CATERPILLAR LOGISTICS SERVICES, INC., Petitioner, v. Hilda L. SOLIS, Secretary of Labor, Respondent, and International Union, United Automobile, Aerospace and Agricultural Implement Workers of America, Intervening Respondent, 674 F.3d 705

EASTERBROOK, Chief Judge. A regulation issued under the Occupational Safety and Health Act requires employers to maintain a log of work-related deaths, injuries, and illnesses. 29 C.F.R. § 1904.4(a). Another regulation defines a death, illness, or injury as work-related if “the work environment either caused or contributed to the resulting condition”. 29 C.F.R. § 1904.5(a). The phrase “contributed to” presumably means an increase in likelihood. How much of an increase is enough neither the regulation, nor any of the Secretary’s decisions, says.

Caterpillar Logistics Services sells parts for Caterpillar tractors, earth-moving machines, and other products. When orders arrive, employees locate the parts, put them in bags (which the parties call totes), and place the totes on a conveyor belt to the packing department (“Consol Pak” in the parties’ jargon). Employees in the packing department remove the parts from the totes, use a scanning gun to read bar codes that reveal which part goes to what destination, and put the parts in shipping containers. Most of the totes sent to the packing department weigh ten pounds or less; none exceeds 50 pounds.

During a normal day, each employee in the packing department fills 12 to 14 shipping containers with parts or bags of small parts— about 650 totes of parts per worker per day must be scanned and placed in the right shipping container. The process requires repetitive hand movements, and turning (pronation) of wrists, elbows, and shoulders. About five weeks after beginning work in the packing department in April 2008, MK began to feel pain in her right elbow. After another five weeks had passed, MK visited Caterpillar Logistics’ medical clinic.

Norma Just, a staff physician, put MK on leave until her condition could be diagnosed. Dr. Just concluded a few weeks later that MK had both medial and lateral epicondylitis in her right arm. Epicondylitis is a painful swelling of the ligaments and tendons around a joint. Lateral epicondylitis is colloquially known as “tennis elbow” and medial epicondylitis as “golfer’s elbow,” although both conditions can arise from other causes, and most instances are unrelated to sports.

MK did not work for the next three months, during which her condition improved (though, while off, she was diagnosed with medial and lateral epicondylitis in her left arm too). She returned to work in the packing department but after a month, when her condition failed to continue improving (as it had while she rested at home), she transferred to a position placing parts in racks. She has recovered from the epicondylitis.

Caterpillar Logistics had to decide whether to log MK’s injury as “work related.” It reviewed guides issued by the National Institute for Occupational Safety and Health, and the American Medical Association, both of which conclude that repetitive motion plus force (weight or impact) can cause epicondylitis, and that pronation plus force also can cause the condition, but that repetitive motion alone does not. Dr. Just concluded that MK’s work activities had not contributed to her epicondylitis.

Caterpillar Logistics convened an internal review panel with five members (three board-certified in musculoskeletal disorders); they agreed with Just’s decision. But the Department of Labor did not and assessed Caterpillar Logistics a $900 penalty for failing to log a work-related injury. After a four-day hearing, Administrative Law Judge Augustine agreed with the Department’s position and sustained the penalty. 2011 WL 2732360, 2011 OSAHRC Lexis 65, 23 O.S.H. Cas. (BNA) 1806 (May 24, 2011).

The full Occupational Safety and Health Review Commission declined to review that decision, which became the Secretary’s final ruling. Robert Harrison, the only physician to testify in support of the Department’s position, provided the basis of the ALJ’s decision. Dr. Harrison, a Clinical Professor of Medicine at the University of California at San Francisco, agreed with Caterpillar Logistics’ experts that the packing department is a light-force environment. He testified that, nonetheless, the combination of moderate repetition plus pronation of the wrist, hand, and forearm must have caused MK’s condition (Dr.

Harrison used the standard phrase: “reasonable degree of medical certainty”). He did not explain, however, why if this is so no other worker in the history of Caterpillar Logistics’ operations has contracted epicondylitis (the packing department has been running the same way for about 10 years, with a staff of 30, for 300 person-years of experience). Nor did he discuss any epidemiological study, pro or con. Caterpillar Logistics’ witnesses did discuss these matters. ALJ Augustine agreed across the board with Dr. Harrison—and, like him, ignored much evidence inconsistent with Harrison’s conclusions.

Caterpillar Logistics’ petition for review by the full Commission did not point out that both Dr. Harrison and the ALJ had disregarded available epidemiological studies showing (at least in the employer’s understanding) that jobs similar to its packing department are not associated with an increased incidence of repetitive-motion injuries. This omission forfeits the point for our purposes; the employer failed to exhaust its administrative remedies.

But Caterpillar Logistics did ask the Commission to reverse the ALJ because he disregarded the employer’s own 300-person-year experience, and the evidence from several witnesses the ALJ did not mention. Harrison’s testimony supplies substantial evidence for the ALJ’s decision. Dr. Harrison was properly qualified as an expert, and an agency is entitled to accept the evidence that it finds most persuasive, even if more witnesses testified to a different view (as happened here). See Universal Camera Corp. v. NLRB, 340 U.S. 474, 71 S.Ct. 456, 95 L.Ed. 456 (1951); NLRB v.

Walton Manufacturing Co., 369 U.S. 404, 82 S.Ct. 853, 7 L.Ed.2d 829 (1962). Harrison discussed the fact that MK developed epicondylitis in her left arm two weeks after leaving the packing department. He thought that the condition’s genesis may have been in the workplace, coupled with a tendency to use the left arm more (at both work and home) after pain began in the right arm. The ALJ accepted this view and was entitled to do so.

