the researchers write. That’s because parents unconcerned with investment equality should in theory allocate all the tickets to the child they believe will generate the higher returns from tutoring.
Before the parents allocated their tickets, they were asked to answer questions about their children’s abilities. How did they think each child would score on the test without tutoring, and how might that score change with tutoring?
After running a test lottery, to ensure the study participants understood the concept of returns maximization, the researchers gave the parents five scenarios for ticket allocation, explaining how each would maximize total earnings, minimize the earnings gap between children, or reduce allocation inequality. For example, in one scenario, the higher-ability child would receive a reward that was 10 times higher, per test point, than the lower-ability child. In another scenario, only the lower-ability child would receive a lump-sum payment after taking the test.
After one child per family received an hour of tutoring, all the children took the test, and payments were distributed on the basis of what parents chose in one of the five scenarios, which the researchers randomly assigned to each family. This ensured that a variety of responses could be measured, even though each family could experience only one scenario.
Parents did show some preference for maximizing total returns, the study finds, with 45 percent of allocations being all or nothing, in which the parents simply guaranteed tutoring to one child or the other. However, they also demonstrated a strong interest in allocation equality, enough so that they sacrificed earnings. A significant 35 percent of choices leaned toward allocation equality.
This aversion to inequality led to average earnings per child of approximately 40 percent less than if the parents had simply sought to maximize the return on their investment. The shortfall amounted to 90 percent of an adult’s daily wage in rural Malawi. A follow-up survey finds that few parents reallocated unequal test earnings among their children, meaning there is no evidence they were concerned about an income gap.
The findings could influence educational-policy design, the researchers suggest. “For example, if a gifted and talented program often only affects one child per household, providing parents with information about the high level of inputs provided to one of their children through the program may encourage parents to spend more on their other children to mitigate the inequality,” they write.