POSNER, Circuit Judge.

This is an appeal from a decision by the district court affirming the denial of social security disability benefits by an administrative law judge, whose decision became final when the Social Security Administration's Appeals Council denied the applicant leave to appeal the Council's decision.

After an automobile accident in 1999 Christine Bjornson began having severe back pains. Three years later she was diagnosed with a "Chiari malformation," which is a protrusion of brain tissue into the spinal canal. The malformation may have been caused by the accident, but probably not; the cause, however, is irrelevant. After three operations on her brain and spine in 2002, the vision and speech problems that the Chiari malformation had caused lessened substantially but she developed hydrocephalus—a buildup of cerebrospinal fluid in the brain—that required the installation of a shunt in her brain, to drain the fluid. It took three installations to place the shunt properly, the last in 2003. To alleviate the severe headaches caused or aggravated by what appears to have been a total of nine brain and spinal operations that she had undergone, she was prescribed a number of powerful pain medications, including OxyContin, Percoset, Lyrica, and methodone, often in conjunction.

She hasn't worked since the auto accident. She was last insured for social security disability benefits in June 2005 (when she was 34 years old), so only if she was disabled from full-time work by that date is she eligible for benefits.

At the first of two disability hearings she testified that since before her last-insured date she has had constant, excruciating headaches four or five days a week, which cause her to vomit when she stands up. She takes her pain medications when she wakes up and then goes back to bed for hours because she "could not do anything else because of her pain medications." She also has severe back pain, aggravated by obesity, but it does not appear that the back pain is disabling in itself, though it compounds the effects of the headaches on her ability to work.

At the first hearing the administrative law judge decided that Bjornson should be examined by a physician hired by the Illinois Department of Human Services, which works with the Social Security Administration in determining eligibility for social security disability benefits. See Illinois Department of Human Services, "Disability Determination Services," www.dhs. state.il.us/page.aspx?item=29979 (visited Dec. 30, 2011). This was done, and the physician, Dr. Muhammad Rafiq, reported that Bjornson "gets frequent severe headaches three to four times per week during which she cannot stand," and that in an 8-hour day she can sit for an hour and a half, stand for an hour, and walk for half an hour—the rest of the time she has to lie down. The limitation on standing may be caused by her back pain as well as her headaches, but this is unclear from Rafiq's report.

Other doctors' reports note Bjornson's "chronic headaches, neck pain, intermittent visual problems, swallowing problems, slurred speech, and bilateral finger numbness." Dr. Ira Goodman, a pain specialist who had treated Bjornson since 2003, noted her complaints of constant headaches and diagnosed her with (among other things) occipital neuralgia, a type of headache that involves piercing, throbbing, or "electric-shocklike" chronic pain in the neck and head. He also remarked her cervical spine pain, lower-back pain, a three-week stretch of nonstop headaches, and daytime somnolence because of her pain medications. The administrative law judge did not mention the diagnosis of occipital neuralgia.

All the evidence we've described thus far, except for Dr. Rafiq's, concerns diagnoses and treatments that Bjornson received before June 2005. The record also contains a good deal of evidence, besides Rafiq's, concerning Bjornson's diagnoses and treatments since then. None of this evidence indicates that her symptoms have worsened—that if she is disabled today, nevertheless she wasn't disabled before her last insured date. Yet the government's brief argues the irrelevance of all post-June 2005 medical data, an argument that both is factually mistaken and violates the Chenery rule, because the administrative law judge ruled that post-June 2005 medical data could be considered—and he was right, as there was no reason to believe that Bjornson's ability to work had declined since then. Ray v. Bowen, 843 F.2d 998, 1004-06 (7th Cir.1988); Potter v. Secretary of Health & Human Services, 905 F.2d 1346, 1348 (10th Cir.1990) (per curiam); Dousewicz v. Harris, 646 F.2d 771, 774 (2d Cir.1981).

One physician, Dr. Chukwuemeka Ezike, testified that while Bjornson has a history of chronic headaches attributable to the Chiari malformation and has been treated for "headaches, nausea, vomiting, and some parasthesia" (numbness or tingling in the limbs), he "did not find enough ... in the medical records" to justify Dr. Rafiq's opinion. He believed that her "pain was not well substantiated after 2003," and he did "not find the evidence that says she cannot sustain [a] sedentary job." His testimony goes on and on, but what we have quoted is the only intelligible portion of it that bears on Bjornson's ability to hold a full-time job. Elsewhere he did say that "because of the persistence of the symptoms and the fact that she was on high dose opiates, that's not in my professional opinion based on reasonable degree of medical certainty that physically she would be unable to sustain unemployment at that time.... [W]e do not have any evidence physically." But we don't know what this passage means.

The rest of Dr. Ezike's testimony, which was continually and confusingly interrupted by the administrative law judge, is epitomized in the following exchange and seems, to the limited extent that it is even intelligible, irrelevant:

One turns with relief to a November 2008 evaluation by Dr. Goodman:

Bjornson testified at the second hearing, consistently with her testimony at the first, that she needed the shunt in her brain replaced but that the neurosurgeon she had consulted had refused for fear that it might bring back her hydrocephalus and require still another brain surgery. (His report was consistent with that testimony.) She testified that "there's nothing more they [the doctors] can do for me," and that the pain medications help "a little" but "usually ... just knock me out and I sleep through the headaches." She testified that she spends most of her day lying down.

The administrative law judge concluded that Bjornson was capable of performing sedentary work (she had been at various times before her accident a waitress, an off-track betting teller, and a clerk in a real estate office), albeit with some limitations—mainly that she be able to sit or stand whenever she wanted to. He asked the vocational expert (an expert who testifies in disability hearings about whether there is a substantial number of jobs in the local economy that the claimant is physically and mentally able to perform) whether given these limitations Bjornson could satisfy the requirements of such a job. The vocational expert testified that Bjornson could work as a clerk in a real estate office because she "could remain in a seated position long enough to perform data entry tasks." The administrative law judge asked whether her opinion would change if Bjornson had to lie down "at least an hour during the workday at irregular intervals every day differently"—to which the vocational expert replied that in that event "there would be no jobs [she] could perform." And in response to a further question from the administrative law judge—what if Bjornson had occasional bad headaches that caused her to miss two days of work?—the vocational expert answered: that too "would knock her out of all work." Nevertheless the administrative law judge concluded that Bjornson was not disabled.

Reading the administrative law judge's opinion, we first stubbed our toe on a piece of opaque boilerplate near the beginning, where, after reciting Bjornson's description of her medical condition, the opinion states: "After careful consideration of the evidence, the undersigned [administrative law judge] finds that the claimant's medically determinable impairments would reasonably be expected to cause the alleged symptoms; however, the claimant's statements concerning the intensity, persistence and limiting effects of these symptoms are not credible to the extent they are inconsistent with the above residual functional capacity assessment." The government's brief describes this passage as a "template," by which it means a passage drafted by the Social Security Administration for insertion into any administrative law judge's opinion to which it pertains.

This "template" is a variant of one that this court (and not only this court) had criticized previously—that "after considering the evidence of record, the undersigned finds that claimant's medically determinable impairments would reasonably be expected to produce the alleged symptoms, but that the claimant's statements concerning the intensity, persistence and limiting effects of these symptoms are not entirely credible." In Parker v. Astrue, 597 F.3d 920, 922 (7th Cir.2010), we called this "meaningless boilerplate. The statement by a trier of fact that a witness's testimony is `not entirely credible' yields no clue to what weight the trier of fact gave the testimony" (emphasis in original); see also Punzio v. Astrue, 630 F.3d 704, 709 (7th Cir.2011); Martinez v. Astrue, 630 F.3d 693, 696-97 (7th Cir.2011); Spiva v. Astrue, 628 F.3d 346, 348 (7th Cir.2010). "Such boilerplate language fails to inform us in a meaningful, reviewable way of the specific evidence the ALJ considered in determining that claimant's complaints were not credible. More troubling, it appears that the Commissioner has repeatedly been using this same boilerplate paragraph to reject the testimony of numerous claimants, without linking the conclusory statements contained therein to evidence in the record or even tailoring the paragraph to the facts at hand, almost without regard to whether the boilerplate paragraph has any relevance to the case." Hardman v. Barnhart, 362 F.3d 676, 679 (10th Cir.2004) (citation omitted).

