Home » 2015 Categories » 2015 Literature Review » Literature Survey: Education Finances and Segregation by Haisi Liu

Literature Survey: Education Finances and Segregation by Haisi Liu

A fascinating, age-old U.S. public policy question surrounds the relationship between educational outcomes and segregation (e.g., income, racial). Over the years, growing economic literature continues to add nuanced perspectives to the issue of the interaction between school finances and the makeup of local communities. Social science researchers emphasize the importance of studying residential segregation, whether by income or race-ethnic groups, because of potential neighborhood effects on the long run educational outlook for young students. Since public education in the U.S. is largely financed by local property taxes, there is a large disparity between funding for schools in different communities.

Thomas Nechyba, “School finance, spatial income segregation, and the nature of communities”:

Although many education policy workers focus almost solely on the impacts of disparate per pupil expenditures across schools, a large body of economic research shows strong evidence that there are broader factors affecting educational inequities, and vice versa. For instance, since public school systems are based in local districts, residential segregation by income is more likely to occur—and over time this greater residential segregation feeds even larger inequalities, thus leading to a relentless cycle. Differences in education quality are capitalized into housing prices (i.e., property values often further decline in poor areas with inadequate schools).

Nechyba (2003) questions why the contemporary educational inequalities discussion is so restricted to per pupil spending gaps; rather, he calls for a more general examination of different school finance institutions and their various equilibrium effects. Thus, the paper sets up a structural model representing a decentralized economy in which households select their place of residence, where to send kids for education, and the degree of support to provide public schools; the model includes the most causal potential factors leading to income segregation.[1] Using observations from New Jersey districts, Nechyba (2003) adjusts the underlying structural parameters until his model simulates realistic features from the data. Then, holding these parameters constant, policy simulations can be conducted. The framework accounts for a couple main sources of residential segregation by income: 1) different neighborhoods are historically endowed with disparate housing quality and neighborhood characteristics; 2) places of residence affect a child’s school quality since public education systems require households to live within exogenously outlined boundaries. However, the inclusion of private schools into this model complicates matters because households that send their children to private educations care less about public school quality near their homes; alternatively, they would be even motivated to live in a subpar public school district with lower housing costs. Since private school households are often wealthier, this counterintuitive phenomenon pushes the local economy toward income desegregation.[2]

Subsequently, Nechyba (2003) use the structural model to find that state financing of school districts indeed dampens residential income segregation in an area. [3] This effect is expected because school systems that are purely locally financed give wealthier households motivation to segregate by income and thus shape better schools. However, two less intuitive findings are that the existence of a private school market leads to considerable residential income desegregation, and that in the presence of private schools, the existence of public schools actually tends to lower residential segregation—even though, by themselves, public school systems are associated with higher spatial segregation. However, a notable caveat to this study is that the model equilibrium overestimates the movement of private school households into poorer neighborhoods simply because the structural model does not account for non-school characteristics of such communities.

Raquel Fernandez and Richard Rogerson, “Keeping people out: income distribution, zoning, and the quality of public education”:

Meanwhile, Fernandez and Rogerson (1993) investigate the effects of community zoning regulations on per pupil spending, particularly with respect to property taxes and the formation of communities with average income differences. The analysis relies on simulations using a two-community model, where each community determines tax rates by majority vote, and households can choose their place of residence. Without zoning, the equilibrium of this model results in a poor neighborhood with a low tax rate (i.e., low per pupil expenditure) and a wealthier community with a higher tax rate.[4] The study found that the existence of zoning regulations, meaning that households have to purchase a minimum level of housing in order to live in a predetermined area, is associated with the wealthy community decreasing in size and becoming richer, while the poorer neighborhood grows larger. The increased exclusivity of the rich neighborhood causes the lowest income households of that neighborhood to enter a poorer one, thus raising average income in both areas.[5] As a result, the poorer community sees greater per pupil spending, but the change in education quality across the two areas is ambiguous.

Sarah Reber, “School desegregation and educational attainment for blacks”:

Although the first two studies mentioned did not delve into racial segregation, it is important to gauge the relationship between segregation and educational attainment for certain race-ethnic groups. Reber (2007) looks into the effects of the desegregation process after the Supreme Court’s 1954 Brown v. Board of Education decision, specifically focusing on schools in Louisiana. Previous studies showed that the desegregation policy had two effects: 1) increasing black students’ exposure to white peers in school and 2) raising the level of federal and state funding such that the average spending in predominantly black schools increased. Given that desegregation essentially eliminated disparities in student-teacher ratios for black and white students within previously segregated districts, Reber (2007) sought to determine whether greater educational attainment by blacks following desegregation resulted more from greater exposure to white peers or increased funding.[6]

