CUDAHY, Circuit Judge.

Daniel Vanskike, an inmate at the State-ville Correctional Center in Joliet, Illinois, has performed various work assignments while in prison. In this appeal we must decide whether Vanskike is entitled to the federal minimum wage for his work by virtue of being an “employee” under the Fair Labor Standards Act.

Vanskike filed a pro se complaint against the Director of the Illinois Department of Corrections (DOC), alleging that the DOC used and continues to use prisoners for work assignments. The complaint alleges that Vanskike has done “forced labor” as a janitor, kitchen worker, gallery worker and “knit shop piece-line worker” while incarcerated at Stateville and Menard Correctional Centers. It charges that the DOC does not compensate working prisoners with “equal minimum fairness employment compensation” ordinarily paid to “any normal employee ... on a[n] hour for hour wage.”

The district court construed the claim as one brought under the Fair Labor Standards Act (FLSA), 29 U.S.C. §§ 201 et seq., and granted leave to proceed in forma pauperis. A magistrate judge denied the *807 plaintiff’s motion for appointment of counsel. The district court then granted the DOC’s motion to dismiss pursuant to Fed. R.Civ.P. 12(b)(6), concluding that prisoners are not “employees” under the Fair Labor Standards Act and that neither the DOC nor the State of Illinois acts as an “employer” with respect to the prisoners. 1

II.

We review the grant of a motion to dismiss de novo, assuming the truth of all factual allegations and drawing reasonable inferences in favor of the plaintiff. Prince v. Rescorp Realty, 940 F.2d 1104, 1106 (7th Cir.1991). Because the plaintiff in this case was proceeding pro se in the district court, his complaint must be liberally construed to ensure that his claims receive fair and meaningful consideration. Haines v. Kerner, 404 U.S. 519, 520-21, 92 S.Ct. 594, 595-96, 30 L.Ed.2d 652 (1972); Matzker v. Herr, 748 F.2d 1142, 1146 (7th Cir.1984).

The FLSA, which was enacted in 1938, requires employers to pay their employees a minimum hourly wage — currently $4.25. 29 U.S.C. § 206(a)(1) (1992 Supp.). The Act defines “employee” in a circular fashion, as “any individual employed by an employer,” 29 U.S.C. § 203(e)(1), including “an individual employed by a State,” 29 U.S.C. § 203(e)(2)(C). The term “employer” includes “a public agency.” 29 U.S.C. § 203(d). The FLSA defines the term “employ” as “to suffer or permit to work.” 29 U.S.C. § 203(g).

Vanskike argues initially that the FLSA’s “clear and unambiguous” language places working prisoners squarely within the scope of the minimum wage requirement. Vanskike relies on Demarest v. Manspeaker, 498 U.S. 184, 111 S.Ct. 599, 112 L.Ed.2d 608 (1991), in which the Supreme Court held that a state prisoner was entitled to a statutory witness fee as a “witness ... in attendance” under 28 U.S.C. § 1821. In Demarest, however, there was no question that the prisoner was literally a “witness” within the plain language of the statute. Here it is simply not so clear. The statute itself provides little assistance, and the term “employee” does not obviously include prisoners who perform work within a prison. 2 When it comes to such appeals to “plain” or “clear” language, perhaps our best guide consists of our common linguistic intuitions, and those intuitions are at least strained by the classification of prisoners as “employees” of the DOC or of the State. Moreover, words have meaning in context, and it is rare that statutory terms may be responsibly applied without any consideration whatsoever of the statutory context , and purposes. See First Chicago Corp. v. Commissioner, 842 F.2d 180, 183 (7th Cir.1988).

