Opinion
21-cv-379-jdp
04-03-2023
PAUL STAPLES, Plaintiff, v. KILOLO KIJAKAZI, Acting Commissioner of the Social Security Administration, Defendant[1].
OPINION AND ORDER
JAMES D. PETERSON, DISTRICT JUDGE
But Staples' argument about the vocational expert's job numbers raises a thorny issue that has bedeviled social security hearings and the courts for several years. The vocational expert derived her job numbers from Job Browser Pro, database software provided by a private company, SkillTRAN. Job Browser Pro might well be, as SkillTRAN contends, a vast improvement over historical methods of calculating job numbers. But its reliability remains controversial, and when a claimant challenges job numbers based on it, the ALJ must address
Plaintiff Paul Staples seeks judicial review of a final decision of the Commissioner of the Social Security Administration, finding him not disabled under the Social Security Act. He contends that administrative law judge (ALJ) Joseph Jacobson erred by (1) accepting the vocational expert's testimony about the number of jobs Staples could perform without ensuring its reliability; and (2) rejecting Staples' subjective complaints, including his allegation that he required a cane for ambulation. Staples' second argument is not persuasive: the ALJ gave good reasons for rejecting Staples' subjective complaints. those challenges. The ALJ did not do so here. Accordingly, I will remand this case to the agency for further proceedings.
BACKGROUND
A. Administrative proceedings
Staples applied for disability insurance benefits on March 7, 2016. He alleged that he had been disabled since January 1, 2012, but he later amended that date to March 7, 2016. As of that date, Staples was 47 years old. He had a fairly extensive history of chronic neck, back and knee pain stemming from a series of injuries and surgeries, including a broken neck sustained in a head-on collision in 2001. He completed the 11th grade and had past work as a shipping and receiving clerk. He stopped working in 2011. When applying for disability benefits, Staples alleged that he had continued to work for many years in spite of his injuries and surgeries by relying on high doses of oxycodone, and that in doing so, he pushed himself too far and can no longer function like he used to. R. 183.
Record cites are to the administrative transcript, located at Dkt. 10.
The local disability agency denied his application, and Staples appealed. In October 2018, he appeared with representation at an administrative hearing. In January 2019, the administrative law judge (ALJ) issued a decision finding him not disabled at any time from his alleged onset date of March 7, 2016, through his date last insured of December 31, 2016. After the Appeals Council denied his request for review, Staples filed suit in this court. Staples v. Saul, 19-cv-925-bbc. On April 21, 2020, the court granted a joint motion by the parties to remand the case to the commissioner for further proceedings. Dkt. 16 in the ‘925 case.
After remand, a second hearing was held telephonically on November 18, 2020, before ALJ Joseph Jacobson. Staples, who was represented by counsel, testified, as did Emily Veith, an impartial vocational expert. The ALJ presented a hypothetical to Veith that asked her to assume an individual of Staples' age, education, and work experience who was limited to sedentary work, with: no climbing of ladders, ropes or scaffolds; occasional use of ramps or stairs; and occasional stooping, crouching, kneeling or crawling. R. 629. Veith testified that such an individual would be unable to perform Staples' past work but could perform the following jobs, as listed in the Dictionary of Occupational Titles: (1) telephone solicitor, DOT No. 299.357-014, 35,000 jobs nationally; (2) appointment clerk, DOT No. 237.367-010, 72,000 jobs nationally; and (3) circuit board assembler, DOT No. 726.684-110, 1,000 jobs nationally. R. 629-30. Veith explained that although the jobs of telephone solicitor and appointment clerk were classified as semi-skilled jobs according to the DOT, she was aware from her experience and the Department of Labor's Occupational Requirement Survey (ORS) that approximately half the jobs nationally are unskilled, and she had reduced the job numbers accordingly. R. 630. The ALJ then asked Veith whether her numbers would change if the individual had the option to alternate between sitting and standing throughout the workday while remaining on task. Relying again on the ORS, Veith said that about half the telephone solicitor jobs, or 17,000, and 60 percent of the appointment clerk jobs, or 43,000, could be performed with a sit-stand option. As for the circuit board assembler, Veith testified that the ORS didn't address assemblers, but in her experience, about half of the jobs allowed for sitting or standing, for a total of 500. R. 630-31. In response to further questioning by the ALJ, Veith added the job of hand packager to the list of available jobs, stating that although the DOT classified the job as medium and unskilled, according to the ORS and SkillTRAN there were sedentary hand packager jobs available nationally, of which four percent-or 3,000-allowed for a sit/stand option. R. 631. Finally, Veith testified that if the individual had to use a cane for uneven terrain and ambulating long distances, her answers would not change, but if the person had to use the cane at all times while standing, all of the jobs would be eliminated. R. 632.
