Jon P. Dray petitions for review of a decision by the United States Railroad Retirement Board ("the Board") denying him a permanent disability annuity under 45 U.S.C. Sec. 231a(a)(1)(v). By a 2-1 majority, the Board affirmed and adopted the decision of the hearing officer, who herself sustained the rejection of Dray's claim by the Bureau of Retirement Claims ("the Bureau"). We deny the petition for review.

On a winter day in 1985, Dray lost his footing while working on a telephone pole. He fell to the frozen ground twenty feet below, landing in a sitting position. Even though his lower back caused him some trouble after the accident, he remained on the job at Conrail. But when Dray went to lift a five-gallon can of gasoline in the summer of 1986, sudden pain shot through his back. After he consulted with a neurological surgeon several times in 1986 and 1987, the regional medical director of Conrail informed him that he was disqualified from his duties as signalman. Since then, in July 1987, Dray has not worked for any other employer.

A number of doctors have examined Dray or had him in their care. On the advice of Dr. Donald L. Myers, a neurological surgeon, Dray underwent a diskectomy1 in June 1988. Dr. Myers examined Dray once before surgery and twice after, on December 1, 1988, and April 20, 1989. Dr. Steven Mandel, a neurologist, saw Dray twice before surgery, and three times afterwards, on December 1, 1988, April 19, 1989, and January 29, 1990. Dr. Mandel also submitted for review three letters containing his findings (dated June 2, 1989, April 5, 1990, and June 15, 1990), including his responses to a Board form detailing Dray's physical limitations. In connection with Dray's application for a disability annuity, the Board sent him for an orthopaedic examination with Dr. Michael Arata on March 2, 1989.

Claiming total disability based on back injury and chronic pain, Dray filed for benefits in November 1988. The Bureau found on April 18, 1989, that although Dray had a back injury that caused some pain and muscle spasms, he was not totally and permanently disabled for all work. On reconsideration of Dray's claim, the Bureau determined that despite Dray's moderately severe condition, he could do medium work.

In July and December 1990 the hearing officer held hearings. At the first hearing Dray submitted a report from a vocational expert and testified about his work experience, medical history, and daily activities. Delana Dray, his wife, testified about her observations of Dray. During the second hearing, a vocational consultant presented expert testimony on jobs Dray could perform despite his impairments. The hearing officer found Dray to be not disabled and denied his application for an annuity.

Dray appealed from that decision to the three-member Board. He also supplemented the record with an examination report from Dr. Arthur C. Warr, an orthopaedic surgeon. The Board affirmed by a 2-1 margin, adopting the decision of the hearing officer.

Appellate review of decisions by the Board closely emulates our approach in appeals concerning Social Security disability benefits. We will sustain a decision of the Board that is supported by substantial evidence and based on the proper legal criteria. Hayes v. Railroad Retirement Bd., 966 F.2d 298, 302 (7th Cir.1992); Aspros v. Railroad Retirement Bd., 904 F.2d 384, 386 (7th Cir.1990). Evidence is insubstantial if it offers only a "mere scintilla" of proof. Richardson v. Perales, 402 U.S. 389, 401, 91 S. Ct. 1420, 1427, 28 L. Ed. 2d 842 (1971). If, however, the Board's finding rests on "such relevant evidence as a reasonable mind might accept as adequate to support a conclusion," id., we will affirm. We may not weigh the evidence a new and decide the ultimate question whether a claimant is disabled. Nor may the court supplant the judgment of the Board with its own reasoning, Lambert v. Railroad Retirement Bd., 929 F.2d 1197, 1200 (7th Cir.1991), even if some evidence in the record contradicts the Board's findings. Soger v. Railroad Retirement Bd., 974 F.2d 90, 92 (8th Cir.1992). Likewise, the court must do more than merely rubber stamp Board decisions. See Ehrhart v. Secretary of HHS, 969 F.2d 534, 538 (7th Cir.1992). When, as here, the Board adopts the hearing officer's opinion without issuing further findings, we evaluate the judgment of the hearing officer. Hayes, 966 F.2d at 302.