When emphasizing that Caterpillar Logistics had not supplied any competing explanation for MK’s condition, the ALJ did not contradict Home Depot # 6512, 2009 WL 3030762, 2009 OSAHRC Lexis 58, 22 O.S.H. Cas. (BNA) 1863 (Sept. 16, 2009), a decision by the full Commission.

Home Depot concludes that an employer is not required to show a non-work cause for an injury and that, when the only evidence of record was that an employee had died in a parking lot (apparently he had fainted and hit his head on the ground when he fell), the Department had not established by a preponderance of the evidence that the death was work-related. ALJ Augustine concluded that Dr.

Harrison’s testimony supplied the sort of evidence missing in Home Depot-, and in evaluating the probability that conditions at work increased the probability of a given injury, the agency may consider the absence of a competing explanation. (There may be such an explanation, but it is not in the record; Caterpillar Logistics did not try to show the genesis of MK’s condition.) Substantial evidence is not enough to sustain an administrative decision, however. The adjudicator also must take account of competing evidence and inferences.

That’s essential to show why the agency credited one witness rather than another. See, e.g., Morgan Stanley Capital Group Inc. v. Public Utility District of Snohomish County, 554 U.S. 527, 552, 128 S.Ct. 2733, 171 L.Ed.2d 607 (2008); Richard J. Pierce, Jr., 2 Administrative Law Treatise § 11.2 (5th ed.2010). This principle does not require elaborate discussion; the goal is not to produce tedious opinions that bury the analysis under an avalanche of detail. See Stephens v. Heckler, 766 F.2d 284 (7th Cir.1985). But it does require the agency to test its hypothesis against competing hypotheses.

It may not simply ignore strong indications that its favored witness got things wrong. See, e.g., Bjornson v. Astrue, 671 F.3d 640 (7th Cir.2012). That’s what happened here. The big consideration missing from the ALJ’s analysis is Caterpillar Logistics’ 300-person-years of experience with its packing department. Epicondylitis occurs at a rate of about 1% to 2% per year in the general population. (Or so the parties tell us.) This implies that Caterpillar Logistics should have encountered between three and six cases of epicondylitis among the staff of the packing department if work played no causal role at all.

It actually had one case (MK’s). If conditions in the packing department do cause or contribute to epicondylitis, the condition should occur at levels exceeding those of the populace at large. The record does not show an elevated incidence. It might require a statistical analysis to determine whether the incidence of epicondylitis among the staff could have been the result of chance, and what frequency would imply a causal role for workplace conditions.

Caterpillar Logistics did not perform tests for statistical significance — but the agency has the burden of proving causation by a preponderance of the evidence, see Director, OWCP v. Greenwich Collieries, 512 U.S. 267, 114 S.Ct. 2251, 129 L.Ed.2d 221 (1994). Maybe a sample of 300 person-years is too small for the numbers to be significant when the background incidence of epicondylitis is so low — but the ALJ, having disregarded the experience at Caterpillar Logistics, did not make a finding one way or the other about statistical significance.

The significance of Caterpillar Logistics’ experience depends in part on what § 1904.5(

a) means in saying that an injury is work-related if working conditions “contributed to” the injury. This could mean “increased the probability, above background levels, by a statistically significant amount.” That’s how we’ve been thinking about it and is why Dr. Harrison’s (and the ALJ’

s) failure to consider Caterpillar Logistics’ actual experience matters. If the requirement of “contribution” is stronger — if, say, it means something like “doubled the probability” — then the experience in the packing department is even more important. But perhaps the agency means a weaker link. Then it would be hard to know whether Caterpillar Logistics’ own experience has any salience. What is certain is that the agency must choose among these possibilities; the judiciary cannot choose for it but must affirm, or not, based on the agency’s rationale. See SEC v.

Chenery Corp., 318 U.S. 80, 88-89, 63 S.Ct. 454, 87 L.Ed. 626 (1943); SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947). The ALJ did not choose, indeed did not appreciate the need for choice, and the Commission as a whole has never discussed the subject. What the work-relatedness requirement is doing in § 1904.4(

a) is a puzzle. At oral argument, counsel for the Secretary suggested that the injury log’s function is to help the Department determine which occupations are hazardous, so that it can concentrate enforcement resources on them and propose regulatory changes that may reduce risks to employees. These purposes can be served, however, only if the log contains all injuries. Then the Department can compare rates of injury in a given job with the background rate in the general population; the difference can be attributed to workplace hazards.

If, however, employers log injuries only after first deciding that each is work-related, the log becomes less useful as an exploratory or investigatory tool. Given the work-relatedness requirement in § 1904.4(a), the log does not show actual risks; it shows whether the employer believes that there is a connection between the working environment and the injuries. The Secretary can get no more information out than the employer puts in: GIGO (garbage in, garbage out).

Eliminating the work-relatedness requirement would make the log more useful and avoid the potentially high costs of evaluation illustrated by this case. An elaborate board of inquiry at Caterpillar Logistics was followed by the Department’s investigation, a four-day trial, an opinion by an ALJ, submissions to the Commission, and then briefs and arguments in a court of appeals. Because saving all of this time and expense might simultaneously improve the log’s usefulness, the Secretary may wish to take another look at § 1904.4(a).

The petition for review is granted, the Secretary’s decision is vacated, and the case is remanded for proceedings consistent with this opinion.