Some racial disparities are a disturbing sign of potential racism. Others reveal nothing of the kind, and aren’t even a warning sign of flawed policies. Focusing on them is just meaningless racial bean-counting.
In 2016, the Obama administration adopted a regulation that measures the latter, meaningless kind of disparity, rather than the former. It’s known as the “significant disproportionality rule.” It assesses whether placement or discipline of children in special education is racially disproportionate, and then redirects 15% of a school district’s special education funding if a “significant disproportionality” is found.
But it uses the wrong yardstick for disproportionality. It assesses disproportionality based on statistical “overrepresentation” rather than better alternative yardsticks such as “over-identification” — even though federal law equates “significant disproportionality” with “groups that were significantly overidentified.”
That is, it finds racial disparities simply when black children are identified as needing special education at higher rates than other races, or placed in special education at higher rates, even if black children actually need special education at higher rates due to growing up in deprived circumstances — as scholars have noted is actually true in the real world. And schools are deemed racially disproportionate when they discipline black students at a higher rate, even when black students actually misbehaved at a higher rate, meaning that individual black and white students were treated the same, and no student was treated disproportionately harshly based on her race.
This is a very defective yardstick for disproportionality. As a result, the significant disproportionality regulation is utterly meaningless as a measure of whether a school district is discriminating against black childen or whether it over-identified them for special education compared to the correct rate. This defeats a goal of the Individuals with Disabilities Education Act, which seeks to find out whether schools are discriminating against minority children in identification and placement, or whether minority children are being “overidentified” for special education versus the correct rate. Perhaps for this reason, the Trump administration attempted to delay the regulation from going into effect in 2018 — although its delay of the regulation was overturned by a judge appointed by Obama in March 2019, in a ruling that will have the effect of incentivizing racial quotas in special education placement and discipline.
Proof that disproportionality was defined by the regulation in a meaningless way is found in a recent study by the Wisconsin Institute for Law & Liberty. It shows how racial discipline ratios as defined by the regulation vary across the country. Higher ratios are deemed “significantly disproportionate” under it. It turns out that these ratios are not just worthless as a sign of potential racism, but actually negatively correlated with it. Across the country, discipline ratios tend to be highest in prestigious, prosperous, or “progressive” school districts (where blacks are often suspended at four times the rate of other races), and lower in districts (such as those in the Deep South) that some people tend to assume are populated heavily by racist rednecks (where discipline ratios are often two or less). So it makes no sense to single out these school districts for a “root cause” analysis under the IDEA of whether racism or discrimination is occurring, as the significant disproportionality regulation does.
Research confirms that such crude measures of disproportionality aren’t a warning sign of racism. As a recent article in the American Journal of Sociology notes, racial gaps are largest not in backwoods districts rife with racism, but in wealthy, sophisticated school districts. This has long been known as the Shaker Heights Effect, after the prosperous, liberal Ohio suburb where the white-black gap in academic achievement is a huge four grade levels. (See Sean F. Reardon, Demetra Kalogrides & Kenneth Shores, The Geography of Racial/Ethnic Test Score Gaps, 124 American Journal of Sociology 1164 (2019)).
That shows that the Obama-era regulation’s approach to disproportionality — comparing the percentage of students disciplined who are of a particular group, to that group’s percentage of the student body, rather than to the percentage of those who violated school rules who are in that group — fails to convey useful information about whether discipline is being imposed in an improper or racist manner. Thus, the regulation is using the wrong yardstick for disproportionality.
As commenters such as attorney Roger Clegg and Paul Morgan, Ph.D argued in 2018 in response to the Education Department’s proposed delay of the significant disproportionality regulation, the significant disproportionality regulation should be replaced with a regulation that focuses on overidentification relative to the correct rate of placement and discipline (such as differential treatment of minority students or punishing black students at a higher rate than their actual rate of misbehavior would warrant) rather than its current, quota-like focus on mere overrepresentation between different demographics in the general population of disabled children. (See, e.g., Clegg April 13, 2018 comment, arguing that under the regulation, “proportionality was based on a comparison with the demographics of the student population, rather than the appropriate rate,” rendering it an unconstitutional racial classification under People Who Care v. Rockford Board of Education, 111 F.3d 528, 537-38 (7th Cir. 1996)); see also Clegg May 14, 2018 comment, arguing that “The text of 20 U.S.C. 1418(d)(2)(B),” which speaks of focusing any remedy on the “over-identified” race, “itself calls into question the regulation’s disregarding “overidentification” relative to the correct rate, in favor of crude “overrrepresentation” compared to the overall population.”).
