FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
KRISTIN COLLARD, No. 25-6334
D.C. No.
Plaintiff - Appellant,
2:24-cv-00754-SKV
v.
FRANK BISIGNANO,
Commissioner of Social Security, OPINION
Defendant - Appellee.
Appeal from the United States District Court
for the Western District of Washington
Sarah Kate Vaughan, Magistrate Judge, Presiding
Argued and Submitted August 4, 2026
Seattle, Washington
Filed September 23, 2026
Before: Michael Daly Hawkins, M. Margaret McKeown,
and Morgan B. Christen, Circuit Judges.
Opinion by Judge McKeown
2 COLLARD V. BISIGNANO
SUMMARY*
Social Security
The panel affirmed in part and reversed in part the
district court’s judgment affirming the denial of Kristin Collard’s application for disability insurance benefits (“DIB”) under the Social Security Act.
Collard applied for DIB based on diagnoses of
fibromyalgia, migraines, and degenerative disc disease. Collard testified to immense pain from those
impairments and ongoing limitations that prevented her from working years later. The administrative law judge (“ALJ”) discredited Collard’s testimony, and compatible third-party evidence from relatives and a co-worker, in part because Collard had worked a strenuous job before her conditions worsened.
The panel held that the ALJ erred in evaluating Collard’s subjective testimony. First, the ALJ erred in his
characterization of Collard’s work history. Although the ALJ pointed to Collard having worked a high-strain job even after developing her impairments to infer that her condition was not disabling, a past ability to work through pain can be entirely consistent with testimony of worsening impairments that become disabling over time. Second, the ALJ erred in analyzing Collard’s activities of daily living. The ALJ merely listed a number of activities that Collard performed, while failing to acknowledge aspects of the daily activity evidence consistent with Collard’s claim. Third, substantial
*
This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.
COLLARD V. BISIGNANO 3
evidence did not support the ALJ’s rationale that Collard could effectively treat her symptoms with medication. Because another ALJ may reach a different
disability determination when fully crediting that testimony, the ALJ’s errors were prejudicial and must be reevaluated in further proceedings.
The panel held that the ALJ also erred in evaluating the lay evidence from Collard’s husband, niece, co-worker, and sister-in-law. Although the ALJ’s evaluation of lay
evidence is held to a lower standard than subjective
testimony, it was improper for the ALJ to reject that
testimony simply because it was consistent with Collard’s testimony given the errors previously identified with the ALJ’s analysis. The panel held that this evidence must be reevaluated as well.
The panel held that with respect to the medical opinion evidence, substantial evidence supported the ALJ’s
determination. The ALJ provided specific and legitimate reasons for rejecting the opinions of Collard’s treating physicians. The panel therefore affirmed as to the treatment provider evidence.
4 COLLARD V. BISIGNANO
COUNSEL
Amy M. Gilbrough (argued), Douglas Drachler McKee &
Gilbrough LLP, Seattle, Washington, for PlaintiffAppellant.
Katherine B. Watson (argued), Attorney; Lars J. Nelson, Special Assistant United States Attorney; Franco L. Becia, Assistant Regional Counsel; Mathew W. Pile, Head of
Program Litigation 1; Social Security Administration,
Baltimore, Maryland; for Defendant-Appellee.
OPINION
McKEOWN, Circuit Judge:
Hanging in there and continuing to work during a
worsening disability ought not mean that a Social Security claimant is disqualified from benefits. Instead, one would expect a consistent work record to be neutral, if not positive evidence. But the Administrative Law Judge (ALJ) did not heed that intuition here.
Kristin Collard applied for disability-insurance benefits based on diagnoses of fibromyalgia, migraines, and
degenerative disc disease. Collard testified to immense pain from those impairments and ongoing limitations that prevent her from working years later. The ALJ, however, determined that Collard was not “disabled” within the meaning of the Social Security Act. To reach that determination, the ALJ discredited Collard’s testimony, and compatible third-party evidence from relatives and a coworker, in part because Collard had worked a strenuous job before her conditions
COLLARD V. BISIGNANO 5
worsened. Because the ALJ’s reasoning was error, we
remand for further reevaluation of that evidence. We affirm in all other respects.
