This case involves a recurring theme that this Court has repeatedly noted. Namely, this case involves another young, not particularly likeable claimant, who also happens to suffer from some form of mental illness. See, e.g. , Largent v. Colvin , 2016 WL 47918, *1 (N.D. Ill. Jan. 5, 2016). And, again, this Court ultimately determines that the case must be remanded despite evidence in the administrative record that supports the administrative law judge's concerns about the claimant.
Plaintiff Javante Pulley filed an application for supplemental security income when he was 20 years old, alleging that he was disabled based on his mental illnesses and cognitive impairments. Over the years and beginning when he was a young boy, he has been diagnosed variously with paranoid schizophrenia, schizoaffective disorder, bipolar disorder, borderline intellectual functioning based on multiple low IQ scores, a learning disability, and attention deficit hyperactivity disorder. After two administrative hearings, at which plaintiff and a medical expert (Dr. Heinemann) testified, the administrative law judge ("ALJ") concluded that plaintiff's schizoaffective disorder constituted a severe impairment (but that his ADHD and other alleged illnesses did not), and then found that plaintiff could perform a full range of work subject to certain limitations. This decision rested on two primary rationales.
First, the ALJ concluded that plaintiff was exaggerating or lying about his symptoms and limitations, as well as about other matters, such as whether he ever drove a car and how much alcohol and marijuana he consumed. This conclusion rested heavily on the observations of consulting psychologist, Dr. John Peggau, and also, to a lesser degree, on a statement by a school psychologist. The ALJ also found that plaintiff's testimony was "articulate" and "responsive," thus reinforcing the conclusion that he was more capable than he was portraying himself to be. In short, this is a malingering rationale.
Second, the ALJ also relied on a non-compliance rationale, a theory arguably at odds with the first one. The ALJ concluded that plaintiff's bad periods-the hospitalizations, the hearing of voices, several alleged suicide attempts, and an arrest-were largely the result of his voluntary decision not to take his medications. See R. 23 ("He was in crisis due to noncompliance.");
*901R. 22 ("The voices go away while he is on his medications."). But when he took his medications, his symptoms "were reportedly in good control, or at least stable with no hallucinations." R. 23.
In this appeal, plaintiff's main argument for remand is that the ALJ failed to give any consideration to the possibility that plaintiff's mental illnesses made it difficult to comply with treatment recommendations and also perhaps even contributed the impression by some that he was malingering. As set forth below, this argument warrants a remand.
The Seventh Circuit has repeatedly held that ALJs must consider the effect mental health illnesses may have on a claimant's ability to comply with treatment. In Kangail v. Barnhart , 454 F.3d 627 (7th Cir. 2006), a case involving a bipolar claimant who had difficulty adhering to treatment recommendations, the Seventh Circuit stated as follows:
The administrative law judge thought the plaintiff's inability to hold a job unimportant because she could work when she took her medicine. And it is true that bipolar disorder is treatable by drugs. But mental illness in general and bipolar disorder in particular (in part because it may require a complex drug regimen to deal with both the manic and the depressive phases of the disease) may prevent the sufferer from taking her prescribed medicines or otherwise submitting to treatment. The administrative law judge did not consider this possibility.
Id. at 630 (internal citations omitted). Subsequent Seventh Circuit cases have relied on a similar rationale. See Jelinek v. Astrue , 662 F.3d 805, 814 (7th Cir. 2011) ("ALJs assessing claimants with bipolar disordermust consider possible alternative explanations before concluding that non-compliance with medication supports an adverse credibility inference.") (emphasis added); Spiva v. Astrue , 628 F.3d 346, 351 (7th Cir. 2010) ("The administrative law judge's reference to Spiva's failing to take his medications ignores one of the most serious problems in the treatment of mental illness-the difficulty of keeping patients on their medications. The drugs used to treat schizophrenia, for example, can make a patient feel drowsy and stunned."); see also Shauger v. Astrue , 675 F.3d 690, 696 (7th Cir. 2012) ("Although a history of sporadic treatment or the failure to follow a treatment plan can undermine a claimant's credibility, an ALJ must first explore the claimant's reasons for lack of medical care before drawing a negative inference"). In all these cases, the Seventh Circuit held that a remand was required.2
The Government acknowledges this line of cases, but argues that the ALJ, in fact, "did explore the reasons for plaintiff's non-compliance," but "ultimately found them unconvincing." Dkt. # 16 at 4. The Court is not persuaded by this argument.