Currently, the regulation explicitly rejects a focus on “over-identification” in favor of focus on “overrepresentation.” (See Assistance to States for the Education of Children With Disabilities, 81 Fed. Reg. 92376, 92380 n.1, 92381 (Dec. 19, 2016)).
But that’s wrong. First, the IDEA statute speaks of what race is “over-identified.” It says that the remedy for disproportionality should focus on the “groups that were significantly overidentified” and thus subjected to the disproportionality. (See 20 USC 1418(d)(2)(B)).
Second, even putting that aside, the IDEA is a civil-rights law, and in civil-rights law, the correct measure of whether a racial disparity exists is whether a minority is being disciplined or placed at a disproportionately inaccurate rate, not just at a rate higher than its percentage of the general population would suggest. (See, e.g., Mozee v. American Commercial Marine Ins. Co.,940 F.2d 1036, 1047-49 (7th Cir. 1992) (comparison in discrimination lawsuit alleging racially disparate impact in discipline should take into account “prior records of discipline,” not just racial breakdown of employer’s workforce); Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 651 (1989) (ratio relevant in disparate impact analysis is that which takes into account the “qualified population,” not the general population)).
“Significant disproportionality” should be defined in a more meaningful way, to reflect that correct measure. Doing so would not, like the current regulation, incentivize racial quotas. Instead, it would incentivize the elimination of racial discrimination and misidentification.
The Trump administration was right to try to delay the regulation. A delay of the regulation would give the Education Department time to come up with such an alternative approach that would better serve the purposes of the IDEA statute and its goal of appropriate, non-discriminatory placement of students. The regulation defines disproportionality in terms of “overrrepresentation” compared to the general child population, rather than “overidentification.” 81 Fed. Reg. at 92380 n.1 (Dec. 19, 2016).
A more sensible alternative methodology would be to define disproportionality in terms of whether placement or discipline decisions were more inaccurate for one group than another, leading to that group’s members being disproportionately “overidentified” compared to the correct rate, unlike members of other groups. The very statutory language that the regulation purports to enforce says just that. (See 20 USC 1418(d)(2)(B) (response to disproportionality in school district should focus on the group that was “overidentified” and as a result subjected to significant disproportionality)).
The meaning of a high racial ratio as computed under the significant disproportionality regulation seems to be the exact opposite of the what many commenters who supported it thought. In the real world, it appears that racism is inversely related to the school-district racial ratios that trigger a finding of significant disproportionality. That’s made clear in the data found in recent study from the Wisconsin Institute for Law & Liberty, in Appendix 1 of its report. It computes racial discipline ratios for black students versus other students, using the regulation’s overrepresentation-based approach, and data taken from the Education Department’s Civil Rights Data Collection.
The wealthiest county in Colorado has one of the highest racial discipline ratios in the whole state, way over the state’s threshold of 2.331. Colorado’s Douglas County has a long-term suspension ratio of 6.996, way over the state significant-disproportionality threshold of 2.331. Similarly, well-to-do, Democratic-run Jefferson County, not known for its racism, has an even higher ratio.
Minnesota’s Edina schools are famed for their quality, the prosperity of its parents, and their schools’ progressive slant on racial issues and institutional commitment to racial “equity” (the subject of coverage in the Minneapolis Star-Tribune). Yet Edina has one of the higher racial ratios in Minnesota (4.569).
Wisconsin school districts not known for having a racist school climate have very high ratios compaed to other school districts in the state — for example, Eau Claire has a ratio of 8.912.
Issaquah, a prosperous, center-left Seattle suburb, has a ratio of 6.915, the highest ratio in Washington State for a major school system. School districts in more reactionary and poorer school districts in eastern Washington have lower ratios, and thus, would ironically be considered less “disproportionate” under the significant disproportionality regulation.
By contrast, in most states in the Deep South, no major school district has a ratio of above 3. (See states colored green on the map on page 10 of the WILL study, in the map titled, “States with Districts Identified as Significantly Disproportionate at Varying RRTs, African American Students”).
