What's Wrong with the OEQ -- Attacking Its Foundation and Methodology

Ah, the Occupational Employment Quarterly (OEQ), it used to be the only game in town. Ridiculed by David Traver so many years ago, superseded by Job Browser Pro (JBP), and obsoleted by the Occupational Requirements Survey (ORS) joined by the Occupational Employment and Wage Statistics (OEWS). I have labeled the OEQ statistical trash, mostly because it is. But the OEQ does have a cadre of devotees that cling to it like a plank of balsa wood adrift in a sea storm. What's wrong with the OEQ? I am so glad that you asked. 

The OEQ is well-known for its use of the equal distribution method of calculating job numbers. The Seventh Circuit labels that methodology as preposterous. Alaura v. Colvin. As Kevin Liebkemann points out in Job Incidence Numbers in Social Security Disability Claims: ACase Study and Analysis, even SkillTRAN publicly derides the OEQ as using a preposterous equal distribution methodology in its SkillTRAN Process for Estimating Employment Numbers (citing the later decision in Hill v. Colvin). 

In Woods v. Bisignano, the Ninth Circuit affirmed the vocational witness's patent use of the OEQ using the equal distribution method. Judge Nelson concurring states that a categorical rule excluding testimony based on the equal distribution method runs afoul of Biestek v. Berryhill. Woods is wrong and so is Judge Nelson. Job numbers in Standard Occupational Classification (SOC) groups with very few DOT codes leads to results that are absurd. Consider telemarketers -- one sedentary semi-skilled DOT code. The Occupational Outlook Handbook disagrees. Telemarketers typically have short-term on-the-job training. The 2018 ORS dataset describes telemarketers as having up to 1 month of training in 50.3% of jobs. Equal distribution should require an explanation -- every single time it is used. The Seventh Circuit is right.

But let us assume that Judge Nelson is right, the equal distribution method is not so inherently flawed that there do exist some circumstances where it might be reasonable to use it. Let's play along. Ask the witness this question:

What is the data source that US Publishing uses to estimate job numbers stated in the OEQ?

The first page of the OEQ II  3.2 states that column 4 sets out the "current employment for this occupation." The last page of the OEQ 3.2 states that  

- All data are estimates from government sources including the U.S. Department of Labor, Division of Occupational Employment Statistics and of the Local Area Unemployment Statistics.

The US Publishing web site invokes the 2010 decennial census. Clearly US Publishing has not updated its page or claim to use the 2020 decennial census. Nowhere does US Publishing claim to use the Current Population Survey or any other data source. Nor the US Publishing recognize that the OES is now the OEWS. A rose by any other name is still a rose and the OEWS and the OEWS data is found at www.bls.gov/oes/. 

Since we know the US Publishing relied on OES/OEWS data from the OEQ and from the web page, we can compare and contrast the gross job number cited by the OEQ to the OEWS. A sample:

Occupation	OEQ total employment 4th Qtr. 2024	OEWS total employment  2024
Credit Authorizers, Checkers, and Clerks SOC 43-4041	63,662	11,960
Order Clerks  43-4151	216,280	83,420
Couriers and Messengers SOC 43-5021	232,941	71,920
Word Processors and Typists  SOC 43-9022	258,841	36,020
Office Clerks, General SOC 43-9061	2,351,948	2,510,550
Electrical and Electronic Equipment Assemblers SOC 51-2022	179,597	261,140 SOC 51-2028 (includes SOC 51-2022 and SOC 51-2023)
Inspectors, Testers, Sorters, Samplers, and Weigher SOC 51-9061	727,005	591,180
Helpers—Production Workers SOC 51-9198	273,294	167,490
Production Workers, All Other SOC 51-9199	813,370	277,060
Cleaners of Vehicles and Equipment SOC 53-7061	395,474	373,960
Packers and Packagers, Hand SOC 53-7064	676,479	601,440
Stock Clerks and Order Fillers SOC 43-5081	2,009,370	2,779,530 Stockers and Order Fillers SOC 53-7065

 The numbers are not reconcilable. Most of the occupations selected off the top of my head are so far off that they are clearly unreliable. 

That is strike two against the OEQ. US Publishing uses equal distribution based on the number of exertion-skill DOT codes resident in the SOC code. US Publishing's stated source for job numbers does not support the job numbers stated. Of the 867 codes in the 2018 SOC, 485 have at least some change from the 2010 SOC. US Publishing and its OEQ have not kept up nor paid attention to the combination of two SOC detailed groups into a single reported group (51-2028) in the current dataset. 

The OEQ as it is currently constituted needs to die.

___________________________



Suggested Citation:

Lawrence Rohlfing, What's Wrong with the OEQ -- Attacking Its Foundation and Methodology, California Social Security Attorney (August 27, 2025) https://californiasocialsecurityattorney.blogspot.com


The author has been AV-rated since 2000 and listed in Super Lawyers since 2008.