Saturday, September 14, 2019

The ALJ Says that the COSS Rejects the O*NET -- What do I do Now?

Our friends at SkillTran post a memorandum from SSA stating that the O*NET is not suitable for use in disability litigation.  The memorandum is in response to a request for information (RFI) dated June 4, 2004.  The question is whether this memorandum licenses the adjudicator to reject O*NET-based information without any other reason. 

As with any question, we start with the law - the statute and regulations.  The statute is silent as to the sources of vocational information.  The regulations provide for administrative notice of reliable job information from governmental and private resources.  Let's assume that in a memorandum that is not published in HALLEX or POMS that the COSS can describe a data set as not reliable.  Let's also assume that this unavailable memorandum survives the Kisor v. Wilkie test for deference or that a court would find it persuasive under Skidmore

The response to the RFI is clear:
SSA has determined through contracted research and through further investigation that O*NET, as it currently exists, cannot be used in SSA’s disability determination process (see Bibliography of research and investigations following).
The critical phrase from the 2004 response to the RFI is as it currently exists.  The question is whether the O*NET exists today as it did in 2004.  The O*NET has updated vocational data over every occupation for work context -- the data that Social Security representative are most interested.  Three occupations were last updated in 2004 and nine in 2005.  The other 1,004 have been updated between 2006 and 2019.  The O*NET does not exist today as it existed in 2004. 

The O*NET changed its taxonomy to version 10.0 in 2006.  The 2019 data set reflects version 24.0.  The O*NET does not exist today as it existed in 2004.  The O*NET now describes:
Occupational Requirements
A comprehensive set of variables or detailed elements that describe what various occupations require.
This domain includes information about typical activities required across occupations. Task information is often too specific to describe an occupation or occupational group. The O*NET approach is to identify generalized work activities (GWAs) and detailed work activities (DWAs) to summarize the broad and more specific types of job behaviors and tasks that may be performed within multiple occupations. Using this framework makes it possible to use a single set of descriptors to describe many occupations. Contextual variables such as the physical, social, or structural context of work that may impose specific demands on the worker or activities are also included in this section.
Work activities that are common across a very large number of occupations. They are performed in almost all job families and industries.
Work activities that are common across many occupations. They are performed in many job families and industries.
Specific work activities that are performed across a small to moderate number of occupations within a job family.
Characteristics of the organization that influence how people do their work
Physical and social factors that influence the nature of work
It is the work context description.  The physical and social factors that influence work are important to the disability analysis.  The data presentation describing the percentages of work within an occupation that require constant contact with others or very important work function of working with a group or team are exactly the kind of data that resolves holes in the DOT. 

And that brings the question back full circle to the Kisor question about deference and the Skidmore question about respect to the ability to persuade.  The Department of Labor is the agency that has the expertise to determine which of its publications constitutes the best source of information about the existence and requirements of work in the national economy.  Labor states about the DOT:
The O*Net is now the primary source of occupational information. It is sponsored by ETA through a grant to the North Carolina Department of Commerce. Thus, if you are looking for current occupational information you should use the O*Net.
Whether the DOT, SCO, O*NET, or ORS answers every question about work represents a different question than whether any of those sources answer one or more questions about the number of jobs, the erosion of the occupational base, and the requirements of work.  Which takes us back to the first Kisor question, the regulation is not ambiguous.  The COSS takes administrative notice of reliable governmental data.  The only question is whether the O*NET is reliable and if so whether is answers the specific question that we have today. 

Friday, September 13, 2019

Answers to Questions from the New Orleans NOSSCR Conference

20           An example of a cross or rebuttal to “no fast pace production or piece rate quota” ?
The O*NET OnLine describes pace requirement in the work setting.  The data for cashier, for example:
Pace Determined by Speed of Equipment — How important is it to this job that the pace is determined by the speed of equipment or machinery? (This does not refer to keeping busy at all times on this job.)
Extremely important
Very important
Fairly important
Not important at all

So cashiers are not pace determined by speed of equipment as extremely important or very important. 

14           Can you please post your slides on your blog?

I posted from:

9             But if you ask the VW, they will say the cashier slides the case of water across the scanner, they will say the VW is not lifting that weight.

The BLS (ORS) states that:
Series ID: ORUP1000066700000663
Not seasonally adjusted
Series Title: % of cashiers; strength is medium work
Requirement: Physical Demands
Occupation: Cashiers
Estimate: strength is medium work

What is your basis for disagreeing with that finding. 

9             What is the citation for SVP 1 is 4hours demonstration?

5             How do counter a vocational witness that you know is not being forthcoming or honest?

I just want an answer to the question that the Supreme Court directs me to ask:  do you have a reliable methodology for extrapolating your local experience to the national economy?

5             In my experience, if my cross of the VE possibly works, the ALJ finds on another issue, so it is difficult to tell how my cross was received. How can you tell if what you are doing is winning more hearings? Are you just preserving the record for appeal?

I can do this all day long.  If the claimant has significant limitations, you can erase any occupation.  Just persevere. 

4             The payment of MEs/VEs is available through FOIA. How do you suggest we use that as a bias argument?

I don’t have to prove bias.  I have to prove that the testimony is feeble and contradicted.  Stick to what the claimant must prove. 

4             the answer VEs give to cashier lifting a case of water is that it is left in the cast & they don't have to lift it.

The BLS states that cashiers engage in medium work in most jobs, do you have a statistical basis to show that the BLS is wrong?