I. Background
Collard applied for Disability Insurance Benefits in
October 2016, alleging an onset of disability that June. Before then, Collard had worked for eleven years as a
hospital patient care technician. This was a strenuous job, requiring Collard to spend 12-hour shifts mostly on her feet and to frequently lift 50 pounds or more. While Collard’s impairments had developed as early as 2007, she stated that they worsened in 2016, causing her to first adjust her work schedule and ultimately to accept a voluntary layoff when she became too “weak and exhausted.”
In her application, Collard asserted impairments from
fibromyalgia, migraines, and degenerative disc disease. Collard raises arguments as to the first and third conditions, so we address only those. Relevant here is the following evidence:
• Collard’s testimony that she suffered
constant muscle “weak[ness],” difficulty
with fine manipulation, and frequent
flare-ups that prevented her from
standing more than “15 minutes,”
6 COLLARD V. BISIGNANO
spending long periods sitting in the same
position, and lifting more than 10 pounds.
• Testimony from Collard’s husband,
niece, coworker, and sister-in-law
attesting to many of the same limitations.
• Opinions from Collard’s treating
physicians Drs. Paul Brown and Kristin
Brecht attesting to Collard’s lifting- and
mobility-based constraints.
In addition, an agency physician examined Collard and
opined that she could tolerate sitting for most of a workday with occasional breaks.
The ALJ found that Collard was not “disabled” within
the meaning of the Social Security Act. The ALJ first
concluded that Collard was eligible for benefits and assessed several “severe impairments” including fibromyalgia, and a shoulder condition. Despite those impairments, the ALJ found that Collard could perform “light work”—involving “sitting most of the time” and lifting no more than “20 pounds”—and that there were such jobs available in the national economy.
To reach that conclusion, the ALJ made several
decisions about how to weigh Collard’s evidence. The ALJ gave reduced weight to Collard’s testimony, finding it was inconsistent with her long history of work, with Collard’s daily activities, with her treatment history, and with certain objective medical evidence. Along the same lines, the ALJ discounted the lay testimony because it reflected the “same allegations made by [Collard]” that the ALJ had already addressed. Finally, the ALJ gave “little weight” to the
COLLARD V. BISIGNANO 7
opinions of Drs. Brown and Brecht, reasoning those opinions were conclusory and lacked support.
The district court affirmed the ALJ’s determination, and this appeal followed.
II. Analysis
We review de novo a district court order affirming an
ALJ’s denial of benefits and for substantial evidence the underlying ALJ decision. Woods v. Kijakazi, 32 F.4th 785, 788 (9th Cir. 2022). Under the latter standard, we ask whether the record “contains ‘sufficient evidence’ to support the agency’s factual determinations.” Biestek v. Berryhill, 587 U.S. 97, 102 (2019) (quoting Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 (1938)) (citation modified). But because the ALJ’s evaluation of significant evidence was “not supported” by the record or was “based upon legal error,” see Luther v. Berryhill, 891 F.3d 872, 875 (9th Cir. 2018), we affirm in part, reverse in part, and remand for a new disability determination.
A. Subjective Symptom Testimony
We begin with Collard’s own testimony about her
symptoms. Because the ALJ assessed severe impairments
and did not find any “evidence of malingering,” the ALJ could only discount Collard’s testimony “based on clear and convincing reasons.” Smartt v. Kijakazi, 53 F.4th 489, 497 (9th Cir. 2022) (quoting Carmickle v. Comm’r, Soc. Sec. Admin., 533 F.3d 1155, 1160 (9th Cir. 2008) (citation
modified)). That is the “most demanding” standard under our precedent. Moore v. Comm’r of Soc. Sec. Admin., 278 F.3d 920, 924 (9th Cir. 2002). It requires that the ALJ provide more than “general findings,” identify “what
testimony is not credible,” and identify what evidence 8 COLLARD V. BISIGNANO
“undermines” that testimony. Lester v. Chater, 81 F.3d 821, 834 (9th Cir. 1995). The ALJ must also state a rationale “clear enough that it has the power to convince.” Smartt, 53 F.4th at 499.
The government points to four supporting rationales in the ALJ’s order: (1) that Collard’s work history undermines her complaints; (2) that Collard’s testimony conflicts with her day-to-day activities; (3) that her statements conflict with treatment records; and (4) that her testimony conflicts with other objective medical evidence. For the following reasons, we conclude those rationales are lacking under the clear and convincing standard.
Starting off, the ALJ erred in his characterization of Collard’s work history. The ALJ pointed to Collard having worked a high-strain job even after developing her
impairments to infer that her condition is not disabling. But a past ability to work through pain can be entirely consistent with testimony of worsening impairments that become
disabling over time.