The simplest specification in this paper is a univariate regression that looks for an association between a county’s initial share of black enrollment and change in educational attainment from before and after desegregation; in this model, the dependent variable of interest is mean attainment for 1970-1975, minus the average attainment for 1960-1965. Later, the paper considers metrics such as the 12th grade continuation rate and adds controls for other characteristics such as change in county’s employment. These regressions on high school grade continuation and graduation rates indicate that greater educational attainment increased more for black students in districts with higher rates of black enrollment following desegregation—thus implying that, following desegregation, increased funding played a larger role than higher exposure to white students in raising attainment.[7]


Growing literature contribute to U.S. policies designed to match educational opportunities of students coming from vastly different racial and socio-economic backgrounds. Whether through investigating the means by which certain communities are able to make use of taxes and zoning as instruments to keep specific segments of the population out of a school district, or by evaluating the effect of private schooling on income segregation, valuable insights can be drawn from these types of analyses. A thorough look at the institutional set-up of education can reveal the role that segregation and various school finance mechanisms play in long-run inequality.


Raquel Fernandez and Richard Rogerson, 1997, “Keeping people out: income distribution, zoning, and the quality of public education,” International Economic Review 38(1): 23-42.

Sarah Reber, 2007, “School desegregation and educational attainment for blacks,” Cambridge, MA: NBER working paper 13193.

Thomas Nechyba, 2003. “School Finance, Spatial Income Segregation and the Nature of Communities,” Journal of Urban Economics 54(1), 61-88, July.

[1] Thomas Nechyba, School finance, spatial income segregation, and the nature of communities, 66: It is notable that although this analysis focuses on income segregation, it can apply to problems involving racial segregation as well, “only to the extent that such segregation is driven by income differences.”

[2] Nechyba, 65

[3] Nechyba, 74: “While it might be expected that state financing will lead to less segregation than local financing, the relatively small magnitude of this effect compared to the huge effect of private schools is surprising, as is the different effect of public schools in a world with and without private school markets.”

[4] Raquel Fernandez and Richard Rogerson, “Keeping people out: Income distribution, zoning, and the quality of public education,” 1

[5] Fernandez and Rogerson, 32

[6] Sarah J. Reber, “School desegregation and educational attainment for blacks,” 3

[7] Reber, 6: To prevent white flight, schools with larger proportions of black students received more funding to “level up to the levels previously experienced only in the white schools”


  1. I found your paper very interesting. I enjoyed learning about how factors like zoning and taxes can impact schools. I thought you looked at thought-provoking models in this paper such as the two-community model where citizens vote on their tax rate.

    I found it interesting that the poorer community in the Fernandez and Rogerson paper see a greater per pupil spending because of zoning regulations but the change in education quality is still ambiguous. I would have that thought that the education quality would rise as per pupil spending increases.

    Also, it would have been helpful if you had spent more time tying the papers together or if you had chosen papers with contrasting viewpoints of the same topic. However, I thought you did a great job of explaining the ideas that were examined in the different papers.

  2. I enjoyed the findings of your literature review, especially because it’s on a topic that is very relevant to our world as students (most of whom have been raised and educated under the American education system).

    I found it very interesting that the existence of a private school market leads to residential income desegregation, and that in the presence of private schools, the existence of public schools actually tends to lower residential segregation—even though, by themselves, public school systems are associated with higher spatial segregation. You mentioned that households that send their children to private educations care less about public school quality near their homes — a statement that I find interesting because where I grew up, having a home in a good public school district is extremely important (even if there are private schools nearby) for the re-sale value of the home.

    I also found it very interesting that the increased exclusivity rich neighborhoods ultimately results in greater per pupil spending in poorer communities but the change in education quality across the two areas is ambiguous. Especially since in the final study that focused on racial segregation, increased funding played the more significant role in raising attainment.

    Overall, the insights you provided through your literature review were both well-developed and interesting. Your selection of papers added a diversity to the conversation of education and segregation. However, it would have been nice to see a more common theme throughout all of them, perhaps had you selected one form of segregation (income, race, etc.) and presented it from various viewpoints.

  3. Haisi,

    I agree with you that the relationship between educational outcomes and segregation is a fascinating topic that unfortunately from the policy perspective we, as a society, have not been able to find a solution. I think that the studies you chose are relevant not only for the U.S. but also for other countries.

    I agree that at first glance Nechyba’s (2003) results are counterintuitive. However, it is possible that private schools choose to locate in areas that were homogeneous in the first place. It will be very interesting to see in the coming years what happens with the relationship between educational outcomes and segregation in the presence of charter schools such as KIPP schools.