On cross-examination, Staples' counsel asked Veith where she got her job numbers from before applying any reductions based on the ALJ's hypothetical. Veith responded that the source of her numbers was SkillTRAN's Job Browser Pro, a computer program that analyzed information from the United States Department of Labor, Bureau of Labor Statistics. After counsel asked Veith how SkillTRAN collects its data, the following exchange took place:
A: Well, like I said, they rely on the information that the U.S. Department of Labor, Bureau of Labor Statistics collects from employers and their estimates about jobs that exists in the nation economy under SOC codes. But because there is no other source that breaks job estimates down to DOT number only, the site at SkillTRAN and Job Browser Pro, that's why I rely on this source. And in terms of their methodology, that is something that's published on their website that's publicly available.
Q: I'm sorry, did you say SkillTRAN's methodology's is published on their website?
A: Yes.
Q. Okay. And which, so, have you actually seen any of the data that SkillTRAN uses?
A: Yes and no. I guess, I have always looked at the original data that the U.S. Department of Labor publishes based on the SOC codes, so I rely on that and I've looked at that for testimony in this and other settings but because this setting requires estimates by Dictionary of Occupational Title number, I need to rely on a different resource, which is SkillTRAN.
Q: Right, but does SkillTRAN rely on census data and employer related occupational statistics?
A: Yes. They rely on those other sources as well, which is all publicly available.
Q: But have you actually seen the data that SkillTRAN collected from these sources?
A: I mean, yes. Like I said, anybody can go look at that data but how they analyze it - if you're asking if I've seen kind of their steps along the way of how they analyze it to come up with their final estimates, no, I haven't seen the processed data....
A: As I mentioned before, SkillTRAN has developed a method to estimate those numbers by DOT number only and that method is publicly available. They are better to speak to that than I am.
Q: Are you feel, do you feel confident relying on their methodology?
A: SkillTRAN and Job Browser Pro are relied upon in our field regularly by my peers and colleagues. It's a well-used resource, so yes, I do feel confident in relying upon their numbers.R. 636-38.
After the hearing, Staples' counsel submitted a detailed post-hearing brief, challenging the reliability of Veith's job estimates. R. 845-50. He raised three objections: (1) Veith's baseline job numbers were not reliable because Veith could not explain how SkillTRAN arrived at its numbers for DOT-titled jobs; (2) Veith was incorrect about the skill level required of the jobs; and (3) the total number of jobs (63,500) was not significant. Staples also raised substantive objections to SkillTRAN's methodology, arguing that its reliance on NAICS codes resulted in double counting of jobs. AR 849.
B. ALJ's decision
On December 9, 2020, ALJ Jacobson issued a written decision denying Staples' request for benefits, following the five-step sequential evaluation process set forth in the regulations.
R. 592-604. At step one, he concluded that Staples hadn't engaged in substantial gainful activity since his alleged onset date. At step two, he found that Staples had osteoarthritis of the right knee and degenerative disc disease of the cervical and lumbar spine, impairments that were severe. But at step three, the ALJ found that Staples' impairments were not severe enough on their own or in combination to meet the severity requirements of a listed impairment. R. 597.
