MEMORANDUM *

Robert Sherman appeals the denial of his application for supplemental security income under the Social Security Act. We “reverse only if the ALJ’s decision was not supported by substantial evidence in the record as a whole or if the ALJ applied the wrong legal standard.” Molina v. Astrue, 674 F.3d 1104, 1110 (9th Cir. 2012). We have jurisdiction pursuant to 28 U.S.C. § 1291, and we affirm.1

Substantial evidence supports the ALJ’s adverse credibility determination. The ALJ offered “specific, clear and convincing reasons” for rejecting Sherman’s testimony about his limitations and pain to the extent that his testimony conflicted with the residual functional capacity as*748sessment. Molina, 674 F.Bd at 1112 (internal quotation marks omitted). First, the ALJ noted Sherman’s sporadic work history before he filed for disability and the fact that his job rehabilitation services case worker closed his file for failure to cooperate and participate in job searching. See Thomas v. Barnhart, 278 F.3d 947, 959 (9th Cir. 2002).

Second, the ALJ relied on physicians’ observations that Sherman exhibited “self-limiting behaviors” and had “poor credibility.” Dr. Hurd, an examining physician, reported that Sherman “refused to do many of the exam maneuvers” and “used less than actual effort.” See id. (holding that a claimant’s failure “to give maximum or consistent effort during two physical capacity evaluations” supported the ALJ’s adverse credibility finding). He also observed Sherman perform multiple tasks that were inconsistent with his alleged limitations. See Carmickle v. Comm’r Soc. Sec. Admin., 533 F.3d 1155, 1161 (9th Cir. 2008) (“Contradiction with the medical record is a sufficient basis for rejecting the claimant’s subjective testimony.”). Dr. Hurd’s findings were consistent with other physicians’ observations. Dr. Goodell, a treating physician, said that Sherman’s “inability to cooperate/participate” and “unwillingness to participate in a physical exam” prevented her from testing his physical abilities. Sherman refused to bear weight during the exam. When Dr. Goodell left the room, however, he was able to climb up on the exam table to take a nap. Similarly, Dr. Mozer, an examining psychologist, described Sherman as “evasive” and commented on his failure to give a “valid effort.”

Third, the ALJ reasonably concluded that Sherman’s medical conditions did not limit his daily activities to the extent that he claimed. No doctor placed any restrictions on Sherman’s activities. See Burch v. Barnhart, 400 F.3d 676, 681 (9th Cir. 2005) (“Although lack of medical evidence cannot form the sole basis for discounting pain testimony, it is a factor that the ALJ can consider in his credibility analysis.”). Further, the ALJ found that Sherman’s daily marijuana use contributed to his lack of activity. Dr. Mozer opined that Sherman had “very marginal motivation” and is “basically doing what he wants to do (sit around and smoke pot).” The ALJ’s conclusion that marijuana use can impact daily activities, social functioning, and concentration is reasonable, so we must accept it. See Tommasetti v. Astrue, 533 F.3d 1035, 1040 (9th Cir. 2008).

To the extent that the ALJ might have erred in relying on Sherman’s failure to get back surgery when he did not have money or insurance, see Orn v. Astrue, 495 F.3d 625, 638 (9th Cir. 2007), any error was harmless in light of the ALJ’s other findings that provide substantial evidence for the adverse credibility determination. See Carmickle, 533 F.3d at 1162.

Sherman argues that the ALJ rejected the opinions of his treating physicians. We disagree. The ALJ considered the opinions of Dr. Verby, Dr. Quenemon, Dr. Goodell, and Dr. Draper, and their opinions are consistent with the ALJ’s finding that Sherman has severe degenerative disc and joint disease in his back and osteoarthritis in his hands. The ALJ’s written decision was not required to discuss every piece of evidence. Howard v. Barnhart, 341 F.3d 1006, 1012 (9th Cir. 2003). Sherman challenges the residual functional capacity assessment, but he has not shown how it is inconsistent with any physician’s opinion. No physician placed any restrictions on Sherman’s activities or diagnosed any physical impairments beyond those found by the ALJ.

*749We also reject Sherman’s claim that the ALJ ignored the opinions of mental health care providers that he was depressed. After considering the evidence that Sherman claims was ignored, the ALJ ruled that Sherman’s alleged depression is “either not medically determinable, due to the lack of a formal diagnosis, or is non-severe.” The ALJ’s determination is supported by substantial evidence. Sherman was never diagnosed with depression. Dr. Mozer said that depression was “questionable” and he ruled it out of consideration by finding that if Sherman does have depression it is “mild,” “certainly not limiting,” and “a natural consequence of an empty lifestyle.” Likewise, Dr. Martin, a non-examining physician, testified that there was not much support in the record for depressive disorder. Sherman’s depression was not assessed by any other acceptable medical source. See 20 C.F.R. § 404.1513(a). The ALJ gave a “germane” reason, Molina, 674 F.3d at 1111, for rejecting Sherman’s other source evidence, noting that the therapist who assessed his global functioning assessment score did not diagnose depression. For the same reason, the ALJ properly discounted Sherman’s answers to mental health questionnaires.