The key problem with the significant disproportionality regulation is that it chose the easiest to measure definition of disproportionality, rather than the most meaningful (but harder to measure) definition of disproportionality, which would take into account students’ actual behavior (for discipline) or need for special education (for placement).
Court rulings don’t necessarily follow the simplest to measure definition of disproportionality, when doing so would make no sense or incentivize racial quotas. For example, the Supreme Court’s Wards Cove decision said that ratios for assessing the existence of disparate-impact discrimination should be based on on the “qualified population,” not the overall population, even when that means taking into account special qualifications that add to the complexity of calculating a disparity ratio. It said that even when skilled workers were mostly white, and unskilled workers were mostly minorities, the disparity didn’t show disparate impact discrimination, because the disproportionality is only relevant when it takes into account the “qualified population.” (See Wards Cove Packing Co. v. Atonio, 490 U.S. 642, 651 (1989)).
So the 2016 final regulation was wrong to prevent states from taking such qualifications into account, by precluding them from being considered in its “standard methodology” and the regulatory straitjacket it imposes on the nation’s school districts. (See 81 Fed. Reg. at 92380 n.1 & 92381 (Dec. 19, 2016)).
Thus, it was unfortunate that a trial court ruling, now on appeal, overturned the Trump administration’s delay of the significant disproportionality regulation. That ruling was wrong to prevent states from coming up with their own, more meaningful measures of disproportionality. That ruling had the effect of forcing all school districts in America to follow the 2016 regulation’s “standard methodology,” which does not take into account qualifications for discipline, such as misbehavior, or qualifications for special education, like actually being disabled and thus needing special education. The standard methodology is in tension with the Supreme Court’s limitation of comparisons to only individuals “qualified” for the classification in question (see Wards Cove, 490 U.S. at 651).
As the Paul Morgan study cited in the Education Department’s July 2018 delay notice pointed out, the prevalence of disability differs for members of different races, since more black students grow up in disadvantaged backgrounds than do white students, meaning that they need special education at a higher rate. Yet the 2016 regulation defined disproportionality solely in terms of a minority group’s “overrepresentation” in special education or discipline, rather than whether minority group members were wrongly “over-identified” as needing special education or deserving discipline, 81 Fed. Reg. 92376, 92380 n.1, even though the Obama administration admitted in 2016, in the regulation itself, that overrepresentation can be the result of minorities in fact needing special education at a higher rate, rather than mis-identification or discrimination. As it conceded, ” overrepresentation of one racial or ethnic group that rises to the level of significant disproportionality may occur for a variety of reasons, including over-identification of that racial or ethnic group…or appropriate identification with higher prevalence of a disability in a particular racial or ethnic group.” (see 81 Fed. Reg. at 92381 (italics added)).
In light of this reality, the judge should not have prevented states from coming up their own, more appropriate definition of disproportionality, such as a definition that actually does guard against overidentification, by taking into account students’ actual status as disabled (when identifying and placing them as disabled), or their actual behavior (for discipline). But that was the effect of the trial judge’s ruling blocking a delay of the significant disproportionality regulation.
That regulation doesn’t just suggest that school districts are racist when they aren’t (which is bad enough). Its crude methodology may also hide racism where it exists, such as against Asians, defeating federal law’s goal of remedying overidentification of minority groups.
There don’t seem to be any school districts where Asians are disproportionately suspended, as defined by the 2016 regulation. In fact, Asians apparently have the lowest suspension rate for any race, by far, across the country. As commenters such as Roger Clegg pointed out to the Education Department in response to the Trump administration’s proposed delay of the regulation, Asians are suspended at about one-fourth the rate of whites in California, and whites in turn are suspended at about one-fourth the black rate. (See Tom Loveless, The 2017 Brown Center Report on American Education: How Well Are American Students Learning?, Brookings Institution, March 2017, pg. 25).