The ALJ found that Staples had the residual functional capacity to do sedentary work with some additional postural limitations, as well as the need for a sit/stand option and the ability to use a cane on uneven terrain or for prolonged ambulation. In reaching this conclusion, the ALJ noted that Staples had not received the type of medical treatment one would expect for a totally disabled individual, there were some gaps in the treatment history, and Staples had generally refused any treatment for symptoms, other than narcotic pain medication. R. 600. In addition, there were no opinions from any treating or examining physician indicating that Staples was disabled or even that he had specific functional limitations. Id. The ALJ specifically considered whether Staples required the use of a cane at all times, but he did not find enough evidence in the medical record to support that limitation. R 601.
At step four, the ALJ found that Staples' limitations prevented him from performing his past relevant work. But at step five, relying on Veith's testimony, the ALJ found that there was a significant number of other jobs in the national economy that Staples could perform, namely telephone solicitor, appointment clerk, circuit board assembler, and hand packager. R. 603. The ALJ noted Staples' counsel's objection to Veith's testimony regarding her job number estimates, but the ALJ overruled the objection, concluding that requiring Veith to explain in detail how SkillTRAN's software calculated job numbers imposed an “impossible burden” on the VE. R. 603 (citing Liskowitz v. Astrue, 559 F.3d 736, 743-44 (7th Cir. 2009)). In addition, the ALJ noted that the detailed methodology used by SkillTRAN's Job Browser Pro was publicly available on the developer's website. Noting that Veith had professional knowledge and experience in job placement, the ALJ concluded that Veith's job information was reliable. R. 604. He did not address any of the substantive objections Veith raised to the SkillTRAN methodology.
ANALYSIS
On appeal, this court's role is to review the ALJ's decision for legal errors and to determine whether the decision is supported by substantial evidence. See Martin v. Saul, 950 F.3d 369, 373 (7th Cir. 2020). The substantial evidence standard is not high and requires only “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Id. But the ALJ's decision must identify the relevant evidence and build a “logical bridge” between that evidence and the final determination. Moon v. Colvin, 763 F.3d 718, 721 (7th Cir. 2014).
Staples contends that the ALJ erred by: (1) accepting the VE's job numbers; and (2) rejecting Staples' subjective complaints, including his allegation that he required a cane for ambulation.
A. The VE's job numbers
At step five of the sequential evaluation for adjudicating disability claims, it becomes the Commissioner's burden to demonstrate with substantial evidence that the plaintiff can make a vocational adjustment to work existing in the national economy, and that those jobs exist in significant numbers. Chavez v. Berryhill, 895 F.3d 962, 964 (7th Cir. 2018). The Seventh Circuit recently summarized how the substantial evidence standard applies to a VE's job numbers:
In the context of job-number estimates, substantial evidence requires the ALJ to ensure that the vocational expert's estimate is the product of a reliable methodology. A methodology is reliable when it is based on well-accepted sources and the vocational expert explains her methodology cogently and thoroughly. And when, as here, the claimant challenges the job-number estimate, the ALJ must compel the vocational expert to offer a reasoned and principled explanation of the methodology she used to produce the estimate. The expert's explanation must be sufficient to instill some confidence that the estimate was not conjured out of whole cloth.Ruenger v. Kijakazi, 23 F.4th 760, 763 (7th Cir. 2022) (internal quotations and citations omitted). The Supreme Court has explained that a reviewing court's inquiry into the reliability of VE testimony “is case-by-case” and should “take[ ] into account all features of the vocational expert's testimony, as well as the rest of the administrative record.” Biestek v. Berryhill, 139 S.Ct. 1148, 1157 (2019).
