Appeal from the United States District Court for the Western District of Wisconsin. Nos. 00 C 619, 00 C 662--Barbara B. Crabb, Chief Judge.
Before Posner, Manion, and Kanne, Circuit Judges.
The opinion of the court was delivered by: Posner, Circuit Judge.
The Individuals with Disabilities Education Act entitles a disabled child to a "free appropriate public education" tailored to his disability, 20 U.S.C. §§ 1400(d)(1)(A), 1412(a)(1)(A); Cedar Rapids Community School District v. Garret F. ex rel. Charlene F., 526 U.S. 66, 68 (1999); Morton Community Unit School District No. 709 v. J.M., 152 F.3d 583, 583-84 (7th Cir. 1998), and expresses a strong preference for "mainstreaming" (the statutory term is "least restrictive environment," § 1412(a)(5); Beth B. v. Van Clay, 282 F.3d 493, 497 (7th Cir. 2002)), that is, for educating the disabled child in classes with nondisabled children rather than in special classes or at home. § 1412(a)(5)(A); Board of Education v. Rowley, 458 U.S. 176, 202-03 and n. 24 (1982); Rome School Committee v. Mrs. B., 247 F.3d 29, 33 (1st Cir. 2001). Efficacy is relevant, obviously, but so is cost, Hartmann by Hartmann v. Loudoun County Board of Education, 118 F.3d 996, 1004-05 (4th Cir. 1997); Seattle School District, No. 1 v. B.S., 82 F.3d 1493, 1500 (9th Cir. 1996), including the disruptive impact on other children when, as in this case, a disability causes antisocial behavior. Oberti by Oberti v. Board of Education, 995 F.2d 1204, 1217 (3d Cir. 1993); Greer By & Through Greer v. Rome City School District, 950 F.2d 688, 697 (1991), withdrawn, 956 F.2d 1025, reinstated, 967 F.2d 470 (11th Cir. 1992) (per curiam); Daniel R.R. v. State Board of Education, 874 F.2d 1036, 1048-50 (5th Cir. 1989). For a good summary see Murray By & Through Murray v. Montrose County School District RE-1J, 51 F.3d 921, 926-27 n. 10 (10th Cir. 1995).
Z.S. has symptoms characteristic of autism. As a kindergartener he displayed abnormally aggressive behavior and it worsened as the years passed. He would kick, hit, and bite students, teachers, and teachers' aides, and when he was not having outbursts he would be depressed and withdrawn. His basic problem was inability to function in a social setting or indeed in any setting that was not highly structured; any sudden movement, any disruption of routine, set him off. He also displayed abnormal sensitivity to touch and sound, another symptom of autism. American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, Text Revision 70-72 (4th ed. 2000); T. Peeters & C. Gillberg, Autism: Medical and Educational Aspects 27, 29 (2d ed. 1999). He received various medications and special educational assistance, all it seems to little or no avail. At any rate he got progressively worse; and in 1999, when he was 10, he was placed in a residential mental health facility (Mendota) which included a school. He did pretty well there, and the staff suggested that he be returned to a regular public school for the 1999-2000 school year. They thought--erroneously, as it turned out--that his experience at Mendota was the kind of transition he had needed to a regular public school. An Individualized Education Program (required by IDEA; see 20 U.S.C. § 1414(d)(1)(A)) was prepared for him. It called for him to spend 70 percent of his school time when he returned to public school in regular classes and the other 30 percent in special-education classes; a special-education program assistant would be present to help him in all his classes although not assigned specifically to him. The program failed; when he returned to public school he was disruptive, violent (the police had to be called on one occasion), and truant. After a few weeks of this, he was placed in another specialized school, called SCAN, which was not, however, residential like Mendota.
He was totally unmanageable in SCAN, and was removed after less than a month. He had done better at Mendota, but his guardian (his grandmother: his father has disappeared and his mother, a drug addict, is unable to care for the child) did not want him sent back there. She wanted him sent back to the regular school but with an aide assigned to him full time to keep him under control. The school district, afraid that sending Z.S. back to school might irreparably damage the prospects of his ever being able to get along with other children, chose instead, after canvassing other alternatives, a program of homebound instruction for him. (It took the district a month to decide on this, during which time Z.S. was at home with no instruction.) It hired a retired special-education teacher to teach him for six hours a week at home, and an occupational therapist to give him another hour's instruction each week, in the hope that after a while it would be possible for him to return to school. This continued until the end of the 1999-2000 school year. His subsequent experiences are not a part of the record.