The 2016 regulation provides no redress for Asians subjected to double standards in discipline, even when they are subjected to discrimination, because their discipline ratio compared to other races is always under one, even when they are being subjected to discrimination.They are just too well-behaved (with much lower arrest and conviction rates than other races). A ratio of less than one cannot trigger any disclosure of discrimination against a minority group, even if it is in fact being discriminated against. The Philadelphia schools discriminated against Asian students, and were “deliberately indifferent” to racial harassment of Asian students by black students. But because Asians misbehave less than students of other races, both in Philadelphia and nationally, the discrimination in Philadelphia would never have triggered a finding of disproportionality as to Asian students, because even treating Asians worse than other races in discipline will not increase their discipline rate up to the rate for other races. That is true even though “Asian Americans have been the subject of sometimes overt discrimination” by schools, as the Wisconsin Institute for Law & Liberty notes.
A more meaningful definition of disproportionality that took into account students’ actual conduct could reveal discrimination against Asians. But the 2016 significant disproportionality regulation rejects that approach. See 81 Fed. Reg. at 92380 n.1, 92381 (focusing on simple “overrepresentation” rather than “over-identification”).
But such better approaches are barred by the trial judge’s ruling against the delay of the significant disproportionality regulation.
There is yet another reason for delaying (and ultimately repealing) part of the significant disproportionality regulation. Specifically, its mistakenly including “disciplinary removals” as a separate, stand-alone, covered category in the portions of the regulation redirecting school districts’ funds and ordering review of their policies based on disproportionality.
The significant disproportionality regulation requires the diversion of “funds” to “coordinated early intervening services” not just for “identification” and “placement,” as the IDEA statute requires, but also for disciplinary actions, which the IDEA statute pointedly does not require (indeed, the IDEA statute implicitly forbids it, based on rules of statutory construction used by judges).
The IDEA statute includes all three categories — identification, placement, and discipline — in the data collection provision, 20 U.S.C. 1418(d)(1). But it only includes the first two categories (identification and placement), and omits discipline, in the following provision, about when school districts must reserve “funds” for “coordinated early intervening services” and engage in “review and revision” of school policies. See 20 U.S.C. 1418(d)(2).
As commenters like lawyer Roger Clegg pointed out, the omission of discipline from the latter provision means that it is improper for the significant disproportionality regulation to shift funding or order changes in policy based on disciplinary disproportionality, or to label local school districts as disproportionate solely based on discipline disparities.
As Roger Clegg pointed out, under a 2018 Supreme Court ruling, judges apply a rule of statutory construction that when “Congress includes particular language in one section of a statute but omits it in another, this Court presumes that Congress intended a difference in meaning.” (citing Digital Realty Trust v. Somers, 138 S.Ct. 767, 777 (2018)). Congress pointedly omitted the language about disciplinary actions from 20 U.S.C. 1418(d)(2), which deals with reservation of school funds and revision of school policies.
Applying that rule of construction to the IDEA statute means that discipline disparities should not be covered at all by the portions of the significant disproportionality regulation that redirect school districts’ funding (by ordering them to reserve the maximum amount of funds to provide coordinated early intervening services). Only identification and placement disparities should be covered by the reservation of funds requirement, not discipline disparities.
But oddly, discipline disparities currently are covered by the regulation’s redirect of funds and revision-of-school-policy provisions. (See 34 C.F.R. 300.646(c)&(c)(1) (review and revision of school policies is required even based solely on disproportionality in “disciplinary removals”); 34 C.F.R. 300.646(d) (applying redirect of funds not just to the identification and placement disparities covered by paragraph (a)(1) and (a)(2) of 34 CF.R. 300.646, but also the discipline disparities covered by paragraph (a)(3), by requiring that any school district identified as having any of the three types of disproportionality in “paragraphs (a) and (b) of this section to reserve the maximum amount of funds…to provide comprehensive coordinated early intervening services”).
That diversion of funds is an incentive for schools to adopt veiled racial quotas. Financial incentives to have veiled racial quotas or “goals” are unconstitutional (see Lutheran Church–Missouri Synod v. FCC, 141 F.3d 344 (D.C. Cir. 1998), even when they are accompanied by a disclaimer like that found in the 2016 regulation that no quotas are required.(See Lutheran Church, 154 F.3d 487 (D.C. Cir. 1998) (rejecting rehearing en banc, and noting that the proportionality hiring goal it invalidated would not have been constitutional even had the agency added boilerplate language that no quota is required). Proportionality requirements are considered quotas in both placement and discipline (see People Who Care v. Rockford Board of Education, 111 F.3d 528, 537-38 (7th Cir. 1997) (striking down requirements that school district not discipline minority students at a higher rate than whites, and that it not place them in remedial education at a higher rate than whites).