At issue in this case is Veith's use of SkillTRAN's Job Browser Pro software as a source of her baseline job estimates. Although this challenge has arisen with increasing frequency in social security disability cases since Job Browser Pro was released in 2008, neither the Seventh Circuit nor this court has conclusively determined that the SkillTRAN software employs reliable methods. To the contrary, in Westendorf v. Saul, No. 19-cv-1019-jdp, 2020 WL 4381991 (W.D. Wis. July 31, 2020), this court reviewed SkillTRAN's 2008 methodology document and found it to be a “black box”: not only was it out of date, but it did not “fully explain how it uses the OES data to produce estimates of the number of positions available within jobs defined in the DOT.” Id. at *3. The document acknowledged that Job Browser Pro was a new “disruptive” approach that had not been thoroughly tested or widely accepted. Id.
The court also noted that the ALJ did not address Westendorf's substantive objections to the software, “nor did the ALJ actually evaluate the method as described in the documentation.” Id. Ultimately, the court acknowledged that “Job Browser Pro might be a useful tool, but the VE needs to be able to explain how Job Browser Pro makes its job number estimates, how she used the software to generate her own estimates and why she believes those estimates are reliable.” Id. at *4.
Veith did not do that here. When asked how SkillTRAN made its job estimates, Veith deflected, stating only that it was based on government job data and that the methodology was publicly available on its website. Veith also did not explain how she used the software to come up with her baseline job numbers. As for why she believed SkillTRAN's estimates were reliable, Veith said it was a source regularly relied on by her peers and colleagues. Finally, as in Westendorf, the ALJ did not address Staples' substantive objections to the software, nor did he actually evaluate the method as described on SkillTRAN's website.
Westendorf was decided three years ago and the court considered an outdated version of SkillTRAN's methodology document. The updated version, published in October 2020, goes further to explain the method the software uses to estimate DOT employment numbers from the OES groupings and explains steps the company has taken “to refine its technology to stand up to a Daubert challenge.” But the acting commissioner does not address the updates, describe SkillTRAN's methodology, or ask the court to revisit the question whether the methodology is reliable. So I decline to do so.
Available at https://skilltran.com/pubs/DOTempnum_method.pdf.
Instead, the acting commissioner argues that Westendorf is distinguishable because Veith said she used other sources such as the ORS and her experience to adjust the numbers she derived from Job Browser Pro to fit the ALJ's hypotheticals. But this does not answer the question whether the numbers Veith obtained from Job Browser Pro in the first place were reliable, which is the focus of Staples' challenge. The acting commissioner also relies on Kenealy v. Saul, No. 19-cv-40-jdp, 2019 WL 6463840 (W.D. Wis. Dec. 2, 2019), which affirmed a step five finding where the VE relied on data from Job Browser Pro. But in that case, the VE began her job number estimates “by cross-referencing the Dictionary job titles with job numbers compiled by the Department of Labor and Bureau of Labor Statistics,” and then considered job-number data from Job Browser Pro. Id. at *6. The acting commissioner does not point to any evidence indicating that Veith engaged in such cross-referencing in this case. Cf. Dolan v. Kijakazi, No. 20-cv-364-wmc (W.D. Wis. July 31, 2020) (Dkt. 34, at 10) (remanding where “VE provided no discussion of other sources, including her own experience, that she relied on as a check on the Job Browser Pro's estimation of job numbers by DOT code.”). Further, as this court noted in Westendorf, the plaintiff in Kenealy “made a general reliability objection, but she didn't take specific exception to the VE's use of Job Browser Pro, so that wasn't the focus of the court's analysis.” Westendorf, 2020 WL 4381991, at *4.
The acting commissioner also contends that Westendorf is distinguishable because the ALJ “explicitly addressed” Staples' objections to the software. She is incorrect. The ALJ in this case ruled that requiring the VE to explain the methodology used by SkillTRAN was an “impossible burden” and in any event, the methodology was publicly available on the developer's website. R. 603. As in Westendorf, the ALJ did not address Staples' substantive objections to the reliability of SkillTRAN's methods, nor did he actually evaluate the methodology described on the website.