The IDEA is enforced in the first instance in state administrative proceedings. An administrative law judge found that Z.S. had, in the 1999-2000 school year, been denied the free appropriate public education to which the Act entitled the child. He found that the school district had failed to diagnose Z.S. as autistic, should not have needed a month to create a new educational program for him after he was removed from SCAN, and shouldn't have placed him in a "restrictive" environment (namely his home) without giving more consideration to the possibility of "main-streaming" him, perhaps returning him to a regular public school but assigning a full-time ("one on one") aide to attend him throughout the school day. The school district sought judicial review in federal district court (Wisconsin has a one-tier rather than the more common two-tier system for administrative review of IDEA claims, see 20 U.S.C. §§ 1415(g), (i)(2); Wis. Stat. § 115.80; compare 105 ILCS §§ 5/14-8.02(h)-(i)), which reversed the administrative law judge, 184 F. Supp. 2d 860 (W.D. Wis. 2001), precipitating this appeal by Z.S.'s guardian on his behalf. The guardian's counterclaim in the district court was for attorneys' fees, to which of course she is not entitled if the judgment in favor of the school district stands.
She argues, to begin with, that the district judge failed to give the administrative law judge's decision the proper deference. Just how much deference that is is unclear. In ordinary cases of judicial review of administrative action, if no pure issues of law (on which judicial review is plenary) are involved except ones the resolution of which the Chevron doctrine commits to the agency, the court must defer to the agency's decision if the decision is supported by "substantial evidence." Some of our cases say--realistically--that this is the same standard as clear error. Thomas v. Chicago Park District, 227 F.3d 921, 926 (7th Cir. 2000), aff'd on other grounds, 534 U.S. 316 (2002); United States v. Hill, 196 F.3d 806, 808 (7th Cir. 1999); Johnson v. Trigg, 28 F.3d 639, 643-44 (7th Cir. 1994); cf. Ingram v. ACandS, Inc., 977 F.2d 1332, 1340 (9th Cir. 1992). Aegerter v. City of Delafield, 174 F.3d 886, 890 (7th Cir. 1999), says that "it is possible, though not always easy," to distinguish among the canonical standards of review, such as substantial evidence and clear error, noting that "this court has expressed skepticism in the past about the ability of judges to apply more than a few standards of review," id. at 889, citing United States v. Boyd, 55 F.3d 239, 242 (7th Cir. 1995).
"Realistically" is the key qualification. The Supreme Court has said that "the court/agency standard [substantial evidence] . . . is somewhat less strict than the court/court standard [clear error]. But the difference is a subtle one--so fine that (apart from the present case) we have failed to uncover a single instance in which a reviewing court conceded that use of one standard rather than the other would in fact have produced a different outcome." Dickinson v. Zurko, 527 U.S. 150, 162-63 (1999), quoted in CAE, Inc. v. Clean Air Engineering, Inc., 267 F.3d 660, 676 and n. 10 (7th Cir. 2001). The courts that think the difference perceptible cannot, despite the passage we have just quoted from the Supreme Court's decision in Zurko, agree on which is the more searching standard; an illustrative comparison is between In re Zurko, 258 F.3d 1379, 1384 (Fed. Cir. 2001) (substantial evidence less searching than clear error), with Aruta v. INS, 80 F.3d 1389, 1393 (9th Cir. 1996) (substantial evidence more searching than clear error), and General Electro Music Corp. v. Samick Music Corp., 19 F.3d 1405, 1408 (Fed. Cir. 1994) (ditto). If the difference in standards were material, courts would know which way the difference cut and would be able to identify cases in which the choice of standard had determined the outcome.
Defenders of the difference will point out that administrative proceedings often involve technical issues on which the agency is expert and the reviewing court is not. But court cases often involve issues on which the trial judge has an equally great advantage over the reviewing court--the factual issues may be esoteric ones that the judge was able to immerse himself in and the reviewing court, which has limited exposure to a case, was not. So it is not obvious that a blanket distinction between the standard for review of agency findings and the standard for review of judicial findings is warranted, and in any event the cognitive limitations that judges share with other mortals may constitute an insuperable obstacle to making distinctions any finer than that of plenary versus deferential review. See Reynolds v. City of Chicago, No. 00-3771, slip op. at 4, 2002 WL 1354689, at *2 (7th Cir. June 21, 2002). Maybe in judicial review two's a company and three's a crowd.