Because the ALJ did not completely rule out the possibility that Sherman has mild depression, we assume without deciding that he was required to consider it in assessing Sherman’s residual functional capacity. “The ALJ is required to consider all of the limitations imposed by the claimant’s impairments, even those that are not severe.” Carmickle, 533 F.3d at 1164. In this case, however, Sherman has not shown that his alleged depression resulted in any functional limitations that the ALJ failed to consider. See Burch, 400 F.3d at 684.

Finally, we reject Sherman’s claim that the ALJ ignored the findings of Dr. McFarland, a state agency reviewing psychologist, and Dr. Mozer that he has difficulties maintaining concentration, persistence, and pace as a result of his anti-personality disorder. Sherman is correct that the residual functional capacity assessment does not specifically mention limitations in concentration, persistence, and pace. This omission, however, does not constitute reversible error. We held in Stubbs-Danielson v. Astrue, that a residual functional capacity assessment “adequately captures” a claimant’s limitations in concentration, persistence, and pace as long as the assessment is “consistent with restrictions identified in the medical testimony,” 539 F.3d 1169, 1174 (9th Cir. 2008).

Here, the ALJ, like the one in Stubbs-Danielson, adopted the only “concrete restrictions” identified by Sherman’s physicians. See id. Both Dr. McFarland and Dr. Mozer concluded that Sherman is capable of unskilled work, despite his functional limitations. Dr. McFarland said that Sherman “would do best at work that does not require dealing with the public or working closely with others” but he is able to “understand, carry out and remember simple instructions,” and “respond appropriately to supervision, coworkers and work situations.” The ALJ gave Dr. McFarland’s opinion “significant weight” and noted that the residual functional capacity assessment is consistent with her findings. We agree. The assessment includes routine unskilled jobs with occasional to frequent new learning and excludes jobs that require constant dealing with the public, large groups of people, distracting situations, constant critical supervision, high constant focus requirements, and high constant stress requirements. To the extent that the ALJ might have erred by also including semi-skilled jobs, this error *750was harmless because the ALJ found that Sherman is capable of performing unskilled jobs that exist in significant numbers in the national economy. See Molina, 674 F.3d at 1115 (“[A]n ALJ’s error is harmless where it is inconsequential to the ultimate nondisability determination.” (internal quotation marks omitted)).

The ALJ gave the vocational expert a hypothetical based on the residual functional capacity assessment. For the reasons discussed above, we conclude that the hypothetical properly included all of Sherman’s limitations that are supported by substantial evidence. See Bayliss v. Barnhart, 427 F.3d 1211, 1217-18 (9th Cir. 2005).

AFFIRMED.

Dissenting:

Substantial evidence does not support the ALJ’s finding that claimant Robert Sherman lacks credibility. Accordingly, I dissent.

I

1. Contrary to the ALJ’s assessment, Sherman’s sporadic work history before his 2009 filing date is entirely consistent with his testimony. Sherman’s back injury occurred in 1992. Although he attempted to keep working after the injury, his back problems worsened in 1998, when he re-injured himself in a car accident. When Sherman protectively filed in 2009, he alleged a disability onset date of 1998. He amended that date because he did not have the medical records to prove his case that far back. But treatment notes from 2001 and 2004 do indicate that Sherman was in serious pain from back problems during this time, and that when he did work, he “exacerbated his injury in the low[er] back.” Thus, Sherman’s sporadic work history before his filing date does not show that factors other than his disability prevented him from working, and is in fact consistent with his account of his back injury.

2. The second reason the panel gives in support of the ALJ’s credibility determination is Sherman’s lack of cooperation with state vocational rehabilitation services. The record does not support such a finding.

The Montana Vocational Rehabilitation Program (“Montana Vocational”) reports as a whole confirms Sherman’s alleged limitations and illustrates that it was his disability that prevented him from finding work. Those records show that Sherman thought he could do short-term, but not long-term driving, as long as lifting was not required. The vocational school “only had ... long term over the road stuff,” and short-run jobs not involving lifting were not available for inexperienced drivers.

Sherman did receive his Commercial Driver’s License, and tried some driving jobs, but they turned out to require lifting as much as 70 pounds, which he could not handle. At a meeting with Sherman, Montana Vocational recognized that the agency had been “unsuccessful in finding a driving job for Robert that does not involve heavy lifting and labor.”

Only after these failures did Sherman invite Montana Vocational to close his case. In doing so, he explained that he did it because “my back’s not in the best of shape right now, and I kinda need to wait until a few months from now;” “I can’t really work 3-5 hours and I can’t really put in a 5-day work week either.”

In short, the vocational program reports support rather than disprove Sherman’s recurring back pain and physical inability to work.

*7513. Sherman’s failure to have back surgery was also an inappropriate ground for rejecting Sherman’s testimony. “[I]f a claimant complains about disabling pain but fails to seek treatment, or fails to follow prescribed treatment, for the pain, an ALJ may use such failure as a basis for finding the complaint unjustified or exaggerated.” Orn v. Astrue, 495 F.3d 625, 638 (9th Cir. 2007). But there are four problems with the ALJ’s doing so here.

