Kerri K., et al. v. California, et al. (Draft)
(June 2, 2019)
` The State of California devotes a great deal of attention to demographic differences in educational and other outcomes, always proceeding under the belief that generally reducing an outcome will tend to reduce (a) relative differences between the rates at which more and less susceptible groups experience the outcome and (b) the proportion more susceptible groups make up of persons experiencing the outcome. This page involves a putative class action brought against California on May 15, 2019, which is based on the mistaken premise that limiting the use of restraints will tend to reduce, rather than increase, the proportion students with disabilities make up of restrained students. Because California is a defendant in the case, the case may cause California to understand that reducing an outcome tends in fact tends to increase, not reduce, (a) and (b) to the outcome. The page also discusses the role of the federal government in causing this issue to be misunderstood.
As explained on many places of this site and materials referenced on it, the Departments of Education, Health and Human Services, and Justice, the Government Accountability Office and many other federal agencies, as well as the social science community, have long promoted the belief that generally reducing an outcome tends to increase (a) relative differences between the rates at which more and less susceptible groups experience the outcome and (b) the proportion more susceptible groups make up of persons experiencing the outcome. All those sharing this belief have simply taken this for it for granted and no one has ever advanced an argument as to why this would be so.
In fact, generally reducing an outcome tends to increase, not reduce, (a) and (b) as to the outcome. That is, reducing an outcome and thereby increasingly restricting it to those most susceptible to it, while tending to reduce relative differences in rates of avoiding the outcome (i.e., experiencing the opposite outcome), will tend to increase relative differences in the outcome itself; correspondingly, reducing the outcome, while tending to increase the proportions groups most susceptible to the outcome make up of persons avoiding the outcome, will tend also to increase the proportions such groups make up of persons experiencing the outcome itself.
I explain the matter fairly comprehensively in “Race and Mortality Revisited,” Society (July/Aug. 2014) and my Comments for Commission on Evidence-Based Policymaking (Nov. 14, 2016). I explain the matter in ways that may are particularly pertinent to the subject of this page in my July 17, 2017 letter to the U.S. Departments of Education, Health and Human Services, and Justice, my December 8, 2017 testimony before the U.S. Commission on Civil Rights, a handout I used at meeting with Department of Education staff on March 22, 2018, and letters of April 12, 2018, and April 17, 2018 to the Government Accountability Office. Those items discuss obligations of federal government agencies to explain to the many entities the agencies have led to believe that generally reducing an outcome will tend to reduce (a) and (b) as to the outcomes is incorrect. Also especially pertinent to the subject of this page is my Restraint Disparities page, which will eventually contain some reference to this page. Some discussions of this obligation may also be found in my “Innumeracy at the Department of Education and the Congressional Committees Overseeing It,” Federalist Society Blog (Aug. 24, 2017), my “United States Exports Its Most Profound Ignorance About Racial Disparities to the United Kingdom,” Federalist Society Blog (Nov. 2, 2017).
While most of those materials are focused on racial disparities issues, the points apply equally when the more susceptible group is students with disabilities. I assume that in the great majority of the many cases I have identified when in fact generally reductions in discipline rates were accompanied by increased relative racial differences in discipline rates – including in subpages to my of the Discipline Disparities page bearing the names or states of local jurisdictions and in the study discussed in “Discipline disparities in Md. Schools,” Daily Record (June 21, 2018) where this occurred in 20 of 23 Maryland school districts – relative differences between the rates of students with and without disabilities also increased. See Tables B1 and B2, slides 66 and 67, of my 2015 University of Massachusetts Medical School seminar, which illustrates the same point about suspension rates in Massachusetts for African American students and students with disabilities. That is, as to both groups, this state with comparatively low suspension rates shows larger than average relative differences in suspension, and smaller than average relative differences in rates of avoiding of suspensions, which is what an informed observer should expect in the circumstances. The tables also show that to the extent that the matter can be measured, racial differences and differences between students with and without disabilities are smaller in Massachusetts than nationally.