But here we must note a difference between the questionable judicial attempt to multiply standards of review and the unavoidable heterogeneity in the application of a given standard across the full range of cases governed by it. Even if the different formulations of deferential review (substantial evidence, clear error, abuse of discretion, etc.) amount to the same thing--a proposition that we are entertaining as a hypothesis, not asserting as a rule for the circuit--the actual amount of deference given the finding of a lower court or an agency will often depend on the nature of the issue. We mentioned technical issues. The more technical the issue resolved by the agency, the less likely the reviewing court is to feel comfortable second-guessing the agency's resolution. As a practical matter, having nothing to do with the precise articulation of the standard of review, the agency's finding will receive greater judicial respect in such a case.
The distinction between standards of review and the application of a given standard in different cases affects judicial review of administrative decisions under IDEA. The reason is that the statute authorizes the district court to receive and consider new evidence, that is, evidence that was not before the administrative law judge, and to reverse his decision if it is contrary to the "preponderance of the evidence." 20 U.S.C. § 1415(i)(2)(B). Both the receipt of evidence by the reviewing court and the preponderance standard of proof are features alien to ordinary judicial review of administrative action, though related to each other. As we explained in Dale M. ex rel. Alice M. v. Board of Education, 237 F.3d 813, 815-16 (7th Cir. 2001), consistent with the discussion in the preceding paragraph of this opinion, a reviewing court that has before it evidence not considered at the administrative level will naturally defer less to the administrative decision, as it has an information advantage over the administrator that it lacks when judicial review is limited to the record that was before him. Judicial review is more searching the greater the amount (weighted by significance) of the evidence that the court has but the agency did not have. Id. at 816; Kerkam v. McKenzie, 862 F.2d 884, 887 (D.C. Cir. 1988); Ojai Unified School District v. Jackson, 4 F.3d 1467, 1471-72 (9th Cir. 1993).
We are near the bottom of this sliding scale in the present case. The school district, the loser at the administrative level, did put in evidence that had not been introduced at that level, and Z.S. did not. But that evidence seems not to have played a significant role in the district court's decision. The only bit of it that the court considered was school records "that relate to Z.S.'s Individualized Education Programs before the 1999-2000 school year," 184 F. Supp. 2d at 874, and it appears from the administrative law judge's opinion that he had before him most of the information contained in the records. Since it thus was essentially a no-new-evidence case, the district judge owed the administrative law judge's decision the usual deference that reviewing courts owe agencies when judicial review is limited to the administrative record. See Dale M. ex rel. Alice M. v. Board of Education, 237 F.3d at 815-16; Morton Community Unit School District No. 709 v. J.M. , supra, 152 F.3d at 587-88. When no fresh evidence is taken, "the fact that [the district judge] disagrees with the [administrative law judge or other administrative hearing] officer is not enough to justify setting aside the latter's order; he must be strongly convinced that the order is erroneous . . . . [H]e owes considerable deference to the reviewing officer." Dale M. ex rel. Alice M. v. Board of Education, supra, 237 F.3d at 815-16; see also Patricia P. v. Board of Education, 203 F.3d 462, 466-67 (7th Cir. 2000). This is just another way of stating the clear-error or substantial-evidence standard. For there is no magic in the particular words used.
The district judge's opinion contains language which suggests that she did the forbidden and, despite not having taken material new evidence, made an "independent" determination that the school district had not violated the IDEA. Nevertheless, if she did apply the wrong standard of review (more likely the statement that she had made an "independent" determination was just a slip of the pen, cf. Lenn v. Portland School Committee, 998 F.2d 1083, 1087-89 (1st Cir. 1993)), it was a harmless error. Winner Int'l Royalty Corp. v. Wang, 202 F.3d 1340, 1347 and n. 5 (Fed. Cir. 2000); Perdue v. Burger King Corp., 7 F.3d 1251, 1254 (5th Cir. 1993); Blizard v. Fielding, 572 F.2d 13, 15-16 (1st Cir. 1978). The administrative law judge's decision could not survive even the deferential review to which it was entitled. The critical issue before him was whether the school administrators were unreasonable (Board of Education v. Rowley, supra, 458 U.S. at 206-07) in thinking it would be a mistake to send Z.S. back to his regular public school to take regular school classes. The one-month delay in figuring out what to do with Z.S. after he had to be removed from SCAN, much harped on by ...