Finally, the acting commissioner cites a number of other decisions, including one from this court, affirming the ALJ's reliance on a VE's job number estimates that were derived from from Job Browser Pro without requiring the expert to explain the methodology behind the software. See, e.g., Zeman v. Kijakazi, 20-cv-591-wmc, dkt. 29; Luna v. Berryhill, No. 1:17CV354, 2018 WL 2316182, at *4 (N.D. Ind. May 21, 2018); Blankenship v. Colvin, No. 1:16CV222, 2017 WL 2805883, at *4 (N.D. Ind. June 29, 2017). Implicit in these holdings, however, is that the SkillTRAN software employs reliable methods. Although I agree that job numbers generated by Job Browser Pro may well be at least as reliable as other estimation methods used by VEs, that has not been established in this case by substantial evidence. As noted, the acting commissioner has made no attempt to do so, for example, by detailing how the software works, or by presenting evidence showing that the software has been widely accepted as reliable by vocational experts. (Veith's testimony that she found the software reliable because it was used by her peers falls short of demonstrating widespread acceptance among experts in her field.) And the ALJ didn't make these findings, either. Instead, “he accepted the VE's conclusory assurances of reliability without explanation or analysis.” Westendorf, 2020 WL 4381991, at *3. As in Westendorf, that was not enough to ensure that the VE's job estimates were the product of a reliable methodology.
Accordingly, I conclude that this case must be remanded. On remand, the ALJ will need to inquire into the VE's methods, whether those methods are reliable, and whether they were accurately applied to the case at hand before relying on the VE's conclusions. The court repeats what it said in Westendorf: “Job Browser Pro might be a useful tool, but the VE needs to be able to explain how Job Browser Pro makes its job number estimates, how she used the software to generate her own estimates, and why she believes those estimates are reliable.” Id. at *4. The ALJ must address any substantive challenges to job numbers derived from Job Browser Pro on the merits. Because the VE may well provide different testimony on remand, I conclude that it unnecessary to address Staples' other objections to the VE's testimony related to the skill level of the jobs and the sufficiency of the job numbers.
B. Evaluation of subjective complaints
Staples' remaining challenges require little discussion. First, he argues that remand is required so the ALJ can properly consider whether and to what extent he relies on a cane. But as he admits, the ALJ explicitly considered this question. The ALJ noted that although Staples testified to consistent use of a cane for years, treatment notes did not support his testimony. R. 601. Reviewing the notes, the ALJ found “few, if any treatment notes prior to the date last insured mention a cane,” while January 2018 records indicated that Staples used a cane or crutches on “bad days.” R. 601 (citing to specific medical records). Staples does not argue that the ALJ misinterpreted any treatment notes or overlooked records stating that he required the use of a cane as opposed to merely having one that he used at times. So he has failed to show that the ALJ's decision in this regard is not supported by substantial evidence.
Second, Staples argues the ALJ erred in citing his failure to seek treatment other than narcotics as a basis for rejecting his subjective complaints because the ALJ did not say what other treatment options were available to him. This is incorrect. The ALJ noted that Staples' health providers had recommended physical therapy, which Staples refused, and referred him to a pain clinic, which he did not follow through with until late 2017 or early 2018, after his date last insured. R. 599. And Staples does not raise any challenge to any of the other reasons the ALJ cited for rejecting his subjective complaints, including his activities, the objective medical record, and the opinions of the state agency physicians, who concluded he was capable of performing a range of light work. Accordingly, Staples hasn't shown that the ALJ erred in assessing his subjective complaints.
C. Conclusion
In sum, Staples hasn't shown that the ALJ erred in his assessment of his subjective complaints or his alleged need for a cane. But he has shown that the ALJ's step five finding is not supported by substantial evidence. So I am reversing the acting commissioner's decision and remanding the case to the agency, for further proceedings consistent with this opinion.
ORDER
IT IS ORDERED that the decision of the Acting Commissioner of Social Security denying Paul Staples's application for disability benefits is REVERSED and REMANDED pursuant to sentence four of 42 U.S.C. § 405(g) for further proceedings consistent with this opinion.
BY THE COURT: