Back to Top

Parental Beliefs about Returns to Different Types of Investments in School Children

By: Orazio Attanasio, Teodora Boneva, and Christopher Rauh

NBER Working Paper No. 25513

Published version

Earlier version (free)

Commentary by Gabriel Heller-Sahlgren

Parental investments in children’s education are generally assumed to be crucial for pupil outcomes. However, such investments can take many different forms. For example, parents may choose to invest time in helping their children complete their homework or prepare for examinations. They could also invest in education by opting to move to a different neighbourhood to secure a place at a well-performing state school. Or they may invest money directly by purchasing private tutoring or a place at a private school.

Differences in parental investments in education of course depend on differential resources, such as the availability of time and money, but their beliefs about returns to different types of investments are also likely to be important for their decisions. Understanding how parents perceive the relative value of different investments, and how these perceptions in turn affect investment choices, is important for understanding obstacles to parental educational investments – and how such obstacles could be overcome.

In this paper, Orazio Attanasio and colleagues collect and analyse a new dataset on parental beliefs about investments, and their actual investment decisions, in the educational process. The dataset is built from a survey of about 2,000 parents in England with at least one child aged 5–16 years, providing the first nationally representative estimates of parental beliefs about returns to educational investments. The survey presents parents with hypothetical investment scenarios that differ in three respects – the level of time investments, the level of material investments, and the quality of the child’s current school – and asks them what they expect the child’s earnings to be at the age of 30 for each scenario.

In the survey, parents are asked to consider two hypothetical English families, the Smiths and the Jones, both of which have one child attending school. They are also told these families live in the same neighbourhood and are identical in all respects, such as income, parental educational levels, and child ability, with the exception that one family’s child attends a school that has been rated ‘Outstanding’ by Ofsted and the other family’s child attends a school that has been rated as ‘Requires Improvement’.

The survey then presents the parents with four different investment scenarios – the order of which is randomly allocated – that vary in terms of the level of time and material investments: (1) low time investments/low material investments, (2) low time investments/high material investments, (3) high time investments/low material investments, and (4) high time investments/high material investments. Parents are also randomly allocated to different groups that vary along further dimensions: the initial level of time and money invested, the child’s initial ability level, the child’s gender, and the age when the investments are made.

In total, each respondent is therefore presented with eight different scenarios – that randomly vary across the respondents – and for each of these they are asked to determine how much they expect the hypothetical child will earn at the age of 30.

The results indicate that parents on average believe that attending an ‘Outstanding school’ rather than a school that ‘Requires Improvement’ increases earnings at the age of 30 by about 10 per cent. They also believe that spending three more hours per week with the child increases earnings by 21 per cent, while spending £30 more per week raises earnings by 15 per cent. Parents therefore believe time and monetary investments are more important for earnings than the quality of the school – and that three extra hours of time spent with the child per week have a higher return than spending £30 per week. They think that the returns to monetary investments are higher if the child attends a higher-quality school, but they do not believe that investments complement each other in other respects.

Parents also believe returns are higher if the level of time investments increases from 0 to 3 hours per week than if it increases from 1 to 4 hours per week, and that the returns are higher if monetary investments increase from £0 to £30 per week than if they increase from £10 to £40 per week – suggesting a belief in diminishing returns – but they do not believe that returns to investments depend on the initial skills or the gender of the child. However, they do believe that school quality matters more for older pupils than younger ones. Importantly, perceived returns do not differ depending on the respondents’ socio-economic background, suggesting that parents from different backgrounds hold similar beliefs about the relative importance of different parental investments. At the same time, perceived returns are positively related to parents’ belief in the malleability of skills, suggesting some parents may choose not to invest simply because they think it will not make a difference.

Finally, parental beliefs about the relative value of different investments do predict their actual investment choices, as reported by the respondents. For example, a 10 percentage-point increase in the perceived return to time investments predicts 46 minutes more time spent with their children per week, whereas the same increase in the perceived returns to monetary investments predicts an increase in monthly education spending by £15.

Overall, the paper provides useful data for policymakers seeking to increase parental investments in education. For example, as respondents’ belief in the malleability of skills positively predicts perceived returns, targeted information may be able to increase investments by improving awareness of the extent to which interventions in fact affect pupil performance. Providing information about the relative effectiveness of different types of investments, which requires more research to determine, may also be important to improve parental decision-making more generally. Testing such ideas through experiments would at the very least be a fruitful venue for future research to pursue.


Gabriel Heller-Sahlgren is CfEE’s Lead Economist and Editor of its Monthly Research Digest, in the February issue of which this commentary first appeared.

You can view it in flipbook, download a pdf, and subscribe here.


Blog Category: 
About the author