To avoid such constitutional problems, the regulation should not define disproportionality (as it currently does) in terms of simple overrepresentation. Even when a definition would otherwise be permissible, it must be rejected in favor of an alternative definition when it would create potential constitutional problems, like encouraging the use of race, or racial quotas. The Supreme Court has made this clear in a a number of cases. (See Miller v. Johnson, 515 U. S. 900, 923 (1995) (refusing to construe race-conscious provisions of Voting Rights Act broadly, as Justice Department advocated, because that raised potential constitutional problems); Northwest Austin Municipal Utility District No. 1 v. Holder, 129 S. Ct. 2504, 2513 (2009) (construing Voting Rights Act narrowly to avoid possible constitutional federalism problems, despite broader reading of the Act advocated by the government and civil-rights groups)).
While it might be permissible to use a crude but simple definition of disproportionality for purposes of statewide data collection (20 U.S.C. 1418(a)&(b)), doing that is unacceptable in identifying local school district as significantly disproportionate (2o U.S.C. 1418(c)&(d)), because such a finding triggers adverse financial consequences for the school district so identified, that incentivize it to use race in a covert manner. The constitutional avoidance principle requires that a law be given a narrower meaning rather than a broader one to avoid a serious constitutional issue, even when rules of grammar, broad construction, or deference to agencies would otherwise lead to a broader reading. (See, e.g., NLRB v. Catholic Bishop of Chicago, 440 U.S. 490 (1979) (adopting very narrow construction, amounting to judge-made exception, to avoid constitutional problem); United States v. X-Citement Video, 513 U.S. 64, 68 (1994) (adopting narrower reading of statute rather than “most natural grammatical reading” to avoid having the statute potentially violate the Constitution); Edward J. DeBartolo Corp. v. Florida Gulf Coast Building & Constr. Trades Council, 485 U. S. 568, 574-575 (1988) (need to avoid potential constitutional problem overcame deference to agency’s broad reading)).
A judge overturned the Trump administration’s delay of the regulation in a court ruling on March 7, citing a lack of clear proof that it would lead to quotas. The judge argued that the Education Department’s finding about the potential risk of quotas was too “drenched in qualification” to warrant a change of course from the Obama administration’s policy in its 2016 “significant disproportionality” regulation.
But the regulation will in fact lead to racial quotas, as the African-American Wall Street Journal columnist Jason Riley predicted, and legal commentaries have explained in the past. It will lead to black children being denied needed special education services, and white and Asian students being subjected to harsher discipline for the same offenses than black students.
It puts financial pressure on schools to adopt veiled quotas: If a school system has “significant disproportionality” based on race in special education placement or discipline, a lot of its special education money gets diverted to other services. “Disproportionality” is defined in a quota-like way. As the Obama administration said in adopting the regulation, disproportionality is calculated based on “overrepresentation,” which can be due to innocent factors such as the “higher prevalence of a disability in a particular racial or ethnic group,” rather than “over-identification.”
There was evidence in the Education Department’s administrative record that veiled quotas can occur from yardsticks similar to that used in the “significant disproportionality” regulations. The Education Department cited a single example: Texas school districts adopted informal quotas on identification of children as disabled, in response to a state numerical indicator that scrutinized districts that identified more than 8.5% of kids as disabled as potentially disproportionate. That experience seems relevant and appropriate to draw lessons from. The judge wrongly saw that example as irrelevant, because it didn’t involve a racial quota, but rather a non-racial quota in response to a numerical indicator. But if such mild numerical indicators can lead to quotas in non-racial contexts, they can presumably lead to quotas in the racial context as well).
Moreover, there was specific evidence of racial quotas, too, in the administrative record, even if the Education Department did not specifically cite it. The Center for Equal Opportunity’s Roger Clegg pointed out that similar state disproportionality rules had led to informal, veiled racial quotas in the past, in comments he submitted to the Education Department in May 2018, supporting its proposed delay. As he said, “In the past, similar rules at the state level have led to quotas. State education agencies and associations of special education officials made this clear back in 2016 in their comments.” For example, he cited the May 2016 comment of the National Association of State Directors of Special Education, submitted by Myrna Mandlawitz, which recounted that under a similar state rule, “state staff heard of districts telling parents that there were now quotas for identification, and no more minorities could be identified.”
