The main Discipline Disparities page of this site discusses interpretations of data on racial differences in school discipline rates in light of the pattern whereby the rarer an outcome the greater tends to be the relative difference in experiencing it and the smaller tends to be the relative differences in avoiding it. In the school discipline context this means that, contrary to the view of the Department of Education and Department of Justice. relaxing discipline standards tends to increase, rather than decrease, relative racial differences in disciplines. I explain the point fairly succinctly in “Misunderstanding of Statistics Leads to Misguided Law Enforcement Policies” (Amstat News, Dec. 2012) and The Paradox of Lowering Standards” (Baltimore Sun, Aug. 5, 2013). This subpage discusses patterns observed in Los Angeles when a program was implemented aimed at generally reducing discipline rates. The fact that relative differences increased following implementation of the Los Angeles program is also discussed in my “Racial Differences in School Discipline Rates” (The Recorder, June 22, 2012). The DOE Equity Report subpage discusses that Department of Education data show that relative racial differences in expulsions are smaller at schools with zero tolerance policies than at those without such policies.
Beginning in the 2007-2008 school year, out of concern about generally high discipline rate and racial disparities in discipline rates, the Los Angeles Unified School District adopted the “School Wide Positive Behavior Support program (SWPBS), which was described as “an evidence-based approach to improving student behavior and learning outcomes by focusing on behavior modeling, corrective responses, and intensive proactive interventions, and by seeking to decrease the use of aversive and exclusionary punishments, such as class removal and suspension.” A report styled Redefining Dignity in Out School: A Shadow Report on School-Wide Positive Behavior Support Implementation in South Los Angeles[i] found that while the program reduced discipline rates it increased racial disparities in discipline rates.
Table 1 below, which is based on the suspension rate information in the report’s page 6, shows the black and white suspension rates for the last school year before the program was implemented (2006-07) and for the most recent year covered in the report (2008-09), along with the ratio of the black suspension rate to the white suspension rate and the ratio of the white rate of avoiding suspension to the black rate of avoiding suspension. The final column shows the difference between black and white outcomes according to the method, discussed on the main Discipline Disparities page and elsewhere, that is unaffected by changes in the prevalence of an outcome. The figure, “EES” for “estimated effect size, is the differences between means of hypothesized underlying distributions derived from each pair of rates.
Table 1 Black and White Suspension Rates Before and After SWPBS, with Measures of Difference [ref b2822 a 2]
In this case, it is clear that the black-white disparity widened in a meaningful sense. One can tell that based on the fact the relative difference in avoiding discipline increased, which is contrary to the direction in which prevalence-related forces typically would drive such difference when discipline rates decreased generally. But one can tell that more effectively by the EES figure.
In circumstances where the figures for both groups being compared are reliable, there might be reason to explore explanations for the increase in the disparity. In that regard, it is useful to keep in mind that the underlying distributions of susceptibility to suspension of each group of which I broadly speak are in fact comprised of varying distributions with respect to different types of offenses. The racial differences between those distributions may well differ by type of offense and it may be easier to dramatically reduce or eliminate suspension for some offenses than other. Such underlying circumstances could yield varying results by race when efforts are made to generally reduce discipline rates.
But in the situation examined in the report, while blacks comprised 18.9% of students in 2008-09), whites comprised only 0.5% of students that year. Thus, the figures for whites are likely based on a very limited number of observations where random variation may easily have had a substantial role.
A more useful appraisal of the patterns of changes would compare black students with Hispanic students who comprised 79.5% of all students in 2008-09. Thus, Table 2 presents the same information for blacks and Hispanics that Table 1 presented for blacks and whites.
Table 2 Black and Hispanic Suspension Rates Before and After SWPBS, with Measures of Difference [ref b2822 a 3]
In Table 2, one observes patterns that are much more in accord with the distributional forces described on the main Discipline Disparities page and other places on this site. The relative difference in discipline rates increased and the relative difference in rate of avoiding discipline decreased. The EES figures indicates a slight widening of the gap, which may or may not be statistically significant.[ii]
[i] The report was jointly prepared by the organizations Community Asset Development Redefining Education (CADRE), Mental Health Advocacy Services, Inc. and the Public Counsel Law Center.
[ii] Even if the difference is statistically significant, there is sufficient uncertainty about the precise shapes of the underlying distribution that it would be mistake to the think the change from .77 to .81 strongly suggests anything meaningful occurred in the comparative situation of black and Hispanic students.