Focusing first on the ALJ's written decision, the Court can find no textual evidence that the ALJ considered this issue or, if she did, what her reasoning was. The ALJ relied heavily on the non-compliance argument, raising it repeatedly throughout the decision.3 The ALJ found that plaintiff's *902failure to comply with treatment recommendations, as well as his lack of urgency in answering IQ test questions, were voluntary choices and even suggested that they were the result of a deeper moral failing. See R. 14 (plaintiff's failure to try harder on IQ tests was "within the claimant's control"); id. (plaintiff's learning disorder was "mostly motivational"); R. 23 ("His presentation at hearings indicated what is clearly [ ] evident in the record, shallowness, immaturity, and a lack of motivation.").4 Yet, in all these discussions, the ALJ never confronted the possible counter-explanation that these problems were caused, at least in part, by plaintiff's mental illnesses. The ALJ also did not consider the possibility that plaintiff's oscillating good and bad periods could have been due to the natural progression of his underlying mental illnesses. As the Seventh Circuit has noted in Kangail , "bipolar disorder is episodic," meaning that a claimant often has "severe" periods following by periods where the claimant was "behaving pretty normally." 454 F.3d at 629 ; see also Jelinek , 662 F.3d at 814 ("bipolar disorder [ ] is by nature episodic and admits to regular fluctuations even under proper treatment ") (emphasis added). Two other issues not considered are that plaintiff's treatment was more complex than simply taking the same pill every day and that some of his improvement came only after a multi-day stay at an in-patient facility. See Harlin v. Astrue , 424 Fed.Appx. 564, 568 (7th Cir. 2011) ("the evidence that the ALJ chose [ ] to rely on-discharge summaries showing Harlin's improved condition at the time of discharge-was hardly remarkable because one would expect a patient with severe mental impairments to improve upon a course of treatment in a structured hospital environment").
The Government's main argument is to note that the ALJ, at the first hearing, asked plaintiff "multiple times" why "he had been off medication or stopped attending treatment." Dkt. # 16 at 4. Although the Government is correct in this factual assertion, it provides little consolation because, first of all, the ALJ never acknowledged or analyzed this testimony in the later written decision. Therefore, the Court cannot be sure how the ALJ reasoned through this issue. The Government attempts to reconstruct what it believes was the implicit reasoning in the ALJ's decision. Specifically, the Government argues *903that the ALJ concluded that plaintiff's assertion at the hearing that he could not remember why he failed to take his medications was suspicious and likely a part of his larger malingering enterprise. As an initial matter, this Court may not rely on this argument because, as noted already, the ALJ never endorsed this theory in the decision. See Parker v. Astrue , 597 F.3d 920, 922 (7th Cir. 2010) ("the [Securities and Exchange Commission v. ] Chenery [Corp. , 318 U.S. 80, 63 S.Ct. 454, 87 L.Ed. 626 (1943) ] doctrine [ ] forbids an agency's lawyers to defend the agency's decision on grounds that the agency itself had not embraced").5
But even if the Court were not limited by Chenery and could piece together an implicit rationale from statements made at the hearings, the Court would still have a number of questions. To explain why, it will be helpful to set out the relevant testimony. The Court will liberally quote from the first hearing because this testimony is germane not only to the non-compliance issue, but also to the separate but related issue of whether plaintiff's testimony was as articulate as the ALJ claimed. Both plaintiff and Dr. Heinemann were asked about the non-compliance issue.
The general issue was first raised with plaintiff in the following exchange:
Q (ALJ) And your records show that when you stay on your medication, and you go to counseling, that you do much better. Do you agree with that?
A Yeah, as long as I stay out of trouble, but not, not mostly all the time, not mostly all the time that I actually do be, be able to stay out of trouble.
Q What does that mean? What kind of trouble do you get in to?
A I, it really is not bad trouble, it's like sometimes it's just like I'll be walking home, like the last time I was walking home. And I had got stopped by the police office for some, for some reason he said that I was jaywalking but I really wasn't jaywalking. But he was really trying to get the kids from behind me because they was running like they was going to go fight. So I don't know, that would be like one of the times I really got in trouble and I really ain't did nothing.
R. 63-64.6 Later in the hearing, the ALJ re-visited the issue in the following colloquy:
Q And then this year, and there's very little that [is] in between 2012 and 2014, but again you had been off your medications when you just went back last month, you had been off for I think three months. Why did you do that?
A It was not why, it was much as they was telling us we had to find a person to go see outside of the Rosecrance Aware Center. I remember that much because it took a long time, and then I ended up having to go back to the crisis beds residential and I got out, when I got out they had people to come to my house so they can see me to do my check-ins. So I had, so I had help to get where the way they can come to me and help me do my check-ins instead of me going down there every day, because I don't, I get to go down.
*904Q Well, how do you forget for three months? This is what I don't understand. Your treatment never ended with Rosecrance Aware Center except that you just stopped going, so why did you do that?
A That's the thing, I don't know how or why I forget.
Q So you just went back again, what was the reason that made you remember to go just last month? Was it because you had a disability hearing coming up?
A No, I-
Q Then why did you go last month, why did you do that?
A I went back because I noticed I was slipping a little worse then I usually was.
Q Why, what was happening?
A Like a lot of situations. I have people yelling at me, people coming to my door knocking on it constantly. I have been haven't slept really too much for long periods of time, then I'm very forgetful of anything I do. Without a reminder I don't remember nothing.
Q So how is it you're able to live by yourself as you've been doing so for a year?
A Because my mom is, right now my mom is there so I'm kind of able to remember every day because she'll sit and remind me before I go or she'll leave a note. But I'm by myself like off, just by myself from off purposes I'm no good. I'm kind of thankful that they come to the house to do my check-ins, but at the same time if I was to just go down there I probably nine times out of ten forget.
R. 65-67.7 At one point, the ALJ asked plaintiff why he could not go to Rosecrance to get treatment, rather than having Rosecrance provider coming to his house to do check-ins, and he stated that he did not "have no way of transportation to getting there." R. 71. (Remember that plaintiff drove to his appointment with Dr. Peggau.)
The ALJ next asked medication side effects in the following exchange:
Q Any side effects from the medication you are taking now?
A No, not right now, but the Depakote that they had prescribed me I had told the lady when I was down there in the crisis beds, I remember that much because I put it down on a note, I had told her that the feels, the Depakote that she had prescribed me don't work. And it's only, it hasn't really worked too much with it of-it hasn't really worked too much because I was at first it was kind of helping but now it just does not help, it makes it worser. And every time I say something about it she acts, she doesn't listen so she prescribed me, so she prescribed me the same thing and then tagged another pill bottle along with it of the same stuff that I was taking to go to sleep. But she put it in a lower dosage and told me to take it throughout the day along with the Depakote. But that's not doing nothing except for making everything worse.
Q Making what worse?
A Like pretty much my sleeping habits is off, my pretty much functioning is not too much good. It is kind of messed up because I scare my mom and it's, it's not really not good at all.
Q What do you mean you scared your mom, what happened?
A To be honest with you I don't remember. All I remember is I was sitting there and she just looked over *905and said I was scaring her and I, and she said I held a whole conversation with her but I don't remember none of the story.
R. 73-74.
After plaintiff testified, the ALJ asked Dr. Heinemann the following:
Q And do you have an opinion as to whether he functions better when he used medications?
A It appears that he does function better when he take[s] his medication.
Q But the fact that he has, appears to be non-compliant at times, does that have anything to do with his impairments?
A Yes. I believe that diminishes his ability to cooperate with the mental health service providers .
Q Providing he received treatment and medicates, compliance with medication he could do a full-time job from a psychological perspective as long as he's on his medications?
A I do, yes.
R. 82 (emphasis added). When plaintiff's counsel questioned Dr. Heinemann, she asked the following question after Dr. Heinemann testified that plaintiff's low GAF scores were explained by his failure to take his medications:
Q Okay. But, Doctor, wouldn't, isn't it typical for somebody who has a condition like my client such as schizoaffective disorder to perhaps not be compliant with medication that often given delusional thoughts, given the nature of the disorder itself?
A Medication compliance can vary widely. Yes, some people are less-like particularly if the support from family is absent and is not thought to be severe, others are able to comply with their medication regimen.
R. 83-84.
After reviewing this testimony, the Court finds that several important questions remain regarding plaintiff's alleged non-compliance and the ALJ's implicit analysis of it. First, the Government's argument-that the ALJ could have concluded that plaintiff's "I can't remember" answers showed he was lying-are premised on the assumption that a claimant with a mental illness has the self-awareness to understand, and then convincingly explain, why he failed to comply with treatment. However, such a premise is often questionable in cases of mental illness. See generally Lewis v. Astrue , 2012 WL 5342669, *7 (N.D. Ill. Oct. 25, 2012) ("Federal courts have long recognized that, in the context of mental illness, insight can play an important role in evaluating the claimant's credibility; and, by definition, a claimant with poor insight cannot be expected to understand the true nature of his impairments."); see also Spiva , 628 F.3d at 351 ("Spiva's being vague or evasive when questioned about his illness could be evidence of malingering, but equally could reflect the effects of his psychotic mentation."). The ALJ should have explored this possibility.