The issue before the hearing officer was whether Dray had "a permanent physical or mental condition [such that he was] unable to engage in any regular employment." 45 U.S.C. Sec. 231a(a)(1)(v). She answered that question in the negative after following the five-step procedure for evaluating disabilities. 20 C.F.R. Sec. 220.100(b) (1991). At step four, the hearing officer found that Dray's impairments were neither listed nor "medically equal" to any on the Listing of Impairments. In addition, the hearing officer determined that Dray was unable to do his past relevant work. Dray does not appear to contest these findings.

The battle lines form instead at the fifth and final step of the disability analysis, where the burden shifts to the Board to show that the claimant can perform some other job. Campbell v. Shalala, 988 F.2d 741, 743 (7th Cir.1993). "If the claimant cannot do other work [i.e., besides his former work], the Board will find him or her disabled. If the claimant can do other work, the Board will find the claimant not disabled." 20 C.F.R. Sec. 220.100(b)(5)(i). The hearing officer found Dray fit for occupations outside of his prior experience at Conrail.

Dray contends that he is disabled and entitled to a total and permanent disability annuity from the date he claimed. First, he argues that the hearing officer improperly rejected the opinion of Dr. Mandel, whom he labels his "treating physician." We agree. Second, he argues that the hearing officer did not give proper consideration to his complaints of pain. We find this contention unpersuasive.

Dray argues that the opinion of a treating doctor normally is entitled to great weight. Petitioner's Brief at 19. This assertion ignores Amax Coal Co. v. Franklin, 957 F.2d 355 (7th Cir.1992), which questions "the traditional but unwarranted preference of some administrative law judges in disability cases" for the opinion of a treating physician over that of a nontreating specialist. Id. at 359. Certainly in some instances the conclusions of the treating physician might be more probative than a specialist's observations--for example, if the treating doctor learns something from having had the opportunity "to study the course of the developing disease." Id. But the treating doctor's knowledge matters only if the " 'ability to observe the claimant over an extended period of time is essential to understanding' his condition ... and [the] treating physician knows something about the disease in question." Id. (citations omitted); see also Peabody Coal Co. v. Director, Office of Workers' Compensation Programs, 972 F.2d 880, 884 (7th Cir.1992); Micus v. Bowen, 979 F.2d 602, 607-09 (7th Cir.1992) (recognizing trier of fact's ability to consider not only the treating physician's possible bias, but also the doctor's opportunity to observe claimant over a long period of time). To summarize: In the case of dueling doctors, it remains the province of the hearing officer to decide whom to believe--a treating doctor whose experience and knowledge about the case may (or may not) be relevant to understanding the claimant's condition, or a consulting specialist who may bring expertise and knowledge about similar cases. At all events, "resolution of evidentiary conflicts lies within the exclusive domain of" the hearing officer, Ehrhart, 969 F.2d at 541, so long as those factual findings are supported by substantial evidence.

The hearing officer noted that Dr. Mandel found Dray incapable of any physical employment since April 19, 1989, and not capable of sedentary work as of January 1990 "due to a permanent neurological impairment referable to his lumbar spine." Decision of the Hearing Officer, No. 91-607, May 31, 1991, at 15. Apparently the hearing officer rejected Dr. Mandel's assessment of Dray's residual functional capacities for two reasons: (1) Dr. Mandel's conclusions conflicted with the opinions of Dr. Arata (the consulting orthopaedist) and Dr. Myers (the surgeon); (2) Dr. Mandel's opinion was not supported by objective medical evidence establishing the degree of limitations he attributed to Dray.

In discounting Dr. Mandel's assessments, the hearing officer concluded that his findings were not consistent with the reports of other physicians. Decision of the Hearing Officer at 15. The hearing officer is by all means entitled to reconcile contradictory evidence in the record. See Herr v. Sullivan, 912 F.2d 178, 181 (7th Cir.1990); Walker v. Bowen, 834 F.2d 635, 640 (7th Cir.1987). The existence of an evidentiary dispute is not a ground for reversing the hearing officer's decision to give credence to one version of the facts over another. See Herr, 912 F.2d at 181 n. 4; Reynolds v. Bowen, 844 F.2d 451, 454 (7th Cir.1988). Be that as it may, the hearing officer must not only offer some principled basis explaining why one account is worthier than another, but also explain with particularity the basis of the decision. Cf. Thomas v. Sullivan, 801 F. Supp. 65, 68 (N.D.Ill.1992). Within "[r]easonable limits, the reasons for rejecting evidence must be articulated if there is to be meaningful appellate review." Young v. Secretary of HHS, 957 F.2d 386, 393 (7th Cir.1992); see also Brown v. Bowen, 847 F.2d 342, 346 (7th Cir.1988). A judgment favoring medical assessments A-L and discrediting M-Z, without further elucidation, is not based on substantial evidence. Zblewski v. Schweiker, 732 F.2d 75, 78-79 (7th Cir.1984) ("In the absence of an explicit and reasoned rejection of an entire line of evidence, the remaining evidence is 'substantial' only when considered in isolation").

All of which raises a rather elementary point: Is there a true evidentiary conflict among the various doctors at all? One supposed inconsistency between the reports of Drs. Myers and Mandel on December 1, 1988, seems not a contradiction at all. Both doctors report that Dray is having continued discomfort or pain. While Dr. Myers does not indicate whether he did an examination, Dr. Mandel's findings discuss Dray's medical condition in greater detail.

Another inconsistency identified by the hearing officer is one between Dr. Arata's report on March 9, 1989, and Dr. Mandel's examination on April 24, 1989. Both doctors state that Dray has difficulty lifting heavy objects. Whereas Dr. Arata reports that Dray has lower spine stiffness and not so much pain, Dr. Mandel tells of "a considerable amount of pain." Dr. Arata mentions "brisk ankle jerks" and Dr. Mandel discusses "absent ankle jerks." With Dr. Arata, Dray exhibited "no problem with toe or heel walking" and bilateral, straight leg-raising to 90 degrees. In contrast, Dr. Mandel observed Dray "throw" both legs when walking and measured positive, straight leg-raising at 60 degrees. Whether these reports contradict one another or merely document an erosion in Dray's condition is beside the point because Dr. Arata concludes that Dray may need to "have restricted activities for the rest of his life" and Dr. Mandel similarly remarks that Dray "should never return to his prior position." All in all, we are hard-pressed to see why the hearing officer issues such a hue and cry over these two sets of reports, whose differences really are in shades.

Once we discount the hearing officer's descriptions of inconsistencies between Dr. Mandel and Drs. Arata and Myers, what remains for consideration are the examinations and reports of Dr. Mandel, Dr. Myers, Dr. Mandel, Dr. Mandel, Dr. Mandel, and Dr. Warr. So there is not much left for Dr. Mandel to contradict, except, of course, himself. His findings dominate the record after April 1989 and chronicle the deterioration of Dray, leading him to conclude that his patient is permanently disabled.2 Even though she asserted that Dr. Mandel's assessments conflicted with the reports of other physicians, in reality the hearing officer rejected his latest and most severe prognoses. In so doing the hearing officer did not reconcile contradictory evidence so much as she outright rejected assessments by Dr. Mandel that are uncontested by any evidence in the record after April 1989. In such a circumstance, "[t]his court insists that the finder of fact explain why he rejects uncontradicted evidence." Stephens v. Heckler, 766 F.2d 284, 287 (7th Cir.1985).

A hearing officer "must mention and discuss, however briefly, uncontradicted evidence that supports the claim for benefits." Id. In light of the fact that the most significant reports by Dr. Mandel were uncontested, the hearing officer seemingly rejected those findings because they did not support the assessment of Dray's restrictions. In Kapusta v. Sullivan, 900 F.2d 94, 97 (7th Cir.1989), the court implied that an ALJ may reject a treating physician's opinion that is not supported by clinical findings. Id. at 96-97. Here, the hearing officer lists several clinical findings that she claims are absent from Dr. Mandel's report and essential for supporting his assessments. Decision of the Hearing Officer at 16. In response, Dray argues that at least two and maybe three of the four allegedly missing findings appear in the evidence. We find his arguments compelling.