Far milder scrutiny of racial numbers than the significant disproportionality regulation can also lead to quotas in discipline. The significant disproportionality regulation can lead to a local education agency having 15% of its special education dollars redirected to coordinated early intervening services (CEIS) — potentially millions of dollars. That’s a big incentive to do whatever it takes to avoid being declared significantly disproportionate.
But even much smaller incentives to use race can trigger the use of quotas, such as the informal use of numbers as a yardstick for whether to investigate a school district for discrimination in discipline. As Brian Miller pointed out in a comment, the Minneapolis schools adopted racial quotas in suspensions at the close of an investigation by the Education Department in 2014. As Miller pointed out, such a quota wasn’t mandated by Title VI, the law under which the Education Department’s Office for Civil Rights investigated the Minneapolis schools. Indeed, Title VI is interpreted by courts as forbidding racial quotas. But “even though there wasn’t any legal requirement for a quota,” “Just the scrutiny of its numbers was enough to lead to a quota.”
Miller cited an article in Reason Magazine reporting that the Minneapolis Public Schools had decided to “aggressively reduce the disproportionality between black and brown students and their white peers every year for the next four years. This will begin with a 25% reduction in disproportionality by the end of this school year; 50% by 2016; 75% by 2017; and 100% by 2018.” (See Robby Soave, “Schools Implement Explicit Racial Bias in Suspensions,” Nov. 11, 2014). As Miller said, such a proportionality requirement is an unconstitutional racial quota, under court rulings like one that struck down a rule that forbade a “school district to refer a higher percentage of minority students than of white students for discipline.” (See People Who Care v. Rockford Board of Education, 111 F.3d 528, 538 (7th Cir. 1997)).
Why did Minneapolis adopt such a quota? The use of a numerical indicator — even though it was milder than those adopted by states under the Obama administration’s 2016 significant disproportionality regulation.
Although Miller didn’t point this out, at the time Minneapolis adopted its quota, the Education Department used a significant-disproportionality ratio as an informal trigger for whether to investigate schools for racism in discipline. If twice as many minorities were suspended as whites, the school district got investigated to see if it was discriminating against minorities in discipline. The Manhattan Institute’s Max Eden described in 2018 how that trigger worked: “If the ‘number of suspended minority students’ / ‘number of minority students’ is more than twice the value of ‘number of suspended white students’ / ‘number of white students,’ then a districtwide investigation proceeds.” (See “OCR’s Approach to the Evaluation, Investigation, and Resolution of Title VI Discipline Complaints” February 2014, quoted in Max Eden, “Enforcing Classroom Disorder: Trump Has Not Called Off Obama’s War on School Discipline,” Manhattan Institute, Aug. 13, 2018).
That numerical trigger was designed to help administer the Obama administration’s supposedly “nonbinding” January 8, 2014 Title VI guidance on racial discrimination in school discipline — guidance that the Trump administration later withdrew and replaced in December 2018 because it allegedly placed undue weight on racial numbers.
The Minneapolis quota happened even though the pressure to use race from Title VI investigations is much milder than from the IDEA’s “significant disproportionality” regulation. It appears that no school district ever lost a penny of its federal funding due to a Title VI racial discrimination investigation of student discipline, even under the Obama administration. In Title VI investigations, “statistics alone” were not considered proof of discrimination, even though they could trigger an inquiry into whether discrimination was occurring.
By contrast, under the IDEA, a finding of significant disproportionality in discipline can lead to 15% of a school district’s special education money being diverted to CEIS. That is a much more serious consequence, that school officials emphatically do not want and will do a lot to avoid.
So if using a two-to-one racial ratio led to a quota under “nonbinding” Title VI guidance, it can definitely lead to quotas under the binding provisions of the IDEA, which has more teeth. States such as Maryland and Tennessee define significant disproportionality as a ratio of roughly 2-to-1 in their “risk ratio threshold” for various forms of disciplinary action. Similarly, Colorado sets a “risk ratio threshold” of 2.31 for “out-of-school suspension” for “10 days” or less. If suspensions exceed that for “children of a particular race,” then that triggers a finding of significant disproportionality, a finding that can lead to large amounts of money being redirected to CEIS and away from what a local education agency would prefer to spend it on. (See Powerpoint presentation, “Colorado’s definition of significant disproportionality SY2018-19 & SY 2019-20”). That is a big incentive to have a veiled racial quota.