Second, the Government's argument rests on another questionable proposition, which is that plaintiff gave a blanket "I can't remember" answer. However, as the above testimony shows, even though plaintiff may not have been articulate or clear at all times, he nonetheless alluded to several possible explanations for why he stopped taking his medications. One was side effects. Plaintiff explained that he told one medical provider that the Depakote was at first "kind of helping" but eventually made things worse. He stated that the medications made his sleeping worse and *906caused behavior problems, including one incident where he scared his mother. Medication side effects are a valid concern for any patient. Voigt v. Colvin , 781 F.3d 871, 877 (7th Cir. 2015) ("Nor did [the ALJ] note the natural reluctance of a person with psychiatric problems (perhaps of any person) to take powerful pain medications, as they can have serious side effects if not carefully used."). Another explanation was the practical problem of finding transportation to get to appointments. The ALJ never acknowledged these possible explanations.8
Third, the ALJ's two rationales are arguably at odds with each other. On the one hand, the ALJ suggested at various points that plaintiff was lying or exaggerating his symptoms. On the other hand, the compliance rationale assumes that plaintiff was genuinely experiencing the symptoms he reported, but that their extreme nature was merely because he failed to take medication and engage in counseling. This rationale rests on an assumption that plaintiff was telling the truth about his symptoms because, if he were not, then it would be difficult to know whether the medication was working. It is possible that the ALJ's two rationales could co-exist in some fashion, but if so, the ALJ must provide a better explanation of this seeming anomaly.
Fourth, the Government's argument is based on the conclusion that plaintiff's "I don't remember" answers about medication non-compliance were suspicious given that he was otherwise articulate in his testimony. The ALJ observed that plaintiff "was articulate and descriptive at the hearing with no indication that he did not understand the numerous questions and with no hesitancy in answering." R. 24; see also R. 14 (plaintiff "used phrasing appropriately"). After reading the transcripts from both hearings, the Court reached a different conclusion, consistent with the Kangail decision, which, one more time, binds this Court. The Court found that plaintiff's testimony was often repetitive and halting, was rambling and unfocused at times, and was occasionally marked by logical gaps. The extensive quotations set forth above illustrate some of these tendencies. It is true, of course, that ALJs generally are in a better position to judge a claimant's credibility than a court later reviewing a transcript. But in this case, the differing assessments are more striking than the typical case seen by this Court. On remand, the ALJ should ask the medical expert to provide an opinion on the cogency of plaintiff's speaking abilities, especially given that disorganized speech is a symptom of schizophrenia. See Mayo Clinic website ("Disorganized thinking is inferred from disorganized speech. Effective communication can be impaired, and answers to questions may be partially or completely unrelated.").
Fifth, whatever assessments the ALJ may have made about plaintiff's ability to comply, the ALJ failed to acknowledge that Dr. Heinemann was specifically asked about this issue and that he gave an answer supporting plaintiff. This is critical medical expert opinion evidence. Specifically, he testified that he believed that plaintiff's mental illness "diminishes his ability to cooperate with the mental health service providers." R. 82. The only slight qualification he gave, which was in a later answer, was the broader assertion that "some people" with mental illness are better *907able to comply, especially if they have family support. R. 84. But Dr. Heinemann never connected these general comments back to plaintiff. Here, based on the Court's reading of the record, the evidence appears mixed on the question of family support. On the one hand, plaintiff's grandmother has shown great concern and care over the years. On the other hand, there is evidence that plaintiff was estranged from some family members who were less inclined to help over time. See R. 16 (plaintiff "reported having difficulty getting along with family"); R. 40 (difficulties with his aunt).