First, the hearing officer states that there are no findings that Dray has had muscle atrophy, weakness, or sensory loss. But on at least two occasions Dr. Mandel found absent ankle jerks. R. 243, 269. Also, he describes "distal weakness in [Dray's] lower extremities which [sic] appears to be in an L5-S1 distribution." R. 279. Moreover, Dr. Mandel stated in a letter that Dray could sit or walk for only ten minutes at a time due to severe limitations of motion. R. 272. Next, the hearing officer states that Dray had no significant impairment of gait. On at least one occasion, however, Dr. Mandel found that Dray "throws both legs when walking." R. 269. In addition, the hearing officer is troubled by no indication of prescribed medication or other treatment to alleviate Dray's symptoms. Dray replies that he was on pain relievers--Halcion, Tylenol, and Percocet--upon his discharge from the hospital after his surgery in 1988. Later, however, Dray took nonprescription drugs only, including several extra-strength Tylenols daily, due to concerns of medical professionals that he might become addicted to painkillers. A claimant is justified in not taking adequate pain medication if there is a realistic chance of addiction. Cf. Stith v. Railroad Retirement Bd., 902 F.2d 1284, 1287 (7th Cir.1990) (citing Dover v. Bowen, 784 F.2d 335, 337 (8th Cir.1986)).

None of the hearing officer's three criticisms of Dr. Mandel's report seems valid. The Board responds weakly that "[t]he fact that one could draw a different conclusion from the conflicting evidence is not dispositive of the case, and the Court is not free to reweigh the record to reach its own conclusion." Respondent's Brief at 17-18. That may be so, but the Board offers no principled reason why Dr. Mandel's findings do not support the degree of restrictions assessed. As suggested by subsection 1, supra, Dr. Mandel's assessments do not contradict those of other doctors. Moreover, after evaluating them in isolation, we do not find them to be supported by the requisite clinical findings identified by the hearing officer. For these reasons, the hearing officer's decision to disregard evidence supplied by Dr. Mandel is not supported by substantial evidence. See Zblewski, 732 F.2d at 78-79.

That is not to say, however, that the objective medical evidence and physicians' reports necessarily establish that Dray suffers from a disability. This court does not, of course, decide whether a claimant is disabled, Ehrhart, 969 F.2d at 538, "unless the record can yield but one supportable conclusion." Campbell, 988 F.2d at 744. Our conclusion here is a narrow one: We merely question the logic of the finding that Dr. Mandel's reports contradict those of other physicians and the determination that his ultimate judgments were unsupported by objective documentation of the extent of Dray's restrictions. But our doubts about the reasoning of the hearing officer need not compel the court to modify or reverse the denial of benefits. Judges--ALJ's and hearing officers among them--must resist the urge to play doctor. Schmidt v. Sullivan, 914 F.2d 117, 118 (7th Cir.1990), cert. denied, --- U.S. ----, 112 S. Ct. 278, 116 L. Ed. 2d 230 (1991); Rousey v. Heckler, 771 F.2d 1065, 1069 (7th Cir.1985). Whether a claimant is disabled is a legal, not a medical, question--an ultimate conclusion for the hearing officer to make. Kapusta, 900 F.2d at 97. The decision of the hearing officer shall be upheld even if reasonable people can differ on the conclusion suggested by the evidence. See Walker, 834 F.2d at 640. Moreover, even if we are certain that the hearing officer arrived at an erroneous result--as distinct from an error in analysis--we are not compelled to reject the opinion in its entirety. Id. at 644 (citing Stephens, 766 F.2d at 287 (court reviews judgments, not opinions)).

In evaluating whether the overall evidence substantially supports the decision of the hearing officer, we take into account many different forms of evidence, including objective medical information, medical evaluations, the claims of the claimant, accounts of his daily activities, and the hearing officer's personal observations. Pope v. Shalala, 998 F.2d 473, 486 (7th Cir.1993). With that in mind, we turn to the issue of credibility.

The hearing officer recognized that Dray has a permanent neurologic impairment related to his surgery and continuing pain. Nonetheless, she found incredible Dray's claims that he can perform no more than six hours of activity in an eight-hour day and that he needs to alternate between sitting, standing, walking, and elevating his feet. Decision of the Hearing Officer at 16. Accordingly, she determined at the fifth step of the disability analysis that Dray retained a residual functional capacity for light work.3 Dray argues that in making this finding, the hearing officer improperly disregarded his complaints of pain.