First, there is no clear prescription during the period at issue that Sherman should follow through with back surgery. Second, there are indications, in two doctors’ reports, that Sherman did not have back surgery when it was previously recommended because he could not afford it, and that he remained interested in doing so if he could obtain funding for it. “[Disability benefits may not be denied because of the claimant’s failure to obtain treatment he cannot obtain for lack of funds.” Id. (quoting Gamble v. Chater, 68 F.3d 319, 321 (9th Cir. 1995)); see Regennitter v. Commissioner of the Soc. Sec. Admin., 166 F.3d 1294, 1297 (9th Cir. 1999). Third, in light of that fact that Sherman has consistently sought treatment for his back pain, his hesitancy to have a major surgery cannot cast doubt on his claim. “It is common knowledge that spinal surgery is often dangerous and entails much pain and suffering.” Schena v. Secretary of Health & Hum. Servs., 635 F.2d 15, 20 (1st Cir. 1980) (reversing Secretary’s decision that claimant’s rejection of spinal surgery barred him from disability benefits “[gjiven the uncertain (and sometimes adverse) consequences of spinal surgery”) (internal quotation marks omitted). While Social Security regulations require claimants to follow “treatment prescribed by [a] physician” to receive benefits, the same regulations make clear that if the claimant has “a good reason” for not following the prescribed treatment, rejection of treatment -will not be held against the claimant. 20 C.F.R. § 416.930(a) & (b); see Molina v. Astrue, 674 F.3d 1104, 1114 (9th Cir. 2012); SSR 96-7p, 1996 WL 374186 (July 2, 1996). The regulations further specify that an example of “a good reason for not following treatment” is that “[t]he treatment because of its enormity (e.g., open heart surgery) ... is very risky for you.” 20 C.F.R. § 416.930(c)(4). Finally, Sherman did purse several treatments for his back pain other than surgery (epidurals, pain medication, physical therapy).

In short, the fact that Sherman has not yet had spinal surgery can have no bearing on his credibility.

4. The ALJ also cited the lack of limitations placed on Sherman by his treating physicians, and his physicians’ recommendation of exercise. However, Sherman suffers from obesity and hypertension. So the exercise recommendation is to be expected, and does not indicate ability to work.

Moreover, examination of the medical reports reveal a recognition that pain might well limit Sherman’s ability to engage in the recommended treatment, as the reports also include recommendations for rest and pain management, rather than uniform recommendations of exercise.

5. Dr. Hurd’s statement that, in his view, Sherman overplayed his limitations and had “poor credibility as an examinee” was based on Dr. Hurd’s observations of Sherman’s behavior during the consultative exam. The ALJ was entitled to consider Dr. Hurd’s view in assessing Sherman’s credibility.

However, Dr. Hurd’s evaluation was equivocal and stated only his “intuitive feeling” that Sherman had a greater ability than he showcased. Moreover, Dr. Hurd was a consulting examiner, so his opinion is entitled to less weight than the opinions *752of treating physicians. See Holohan v. Massanari, 246 F.3d 1195, 1202 (9th Cir. 2001); 20 C.F.R. § 416.927(c). Not a single treating physician suggested that Sherman was malingering or inventing pain. If they had, one would think they would not have prescribed him serious pain medication and considered him a candidate for spinal surgery. My colleagues in the majority refer to reports from Drs. Goodell and Mozer. But Dr. Goodell cited Sherman’s “inability to cooperate/participate,” (emphasis added), and Dr. Mozer was a consulting psychologist, who did not evaluate Sherman’s physical disabilities.

Furthermore, the ALJ himself does not attribute significant weight to Dr. Hurd’s finding, stating only that Dr. Hurd’s assessment “does not enhance the claimant’s credibility,” and does not cite to either Dr. Goodell’s or Dr. Mozer’s reports as supporting the credibility determination.

Finally, the ALJ rejected Sherman’s narration of his limited daily activities as evidence of his disability, because his account was not objectively verifiable, and because his limited daily activities could have been attributed to his marijuana dependence. These remaining considerations do not constitute substantial evidence supporting the ALJ’s credibility determination, and so the ALJ’s reliance on these considerations does not make the aforementioned errors harmless. See Carmickle v. Commissioner Soc. Sec. Admin., 533 F.3d 1155, 1162 (9th Cir. 2008) (error is harmless if “ALJ’s remaining reasoning and ultimate credibility determination were adequately supported by substantial evidence in the record” (emphasis omitted)).

II.

Had the ALJ credited Sherman’s testimony as true, he would have been obliged to include Sherman’s testimony as to his limitations in determining Sherman’s residual functioning capacity, including Sherman’s testimony that he has to lie down to two or three times a day, throughout the day, for up to an hour or two at a time to take the pressure off his back. According to the vocational expert’s evidence at the ALJ hearing, including this limitation in the residual functioning capacity results in a finding that Sherman was disabled. Therefore, I would remand to the agency for calculation and an award of benefits. Lingenfelter v. Astrue, 504 F.3d 1028, 1041 (9th Cir. 2007).

For all these reasons, I respectfully dissent.