Unlike Ahearn v. Saul, on which the government relies, this is not a case where the claimant experienced consistent symptoms “for much of his life” and “was terminated due to funding issues.” 988 F.3d 1111, 1117 (9th Cir. 2021). Work history may be relevant where a claimant’s symptoms
remain constant over time. But work is not dispositive of disability where a claimant alleges a change in
circumstances like Collard does here. Cf. Lester, 81 F.3d at 827 (relying on similar logic to conclude that an ALJ is “preclude[d]” from relying on an earlier disability
determination where claimant asserts “[a]n increase in the severity of [an] impairment”). Nor is it accurate to state, as the ALJ did, that Collard was “laid off.” The portion of the
COLLARD V. BISIGNANO 9
record cited by the ALJ clarifies that Collard accepted a voluntary layoff because of her impairments in lieu of taking medical leave. No part of the record suggests she was, or would have been, fired for another reason. The
circumstances in Ahearn thus do not support the
government’s argument.
With that context, we cannot agree that the absence of acute symptoms in years past constitutes a “clear and
convincing reason” to reject Collard’s testimony now. See Smolen v. Chater, 80 F.3d 1273, 1285 (9th Cir. 1996). Were it otherwise, claimants with long-term impairments like Collard would be caught between a rock and hard place. ALJs often rely on inconsistent or spotty work history in denying claims, reasoning that something other than a
disability prevents a claimant from working. Thomas v. Barnhart, 278 F.3d 947, 959 (9th Cir. 2002). But if a
consistent work history enables an inference of no disability, all claimants would be disadvantaged from the outset.
Insurance claimants should not be made to face that logical trap. Where, as here, Collard testified to worsening
symptoms, the ALJ must point to more than the mere fact that a claimant previously held a job to discredit her testimony.
The ALJ similarly erred in analyzing Collard’s activities of daily living. “The mere fact that a plaintiff has carried on certain daily activities,” of course, does not “detract from her credibility.” Benecke v. Barnhart, 379 F.3d 587, 594 (9th Cir. 2004) (quoting Vertigan v. Halter, 260 F.3d 1044, 1050 (9th Cir. 2001)). Instead, to discredit testimony on that basis, the ALJ had to connect Collard’s activities to the conclusion that she could spend “a substantial part of h[er] day” engaged in “pursuits . . . transferable to a work setting.” Ghanim v. Colvin, 763 F.3d 1154, 1165 (9th Cir. 2014) (quoting Orn v. 10 COLLARD V. BISIGNANO
Astrue, 495 F.3d 625, 639 (9th Cir. 2007)). The ALJ,
however, merely “listed a number of activities” that Collard performs, while “fail[ing] to acknowledge” aspects of the daily activity evidence consistent with Collard’s claim. See Revels v. Berryhill, 874 F.3d 648, 667–68 (9th Cir. 2017). That analysis thus “failed to meet the high bar” set out by our caselaw. Id.
Substantial evidence does not support the ALJ’s next
rationale—that Collard could “effective[ly]” treat her symptoms with medication. Collard tried several forms of treatment—from therapies, to cortisone shots, to opioids— with none providing lasting relief. Even on opioids, the ALJ acknowledged that Collard experienced only “partial pain control,” with providers finding that she still experienced physical limitations. In the context of Collard’s allegations, that evidence at best points to “some improvement,” but it does not warrant an inference that Collard was overstating her pain. See Holohan v. Massanari, 246 F.3d 1195, 1205 (9th Cir. 2001).
The ALJ’s reliance on Collard’s attendance at doctor’s appointments, and her “unremarkable” appearance and test results at those appointments is misplaced. As to attendance specifically, there are multiple record examples of missed or “rescheduled” appointments. More generally, the ALJ’s
reasoning is inconsistent with our precedent regarding testimony of fibromyalgia-related pain—which forms the core of Collard’s complaint. Because the condition is
“diagnosed entirely on the basis of patients’ reports of pain and other symptoms,” Benecke v. Barnhart, 379 F.3d at 590, it is error to discount subjective testimony solely because a claimant exhibited little to no abnormality at doctor’s appointments or in objective testing. Revels, 874 F.3d at 666; see also Benecke, 379 F.3d at 594 (holding ALJ “erred by
COLLARD V. BISIGNANO 11
‘effectively requiring objective evidence for a disease that eludes such measurement’” (quoting Green-Younger v.
