¶ 95. {concurring). I write separately to highlight my concerns with the majority opinion. Once again, the majority reaches out and needlessly addresses an issue that was neither raised by the parties nor briefed and argued. If the majority is going to go down this path, it is best that it get it right.

¶ 96. Here the majority reaches out and introduces a concept foreign to our jurisprudence: that judge-made common law plays a role in worker's compensation decisions. Neither of the parties advance the issue nor make the argument that the odd-lot doctrine explained in Balczewski v. DILHR, 76 Wis. 2d 487, 251 N.W.2d 794 (1977), is anything other than a judicial interpretation of the relevant provision in the worker's compensation act. Nevertheless, the majority tackles the issue.

¶ 97. I believe that the rule of law is generally best developed when issues are raised by the parties and then tested by the fire of adversarial briefs and oral *191arguments. Indeed, "[t]he fundamental premise of the adversary process is that these advocates will uncover and present more useful information and arguments to the decision maker than would he developed by a judicial officer acting on his own in an inquisitorial system." Adam A. Milani & Michael R. Smith, Playing God: A Critical Look At Sua Sponte Decisions By Appellate Courts, 69 Tenn. L. Rev. 245, 247 (2002) (citing United States v. Burke, 504 U.S. 229, 246 (1992) (Scalia, J., concurring)).

¶ 98. Additionally I am suspect of the majority's new concept. As aptly described in the concurrence above, worker's compensation is not an outgrowth of our judge-made common law. Rather it is a carefully crafted and uniquely balanced legislative act. All of our judicial decisions represent the court's application and interpretation of the act together with its attendant administrative regulations. There exists no common law in our worker's compensation jurisprudence.

¶ 99. For the above stated reasons, I respectfully concur.

¶ 1. In this case we revisit the so-called "odd-lot doctrine," an aspect of worker's compensation law dealing with cases of permanent total disability. The issue presented is whether a prima facie case for permanent total disability under the "odd-lot" doctrine must include evidence that the injured employee has made a reasonable effort to find suitable post-injury employment.

*145¶ 2. The odd-lot doctrine is a judge-made adjunct to the law of worker's compensation.1 It represents a modification of the general rule that benefits for permanent total disability compensate an injured worker for loss of earning capacity. Whereas most recipients of permanent total disability benefits have lost all capacity to earn income, claimants under the odd-lot doctrine may qualify for benefits even though they retain a small, residual capacity to earn income. In essence, the odd-lot doctrine provides that some injured workers should be characterized as permanently, totally disabled even though they are still capable of earning occasional income.

¶ 3. We recognized the odd-lot doctrine in Balczewski v. DILHR, 76 Wis. 2d 487, 251 N.W.2d 794 (1977):

"Total disability" in compensation law is not to be interpreted literally as utter and abject helplessness. Evidence that claimant has been able to earn occasional wages or perform certain kinds of gainful work does not necessarily rule out a finding of total disability nor require that it be reduced to partial.

Id. at 493 (quoting 2 Arthur Larson, Workmen's Compensation Law, § 57.51, at 10-107).2 The doctrine operates as a rule of evidence:

[W]here a claimant makes a prima facie case that he has been injured in an industrial accident and, because of his injury, age, education, and capacity, he is unable *146to secure any continuing and gainful employment, the burden of showing that the claimant is in fact employable and that jobs do exist for the injured claimant shifts to the employer.

Id. at 495.

¶ 4. The Labor & Industry Review Commission (LIRC) argues that two enactments since our adoption of the odd-lot doctrine in Balczewski require us to re-evaluate how the Balczewski burden-shifting framework is applied. In 1985, the legislature enacted Wis. Stat. § 102.17(7)(a) (2001-02),3 which authorizes the Department of Workforce Development (DWD) to receive and consider expert evidence on loss of earning capacity. In 1982, an administrative rule, DWD § 80.34, was adopted; it specifies factors the DWD is to consider in determining loss of earning capacity. One of these factors is the claimant's "efforts to obtain suitable employment."

¶ 5. LIRC does not explain precisely how the enactment of Wis. Stat. § 102.17(7)(a) requires this court to modify the Balczewski odd-lot doctrine. With respect to DWD § 80.34, LIRC argues that because the administrative rule lists "efforts to obtain suitable employment" as one of the factors the DWD considers in determining loss of earning capacity, we should modify the formula for a prima facie odd-lot case under Balc-zewski so that the claimant is required to demonstrate that he has made a reasonable job search as part of his prima facie cáse.

¶ 6. We disagree, and hold that a claimant is not required to present evidence of a job search as part of prima facie case of odd-lot unemployability, provided *147the claimant shows that because of his injury and the other Balczewski factors such as age, education, capacity, and training, he is unable to secure continuing, gainful employment. If the claimant succeeds in putting himself within the odd-lot category, it falls to the employer to rebut the prima facie case by demonstrating that the claimant is employable and that jobs exist for him. In this case, LIRC concluded that the employee, Ralph Beecher, failed to make a prima facie case for odd-lot unemployability in part because he did not make enough effort to find a new job. In reaching this conclusion, LIRC placed an evidentiary burden on Beecher that was not required of him, and we therefore affirm the court of appeals' reversal of LIRC's decision.

I. FACTS AND PROCEDURAL HISTORY

¶ 7. Our recitation of the facts is based upon the agency record, including the Findings of Fact and Conclusions of Law of the DWD Administrative Law Judge, Leonard E. Martin, and the subsequent LIRC decision. Beecher was born in 1942 and has a ninth-grade education. He worked for Outokumpu Copper Kenosha, a foundry, or its corporate predecessors, for 29 years. At the time of his injury, Beecher was working on a "Z-Mill" machine, which runs sheets of metal from one large roll to another roll or a spool. The work required Beecher to lean over the first roll and thread the metal into a slit on the second roll. The job required bending over to pick up the sheets of metal, and then pulling them in order to thread the sheets onto the rolls. By all accounts, the work was strenuous.

¶ 8. On April 7, 1997, Beecher developed sharp pains in his lower back, and sought medical treatment with Dr. Christopher Noonan, an orthopedist. Dr. Noonan had treated Beecher for back problems in the past, and had performed two prior surgeries on Beecher. *148Beecher's back pains continued and Dr. Noonan eventually performed a third surgical procedure on Beecher's lower back on September 10, 1997.

¶ 9. Beecher returned to light duty work in April 1998, but after two weeks Outokumpu ran out of work for him and he ceased working. Beecher has not returned to work since that time. Outokumpu moved its operations out of Wisconsin and did not offer to relocate Beecher.

¶ 10. In September 1999, Beecher filed an application for a hearing at DWD pursuant to Wis. Stat. § 102.17, alleging that his day-to-day work activities up to April 7,1997, caused a progression of his pre-existing back condition, leading ultimately to his surgery. He sought temporary total disability benefits from October 14,1998, to May 14,1999, permanent partial disability benefits on a functional basis at 15 percent to the body as a whole, permanent disability on a vocational basis for loss of earning capacity, and payment of medical expenses.

¶ 11. Administrative Law Judge Martin evaluated Beecher's claims on the basis of reports and testimony from several medical and vocational experts. Beecher himself also testified at the hearing. The expert evidence in this case is complex. While we may safely pass over many of its details, the following is a brief description of the experts' conclusions.

¶ 12. Beecher relied upon two medical experts: Dr. Noonan, his orthopedist, and Dr. Richard Karr, who examined Beecher in April and July 1998. Dr. Noonan assessed Beecher at 15 percent permanent partial disability based upon his injury. Dr. Karr assessed Beecher at a minimum of ten percent permanent partial disability. Outokumpu offered the report and testimony of Dr. Thomas O'Brien, who examined Beecher in July 1997 and again in September 1998. Dr. O'Brien opined that *149Beecher suffered no new injury in April 1997, but only a manifestation of symptoms from his pre-existing condition.

¶ 13. Beecher and Outokumpu each also presented evidence from vocational experts. Beecher's expert was Charles McReynolds. McReynolds testified that Beecher tested at the sixth-grade level for math, the seventh-grade level for reading, and the fifth-grade level for spelling. McReynolds testified that Beecher is not a candidate for retraining, based on his aptitude. McReynolds also testified that if Dr. O'Brien's medical assessment was correct, Beecher's loss of earning capacity would be approximately 20 percent. But if Drs. Karr and Noonan were correct, McReynolds concluded, then at best Beecher could do light duty work. McReynolds opined that given Beecher's age, educational level, previous training, and non-suitability for retraining, no stable labor market existed for the limited type and amount of work Beecher was able to do. McReynolds therefore concluded that Beecher was permanently and totally disabled from a vocational standpoint.

¶ 14. Outokumpu's vocational expert was Leanne Panizich, who tested Beecher at the "high school level" for reading and the sixth-grade level for arithmetic and spelling. Panizich opined that if Drs. Noonan and Karr were correct in their medical assessments, then Beecher would suffer only a 25 to 35 percent reduction in earning capacity, as he could perform light duty work for employers in Wisconsin. If O'Brien's medical assessment was accepted, however, the reduction would be ten to 20 percent. Panizich said it was rare to encounter an employer these days who would not accommodate Beecher's disability.

¶ 15. Judge Martin found Beecher's medical experts more credible than Outokumpu's, and awarded *150Beecher temporary total disability benefits for the period May 14, 1998, to May 19, 1999. As for Beecher's claim for permanent total disability on a vocational basis, Judge Martin found that McReynolds' testimony was more credible than Panizich's, and he concluded that Beecher had sustained permanent total disability on a vocational basis as a result of the April 1997 injury. Under Wisconsin law, when an applicant is found to be permanently and totally disabled, benefits are awarded for life. Wis. Stat. § 102.44(2).

¶ 16. Outokumpu appealed the awards of temporary and permanent total disability, and LIRC reversed the award for permanent total disability.4 LIRC's reversal was based upon its contrary assessment of the vocational expert evidence presented to the agency. McReynolds' opinion that Beecher was permanently and totally disabled was premised on work restrictions imposed on Beecher by Dr. Noonan in March 1998, approximately one year before the end of Beecher's healing period. However, in December 2000, Dr. Noonan adopted the findings of occupational therapist Ruth Meehan's Functional Capacity Evaluation (FCE) of Beecher. In adopting these findings, Dr. Noonan did not reiterate his earlier part-time work restrictions.

¶ 17. LIRC then applied the "odd-lot" doctrine to the facts of Beecher's case. Citing Balczewski, 76 Wis. 2d at 495, LIRC explained that "a worker makes a prima facie case of permanent and total disability on an odd-lot basis by showing that he or she is unable to obtain any continuing or gainful employment because of the impairment from his work injury and other factors such as age, training and education." If the worker makes a prima facie case, LIRC continued, *151again citing Balczewski, the burden shifts to the employer to show that some kind of work is "regularly and continuously available to the worker."

¶ 18. LIRC was struck by McReynolds1 failure to explain why he did not adjust his vocational conclusions to account for Dr. Noonan's adoption of the FCE. Then, citing Larson's treatise on worker's compensation law,5 LIRC noted:

While the applicant has made some effort to find work, the work restrictions set out in the November 2000 FCE suggest he could have made more of an effort, a factor that may be considered against him in determining whether he has established a prima facie case of odd-lot unemployability.

Accordingly, LIRC concluded that Beecher had failed to establish a prima facie case for permanent total disability.

¶ 19. Even though LIRC did not accept McRey-nolds' opinion and concluded, based in part on the job search evidence, that Beecher failed to carry his eviden-tiary burden, the agency did not accept Panizich's expert opinion either. Unlike McReynolds, Panizich's conclusions were based upon the results of the FCE, but LIRC concluded that in light of Beecher's age, education, skill level, and diminished capacity to retrain, Panizich's ultimate opinion that Beecher suffered a 35 percent loss of earning capacity was too conservative. Based on all the evidence presented, LIRC concluded that Beecher sustained a loss of earning capacity of 60 percent, and awarded benefits accordingly.

¶ 20. Beecher timely appealed LIRC's decision to the Kenosha County Circuit Court pursuant to Wis. Stat. § 102.23. The scope of a circuit court's review of an *152order of LIRC is narrow; the court may only confirm or set aside an order or award, it may not amend it or substitute the court's judgment for the commission's. Wis. Stat. § 102.23(1)(e); Columbia Cas. Co. v. Ind. Comm'n, 254 Wis. 310, 35 N.W.2d 904 (1949). The circuit court, the Honorable Michael S. Fisher, affirmed LIRC's findings as supported by substantial and credible evidence.

¶ 21. Beecher appealed to the court of appeals under Wis. Stat. § 102.25, contesting LIRC's conclusion that he had failed to establish a prima facie case of permanent total disability. Beecher v. LIRC, 2003 WI App 100, ¶ 1, 264 Wis. 2d 394, 663 N.W.2d 316. The court of appeals reversed, holding that LIRC improperly applied the odd-lot doctrine to the facts of Beecher's case. Id., ¶ 36. In particular, the court held that LIRC improperly denied Beecher's claim for permanent total disability by concluding that Beecher had not demonstrated that he had conducted an adequate search for a new job. Id. The court noted that Beecher was not required under the Balczewski odd-lot doctrine to bring forward any evidence relating to job search as part of the prima facie case. Id., ¶ 25. The court of appeals concluded that Beecher had, in fact, established a prima facie case of permanent total disability under Balcze-wski, and reversed LIRC's order and remanded the case to the agency for the purpose of permitting Outokumpu to present evidence in rebuttal of Beecher's prima facie case. We accepted review.

II. STANDARD OF REVIEW

m

¶ 22. In an appeal following an administrative agency decision, we review the agency's decision, not the circuit court's. Marshfield Elec. & Water Comm'n v. *153WERC, 2002 WI App 68, ¶ 16, 252 Wis. 2d 656, 643 N.W.2d 122. The issue in this case is whether LIRC properly concluded that Beecher failed to present a prima facie case for permanent total disability on an odd-lot basis. Whether a litigant has established a prima facie case is a question of law. Petrowsky v. Krause, 223 Wis. 2d 32, 36, 588 N.W.2d 318 (1998)(citing Burg v. Miniature Precision Components, 111 Wis. 2d 1, 330 N.W2d 192 (1983)).

¶ 23. In the agency review context, we have recently stated that "labeling an issue as a question of law does not mean that a court may disregard an agency's determination." Brown v. LIRC, 2003 WI 142, ¶ 12, 267 Wis. 2d 31, 671 N.W.2d 279. We will accord an agency's interpretation of a statute great weight deference when: (1) the agency is charged with administration of the particular statute at issue; (2) its interpretation is one of long standing; (3) it employed its expertise or specialized knowledge in arriving at its interpretation; and (4) its interpretation will provide uniformity and consistency in the application of the statute. Id., ¶ 16 (citing Harnischfeger Corp. v. LIRC, 196 Wis. 2d 650, 660 n.4, 539 N.W.2d 98 (1995), and Lisney v. LIRC, 171 Wis. 2d 499, 505, 493 N.W.2d 14 (1992)). A lesser degree of deference, "due weight" deference, "is appropriate when an agency has some experience in the area but has not developed the expertise that necessarily places it in a better position than a court to interpret and apply a statute." Brown, 267 Wis. 2d 31, ¶ 15. No deference is owed to an agency interpretation where the issue is one of first impression, where the agency has no special expertise, or where the agency's position has been so inconsistent that it provides no real guidance. *154Brown, 267 Wis. 2d 31, ¶ 14; Am. Mfrs. Mut. Ins. Co. v. Hernandez, 2002 WI 76, ¶ 14, 252 Wis. 2d 155, 642 N.W.2d 584.

¶ 24. The court of appeals gave LIRC's decision great weight deference, and evidently regarded this choice as too obvious to require explanation. Beecher, 264 Wis. 2d 394, ¶ 13. We think some discussion is in order. LIRC's decision was grounded on its interpretation of the Balczewski case, which the agency supplemented by incorporating § 84.01[4] of the Larson treatise. Great weight or due weight deference may be given to agency interpretations of statutes or administrative rules because the basis for judicial deference is a sense of respect for the legislature's prerogative in conferring power on an agency. By according less than the appropriate level of deference, a court invades, albeit indirectly, the province of the legislature.

¶ 25. The legislature empowered DWD and LIRC to administer the worker's compensation statutes. See Wis. Stat. § 102.01(2)(a), (ap); Wis. Stat. § 102.14. In Balczewski, this court adopted the odd-lot doctrine as a basis for the award of permanent total disability benefits to injured workers who fall within this judicially-recognized category. It is clear that DWD and LIRC have used and applied the principles of Balczewski many times.6

¶ 26. But while DWD and LIRC apply Balczewski in furtherance of their statutorily-conferred duties, it does not follow that we must treat LIRC's decision in this case as we would its interpretation of a statute. *155LIRC's decision in this case was based upon an expansion of the Balczewski odd-lot doctrine by incorporation of a passage from the Larson worker's compensation treatise. The agency's decision does not purport to interpret a statute or administrative rule. This court retains the power to explain, modify, or overrule its own precedents; we need not defer to agency interpretations of our own decisions. See also Local 60, Am. Fed'n of State, County & Mun. Employees, AFL-CIO v. WERC, 217 Wis. 2d 602, 608, 579 N.W.2d 59 (Ct. App. 1998) (no deference owed to an agency interpretation of a statute that conflicts with an appellate decision.) Therefore, we will review de novo LIRC's decision that Beecher failed to establish a prima facie odd-lot case under the Balc-zewski doctrine because his post-injury job search was inadequate.7

*156III. DISCUSSION

A. Introduction

¶ 27. Worker's compensation in Wisconsin is governed primarily by the Worker's Compensation Act (WCA), chapter 102 of the Wisconsin Statutes, and is administered by the DWD. The WCA compensates workers who are injured in the course of their employment. State v. LIRC, 136 Wis. 2d 281, 288, 401 N.W.2d 585 (1987). Benefits fall into two categories: temporary disability benefits, which are payable during an injured worker's healing period, and permanent disability ben*157efits, which are payable if a worker remains disabled after the healing period has ended. Mireles v. LIRC, 2000 WI 96, ¶¶ 7-8, 237 Wis. 2d 69, 613 N.W.2d 875; John D. Neal & Joseph Danas, Jr., Worker's Compensation Handbook § 5.1, at 3 (5th ed. 2003).

¶ 28. The WCA distinguishes two categories of permanent disability benefits: benefits for "scheduled" injuries and benefits for "unscheduled" injuries. Mireles, 237 Wis. 2d 69, ¶ 9. As suggested by the terminology, scheduled injuries are more amenable to standardized benefits calculations. For example, for the loss of an arm at the shoulder, the WCA mandates payment of 500 weeks of benefits indexed to the worker's average pre-injury earnings. Wis. Stat. § 102.52(1). There are many injuries, however, that do not fall into any of the statutory schedules. Such "unscheduled" injuries are primarily injuries to the head and torso, as well as injuries to the mental faculties other than hearing and sight. Neal and Danas, supra, § 5.20, at 15; Wis. Stat. § 102.44(3). The calculation of benefits for unscheduled injuries requires a more individualized approach than is necessary for scheduled injuries. Mireles, 237 Wis. 2d 69, ¶ 13. Back injuries such as Beecher's are unscheduled injuries.

¶ 29. In theory, all permanent disability benefits compensate an employee for the loss of future earnings. Neal and Danas, supra, § 5.15, at 10. This idea is well-rooted in our case law. "[S]ince an award for permanent disability is to be made for all time... it must be based upon some sort of prediction as to impairment of earning capacity." Northern States Power Co. v. Indus. Comm'n, 252 Wis. 70, 76, 30 N.W.2d 217 (1947). If a non-scheduled injury causes a total loss of *158wage earning capacity, the employee is entitled to permanent total disability payments.

¶ 30. Determining loss of future wage earning capacity in non-scheduled injury cases is not an easy matter. Given the highly individualized nature of such injuries, and a job market that is constantly transformed by economic and technological change, predicting how an injury will affect future earning capacity is not an exact science. For this reason, worker's compensation law has evolved to give claimants in this type of case more flexibility to build a case for total permanent disability, and to give agency judges more discretion to rule on the merits of such claims.

B. Balczewski and the Odd-Lot Doctrine

¶ 31. One important change in this direction was this court's recognition in Balczewski of the so-called odd-lot doctrine of permanent total disability. Balczewski, 76 Wis. 2d at 495-96. As Balczewski explained, the odd-lot doctrine probably originated in an opinion by Judge Moulton in the King's Bench decision of Cardiff Corp. v. Hall, 1 K.B. 1009 (1911). Cardiff Corp., 1 K.B. at 494; see also 4 Arthur Larson and Lex K. Larson, Larson's Workers' Compensation Law § 83.02, at 83-3 (2003). The essential idea behind the doctrine is that total disability under worker's compensation law should not be taken literally to mean complete and utter helplessness, because some injured workers find themselves, because of their age, education, training, and overall capacity, "incapable of becoming [] ordinary work[men] of average capacity in any well known branch of the labour market." Cardiff Corp., 1 K.B. at 1020. In the colloquial language that has come to *159represent this doctrine, some injured workers are fit only for the "odd lot" job that appears occasionally and for a short time.

¶ 32. As Balczewski itself made clear, the odd-lot doctrine operates as a rule of evidence. Balczewski, 76 Wis. 2d at 495; see also Neal and Danas, supra, § 5.31, at 21. The doctrine creates a burden-shifting framework that determines which party in a nonscheduled injury worker's compensation case is responsible for producing evidence sufficient to go forward with a claim for permanent total disability on an odd-lot basis. Paraphrasing the Larson treatise, Balczewski held:

[Wjhere a claimant makes a prima facie case that he has been injured in an industrial accident and, because of his injury, age, education, and capacity, he is unable to secure any continuing and gainful employment, the burden of showing that the claimant is in fact employable and that jobs do exist for the injured claimant shifts to the employer.

Balczewski, 76 Wis. 2d at 495.

¶ 33. As noted above, LIRC held that Beecher failed to make a prima facie case of odd-lot disability because, in part, he did not demonstrate sufficient efforts to find suitable post-injury employment. LIRC determined that Beecher's insufficient job search was "a factor that may be considered against him in determining whether he has established a prima facie case of odd-lot unemployability." In support of this proposition, LIRC relied on a passage in Larson's treatise not cited in the Balczewski decision. LIRC also argues on appeal that the enactment of Wis. Stat. § 102.17(7), regarding receipt of expert evidence on loss of earning capacity, and DWD § 80.34, setting forth factors that the DWD is *160to consider in determining loss of earning capacity, support a modification of the Balczewski prima facie odd-lot formula to require the applicant to show reasonable job search efforts. Outokumpu disagrees, and concedes that Beecher met his burden of demonstrating prima facie odd-lot status; the employer instead argues that it successfully rebutted Beecher's prima facie case. Beecher argues that the court of appeals properly reversed LIRC's conclusion that he had not met his prima facie odd-lot burden because of an insufficient post-injury job search.

C. The Larson treatise, § 84.01 [4]

¶ 34. LIRC relied on § 84.01[4] of the Larson worker's compensation treatise to conclude that Beecher was required to demonstrate that he had engaged in a reasonable post-injury job search as part of his prima facie odd-lot case under Balczewski. In fact, however, this passage from Larson is entitled "Burden of Proof of Work Search on Employee in Non-Odd-Lot Cases." Larson, supra, § 84.01[4] (emphasis added). This section of the treatise describes an applicant whose medical impairment "is so limited or specialized in nature that he or she is not obviously unemployable or relegated to the odd-lot category." Id., § 84.01 [4], at 84-8 (emphasis added). Larson proposed that in such cases, it is not unreasonable to place the burden of proof on the employee to establish that he or she "has made reasonable efforts to secure suitable employment." Id.

¶ 35. LIRC argues on this review that since Bal-czewski adopted the odd-lot doctrine from Larson's treatise, all sections of the treatise dealing with the doctrine have been implicitly incorporated into Wisconsin law, whether or not the Balczewski opinion specifically mentioned them. There are a number of problems *161with this argument. First, there was no discussion in Balczewski of the issue of whether the claimant had conducted an adequate post-injury job search. LIRC argues that had the job search issue been raised in Balczewski, this court would have explicitly adopted the passage from Larson that LIRC urges upon us in this case.

¶ 36. We find this suggestion too speculative. This court and the court of appeals have adopted various elements of Larson's treatise in the past, and the normal practice has been to specifically identify which aspects of the treatise are being adopted.8 It is true that the Balczewski opinion incorporated large passages of the Larson text into its analysis. But this does not warrant the conclusion that the Balczewski court in*162tended to adopt everything Larson said about the odd-lot doctrine elsewhere in his book.

¶ 37. Second, the passage in question is not really part of Larson's treatment of the odd-lot doctrine but is rather the author's suggested "corollary" to the "general-purpose principle" stated in the previous sections of the treatise — the sections quoted and relied upon in the Balczewski opinion. There was no mention of this "corollary" in the Balczewski opinion, nor was there any suggestion that an exception existed to the burden-shifting framework established by the court. We therefore conclude that LIRC misread and misapplied Balczewski, erroneously expanding its holding to incorporate elements of Larson's treatise not adopted by this court.

D. Wisconsin Stat. § 102.17(7) and DWD § 80.34

¶ 38. LIRC also argues that certain post-Balczewski enactments require a re-evaluation of the components of an odd-lot prima facie case. Specifically, LIRC contends that Wis. Stat. § 102.17(7)(a), pertaining to the admission of expert evidence in loss of earning capacity cases, and DWD § 80.34, specifying factors the DWD is to consider in determining loss of earning capacity, require a modification of the prima facie odd-lot showing under Balczewski to include evidence of a reasonable job search by the claimant.

¶ 39. Balczewski was decided in 1977. In 1985, the worker's compensation statutes were amended to specify the role of vocational experts in loss of earning capacity cases:

Except as provided in par. (b), in a claim [for partial or total permanent disability], testimony or certified reports of expert witnesses on loss of earning capacity *163may be received in evidence and considered with all other evidence to decide on an employee's actual loss of earning capacity.

1985 Wis. Act 83, creating Wis. Stat. § 102.17(7)(a).

¶ 40. Three years earlier, DWD § 80.34 was adopted. This administrative rule provides a set of factors that the DWD is to consider when assessing the extent of lost earning capacity in cases of permanent disability, both partial and total:

Any department determinations as to loss of earning capacity for injuries arising under s. 102.44(2) and (3), Stats., shall take into account the effect of the injured employee's permanent physical and mental limitations resulting from the injury upon present and potential earnings in view of the following factors:

(a) Age;

(b) Education;

(c) Training;

(d) Previous work experience;

(e) Previous earnings;

(f) Present occupation and earnings;

(g) Likelihood of future suitable occupational change;

(h) Efforts to obtain suitable employment;

(i) Willingness to make reasonable change in a residence to secure suitable employment;

(j) Success of and willingness to participate in reasonable physical and vocational rehabilitation program [sic]; and

(k) Other pertinent evidence.

*164DWD § 80.34(1). The rule makes no mention of odd-lot disability.

¶ 41. LIRC does not explain why it believes the enactment of Wis. Stat. § 102.17(7)(a) should impel us to modify the prima facie case requirements or the burden-shifting evidentiary principle of Balczewski. We decline to develop the agency's argument for it. As to § 80.34(1), LIRC argues very generally that the rule "expands the analysis" required under the odd-lot doctrine as originally set forth in Balczewski. But LIRC does not explain why evidence of a job search should now be considered a necessary element of a prima facie odd-lot case, except to point out that § 80.34 lists it as one of many factors to be considered in the agency's determination of loss of earning capacity. LIRC's argument on this point is almost entirely conelusory.

¶ 42. We see no reason why the adoption of § 80.34 requires alteration of the odd-lot doctrine recognized in Balczewski. The administrative rule sets forth a list of ten factors to be considered by DWD in assessing loss of earning capacity, plus a catch-all category for "other pertinent evidence." DWD § 80.34(1)(a)-(k). The rule applies generally to "[a]ny department determinations as to loss of earning capacity" for injuries causing permanent partial or permanent total disability. It is, however, only a list of factors to be considered in evaluating loss of earning capacity. Section 80.34 is not an evidentiary rule; neither does it purport to impose a burden of proof, establish a set of evidentiary requirements, or mandate the substantive elements of a claim.

¶ 43. Taking the first factor, the applicant's age, as an example, § 80.34 does not eliminate applicants who are below a certain age or otherwise establish a particular evidentiary requirement regarding age that *165must be met in order to qualify for loss of earning capacity benefits. Similarly, the fact that § 80.34 lists "education" and "training" as factors to be considered does not establish any particular evidentiary threshold regarding the level or type of education or training that would be considered disqualifying. The rule lists "[w]ill-ingness to make reasonable change in a residence to secure suitable employment" as a factor to be considered; it does not follow, however, that the rule's inclusion of this factor means that "willingness to relocate" is an element of the claim or that lack of evidence of willingness to relocate is by itself disqualifying. The catch-all "[o]ther pertinent evidence" obviously does not impose an elemental evidentiary requirement. The relative weight of any one of the § 80.34 factors depends upon its consideration in context with the others. The factors listed in the rule thus cannot be considered as necessary, independent elements of a claim for loss of earning capacity.

¶ 44. In any event, and more importantly, the odd-lot doctrine is a judge-made exception to the general rule that permanent total disability awards under worker's compensation law are based on proof of total loss of earning capacity. If an injured employee can show that "because of his injury, age, education, and capacity, he is unable to secure any continuing and gainful employment," he has prima facie placed himself in the odd-lot category, and the burden shifts to the employer to show employability and the availability of jobs. The import of this evidentiary burden-shifting rule is that if the employee can make out a prima facie case for odd-lot disability, then he need not produce evidence that he has attempted to secure suitable alternate employment; he has prima facie established *166his inability to secure continuing and gainful employment, and requiring him to produce evidence of an actual job search is an unnecessary exercise under this doctrine. It is up to the employer under these circumstances to demonstrate that the injured employee is actually employable and that there are actual jobs available to him:

If the evidence of degree of obvious physical impairment, coupled with other factors such as claimant's mental capacity, education, training, or age, places claimant prima facie in the odd-lot category, the burden should be on the employer to show that some kind of suitable work is regularly and continuously available to the claimant.

Balczewski, 76 Wis. 2d at 495, quoting Larson, § 57.61, at 10-136-37.9

¶ 45. LIRC cites Advance Die Casting Co. v. LIRC, 154 Wis. 2d 239, 453 N.W.2d 487 (Ct. App. 1989) as proof that the enactment of DWD § 80.34 altered the formula for establishing a prima facie case of odd-lot disability. Advance Die Casting does not support this proposition.

¶ 46. In Advance Die Casting, the LIRC decision on review referred to DWD § 80.34 in the course of concluding that the injured claimant in the case was permanently, totally disabled on an odd-lot basis. The court of appeals upheld the circuit court's affirmance of LIRC's decision:

The agency's finding that the permanent total disability for Niecko's back injury, coupled with its finding that such injury was permanent due to his loss of earning *167capacity under the sec. Ind. 80.34 factors, rendered that Niecko could not perform any services "other than those which are so limited in quality, dependability, or quantity that a reasonable stable market does not exist" is abundantly clear in this record. At the time of his injury, Niecko was sixty-three years old; he had the equivalent of a high school education; he had worked in [the] Advance foundry for thirty-five years; he was trained only to do heavy work as a die cast operator and setup worker for all of those years in the employ of Advance; and he had no other work training experience. Those facts clearly support the agency's finding of an odd-lot permanent total disability due to the loss of earning capacity.

Id. at 254. The court also concluded that "[t]he agency was also correct in finding that Niecko met the substantial evidence test since he made a prima facie showing that the injury coupled with the odd-lot factors of sec. Ind. 80.34 rendered him permanently totally disabled for regular continuous employment." Id.

¶ 47. On the basis of these generalized references to DWD § 80.34 in an odd-lot case, LIRC now argues that the § 80.34 factors must now he considered elements of the prima facie case under the odd-lot doctrine. This is a far too sweeping reading of Advance Die Casting. The court of appeals in Advance Die Casting said only that the agency had properly considered the factors listed in § 80.34; it did not declare that the claimant was required to bring forward evidence on each factor as part of his prima facie case. More specifically, since there is no reference to a job search in the Advance Die Casting opinion, the court of appeals clearly did not regard this factor as a necessary element of a claimant's prima facie case under the odd-lot doctrine, and neither do we.

*168E. The "Prima Facie Case"

¶ 48. Outokumpu offers a more nuanced view of LIRC's decision. Outokumpu concedes on this review that Beecher did, in fact, present a prima facie case for permanent total disability on an odd-lot basis under Balczewski, but argues that it successfully rebutted it. Outokumpu contends that LIRC simply misspoke when it said that Beecher failed to present a prima facie case, and that when LIRC concluded that Beecher's job search was inadequate, it was already at the "second stage" of the analysis, having concluded (silently, as it were, or contrary to the express words of its decision), that Beecher presented a prima facie case and that Outo-kumpu rebutted it. We cannot accept this interpretation of the LIRC decision's treatment of the job search issue. The language of the decision makes it clear that the agency regarded job search evidence as part of Beecher's prima facie odd-lot burden, a legal proposition that we have rejected as inconsistent with Balczewski.

¶ 49. We take this opportunity, however, to clarify the prima facie case and burden-shifting evidentiary rule of the Balczewski odd-lot doctrine. The expression "prima facie case" is widely invoked in our law but rarely explained. McCormick on Evidence states that courts use the concept of "prima facie case" in two senses: 1) evidence that is sufficient to get to the jury (that is, to survive a directed verdict); and 2) evidence that is sufficient to shift the burden of producing evidence. 2 John W. Strong, McCormick on Evidence § 342, n.4 (5th ed. 1999). It is often difficult to determine which of these two senses is intended in a given court opinion. Id.; see also 9 John Henry Wigmore, Evidence in Trials at Common Law § 2494, at 378 (Chadbourn rev. ed. 1981).

*169¶ 50. In Walter Kassuba, Inc. v. Bauch, 38 Wis. 2d 648, 158 N.W.2d 387 (1968), we characterized a prima facie case as one "established only when evidentiary facts are stated which if they remain uncontradicted by the opposing party's affidavits resolve all factual issues in the moving party's favor." Id. at 655. Thus Wisconsin's use of the term prima facie case falls closer to the second sense of the term noted above. Once a claimant brings forward evidence sufficient to establish a prima facie case, the burden is on the opponent to produce sufficient evidence to go forward with its case.

¶ 51. Perfect eongruity between DWD § 80.34 and Balczewski is impossible because the odd-lot doctrine is in reality an exception to the general rule that a permanently, totally disabled employee has no future earning capacity. The odd-lot worker may indeed have some residual earning capacity after the injury, but the odd-lot doctrine essentially provides that in certain cases the law regards that residue as insignificant. Permanent total disability on an odd-lot basis under Balczewski is a judicially-created category distinct from the general rules governing permanent total disability cases, which otherwise depend upon a determination of total loss of earning capacity.

¶ 52. As a general matter, the burden of proof in a worker's compensation case lies with the claimant. heist v. LIRC, 183 Wis. 2d 450, 457, 515 N.W.2d 268 (1994). Burden of proof can be distinguished into the burden of production and the burden of persuasion. Daniel D. Blinka, Wisconsin Evidence § 301.1 at 64 (2d ed. 2001). As the Kassuba holding makes clear, Wisconsin adheres to the view that a prima facie case repre*170sents a burden of production that, in the absence of adequate rebuttal evidence, satisfies the burden of persuasion on the issue involved.

¶ 53. The distinction between burdens of production and burdens of persuasion is implicated most often in the context of evidentiary presumptions. Balczewski's burden-shifting framework can be analogized to the law of presumptions in civil cases, which assists in sorting through the inter-relationship between § 80.34 and Balczewski. In Wisconsin, presumptions in civil cases are governed by Wis. Stat. § 903.01, which states:

[A] presumption recognized at common law or created by statute, including statutory provisions that certain basic facts are prima facie evidence of other facts, imposes on the party relying on the presumption the burden of proving the basic facts, but once the basic facts are found to exist the presumption imposes on the party against whom it is directed the burden of proving that the nonexistence of the presumed fact is more probable than its existence.

Wis. Stat. § 903.01.

¶ 54. Balczewski holds that certain basic facts— the claimant's injury, age, education, capacity, and training — may in combination demonstrate an inability to secure continuing, gainful employment such that these basic facts constitute prima facie evidence of another (presumed) fact, namely that the claimant is permanently and totally incapable of earning a living. Ordinarily this is accomplished through expert testimony. The employer may introduce expert evidence in contradiction of the basic facts of the employee's prima facie case in order to prevent the presumption from arising. Under Balczewski, however, if the claimant brings forward the basic facts sufficient to satisfy the *171DWD that a prima facie odd-lot case has been made, the presumption is triggered and an obligation is imposed upon the party against whom the presumption runs— here, the employer. That obligation is the burden of proving that the non-existence of the presumed fact is more probable than its existence, or in other words, that it is more probable that the claimant is not permanently and totally incapable of earning a living. Balczewski holds that this burden requires the employer to show that there is an actual job that the claimant can do.

¶ 55. Accordingly, we conclude that the burden that shifts from the claimant to the employer under Balczewski is a burden of persuasion, but only as to the sub-issue of whether a job exists that the claimant can do. The burden of persuasion on the other aspects of the claimant's case for permanent total disability benefits remains, as always, with the claimant. The claimant must prove the industrial injury and medical aspects of his claim, and if the claimant makes a prima facie case for odd-lot unemployability based upon his injury, age, education, training, and capacity, then it falls to the employer to show that there exists suitable employment for the claimant. The employer does this by bringing forward evidence of actual job availability, making it more probable than not that the claimant is able to earn a living.10 The claimant may respond with *172evidence of an actual, futile job search or rely on his expert evidence to defeat the employer's attempted rebuttal. The DWD then determines whether the prima facie odd-lot case under Balczewski has been successfully rebutted. The factors enumerated in DWD § 80.34 may come into play in the agency's ultimate determination of eligibility for benefits, to the extent that they overlap with the odd-lot doctrine, or in a broader sense to the extent that the odd-lot case fails and traditional eligibility rules prevail.

IV CONCLUSION

¶ 56. By recognizing the odd-lot doctrine as a part of Wisconsin's worker's compensation law, Balczewski injected a large dose of realism into the adjudication of disability cases. From the general rule that permanent total disability benefits compensate workers for the total loss of future earning capacity, Balczewski created a narrow exception for seriously-injured workers whose ability to earn occasional, unsteady income does not warrant a denial of permanent total disability benefits or a reduction to partial disability.

¶ 57. To establish a prima facie case for permanent total disability on an odd-lot basis, a claimant *173must present, usually by way of expert reports and testimony, evidence demonstrating to the satisfaction of the DWD that the claimant's injury, combined with his age, education, training, and capacity shows that he is unable to secure continuing, gainful employment, putting him into the odd-lot category. This evidence need not include evidence of a post-injury job search. Once a prima facie odd-lot case is made, it falls to the employer to prove that there is continuous and regular work available to the claimant in order to rebut the prima facie case of odd-lot unemployability. In making the ultimate determination of eligibility for benefits, the agency considers any factors enumerated in DWD § 80.34 that are consistent with the odd-lot doctrine, or more broadly if the agency concludes that the claimant is not prima facie odd-lot or the odd-lot prima facie case has been rebutted.

¶ 58. In this case, LIRC considered the job search factor to be part of Beecher's prima facie odd-lot burden under Balczewski. We have concluded that this was legally improper. Outokumpu has conceded in this court that a prima facie odd-lot case has been made, and asks that we conclude that the prima facie case was successfully rebutted.

¶ 59. We do not decide here whether the now-conceded prima facie odd-lot case has been rebutted. Judicial review of a worker's compensation award is limited to confirming or setting aside the award; modification, amendment or a substitution of the court's judgment for the agency's is improper. See ¶ 20, supra. Accordingly, we set aside the agency's decision and remand this case for rebuttal and any further proceedings consistent with this opinion.

*174By the Court. — The decision of the court of appeals is affirmed.

¶ 60. {concurring) . The decision of the court of appeals should be affirmed. The court of appeals correctly concluded that the Labor & Industry Review Commission's (LIRC's) interpretation of Wisconsin's odd-lot doctrine did not comply with Wisconsin law and therefore was erroneous.

¶ 61. I write separately to express my disagreement with the reasoning of the majority opinion on two issues:

(1) The majority opinion errs in characterizing the odd-lot doctrine explained in Balczewski v. DILHR1 as a "judge-made adjunct to the law of worker's compensation,"2 as "a judge-made exception to the general rule that permanent total disability awards under worker's compensation law are based on proof of total loss of earning capacity,"3 and as a court-created (apparently out of whole cloth and not derived from the Worker's Compensation Act) rule of evidence to be used by LIRC.41 conclude that the Balczewski case adhered to prior case law and correctly characterized the odd-lot doctrine (and its application, including the burden-shifting framework) as an interpretation of the meaning of the *175phrases "total disability"5 and impliedly "loss of earning capacity" as used in Wisconsin's Worker's Compensation Act (WCA).

(2) The majority opinion errs by characterizing LIRC's interpretation of the odd-lot doctrine as an interpretation of the Balczewski case not entitled to any deference. I conclude that LIRC's interpretation of the odd-lot doctrine is an interpretation and application of the WCA (including this court's interpretations of the WCA) and is entitled to great weight deference.

¶ 62. Although the agency's determination in this case should be given great weight deference, I conclude, for many of the reasons set forth in the majority opinion and in the opinion of the court of appeals,6 that LIRC's interpretation of the law is erroneous and therefore is unreasonable. Accordingly, this court should not affirm the agency's interpretation of the law, but should, as the court of appeals did, reverse and remand this matter.

¶ 63. The majority opinion errs in characterizing the odd-lot doctrine explained in Balczewski as a "judge-made adjunct to the law of worker's compensation,"7 as "a judge-made exception to the general rule that permanent total disability awards under worker's compen*176sation law are based on proof of total loss of earning capacity,"8 and as a court-created (apparently out of whole cloth and not derived from the WCA) rule of evidence to be used by LIRC.9

¶ 64. A careful examination of the Balczewski decision demonstrates that Balczewski's adoption of the Larson treatise language did not create a new judge-made common-law rule for the WCA. The Balczewski decision interpreted the WCA, identified and labeled prior case law as the famed "odd-lot" doctrine, and distilled and developed prior case law into a two-step, burden-shifting interpretation of the WCA, a more clearly enunciated interpretation of the WCA that the court had adopted in prior cases. The Balczewski court clearly understood, and explicitly stated, that it was taking this approach.

¶ 65. The Balczewski court unambiguously viewed the odd-lot doctrine as an interpretation of the WCA, stating as follows: "We think it clear that what Larson refers to as the 'odd-lot' doctrine is a statement of the Wisconsin law as it has existed at least since the 1923 amendments to the Worker's Compensation Act."10 " 'Total disability' in compensation law is not to be interpreted literally . .. ."11

¶ 66. The existing Wisconsin case law interpreting the WCA and repeatedly referred to by the Balcze-wski court12 began in the 1917 McDonald v. Industrial *177Commission13 case, one of the earliest Wisconsin cases considering whether claimants with injuries that permit them to work sporadic, odd-lot jobs nonetheless qualify for permanent total disability benefits under the WCA. The McDonald court affirmed a finding of permanent total disability under the WCA despite medical testimony that the injured worker could perform work if it did not involve much walking or stooping.14 In deciding whether to affirm such a finding, this court considered whether the claimant's substantial but not total loss of earning capacity qualified as a compensable injury under the then permanent total disability statute.15 This court concluded that it did.

¶ 67. In 1944, in Milwaukee Western Fuel Co. v. Industrial Commission, this court continued looking at loss of actual earning capacity, although the opinion did not cite McDonald or the statute interpreted therein.16 This court affirmed a permanent total disability award even though the claimant was able to shovel snow, run errands, and work for short periods of time until dizziness and weakness took hold.17

¶ 68. In 1972, in Kurschner v. DILHR, 18 a case expressly relied upon by Balczewski, this court again recognized that, in cases of non-schedule industrial *178injuries, the crucial factor in establishing permanent total disability is proof of impairment of actual earning capacity in the industrial labor market.19 The Balcze-wski court noted that Kurschner does not stand alone. Rather, stated the Balczewski court, the Kurschner case is merely "the culmination of a long line of cases interpreting the basis of compensation for nonschedule injuries under the Wisconsin Workmen's [now Worker's] Compensation Act."20

¶ 69. Finally, in 1972, in Transamerica Insurance Co. v. DILHR,21 this court again spoke on the odd-lot doctrine, although again not specifically labeling the doctrine as such. In Transamerica, the court affirmed a finding of permanent total disability because the claimant was permanently and totally disabled from performing labor at his trade, as well as being permanently and totally disabled from performing manual or other labor in another "suitable" employment22 Furthermore, this court declared loss of earning capacity as "one important measuring stick as to degree of disability."23 For both propositions, the Transamerica court cited the cases discussed above.

¶ 70. Not only did the Balczewski court view prior cases as adopting the rationale of the odd-lot doctrine as a matter of statutory interpretation, but also both the claimant and the Attorney General (representing the *179agency) agreed in Balczewski that the odd-lot doctrine was part of the WCA, just not by that name. The claimant's brief in Balczewski stated that "it has long been recognized in Wisconsin Workers' Compensation law that a finding of total disability for industrial purposes can be proper despite the fact that a claimant is capable of... miscellaneous light work."24 The Attorney General's brief in Balczewski states that DILHR and the State "agree with the claimant that [the odd-lot doctrine] is implicit" in the WCA, that the doctrine is but another statement of existing Wisconsin law, and that DILHR has made findings of permanent total disability based on the odd-lot doctrine.25

¶ 71. By identifying, distilling, and labeling the developing statute-based rule of the McDonald line of cases, the Balczewski court was interpreting the foundational legal terms of art in the WCA: "total disability" and "loss of earning capacity." The Balczewski court concluded that" '[t]otal disability' in compensation law is not to be interpreted literally as utter and abject helplessness. Evidence that claimant has been able to earn occasional wages ... does not necessarily rule out a finding of total disability nor require that it is to be reduced to partial."26

¶ 72. Balczewski is, without doubt, a statutory interpretation case. Indeed, even the cases from which Professor Larson's odd-lot doctrine emerges are cases interpreting workers' compensation acts.

*180¶ 73. For example, the Balczewski court referenced Lee v. Minneapolis Street Railway Co.,27 a case Professor Larson thinks exemplary and representative of odd-lot doctrine cases.28 That case expressly interpreted a phrase of Minnesota's Workers' Compensation Act: "working at an occupation which brings him an income."29 The Minnesota Supreme Court held that the phrase implies "at least a reasonable degree of continuity of occupational capacity," and that sporadic work for short periods of time is not enough, on its own, to negate a permanent total disability determination.30

¶ 74. Again citing the Larson treatise, the Balcze-wski court described the origins of the name "odd-lot."31 Professor Larson attributes the doctrine name to the British case of Cardiff Corp. v. Hall, a case discussed in the majority opinion.32 Again, this case interpreted a statute, this time the British Workmen's Compensation Act of 1906 and the phrase "earning or.. . able to earn."33

*181¶ 75. Finally, the Balczewski court looked to Judge Benjamin N. Cardozo although not cited by the Larson treatise for further guidance in expressing the odd-lot doctrine.34 That case, Jordan v. Decorative Co.,35 was also a statutory interpretation case, discussing the parameters of the phrase "wage earning capacity" in the New York workers' compensation statute.36

¶ 76. I therefore conclude that Balczewski and the odd-lot doctrine are judicial interpretations of the WCA, not judge-made rules that are adjunct to the WCA or an exception to the WCA.

