UC Berkeley

In this dissertation, I explore factors that contribute to disparities in parental beliefs about children's academic proficiency across socioeconomic lines and by child gender, drawing on evidence from India, Kenya, and the United States. In all three contexts, I find a strong correlation between socioeconomic status and parental beliefs, where high income parents and those from more privileged social groups are more likely to believe their children are above average academically compared to their lower income or less privileged peers. That these patterns exist after accounting for actual test performance suggests that disparities in beliefs outpace any gaps in performance along these same lines. Parental beliefs could be consequential to the extent they guide educational investments and so shape eventual outcomes. Viewed in this way, disparities in parental beliefs along socioeconomic lines could play some role in the persistence of disparities in educational and other outcomes along these same lines.

The first chapter of this dissertation explores the link between poverty and parental beliefs in the Indian context. I leverage the India Human Development Survey (IHDS), a nationally-representative panel dataset, to explore the link between poverty or social disadvantage and parental beliefs in detail. I find that parents in poor households and those belonging to one of India's more disadvantaged caste groups are substantially less likely to believe their children are above average academically. I also find that context plays an important role; parental beliefs tend to be more positive in high-mobility or low-poverty districts, and even more so for wealthy than for poor parents. Finally, I find that parental beliefs respond negatively to exogenous negative income shocks driven by adverse rainfall events. These findings suggest that the observed link between socioeconomic status and parental beliefs may be more than purely correlational. Instead, poverty and other forms of disadvantage may fundamentally shape beliefs, leading beliefs to take on a negative bias.

In the second chapter, I replicate, validate, and extend these findings in the Kenyan contexts. Using a mix of existing data and original data collected within from the Kenya Life Panel Survey (KLPS), I explore further the link between household income and parental beliefs. I again find that parents in higher-income households are more likely to hold above-average beliefs. I also find that parental beliefs are higher among recipients of a randomized early-life health intervention that itself led to substantial improvements in household economic circumstances. Together, these findings suggest that parental beliefs may not only respond negatively to negative shocks to income, but may also respond positively to exogenous improvements in economic circumstances. I also explore other factors that correlate with parental beliefs, finding that self-efficacy and elevated aspirations correlate positively, while depression and stress correlate negatively.

The third chapter explores related patterns among parents in the United States. Here I draw on data from the Early Childhood Longitudinal Study-Kindergarten Class of 1998-99 (ECLS-K) study. I again find that household income correlates positively with parental beliefs, consistent with the evidence from India and Kenya. However, I also find evidence for disparities in parental beliefs across math and reading domains for male and female children. Specifically, while parents are more likely to believe their female children are above average overall, they are also more likely to believe their male children are above average in math and their female children are above average in reading, patterns that closely mimic prevailing gender-based stereotypes and norms. I also find suggestive evidence that mixed-sex sibling compositions within families may foster even stronger gender-based patterns. To the extent that parental beliefs shape the nature of investments parents make and which fields they encourage their children to pursue, these early-life gender-based disparities could play some role in shaping later-life gaps across males and females in terms of participation in Science, Technology, Engineering and Math (STEM) fields.

Taken together, this research sheds light on how disparities in economic circumstances and gender-based social norms may fundamentally shape parental beliefs. To the extent that these parental beliefs serve as a key input to subsequent educational investment decisions or directly influence child motivation and effort, such disparities could contribute to deepening disparities in outcomes across socioeconomic or gender lines.