Poverty and Education Outcomes: Place Matters

Details

Hits: 278

To understand the relationship between neighborhood level poverty indicators and student trajectories we developed three maps. When considering the first map of on-track African American male students and food stamp rates, there is a very clear association between the neighborhoods with lower levels of food stamps and higher percentages of on-track students -- wealthier communities have more successful students in this case.

However, when we consider the map showing the food stamp rates and the percentage of at-risk students, we see a markedly different picture -- the neighborhoods with higher levels of poverty (higher food stamp rates) also typically have higher rates of at-risk students. When we examine the map showing off-track students and food stamps, the picture is less clear -- there are higher levels of off-track students in both low and high poverty neighborhoods at similar rates.

This may seem counterintuitive, but is perhaps not unlike the dropout effect: many school indicators like chronic absence appear to improve in grades 11 and 12, but the reality is that the most-absent students often have dropped out by those grades, falsely improving the absenteeism statistics. In the case of the off-track students versus food stamps something similar may be present and this relationship bears further investigation. The result from this analysis, however, is that across the city there is a solid relationship between high poverty and the proportion of at-risk students.

When statistically comparing the rate of food stamp recipients in each Census tract in 2010 with the percentage of African American boys considered at-risk, the data showed a strong positive correlation with a coefficient of 0.597. For those students considered on-course, a negative coefficient of -0.431 was found, both significant at the p=0.001 level. What these statistics demonstrate is that at a neighborhood level, higher rates of poverty are strongly tied to a higher proportion of students at-risk while neighborhoods with lower poverty are correlated with more students on-course. This again illustrates the impact that living in a poor neighborhood can have on a student's well-being. We did not find any significant correlations between the students off-course and poverty. As so many national groups are now adopting as a mantra: place matters. It always has, and it's clear in Oakland when you map any number of social outcomes and neighborhood conditions.

African American Male Suspensions by Neighborhood

Details

Hits: 639

As part of our investigation into the variations in suspension practices and outcomes across OUSD schools, we mapped the location of each student and aggregated their data to the census tract they live in.  We then symbolized each tract (darker color indicates higher rate) with the rate of suspension for the African American males in that tract for the 2010-2011 academic year.  The second map below shows the rate of suspension for ALL students per school using the exact same break values in the tract coloring. We think these maps clearly illustrate the massive differential in suspension rates for African American male students compared to students overall and also serve to show how outcomes vary widely across the city.  There were only three tracts with overall suspension rates in the second highest range as the rates for African American males. Again we see an east/west bi-modal distribution: African American males are suspended at high rates across the city, but the overall rates in the central flat-lands are quite positive in comparison to the rest of the city. From the first map, the most striking outcome is that in almost half the city's neighborhoods, African American males are suspended at a rate of at least one in five students in a single year.

As with most outcome data it is important to consider the data from both a facility (school) and a community (tract) perspective.  To view the maps of suspension rates by school read this post. To read the full attendance report, click here.

Note that for relevance we removed tracts with fewer than 10 African American males as these neighborhoods' data can easily skew the results and create misleading figures that are not statistically stable (numbers that cannot be trusted for interpretation).