4             Please zoom in!

Get your free 30 days at 

4             Has SSA formally accepted the ORS data for use in administrative hearings?

ORS provides information about the requirements of work in the national economy conducted under an interagency agreement with the Social Security Administration (SSA).

2             How do you know that JBP uses an equal distribution method? Where does it say that?

2             With the cashier issue and 60% being medium, couldn't the VE just say they could do one of the 40% that is not medium and that's still a significant number of jobs?

The 60% medium is a response to “as generally performed.”  AGP, cashiers are medium. 
If your client can perform the full range of light work, including stand/walk 7+ hours per day, the client is not disabled.  To prove disability for the sub-55 claimant, the record needs to prove a stand/walk limitation, an interaction with others limitation, or something else. 

2             Do you need to buy a subscription to Occu Collect to get all the screens you showed? reproduces in one location the DOT, SCO, O*NET work context, O*NET work abilities, and the ORS.  You can go to four different websites and cobble the data together.  I have done it.  It takes about two hours per occupation.  If you use occucollect, it takes about 20 minutes per occupation once you have done it a handful of time.  The reports that are not available in a report format are the specialty reports that I did not cover, the Sedentary Compendium in the store, and the O*NET education training and experience report.   

2             When a VW says, "this is how the job is generally performed," when we ask about reliable method for extrapolating local experience to national economy, the VW inevitably says it is discussed at conferences, their VW chat groups, etc... how do we knock it out at that point?

What was the reliable methods used in the group?  Did you ask your chat club members about their reliable methods? 
Remember that 100 scientists got together and told the world that Einstein was wrong about relativity. 

2             If you don’t have a letter from the treating physician and the CE puts the claimant at medium. How do you come up with hypotheticals to the VE? How do we assume the amount off task in a day and/or any supportive hypothetical

The CE will typically describe a limitation to six hours of standing/walking.  Medium work requires seven or eight hours of standing/walking.  The CE might have described a limited range of motion for forward flexion.  This is sometimes a fruitful avenue.  Look for gait disturbance in the record, that might give you more. 

2             Is there a Federal Court case that you can cite where the court said the equal distribution method is not a valid way to calculate job numbers?

Chavez v. Berryhill, 895 F.3d 962 (2018)
Brault v. SSA Comm’r, 683 F.3d 443 (2012)

2             Which website was it in which the BLS writes that they maintain DOT only because of SSA, and that if you want reliable current data, go to the O*Net? I'd love to have that webpage so I can print it and submit in every case.

1             What is the BLS?

Bureau of Labor Statistics

1             Good case law on VE - DOT conflict in 11th Circuit (Washington) & 4th. Extend to VE conflict with O-NET etc?

Wait for it.  This depends on Kisor v. Wilkie application to SSR 00-4p.  See the blog at:

Sunday, September 1, 2019

What is Your Reliable Methodology for Extrapolating Your Experience?

That is the question that we should all ask in the course of cross-examining vocational experts in Social Security disability hearing.  What is your reliable methodology for extrapolating your local experience to the national economy?  The reason that this question should be deeply ingrained into the practice of every representative that does this kind of work is simple, that is the question that the Supreme Court directs us to ask.  Biestek v. Berryhill says:
Now say that she testifies about the approximate number of various sedentary jobs an applicant for benefits could perform. She explains that she arrived at her figures by surveying a range of representative employers; amassing specific information about their labor needs and employment of people with disabilities; and extrapolating those findings to the national economy by means of a well-accepted methodology.
(Emphasis added). 

The well-accepted methodology could take the form of using Job Browser Pro because the vocational expert could say that it is accepted by the community of vocational experts as reliable.  The  First Circuit accepted that testimony in Purdy v. Berryhill.  Well-accepted should not include the equal distribution method of calculating job numbers, at least not when conflicting evidence is present.  The Seventh Circuit suggested that result in Chavez v. Berryhill.  Here is what Biestek says about the Chavez line of questioning:
Even without specific data, an applicant may probe the strength of testimony by asking an expert about (for example) her sources and methods—where she got the information at issue and how she analyzed it and derived her conclusions. See, e.g., Chavez v. Berryhill, 895 F.3d 962, 969-970 (CA7 2018).
Most vocational experts either use Job Browser Pro, the equal distribution method (with or without knowing it), or have no discernible methodology at all.  This results in identifying the entire SOC/OES number of jobs as applicable to a single DOT code.  That methodology is worse by equal distribution; it is just lazy and ignorant. 

Here is a working list after hearing the vocational expert identify jobs and job numbers:

  1. Confirm the DOT code.
  2. Obtain the SOC/OES code.
  3. Obtain the number of jobs in the SOC/OES code (most will not know).
  4. Obtain the number of DOT codes within the SOC/OES code (most will not know).  
  5. Ask for the reliable and well-accepted methodology for extrapolating the local experience to the national economy.  
With that information, we can disassemble the vocational expert testimony post-hearing.  If we use the O*NET and the Occupational Requirements Survey during the hearing (use, we can ask after an offer of proof of what the Department of Labor says:
6.  Who as greater resources for accumulating job requirements data (or job numbers data) in the national economy, you or the Department of Labor?
We still have the industry-occupation matrix used by Job Browser Pro, the BLS employment projections, and the Occupational Employment Statistics.  If that route is necessary ask:
7. In what industries, by NAICS code and name, does this occupation work?
We must have the first five.  Number six is the concession by which we hang the witness testimony.  Number seven tightens the knot.  We must use the data from Labor for the sake of our client's benefit entitlement.