Barnhart, 335 F.3d 99, 108 (2d Cir. 2003)) (citation
modified)). To be sure, objective evidence can undermine claims of fibromyalgia where it contradicts alleged
limitations. See Nadon v. Bisignano, 145 F.4th 1133, 1137 (9th Cir. 2025). But the ALJ did not articulate any such inconsistencies with the required level of specificity.
Accordingly, the ALJ erred in its evaluation of Collard’s subjective testimony. Because another ALJ may reach a
different disability determination when “fully crediting that testimony,” the ALJ’s errors are prejudicial and must be reevaluated in further proceedings. See Robbins v. Soc. Sec. Admin., 466 F.3d 880, 885 (9th Cir. 2006) (citation
modified).
B. Lay Testimony
The ALJ also erred with respect to the third-party
evidence from Collard’s husband, niece, co-worker, and sister-in-law.1 The ALJ concluded that evidence paralleled Collard’s own testimony and the ALJ relied on the same rationale for discrediting it. Although the ALJ’s evaluation of lay evidence is held to a lower standard than subjective testimony, it was improper to reject that testimony simply because it was consistent with Collard’s testimony given the errors identified above. See Lewis v. Apfel, 236 F.3d 503, 511 (9th Cir. 2001) (noting ALJ need only provide
“germane” reasons for discrediting lay testimony). The government makes no argument that the standard of review
1
To the extent Collard’s district court briefing did not develop this argument, we exercise our discretion to reach it because the district court ruled on the merits and the issue is now fully briefed. See Kaufmann v. Kijakazi, 32 F.4th 843, 847 (9th Cir. 2022).
12 COLLARD V. BISIGNANO
warrants a different result. The government suggests we may affirm because the ALJ identified certain “nonmedical
source statements” that appear vague or conclusory. Yet, mindful of our guidance that we may not affirm “simply by isolating a ‘specific quantum of supporting evidence,’” this evidence must be reevaluated as well. See Robbins, 466 F.3d at 882 (quoting Hammock v. Bowen, 879 F.2d 498, 501 (9th Cir.1989)).
C. Treating Physician Opinions
With respect to the medical opinion evidence, we
conclude that substantial evidence supports the ALJ’s
determination. Where, as here, treating-physician opinions are contradicted by other medical opinions, the ALJ may reject the opinions for “specific and legitimate reasons.” Bayliss v. Barnhart, 427 F.3d 1211, 1216 (9th Cir. 2005). The ALJ provided multiple grounds applicable to both Dr. Brown and Dr. Brecht.
Dr. Brown provided three relevant opinions: (1) a short statement that Collard was “completely and totally
disabled,” even “in a sedentary position”; (2) a check-box assessment that Collard could lift no more than 10 pounds and walk no more than two hours per day; and (3) a longer assessment that Collard’s impairments prevented her from completing “repetitive tasks, [like] prolonged sitting and walking.”
The ALJ had sufficient reasons to reject each one.
Collard does not dispute that the first opinion is entitled to no weight because it is conclusory. And the ALJ explained that the other two opinions were deficient because they relied on a finding that Collard suffered from “inflammatory polyarthritis”—a condition the ALJ determined lacked
support. Collard does not challenge the evaluation of the
COLLARD V. BISIGNANO 13
arthritis diagnosis. Collard’s suggestion that the ALJ should have contacted Dr. Brown to clarify his opinion has no traction. We have rejected arguments that an “ALJ should have recontacted” a physician where the ALJ found the
physician’s descriptions “inadequate for determining” a claimant’s functional capacity. See Ford v. Saul, 950 F.3d 1141, 1156 (9th Cir. 2020).
The ALJ also presented legitimate reasons for rejecting Dr. Brecht’s assessment that Collard could lift no more than five pounds, sit for no more than 30 minutes, and the like. The ALJ viewed several of Dr. Brecht’s statements as
conclusory and noted that none cited either objective
evidence or findings in the longitudinal record. Notably, when asked what support there was for “the existence of [Collard’s] conditions,” Dr. Brecht provided no explanation at all. The ALJ had no obligation to credit such opinions. Bayliss, 427 F.3d at 1216. We therefore affirm as to the treatment provider evidence.
AFFIRMED in PART; REVERSED in PART; and
REMANDED for further proceedings. The parties shall
bear their own costs.