    In general, research studies about education and inequality focus in the supply side. What I believe will also be an interesting discussion is understanding the demand side. For instance, how and why some families are willing to invest more in their children than others, or is it possible that families are choosing to invest more in one child than in other? I recently read a paper by Marcos Rangel, and thought that analyzing the parent’s perspective was a very interesting approach (Is parental love colorblind? Allocation of resources within mixed families: http://faculty.ucr.edu/~jorgea/econ261/colorblind.pdf )

  4. This literature review presents some really interesting insight into the schooling disparity between neighborhoods. Nechyba’s paper suggests that using local financing for schools causes a larger rift between high income and low income neighborhoods. Is this the only significant finding that his model offered? The paper also seems to mention the issues of housing quality and neighborhood characteristics. It makes sense that state level funding is a much more equal method of spreading education money since both high and low income areas receive the same amount of funding per head. However, I’m not sure if this is right way to go. Yes, it help equality, but I can’t help but think back to the No Child Left Behind policy instituted by the Bush administration which ended up regressing education rather than progressing it. It would be nice to make sure every community had a minimum level of funding provided by the state or local community, but if higher income places are willing to draw more funds from taxes to spend on education, they should still be able to.

    Nechyba also studies the effect of private school in a neighborhood. I think even considering this effect is dubious at best, since as you noted, it complicated the model since residents whose children attend private school have different incentives than residents whose children attend the local public school. Also many private schools draw students from all over the state or region, not necessarily from just that neighborhood, so the relationship between the existence of a private school, and the income distribution within a community might not even be substantial.

    Several of the papers mentioned using simulation models. I’m quite curious about the structure of these models since they also seem to incorporate real data as well. Are the researchers simulating people’s choices and movements given a certain legislative or budgeting decision? How do they measure uncertainty generated by such a model? I feel like there are so many factors that affect both the obligatory and freely made decisions that residents make that it’s really hard to make causal connections. Therefore, education equality can a tricky question and the results of this type of research is not easily translated into effective policy.

  5. I enjoyed reading your paper, thanks for posting! The research that you examine here makes some very interesting points about the ways in which zoning and state financing influence residential income segregation and school quality. It also makes me continue to question the ways in which local property taxes are tied to school funding throughout the U.S. It seems obvious that the predominant system of school funding today is systematically providing unequal access to high quality public education. Wealthier families have significant incentives to choose schools attended by other wealthy families, as those schools will be better funded – as a result, there is grouping of high-income households. Meanwhile, the families of students from more humble backgrounds cannot afford to live in those wealthier neighborhoods, and so there is also grouping among low-income families. There is no better illustration of this than the family choices made by Duke professors. While I only have a few data points, every Duke professor with children that I have spoken with chooses to live in Chapel Hill rather than Durham. While this choice is likely influenced by a number of factors, I think access to high quality schools is certainly one of them.

    Given this background, the conclusions reached in each of the papers you examine do not seem particularly surprising. Reber’s conclusion that black students had higher educational attainment rates in integrated schools because of increased funding rather than exposure to white students makes sense (in most cases, integration of schools involved bussing in students from lower income neighborhoods – the spending gaps between the schools these students previously attended schools and white schools were generally very large). Nechyba’s conclusions that increased state funding and the presence of private schools lower residential income segregation are also expected, given that initiatives such as these link school funding to sources other than local property taxes (hence, there is less incentive for wealthier families to relocate in order to have access to better funded/higher quality schools).

    I think the next step is figuring out how to fix the system. Local property taxes clearly create residential income segregation (and more importantly, are directly determining access to high quality public education). While a flat school tax would probably be the easiest fix, it is also highly unlikely that that sort of policy would ever be passed in the U.S. Increased government spending may be a temporary solution, as it currently accounts for less than 10% of education expenditures. Regardless, something needs to be done in order to address this problem of inequitable school funding. Memphis is a great current example of the public school funding issues, as an attempted 2013 merger of Shelby County (relatively high income) and Memphis City Schools (relatively low income) has been a complete disaster – over the past year, Shelby County has created six separate municipal school districts for each of the major suburbs.

  6. Great review, Haisi! Very interesting paper on this usually less-than-adequately addressed topic. You mentioned that Nechyba called for a more holistic measure than per pupil spending, and I wonder what your thoughts are on what potential indicators should be consider that are both useful and cost-effective to get data on (a quick search suggested expenditure per average daily attendance, but that seems to mask the problem of lower attendance schools). It was only after coming to Duke did I realize the great variance in the caliber of resources available in high school students across the country, but I suppose quality and accessibility of resources may be harder to quantify. Going forward, it might be interesting to contrast these findings suggesting structural push towards segregation with agent-based models (household-choice centric) of segregation, perhaps Shelling’s model of segregation and neighborhood formation which suggests that even the slightest preference to be around agents (households) with similar characteristics could lead to heavy segregation given enough time.

Leave a Reply to Ariadne Rivera Cancel reply

Your email address will not be published. Required fields are marked *