Indeed, for a state to use a ratio of two-to-one or even three-to-one (under the significant disproportionality regulation) as a risk ratio threshold should be deemed an unconstitutional incentive to use race, as a matter of law. When racial proportionality rules incentivize the use of race, they can violate the Constitution, even if there is no evidence of a racial quota being adopted yet, or producing any racial discrimination yet. The mere existence of the incentive is enough by itself to violate the Constitution.
For example, the D.C. federal appeals court ruled that an agency’s diversity regulation that led to audits of broadcasters whose workforce contained a smaller percentage of minorities or women than the regional labor force, was an unconstitutional incentive to use race. That was true even though the regulation did not base violations solely on statistics, and even though the court did not cite any actual real world example of a white male denied a job due to the agency’s regulation, which, indeed, contained language banning racial discrimination. (See Lutheran Church–Missouri Synod v. FCC, 141 F.3d 344 (D.C. Cir. 1998)).
Defining significant disproportionality as discipline ratios of 4-to-1 or less incentivizes the use of race, because misbehavior rates for students of different races vary dramatically, commonly 4-to-1 or more. As an anonymous Asian parent said in a comment to the Education Department, in the State of California, Asians are suspended at one-fourth of the white rate, while whites in turn are suspended at one-fourth the black rate. (See Tom Loveless, The 2017 Brown Center Report on American Education: How Well Are American Students Learning?, Brookings Institution, March 2017, at pg. 25).
White racism cannot possibly explain the lower Asian suspension rate, since a racist white teacher is not going to favor Asians over whites in discipline. So obviously, racial differences in student misbehavior rates are real and profound, as a 2014 study in the Journal of Criminal Justice found. Similarly, education expert Michael Petrilli notes that black students themselves admit to getting into fights and engaging in truancy at much higher rates, in federal surveys. Courts have rejected the idea that “misbehavior” is “a randomly distributed characteristic among racial groups,” and thus ruled that “disparity” in discipline rates “does not constitute discrimination.” (See Coalition to Save Our Children v. State Board of Education of the State of Delaware, 90 F.3d 752, 775 (3d Cir. 1996); Belk v. Charlotte-Mecklenburg Board of Education, 269 F.3d 305, 332 (4th Cir. 2001)).
Given that misbehavior rates can differ by a ratio of 4-to-1, any risk ratio threshold of less than 4-to-1 — and that includes many if not most risk ratio thresholds used nationally — will incentivize racial quotas. This is powerful evidence that the 2016 significant disproportionality regulation will in fact lead to racial quotas.
The Education Department should point this out to support its delay of the regulation. It should also cite the comments of Clegg and Miller, which illustrate that the quotas do in fact happen in the real world, as a result of regulations similar to, or even milder than, the significant disproportionality regulation. That shows the specter of quotas is very real, contrary to what the trial judge thought.
Once sued, the Education Department may have erroneously soft-peddled the risk of quotas, out of a belief that that was necessary to justify its decision to allow states to voluntarily comply with the Obama-era rule during the period of the delay. The Trump administration delayed mandatory compliance with the 2016 regulation, which otherwise would have gone into effect in 2018, for two years, but it let states voluntarily comply with the Obama-era rule if they wanted to.
But allowing voluntary compliance doesn’t undercut the basis for delaying mandatory compliance: It is one thing to force regulated entities to violate the Constitution, and another to merely permit them to do so. A federal agency has an obligation not to violate the Constitution, and not to force other people to violate the Constitution; but it doesn’t have any constitutional duty to keep others from violating the Constitution or racially discriminating.
In any event, even if the delay solved only part of the problem of quotas, by eliminating federally-coerced quotas, while permitting quotas voluntarily adopted by states, it still reduced the pressure for quotas. So it was justifiable as an improvement over the prior status quo. As courts have explained, “Agencies need not address all problems in one fell swoop,” and “Reform may take place one step at a time.” In short, “an agency does not have to ‘make progress on every front before it can make progress on any front.’”