The present "template," which adds at the end of the previous one "... to the extent they are inconsistent with the above RFC assessment," is even worse, though the government's brief defends it with great vigor—while at the same time both mistakenly describing it as the identical boilerplate criticized in the previous cases and confusing it with form orders that the Social Security Administration has authorized when an administrative law judge "chooses to make a wholly favorable [to the applicant] oral decision at the hearing." Social Security Administration, "Oral (Bench) Decision Procedures," www.ssa. gov/OP_Home/hallex/I-05/I-5-1-17.html (visited Dec. 24, 2011), and "Findings Integrated Templates (FIT), Social Security Online," www.ssa.gov/appeals/fit (visited Dec. 24, 2011).

The government regards the "template" as an indispensable aid to the Social Security Administration's overworked administrative law judges. Yet when we asked the government's lawyer at argument what the "template" means, he confessed he did not know.

One problem with the boilerplate is that the assessment of the claimant's "residual functional capacity" (the bureaucratic term for ability to work) comes later in the administrative law judge's opinion, not "above"—above is just the foreshadowed conclusion of that later assessment. A deeper problem is that the assessment of a claimant's ability to work will often (and in the present case) depend heavily on the credibility of her statements concerning the "intensity, persistence and limiting effects" of her symptoms, but the passage implies that ability to work is determined first and is then used to determine the claimant's credibility. That gets things backwards. The administrative law judge based his conclusion that Bjornson can do sedentary work on his determination that she was exaggerating the severity of her headaches. Doubts about credibility were thus critical to his assessment of ability to work, yet the boilerplate implies that the determination of credibility is deferred until ability to work is assessed without regard to credibility, even though it often can't be. In this regard we note the tension between the "template" and SSR 96-7p(4), www.ssa.gov/OP_Home/rulings/di/ 01/SSR96-07-di-01.html (visited Jan. 4, 2012), which states that "an individual's statements about the intensity and persistence of pain or other symptoms or about the effect the symptoms have on his or her ability to work may not be disregarded solely because they are not substantiated by objective medical evidence." The applicant's credibility thus cannot be ignored in determining her ability to work (her residual functional capacity, in SSA-speak).

The Social Security Administration had better take a close look at the utility and intelligibility of its "templates."

The administrative law judge based his doubts about Bjornson's credibility on his assessment of the medical reports or testimony of the three doctors whom we've mentioned, Goodman, Rafiq, and Ezike. He remarked that Dr. Goodman's treatment notes report that Bjornson's complaints about headaches had become less frequent—that she had complained about them only four times in a nine-month period—and that although her headaches "have been an ongoing problem since before she started coming to see me in 2003... she can deal with that," although she "must sit or lie down several times per day to control pain." The administrative law judge appears to have overlooked Bjornson's testimony that she had concluded that the medical profession couldn't do anything more for her headaches—by this time she had had multiple brain surgeries—except give her painkillers that knock her out; if that testimony is credited, it would explain why her complaints to doctors about headaches diminished over time. The administrative law judge did not mention Dr. Goodman's further statements that the headaches "were always present" and that Bjornson had complained of "non-stop headaches" that had lasted for three weeks.

And it was a mistake to infer from Goodman's quoted statement, as the administrative law judge did in ruling that Bjornson can work as long as she has a sit-stand option, that sitting down several times a day is all that she would have to do in order to be able to hold a nine to five job. Apart from the ambiguity not explored by the administrative law judge of the term "control pain," Dr. Goodman's statement that Bjornson "must sit or lie down several times per day to control pain" is consistent with Bjornson's testimony that often she must lie down, not stand or sit down, to alleviate the pain. One does sedentary work sitting (the word "sedentary" is from the Latin word "sedere," which means "to sit"), but not lying down.

