Reframing Poverty in Our Schools: Using GIS to Map Inequality

by Jesse Senechal


In our highly polarized political environment – where every policy debate becomes an opportunity for harsh ideological clashes and brinkmanship – there seems to be one consistently common ground issue, public school policy.  From left to right, the argument is familiar.  It begins with the crisis rhetoric of failing schools, achievement gaps, and bloated bureaucracies and then moves to a demand for “reforms,” a term which has come to mean some ensemble of policies that includes standardization of curriculum, continued use of high-stakes tests, introduction of performance-based pay for teachers, and expanded school choice through charters and vouchers.  These policies, it is commonly argued, will “shake up the status quo” and ultimately lead to more efficient, effective and equitable systems of public schooling. 
In a certain respect, this popular political consensus should be reassuring.  In this divisive age at least this is something on which we can all agree.  However, as a long time public school teacher and parent of a public school student, I can’t help but feel discouraged.  While my discouragement is, in some part, due to my first-hand experiences with the negative effects of testing, accountability and choice initiatives on students and schools; the main reason for my lack of faith in these “reforms” is that they continue to ignore the critical connections between the system of public education and the other important social systems that constitute the communities in which schools exist.  Instead of developing policies that acknowledge and respond to the impact of, for example, housing, healthcare, transportation and food policy on educational outcomes, these “reforms” tend to focus myopically on school specific solutions.  The phenomenon recalls the quote by Henry Louis Menken, ‘for every problem there is a solution that is simple, clean, and wrong.’
What is most surprising about this myopic educational policy perspective is that it has no foundation in research.  In fact, if there is one powerful and indisputable finding from the body of educational research over the past several decades, it is that, in terms of student outcomes, what happens out of schools matters just as much if not more than what happens in schools.  The clearest example of this is the strong correlation that has been established again and again between achievement and socio-economic status (SES).  Students from middle class and upper middle class backgrounds score consistently higher than students of working class and poor backgrounds.  While this certainly suggests some school-specific policy problems (e.g. unequal school funding or culturally disconnected curriculum), it seems that this finding would lead to a concerted effort among policy makers and educational leaders to understand the complexity of the social context in which schools operate and to develop policy that responded to it.  The reaction instead has been to move forward with the typical policy agenda and simply “control for” the out of school factors.
One way of illustrating the depth of this problem is to consider the how the system currently measures poverty in its schools: the percentage of students receiving Free and Reduced Lunch (FRL).  The justification for this lies in the fact that a student’s eligibility for FRL is determined in relation to the federal poverty line, however when considering FRL as a measure of poverty what becomes clear is how incredibly inaccurate this measure is.  Not only have studies shown that FRL is inaccurately awarded 20% of the time, but it is also has declining reliability as students get older and feel the stigma of receiving a free lunch.  There is also the problem that as a dichotomous measure it makes no distinction between a student who just misses the FRL eligibility line (for example, coming from a family of four that has a $40,000 annual income) and one who is far beyond the base line (from an upper middle class family).
To better understand this issue, I constructed a map using GIS software that examined poverty levels at a number of elementary schools in the Richmond area.  The goal was to compare school FRL percentages with median income data by census tract from the 2000 census.[1]   For each selected school, the attendance zone was overlaid onto a census tract map and the average of the attendance zone median incomes were taken.  The justification for this analysis rests on several ideas.  First, elementary schools are generally neighborhood schools, meaning that they draw from and reflect the populations of their attendance zones.  Second, unlike the FRL measure, this measure is not dichotomous and would thus allow a more accurate comparison of the broad range of SES between schools.  And finally, even if there is a discrepancy between the true student SES and the school zone income measure I have created, I would argue that the relative wealth of a neighborhood in which a school lies (and the access to certain systems and institutions that might afford) has a relationship to the SES of students.
The maps presented here show four examples of how elementary schools rate on this new measure.  The first, Bettie Weaver Elementary is located in the northwest part of Chesterfield County in an attendance zone that borders the James River and Goochland County.  It enrolls approximately 900 students (93% White, 4% Asian, 2% Black, and 1% Hispanic).  Out of the 184 census tracts in the three district area, Weaver's attendance zone draws from three of the five most affluent – all with median incomes above $100,000 a year.  Consequently the school has a 0% Free and Reduced Lunch (FRL) population.  This is an example of a school whose student body is likely to reflect the high SES of the surrounding community. 

The second, Fairfield Court Elementary is located in the east end of Richmond along the Henrico county border.  It enrolls approximately 500 students (99% Black and 1% Hispanic).  Its zone draws from two of the five poorest census tracts in the region.  Like Bettie Weaver, it is a school where there is a strong correspondence between the FRL measure (95%) and the tract median income ($15,419). Unlike, Bettie Weaver, this is a school that serves the least, rather than most, affluent students in the region. 

The third, Maybuery Elementary is located in the west end of Henrico.  It enrolls approximately 550 students (72% White, 16% Black, 8% Asian, and 5% Hispanic).  While the school zone includes the wealthiest of the region's 184 census tracts (a tract with a median income of $170,552 a year) it also includes portions of one of Henrico's poorer tracts ($39,375 a year).  Although the averaged median income in this attendance zone is high, the FRL population is also relatively large (23%).  This raises questions about the SES of the students at the school.  Is the range of student SES as dramatic as the community's, or do the more affluent students from the southern census tract opt for private school options (several are located in the area)? 

Finally, Westover Hills Elementary is located on the south side of Richmond.  It enrolls approximately 400 students (95% Black, 4% Hispanic, and 1% White).  While the school is located in a relatively high income census tract ($53,800 a year), it draws from several less affluent tracts.  It is likely that, as with the case of Maybuery, the broad range of tract income levels tends to push the SES of the student body lower.  It is likely that many school age children from more affluent families within Westover Hills’ zone do not attend the school, but are rather enrolled in open enrollment or private schools. 

The goal of this analysis is not to suggest that looking at census tract data is a better indicator of school SES level – in many cases, as illustrated above, it is much less accurate.  However, the benefit of this type of analysis is that it raises questions and it points to the fact that each school exists within a unique context that relates not only to issues of wealth, but also housing, zoning, school enrollment policies, transportation and development.  Along these lines, think how this map might become even more revealing if it included layers that added bus routes, grocery stores, banks and parks.  As I argued above, these are all interconnected systems that are critically important to school outcomes.
In conclusion, I would suggest that we need to reframe the issue of poverty in schools.  For too long we have moved forward with school policy that considers poverty as a condition that we control for as we strive for higher achievement, rather than an outcome that we can affect.  While we have ambitiously declared within the context of schools that no child should be left behind, we have failed to advocate the No Child Left in Inadequate Housing policy or the No Child Left Without a Park policy.  And part of this call for reframing poverty in schools will involve establishing a more sensitive measure than Free and Reduced Lunch.  If we invested even a fraction of the money we put into standardized testing into creating an accurate measure of poverty in schools, we could have a much better way to assess the scope of the problem and develop policies that account for how the interconnected systems of communities affect the equity and justice of our schools. 



[1] At the time of this analysis the 2010 census data was not yet released by census tract.