On November 14, 1991, the Secretary of Health and Human Services began to apply revised regulations for use in evaluation of a claimant's allegations of pain: "[W]e will not reject your statements about the intensity and persistence of your pain or other symptoms or about the effect your symptoms have on your ability to work solely because the available objective medical evidence does not substantiate your statement." 20 C.F.R. Sec. 404.1529(c)(2) (1992). This regulation, which clarifies the one it replaces, see Pope, 998 F.2d at 482, was in effect when the Board affirmed and adopted the hearing officer's decision in May 1992, but not when the hearing officer issued her opinion in May 1991. Because the court evaluates decisions of the Board to determine whether they are based on proper legal criteria, Hayes, 966 F.2d at 302, we must consider its adoption in light of the clarifying regulation, even though prior interpretations were in effect at the time that the hearing officer rendered her decision.

The hearing officer found that Dray's characterizations of his limitations were unsupported by (1) the fact that he does not use painkillers, and (2) Dray's testimony regarding his activities both prior to and after surgery. The court need not consider Dray's use of medications because Sec. 404.1529(c)(2) makes the absence of objective medical evidence a necessary but not sufficient condition to support a finding of not disabled. But, even under the new regulation, the petitioner's credibility figures decisively in our decision whether a denial of benefits is supported by substantial evidence. "Although we cannot discredit a complaint of pain simply because objective medical evidence was not introduced to support the extent of pain, neither are we required to give full credit to every statement of pain, and require a finding of disabled every time a claimant states that she feels unable to work." Pope, 998 F.2d at 486.

We will affirm credibility determinations unless the petitioner demonstrates that they are "patently wrong." Herr, 912 F.2d at 182 (quotations and citation omitted). A reviewing court generally will not disturb a credibility judgment because the trier of fact is in a unique position to observe and to evaluate witnesses. See Anderson v. Bessemer City, 470 U.S. 564, 575, 105 S. Ct. 1504, 1512, 84 L. Ed. 2d 518 (1985); accord Clark v. Sullivan, 891 F.2d 175, 178 (7th Cir.1989); Walker, 834 F.2d at 641. But when "objective inconsistency or fundamental implausibility is at issue," the hearing officer "has no special advantage over a reviewing court in determining credibility." Thomas, 801 F. Supp. at 70 (citing Anderson, 470 U.S. at 575, 105 S.Ct. at 1512). In contrast, credibility determinations also can be based on "intangible and unarticulable elements which impress the [fact finder and] unfortunately leave 'no trace that can be discerned in this or any transcript that we review.' " Kelley v. Sullivan, 890 F.2d 961, 964 (7th Cir.1989) (citation omitted). In that case, the court should not disturb a credibility determination. See also Pearce v. Sullivan, 871 F.2d 61, 64 (7th Cir.1989) (credibility determinations are essentially unassailable).

Dray at no time indicated during the hearing that he was having trouble sitting for any significant period of time. His demeanor there was significant because the hearing officer can rely "on his own observations during a hearing concerning the severity of a claimant's pain. Such observations are credibility determinations and are entitled to considerable weight." Whitney v. Schweiker, 695 F.2d 784, 788 (7th Cir.1982). He testified that he spent his days pursuing his education, running errands, mowing the lawn, visiting, checking on rental properties, doing light housework and paying bills. Also, Dray stated that he needed to put his feet up daily. His wife, however, could not recall witnessing Dray do so on a regular basis. Moreover, Dray reported being able to drive for an hour at a time.

In light of this evidence, the hearing officer surmised after consulting with a vocational expert that Dray was able to do light work. In so concluding, she found that Dray experienced some degree of pain yet, given his daily regimen, was not disabled. There is nothing anomalous about a hearing officer finding a claimant's testimony to be generally credible but still determining that he is not impaired in the manner he claims to be. Cf. Ehrhart, 969 F.2d at 539; Imani on behalf of Hayes v. Heckler, 797 F.2d 508, 511-12 (7th Cir.), cert. denied, 479 U.S. 988, 107 S. Ct. 580, 93 L. Ed. 2d 583 (1986). Even if objective medical evidence partially supports complaints of pain, the hearing officer is free to conclude that the claimant is less than credible and may be exaggerating the extent to which his discomfort prevents him from performing certain kinds of jobs or restricts his activities apart from work. Lee v. Sullivan, 988 F.2d 789, 793 (7th Cir.1993). Dray has not borne his burden that this conclusion is patently wrong.

Because the decision of the Board is based on substantial evidence, we DENY the petition for review.