One area where, with a little thought, observers ought to be able to recognize why reducing an outcome will tend to increase the proportion more susceptible groups make up of persons experiencing the outcome involves the use of physical restraints on elementary and secondary school students. It should be evident that the use of restraints will be more necessary among students with behavioral disorders than students without such disorders. It should also be evident that the more the use of restraints is limited to circumstances where use is unavoidable, and the more schools improve techniques for interacting with agitated students that minimize the need to resort to restraints, the more the usage will be restricted to students with behavioral disorders. In a system that severely restricts the use of restraints and that has especially effective techniques for interacting with agitated students in ways that avoid resort to restraints, very few students would ever be restrained and those restrained could be entirely composed of students with behavioral disorders.
Some people concerned about disparities may find it difficult to accept this reality. In considering the matter, however, they should not think in terms whereby the extreme concentration of the use of restraints among students with disabilities might be an acceptable price to pay for reducing the number of disabled and other students subjected to restraints. Rather, the concentration should be a matter of no concern whatever, though often a useful, if imperfect, indicator of success in limiting the use of restraints. To put the matter rhetorically, why would a humane society, which would naturally want to restrict the use of restraints as much as possible, prefer a situation where it sometimes restrains students even without behavioral disorders to one where it never finds it necessary to restrain students without such disorders?
Nevertheless, the US Department of Education, which at least since 2012 has discouraged the use of physical restraints on a student other than in circumstances where necessary to avoid harm to the student or to others, monitors the perceived disproportionate use of restraints (and seclusion) on student with disabilities in terms of the proportion students with disabilities make up of restrained students. The agency does so while regarding a high proportion to be negative indicator rather than positive indicator. A March 2014 Department of Education Issue Brief titled “Data Snapshot: School Discipline” highlights (at 11) that students with disabilities, while making up 12 percent of students nationwide in the 2011-12 school year examined, made up 75 percent of students subject to restraints. And it treats as the worst disparities the situations in Nevada, Florida, and Wyoming, where, while comprising less than 15 percent of students, students with disabilities make of between 96 and 93 percent of students subject to restraints. The document treats as the states with the smallest disparities Mississippi Arkansas, and Louisiana, where students with disabilities, while comprising between 10 and 12 percent of students, make up between 40 and 43 percent of restrained students. See the Issue Brief’s Table 3 (at 18-19.
A variety of demographic factors that vary from state to state can affect these proportions. But it remains a fair assumption that restraint rates for both students with and without disabilities are lower in the states where students with disabilities make up a comparatively high proportion of restrained students than in states where students with disabilities make up a comparatively low proportion of restrained students. It is also a fair assumption that the former states do more than the latter states to comply with Department of Education guidance to limit the use of restraints. See Table 4 of the handout for the March 22, 2018 meeting with Department of Education staff.
I can see no obvious reasons why it is in Nevada, Florida, and Wyoming that students with disabilities make up the highest proportions of restrained students of all states. But it would be reasonable to assume that the fact that such proportions are comparatively low in Mississippi, Alabama, and Louisiana has something to do with the fact that they are comparatively poor states with fewer resources to devote to training and other things that would limit resort to restraints and possibly something to do with the fact that they are conservative states that might be less ill-disposed toward the use of restraints on disruptive students than other states.
The situations of these states can be contrasted with that of California, a wealthy and very progressive state that would have more resources available for purposes of reducing the need for restraints and where limiting the use of restraints might be regarded as more important than in less progressive areas. Also, while per-pupil expenditures appear not to vary widely from state-to-state, being a very large state, with a budget of $26 billion for support services, California, has more resources than other states for developing humane techniques for control of agitated students that can be used throughout the state.
In any case, Table 3 of the 2014 issue brief showed that in California, while students with disabilities made up 10 percent of total students (one of the lowest figures of any state), they made up 81 percent of restrained students. That figure is significantly above the 75 percent national figure and the thirteenth highest of any state. For reasons discussed on my IDEA Data Center Disproportionality Guide page and page 3 n. 5 of the April 12, 2018 Government Accountability Office letter, there are problems with quantifying differences between the proportion a group makes up of students and the proportion it makes up of persons experiencing an outcome. The degree of concentration is better reflected by the ratio of the rate of the disadvantaged group to the advantaged group, which can be derived from proportion a group makes up of the students and the proportion it makes up of suspended students (which is not say that degree of concentration is a useful concept, but simply that this is the best way to illustrate it) The California ratio, at 38.4, is the sixth highest among states and represents what would likely be deemed a larger departure from a national ratio of 22.0 than the difference between the 81 percent and 75 percent representational figures.