¶ 77. LIRC perceived Wis* Stat. § 102.17(7)(a), Wis. Admin. Code § DWD 80.34, and Professor Larson's text (§ 84.01[4]) as altering the Balczewski line of cases interpreting the WCA. I therefore turn to the question of whether LIRC's interpretation of the odd-lot doctrine of the WCA is entitled to great weight deference.

t — 1 Í-H

¶ 78. I agree with the court of appeals that LIRC's interpretation of the odd-lot doctrine is entitled to great weight deference. Thus this court will not overturn LIRC's determination unless it is unreasonable. I agree with the court of appeals that because LIRC did not adhere to the WCA as interpreted by the court, LIRC's determination was unreasonable.

A

¶ 79. Wisconsin courts must afford great weight deference to LIRC's statutory interpretation of the odd-lot doctrine in the present case because the follow*182ing conditions have been met:37 (1) the agency is charged with administration of the particular statute at issue;38 (2) its interpretation is one of long standing;39 (3) it employed its expertise or specialized knowledge in arriving at its interpretation;40 and (4) its interpretation will provide uniformity and consistency in the application of the statute at issue.

¶ 80. The majority opinion concludes that no deference at all is owed to LIRC in the present case. The majority opinion reaches this conclusion because it erroneously conflates a court's deference to the agency's interpretation of a statute with a court's determination of the reasonableness of the agency's interpretation of a statute. A court may give an agency great weight deference and yet conclude that the agency's determination of law is unreasonable and should be reversed.41

*183¶ 81. The majority opinion's discussion of the appellate standard of review for LIRC's decision in the present case42 is apparently based on the premise that the odd-lot doctrine is a judge-made doctrine created by this court in Balczewski and that the odd-lot doctrine is not based on the WCA or on prior case law interpreting the WCA, but is based on the common practice of other states (as set forth in Professor Larson's Workmen's Compensation Law).43

¶ 82. Using this faulty underlying assertion, the majority opinion summarily concludes that even if LIRC's interpretation of Balczewski is of long standing, no deference is owed to LIRC because LIRC is interpreting a decision of this court, not a statute or administrative rule.44 Not only does the majority opinion refuse to accord the agency's interpretation of the WCA any deference, but the majority opinion also goes even further and expansively asserts that "we need not defer to agency interpretations of our own decisions."45

¶ 83. As support for its erroneous broad legal proposition, the majority cites Local 60, American Federation of State, County and Municipal Employees v. WERC.46 The court of appeals in Local 60 opined that "it is well established that the general deference given to an agency's application of a particular statute does *184not apply when the agency's determination conflicts with prior case law established by our supreme court." The Local 60 court of appeals cited Doering v. LIRC47 as authority for this proposition, and Doering in turn cites the Pabst48 and Klusendorf49 cases.

¶ 84. Neither Pabst nor Klusendorf supports the statement of law set forth in Local 60, in Doering, and in the majority opinion. Rather, both Pabst and Klusen-dorf (cases decided before this court adopted the levels of deference) stated the law correctly: "[C]ourts should not substitute their judgment for the agency's application of a particular statute to the found facts if a rational basis exists in law for the agency's interpretation and it does not conflict with the statute's legislative history, prior decisions of this court, or constitutional prohibitions."50

¶ 85. I agree that if an agency's interpretation of law conflicts with prior decisions of this court interpreting the law, then, even with great weight deference, the agency's interpretation is not rational and a court need not follow the interpretation regardless of the deference accorded the agency's determination.51

*185¶ 86. The court of appeals recently refused to follow the "no deference rule" pronounced in Local 60 and Doering and granted great weight deference to a DWD determination based on a court case, reasoning as follows:52

The fact that the [DWD] also looked to the supreme court's discussion in [.Richland School District v. DILHR, in addition to a statutory provision] in deciding this [claim] does not mean that we may deprive the department of the deference it is due on account of its expertise, experience and legislatively delegated authority.53

¶ 87. The standard of review in the majority opinion and Local 60 is not supported by Wisconsin law and is simply illogical and wrong. Agencies interpret and apply court interpretations of statutes all the time. By failing to afford LIRC the proper level of deference in such cases, the majority opinion enables the courts of this state to usurp legislative power.54 The majority *186opinion's change in the level of deference afforded an agency's interpretation of a statute undermines the stability and uniformity that is necessary in the administration of the WCA.

B

¶ 88. The majority opinion's characterization of LXRC's determination as one based purely on Balczewski and § 84.01[4] of Larson's Workers' Compensation Law,55 and one that "does not purport to interpret a statute or administrative rule,"56 is erroneous. LIRC,57 Beecher,58 the court of appeals,59 and I interpret the LIRC decision as relying on Wis. Stat. § 102.17(7)(a), Wis. Admin. Code § DWD 80.34, case law interpreting *187the WCA (Balczewski), and the Larson treatise. The majority opinion protests.60

¶ 89. The majority opinion parses the LIRC decision to reinterpret it so that it does not rely on anything but Balczewski and the Larson text to justify according the LIRC decision no deference.

¶ 90. Not only does a reading of the LIRC decision itself demonstrate the majority opinion's error,61 but numerous LIRC cases that preceded the present case regularly rely on Wis. Admin Code § DWD 80.34, Balc-*188zewski, and Larson,62 as supporting the proposition that "to make the prima facie case of odd-lot unemploy-ability, the commission requires applicants undertake reasonable efforts to find work."63

*189c

¶ 91. Under the great weight deference standard of review this court's task should have been to determine whether LIRC's conclusion of law that an injured worker must demonstrate that he or she made a reasonable post-injury job search as part of his or her prima facie odd-lot case is a reasonable interpretation of Balczewski, Wis. Stat. § 102.17(7)(a), and Wis. Admin. Code § DWD 80.34. Great weight deference requires this court to sustain LIRC's interpretation of the law "even if an alternative view of the law is just as reasonable or even more reasonable," but requires this court to overturn LIRC's determination if LIRC's interpretation of the law is unreasonable.64

¶ 92. Great weight deference is not a result-oriented standard of review and its use does not mean that a court must rubber-stamp an agency decision. *190Applying the "great weight" deferential standard of review, I agree with the court of appeals and the majority opinion that LIRC's interpretation of the odd-lot doctrine cannot be sustained. For substantially the same reasons as the court of appeals and the majority opinion set forth, I conclude that LIRC's interpretation is unreasonable.

¶ 93. For the reasons set forth, I would affirm the court of appeals decision. I write separately in order to clarify the statutory basis of the odd-lot doctrine and the level of deference that the majority opinion should have afforded LIRC's statutory interpretation.

¶ 94. I am authorized to state that Justice ANN WALSH BRADLEY joins this concurrence.