The administrative law judge rejected Dr. Rafiq's evidence (remember that he opined that Bjornson could sit, walk, or stand for a total of only three hours in an eight-hour workday—90 minutes (sitting) + 60 minutes (standing) + 30 minutes (walking)—which means that she would have to be able to lie down for the other five hours in the workday), saying that "those limitations are not supported by [Dr. Rafiq's] own clinical observations." Well, obviously Dr. Rafiq didn't conduct an eight-hour examination of Bjornson. The administrative law judge faulted Rafiq for "rel[ying] on collateral evidence in concluding that Bjornson's functioning was so markedly limited." He did not explain what he meant by "collateral evidence," but probably he meant the other treatment notes in the record—yet Dr. Rafiq would have been remiss not to consult them and to weigh them in forming his own judgment.

Out of the blue the administrative law judge remarked of Dr. Rafiq that "doctors sometimes express an opinion in an effort to assist a patient with whom they sympathize. While it is difficult to confirm the presence of such a motive, it is more likely in situations where the opinion in question departs substantially from the rest of the evidence of record, as in the current case." Actually the doctor's evidence was consistent with most of the rest of the evidence. And the suggestion that his evidence was based on sympathy for Bjornson is both unsupported and implausible. She is not his patient; the Illinois state agency that works with the Social Security Administration retained and paid him to conduct a single examination of a total stranger. His first medical report (of two reports) states that Bjornson "was informed that this examination was solely for providing information to the Bureau of Disability Determination Services and did not constitute a patient/physician relationship."

The administrative law judge expressed doubt about Bjornson's credibility on the further ground of her "activities of daily living," notably that she can walk up to one block, sit or stand for up to 15 minutes, lift 10 pounds, bathe and dress normally, and even drive and shop. But she had never testified that she was immobilized, and indeed she had testified that she had one or two good days each week—for all that appears, the activities recited by the administrative law judge are concentrated in those days. Doubtless she dresses and bathes more frequently than twice a week, but one can have awful headaches yet still dress and bathe. The critical differences between activities of daily living and activities in a full-time job are that a person has more flexibility in scheduling the former than the latter, can get help from other persons (in this case, Bjornson's husband and other family members), and is not held to a minimum standard of performance, as she would be by an employer. The failure to recognize these differences is a recurrent, and deplorable, feature of opinions by administrative law judges in social security disability cases. See Punzio v. Astrue, supra, 630 F.3d at 712; Spiva v. Astrue, supra, 628 F.3d at 351-52; Gentle v. Barnhart, 430 F.3d 865, 867-68 (7th Cir.2005); Draper v. Barnhart, 425 F.3d 1127, 1131 (8th Cir. 2005); Kelley v. Callahan, 133 F.3d 583, 588-89 (8th Cir.1998); Smolen v. Chater, 80 F.3d 1273, 1284 n. 7 (9th Cir.1996).

The administrative law judge gave decisive weight to testimony by Dr. Ezike, even though he is not a pain specialist, like Goodman, and, unlike both Goodman and Rafiq, had not examined Bjornson. (So all the evidence Dr. Ezike relied on was "collateral," unlike Dr. Rafiq, whom the administrative law judge criticized for relying on such evidence.) He noted with approval Dr. Ezike's testimony that the medical evidence concerning Bjornson's condition after 2003 "supported a sedentary residual functional capacity finding." Yet on the previous page of his opinion he had rebuked Dr. Goodman for saying that Bjornson "remained unable to work," remarking that "statements that a claimant is disabled or unable to work are not medical opinions but are dispositive administrative findings ... reserved to the Commissioner" of social security. The remark is imprecise. The pertinent regulation says that "a statement by a medical source that you are `disabled' or `unable to work' does not mean that we will determine that you are disabled." 20 C.F.R. § 404.1527(e)(1). That's not the same thing as saying that such a statement is improper and therefore to be ignored, as is further made clear when the regulation goes on to state that "the final responsibility for deciding" residual functional capacity (ability to work—and so whether the applicant is disabled) "is reserved to the Commissioner." § 404.1527(e)(2) (emphasis added). And "we will not give any special significance to the source of an opinion on issues reserved to the Commissioner." § 404.1527(e)(3) (emphasis added).

