Does The Rise of Private Tutoring Centers Contribute to Economic Inequality in U.S. Education?

This research investigates the rapid growth of private tutoring centers in the U.S., focusing on economic and demographic predictors of their location. Using 25 years of geocoded data on U.S. tutoring centers, researchers found that the number of centers tripled from about 3,000 in 1997 to 10,000 in 2022. The analysis shows these centers are predominantly located in affluent areas, with 60% in the highest income quintile. A notable finding is that areas with a significant Asian American population, even after controlling for income and education, have more tutoring centers, highlighting cultural factors alongside economic disparities in access to tutoring.

Central Research Question

This study examines the rise of private tutoring centers in the United States, aiming to understand the factors driving their rapid growth and geographic distribution. The central question explores how economic, demographic, and cultural elements influence the prevalence of tutoring centers, particularly in high-income and education-focused areas. The research investigates the potential implications of these factors on educational inequality, considering whether private tutoring further amplifies advantages for affluent families or addresses broader educational needs.

Previous Literature

The literature on private tutoring has traditionally focused on its prevalence in Asian countries, where it has become integral to educational systems and competitive academic environments. Research in countries such as South Korea and China indicates that private tutoring is a response to competitive exams and a means for high-income families to secure academic advantages for their children. Studies in the U.S. have documented the popularity of tutoring for test preparation among Asian American communities, noting that cultural expectations and educational aspirations drive demand for supplemental education. Additionally, past research has shown a positive impact of tutoring on student outcomes, with evidence suggesting modest improvements in standardized test scores and college admissions. However, the U.S. literature remains limited in understanding the spatial and demographic dynamics of private tutoring centers.

Data

This study relies on a comprehensive dataset spanning over 25 years (1997–2022) and includes geocoded information on approximately 33,000 private tutoring center locations across the United States. The dataset, sourced from Data Axle’s Historical Business Data, includes business names, addresses, and industry classifications, covering firms registered under “Tutoring” and “Test Preparation Instruction” codes. The data enables the tracking of individual tutoring center locations over time and captures the growth in number and geographic distribution of these centers.

To analyze predictors of tutoring center locations, researchers combined this data with school district-level demographic and economic indicators from the U.S. Census and the American Community Survey (ACS). These indicators included variables such as income per capita, educational attainment, racial composition, and private school enrollment rates. Additional data from the Common Core of Data (CCD) was used to include school district characteristics, such as student enrollment in charter and magnet schools and socioeconomic indicators like free or reduced-price lunch eligibility.

Methods

The study employs a multi-stage analytical approach, starting with descriptive analyses to document the overall growth and geographic distribution of private tutoring centers. Researchers then conducted a regression analysis using the LASSO (Least Absolute Shrinkage and Selection Operator) variable selection method, which helped identify significant predictors of tutoring center prevalence by minimizing overfitting from a large pool of covariates. This process yielded a concise set of predictors focused on income, educational attainment, racial demographics (notably Asian American population share), and private school enrollment.

Two models were developed: a cross-sectional model for 2020, examining current predictors of tutoring center prevalence, and a predicted-change model for the period 2000 to 2020, assessing changes in tutoring center distribution over time. A non-parametric regression analysis further explored non-linear relationships between predictors and tutoring center prevalence.

Findings and Size Effects

The study’s findings underscore a substantial increase in the number of private tutoring centers, which tripled from approximately 3,000 in 1997 to nearly 10,000 in 2022. The growth trajectory was relatively stable until 2015, followed by a slight plateau and resumed growth through 2022. The prevalence of tutoring centers is notably concentrated in affluent areas, with 60% of centers located in regions within the top income quintile. The analysis also shows that areas with high levels of parental education are more likely to have tutoring centers, suggesting that educational attainment drives demand for supplementary educational resources.

A significant demographic finding is the strong correlation between Asian American population density and tutoring center prevalence. Even after accounting for income and education, school districts with higher Asian American populations tend to have more tutoring centers. This pattern suggests that cultural factors, including high educational expectations and familiarity with tutoring from countries of origin, contribute to the demand for private tutoring. The highest income school districts in 2020 had an average of 0.133 tutoring centers per 1,000 students, compared to just 0.017 in median-income districts, indicating a considerable size effect based on economic status.

The findings also reveal a limited relationship between the presence of charter or magnet schools and the prevalence of tutoring centers, suggesting that private tutoring demand may be more a function of family-level educational aspirations than school choice availability. Additionally, while private school enrollment did correlate with tutoring center presence in some analyses, this association was relatively minor, supporting the idea that private tutoring demand exists alongside other high-cost educational options rather than serving as a substitute.

Conclusion

This study offers a pioneering exploration of the private tutoring industry in the U.S., presenting a nuanced understanding of its growth and geographic distribution. The findings indicate that private tutoring centers are primarily a resource for high-income, highly educated families, with particularly high demand in areas with large Asian American populations. The concentration of tutoring centers in affluent areas raises questions about the implications for educational equity, as families in wealthier districts access more academic resources than those in less affluent areas.

The authors suggest that the rise in private tutoring parallels broader societal trends in high-income families’ investment in children’s education, similar to patterns observed in other competitive and resource-intensive activities. The study concludes that while private tutoring may provide additional learning opportunities, it is likely to contribute to existing disparities in educational access and outcomes. Future research is encouraged to assess the direct impact of private tutoring on student achievement and to explore potential policy responses, particularly in relation to addressing educational inequalities in the U.S.

Citation

Kim, E., Goodman, J., & West, M. R. (2024). The Recent Rapid Rise of Private Tutoring Centers. Education Finance and Policy. https://doi.org/10.1162/edfp_a_00438