In 2018, apparently due to concerns both that restraints are harmful and that they are disproportionately used on students with disabilities and African American students, California enacted a law severely limiting the use of restraints in its schools. That law is likely to increase the standard measures of disproportionality as to both groups of concern, perhaps even causing the proportion students with disabilities make up of restrained students to approach 100 percent.
In December 2016, the same month the Department of Education issued the regulation on the measurement of racial disproportionality in the identification of students with disabilities and discipline among students with disabilities that is the subject of the COPAA v. DeVos
page of this site, agency issued a Dear Colleague letter reaffirming earlier guidance on the limitation of use of restraints or seclusion. But the principal focus of the document was on the potential discrimination issues which the letter deemed to be reflected by the fact that students with disabilities, while making of 12 percent of students in the 2013-14 school year, made up 67 percent of students secluded or restrained.
In December 2018, the Departments of Education and Justice rescinded 2014 guidance on school discipline that was premised on the belief that generally reducing discipline rates would tend to reduce relative differences in discipline rates. In doing so, however, the agencies failed to alert the public that the premise was incorrect and that in fact generally reducing discipline rates tends to increase relative differences in discipline rates (or that generally reducing discipline rates tended to increase the proportion more susceptible groups make up of disciplined students). That failure, in conjunction with the many Department of Education and Justice policies or documents that continue to promote the mistaken belief that reducing an outcome would tend to reduce the aforementioned (a) and (b) as the outcome, effectively further promoted that belief.
In January 2019, the Department of Education announced an initiative aimed at eliminating the inappropriate use of seclusion and restraints. The initiative is a joint undertaking of the agency’s Office of Civil Rights and its Office of Special Education Rehabilitate Services. While the announcement is principally focused on students with disabilities, it makes to mention of statistical disparities. It remains to be seen whether initiative activities regarding the inappropriate use of restraint or seclusion will essentially become a disparities program, though discussions of the program seem to regard it as such, as in this article that cites the high proportion students with disabilities make up of retrained and secluded students. In any case, regardless of the extent to which the program focuses on the perceived overrepresentation of students with disabilities among restrained and secluded students, assuming the program is effective in reducing the use of restraints and seclusion, the proportion students with disabilities make up of restrained and secluded students should increase. That would likely be the case even if the initiative actually identified some sort of bias against students with disabilities with respect to restraints or seclusion and caused the bias to be reduced or eliminated. That is, when bias plays a role in observed differences, reducing bias reduces all measures of disparity. But a substantial general reduction in the outcome still can cause the rates at between rates at which two groups experience the outcome to increase.
On May 15, 2019, students at the Floyd L. Marchus school in Contra Costa County, California brought Kerri K., et al. v. California, et al. in a state court. The putative class action alleges both that restraint and seclusion are used too often and that they are disproportionately used on students with disabilities. With regard to the latter matters, the complaint (at 10) cites the previously mentioned fact that students with disabilities make up 81 percent of restrained students in California compared with a 75 percent national figure.
The Marchus School is a school specifically for students with disabilities. So limiting the use of restraints and seclusion at the school will tend to reduce all measures of difference between the use of restraints and seclusion on students with and without disabilities in California. But limiting the use of restraints and seclusion at all schools will tend to increase the proportion students with disabilities make up of restrained and secluded students, as well as the ratio of the rates of students with disabilities to the rate of students without disabilities that the plaintiffs may eventually regard as a more effective illustrator of the severity of the disparity in California. With regard to restraints in particular, the recent California law may well make the case regarding disproportionality seem stronger by the time the case is tried than it seems today.
Assuming the defendants, including the state, can be made to understand the issue, the case may serve as a vehicle to promote an understanding of the way that many government policies are based on an incorrect belief about the effects of policies on measures of demographic differences. But no defendant has yet understood this issue in any other suit regarding demographic differences.
***
For further information on perceptions about and realities of the effects of policies on demographic differences in California see my California Disparities, Kern County (CA) Disparities, Los Angeles SWPBS, Oakland (CA) Disparities, subpages of my Discipline Disparities page. The Oakland page involves a situation where it has been widely reported that restorative justice programs at the Oakland Unified School District reduced relative racial differences in suspension – something also apparently believed by Oakland School officials – even though the relative racial difference in suspensions in fact increased.