Sixth and finally, as to the larger issue of malingering, the ALJ's conclusions rested heavily on observations from Dr. Peggau, which were based on one 75-minute interview, and also on a brief statement in a report by a school psychologist in 2010.9 R. 389. The ALJ drew a larger inference that plaintiff "has had a history of [ ] lying and malingering." R. 13. Although this evidence supports the ALJ's malingering theory, the ALJ still should have acknowledged and considered more carefully the contrary evidence. Insofar as this Court can tell, none of plaintiff's many treating doctors (including Dr. Jafry, Dr. Uma Srivastava, Dr. Dubois, and Dr. Kuna) ever raised a concern about malingering even though they observed plaintiff over many days in a hospital setting. R. 20-21. On remand, the ALJ should provide an explanation for how she viewed this competing line of evidence, which seems to conflict with her and Dr. Peggau's malingering theory.10
Having found that a remand is required based on plaintiff's primary argument, the Court need not address plaintiff's remaining arguments. These include the claim that the ALJ "failed to give proper consideration to plaintiff's longitudinal records," including treatment records and IQ tests "as early as the 5th grade" (one implication being that plaintiff would not have been malingering for disability benefits at such a young age); the ALJ failed to resolve conflicts in the medical opinion evidence (e.g. Dr. Heinemann versus Dr. Peggau on IQ scores); the ALJ failed to accurately assess all the requirements of Listing 12.05; and the ALJ "played doctor" in assessing IQ scores. Dkt. # 9 at 4-10. However, on remand, the ALJ should consider all these issues with a fresh eye and should provide a thorough analysis of these and other relevant issues.
CONCLUSION
For these reasons, plaintiff's motion for summary judgment is granted, the government's motion is denied, and this case is remanded for further consideration.
This Court recognizes the difficulty that Kangail and its progeny place on ALJs. But despite this Court's grumblings about Kangail , see Lewis v. Colvin , 2016 WL 4530338, *4 n.3 (N.D. Ill. Aug. 30, 2016), that decision controls in this circuit.
See R. 24 (plaintiff's mental health problems are "mostly due to noncompliance with medications"); id. (plaintiff's more severe problems are "transient issues due to non-compliance"); id. (plaintiff's symptoms are not as severe as he claims "if [he is] compliant with medications"); R. 26 ("he has been largely non-compliant with medications").
The ALJ's malingering rationale borrows heavily from Dr. Peggau's report. Ex. 12F. On March 16, 2015, Dr. Peggau interviewed plaintiff and administered an IQ test, seeing him for a total of 75 minutes. Dr. Peggau concluded that "there was a question about [plaintiff's] honesty" and whether he had been "forthcoming with information." R. 1032. Dr. Peggau made the following observations, which appear to be the basis for his assessment about plaintiff's dishonesty: plaintiff drove away in a car after the appointment when he had told Dr. Peggau during the interview that he did not drive; plaintiff "cracked his neck and had a cool façade" and made "little direct eye contact"; he "constantly began or finished sentences by saying, 'Pretty much,' "; he stated that he heard voices every hour but Dr. Peggau did not observe him hearing voices; he denied ever being hospitalized; and he denied having a behavioral problem. Id. at 1032-1037. Plaintiff had a full-scale IQ score of 71 on the test administered by Dr. Peggau, but Dr. Peggau noted that plaintiff "did not respond to encouragement" and "had no sense of urgency" when taking the test, thus making his score lower than it should have been, according to Dr. Peggau. R. 1037. Notably, Dr. Peggau found that plaintiff did not have any type of schizophrenia, a conclusion that differs from the diagnoses of a number of treating doctors. Instead, Dr. Peggau diagnosed plaintiff with an "Unspecified Bipolar Disorder (Rule out drug use)" and an "Unspecified Personality Disorder (Shallow/immature)." Id.
The Court has repeatedly expressed its frustration with the application of the Chenery doctrine in Social Security appeals. See Tucker v. Colvin , 2015 WL 6736700, *5 n.1 (N.D. Ill. Nov. 4, 2015). But, again, this Court is bound by Seventh Circuit precedent. The Court wished the Social Security Administration shared its views of stare decisis and judicial hierarchy.
This line of questioning continued, but did not lead to any clearer answers than in the material quoted above.
This line of questioning also continues on after the quoted material, but the Court again has chosen to end the quotation earlier out of concerns for brevity.
More broadly, the ALJ gave no consideration to the Seventh Circuit's observation that treatment for mental illnesses such as bipolar disorder"may require a complex drug regimen" that is not easy to follow. Kangail , 454 F.3d at 630. Here, in addition to having mental illnesses, plaintiff also had borderline intellectual functioning.
The school psychologist listed plaintiff's diagnosis as ADHD, and did not mention schizoaffective disorder. R. 389. This raises a question whether the psychologist had a complete understanding of plaintiff's medical history.
For example, Dr. Srivastava personally observed plaintiff "responding to voices." R. 22. By contrast, Dr. Peggau did not believe that plaintiff had schizophrenia apparently based on the fact that, during the 75-minute, he was "not actively psychotic." R. 1036.