A second medical report that Dr. Rafiq submitted was a form that the Illinois disability determinations agency had asked him to fill out, and one of the questions was how long Bjornson could sit, stand, and walk and what she did when she couldn't do any of those things. When he stated on the form that she could sit, stand, and walk for a total of only three hours in an eight-hour workday, he was not invading any prerogative reserved to the Social Security Administration. But his statement inescapably implied that she can't work full time, for what employer would hire for a full-time job someone who has to lie down for five hours during the workday?

The administrative law judge not only forgot his dismissive view of physicians' testimony relating to issues "reserved to the Commissioner" when he came to Dr. Ezike, but compounded the inconsistency by adding that he was "assign[ing] substantial weight to Dr. Ezike's opinions, as he is familiar with the Social Security disability program, reviewed all available medical evidence, listened to Bjornson's testimony [at the second administrative hearing] regarding her symptoms and functional limitations, and issued opinions consistent with the rest of the medical evidence record." Apart from the fact that Dr. Ezike's testimony was not "consistent with the rest of the record," his familiarity with the social security disability program could be relevant only if it permitted him to offer an opinion concerning Bjornson's eligibility—which the administrative law judge had just said was the prerogative of the Social Security Administration. The regulation does state that "the amount of understanding of our disability programs and their evidentiary requirements that an acceptable medical source has ... [is among the] relevant factors that we will consider in deciding the weight to give to a medical opinion." § 404.1527(d)(6). But the administrative law judge seems to have forgotten that when he dismissed Dr. Rafiq's evidence. And remember that it was a sister government agency (in effect) of the Social Security Administration that had hired Dr. Rafiq to examine Bjornson, which implies that Rafiq was believed to know as much as he needs to know about the social security disability program in order to be able to give an informed professional opinion about the physical condition of an applicant for disability benefits.

It is impossible to tell, moreover, whether Dr. Ezike based his skepticism about the severity of Bjornson's pain (her "pain was not well substantiated after 2003") on the absence of corroborating objective medical evidence, which if he did would be in tension with SSR 96-7p(4) ("an individual's statements about the intensity and persistence of pain or other symptoms or about the effect the symptoms have on his or her ability to work may not be disregarded solely because they are not substantiated by objective medical evidence"), or on his interpretation of her report of her pain as quoted or paraphrased in the notes of the other doctors (another possible interpretation of "[her] pain was not well substantiated after 2003"). If the former, his opinion is both implausible—for there was plenty of corroborating medical evidence—and in tension with SSR 96-7p(4), as we said; and if the latter it adds nothing to those notes and those doctors' conclusions and does not undermine Bjornson's testimony that she had abandoned hope of being able to deal with her headaches other than by a combination of painkillers that made her somnolent with (what goes with somnolence) lots of lying down. Dr. Ezike may have based his testimony on the fact that the other doctors' treatment notes indicated that Bjornson's complaints about headaches were intermittent.

The Social Security Administration's administrative law judges are overworked, but if one may judge from the transcript in this case, the two hearings proceeded in a leisurely, even meandering, fashion. Whatever the cause, the administrative law judge's opinion failed to build a bridge between the medical evidence (along with Bjornson's testimony, which seems to have been fully consistent with that evidence) and the conclusion that she is able to work full time in a sedentary occupation provided that she can alternate sitting and standing. See Scott v. Astrue, 647 F.3d 734, 740 (7th Cir.2011); Terry v. Astrue, 580 F.3d 471, 475 (7th Cir.2009) (per curiam); cf. Hardman v. Barnhart, supra, 362 F.3d at 678-79.

The judgment of the district court is reversed and the matter returned to the Social Security Administration for further proceedings consistent with this opinion.

REVERSED AND REMANDED.