The Supreme Court has instructed the courts to construe the terms “employee” and “employer” expansively under the FLSA. Nationwide Mutual Ins. Co. v. Darden, — U.S. -, -, 112 S.Ct. 1344, 1350, 117 L.Ed.2d 581 (1992); Rutherford Food Corp. v. McComb, 331 U.S. 722, 730, 67 S.Ct. 1473, 1476, 91 L.Ed. 1772 (1947). Nevertheless, courts have generally declined to extend the FLSA’s minimum wage provision to prisoners who work in prison. See Miller v. Dukakis, 961 F.2d 7, 8 (1st Cir.1992); Gilbreath v. Cutter Biological, Inc., 931 F.2d 1320, 1328 (9th Cir.1991); Alexander v. Sara, Inc., 721 F.2d 149, 150 (5th Cir.1983); Wentworth v. Solem, 548 F.2d 773, 775 (8th Cir.1977); Emory v. United States, 2 Cl.Ct. 579, 580 *808 (1983), aff'd, 727 F.2d 1119 (Fed.Cir.1983); Worsley v. Lash, 421 F.Supp. 556, 556 (N.D.Ind.1976); Sims v. Parke Davis & Co., 334 F.Supp. 774, 787 (E.D.Mich.1971), aff'd, 453 F.2d 1259 (6th Cir.1971), cert. denied, 405 U.S. 978, 92 S.Ct. 1196, 31 L.Ed.2d 254 (1972); Hudgins v. Hart, 323 F.Supp. 898, 899 (E.D.La.1971); Huntley v. Gunn Furniture Co., 79 F.Supp. 110, 116 (W.D.Mich.1948).

Because status as an “employee” for purposes of the FLSA depends on the totality of circumstances rather than on any technical label, courts must examine the “economic reality” of the working relationship. Goldberg v. Whitaker House Cooperative, Inc., 366 U.S. 28, 33, 81 S.Ct. 933, 936, 6 L.Ed.2d 100 (1961); Rutherford Food Corp., 331 U.S. at 730, 67 S.Ct. at 1476; Secretary of Labor v. Lauritzen, 835 F.2d 1529, 1535 (7th Cir.1987), cert. denied, 488 U.S. 898, 109 S.Ct. 243, 102 L.Ed.2d 232 (1988). Vanskike places heavy reliance on three recent cases in which courts applied the “economic reality” standard to prisoners and found that they could qualify as “employees” under the FLSA. In Carter v. Dutchess Community College, 735 F.2d 8 (2d Cir.1984), the plaintiff was a prisoner who worked as a teaching assistant at (and was paid by) a local community college. The court held that prisoners were not categorically excluded from the FLSA minimum wage provision and remanded the case for an inquiry into the circumstances of employment. Id. at 15. In Watson v. Graves, 909 F.2d 1549 (5th Cir.1990), the court held that the plaintiffs, who worked for a construction company outside the prison under a work release program, were “employees” of the company and were entitled to the federal minimum wage. Id. at 1554-56. And in Hale v. Arizona, 967 F.2d 1356 (9th Cir.1992), the court applied the FLSA to prisoners who worked for a state entity making products in prison for sale or use outside the prison. Each of these three cases, in examining the “economic reality” of the putative employment relationship, considered four factors set forth in Bonnette v. California Health & Welfare Agency, 704 F.2d 1465, 1470 (9th Cir.1983): “whether the alleged employer (1) had the power to hire and fire the employees, (2) supervised and controlled employee work schedules or conditions of employment, (3) determined the rate and method of payment, and (4) maintained employment records.”

We do not question the conclusions of Carter, Watson and Hale that prisoners are not categorically excluded from the FLSA’s coverage simply because they are prisoners. We must nevertheless reject Vanskike’s contention that he is an “employee” for purposes of the FLSA. We also decline to apply Bonnette’s four-factor standard in this situation. Carter and Watson involved situations quite different from the one here. In both cases the prisoners performed work for private, outside employers. In addition, they were given the choice to work rather than being assigned to do so. Here, in contrast, there is no suggestion of an outside employer. Vanskike alleges only that he worked for the Department of Corrections, and that his labor was “forced.” We do not think that this is a borderline case like Carter or Watson. The courts have not extended the FLSA’s definition of “employee” to cover prisoners who are assigned to work within the prison walls for the prison. Indeed, cases like Carter and Watson appear to have proceeded under the assumption that the FLSA would not apply in such circumstances. 3 The Ninth Circuit’s recent decision in Hale presents a situation different from Carter and Watson: the prisoners in Hale worked within the prison walls and for a state agency. That case does not persuade us, however, that Vanskike should be considered an “employee” of the *809 DOC. First, the case appears to be distinguishable, since the prisoners in Hale were actually employed by entities that were deemed “private enterprises” under state law. Second, the decision is in tension with the Ninth Circuit’s earlier Gilbreath decision, in which a different panel of that court held that the FLSA does not apply to prisoners working within the prison for a private plasma center (a situation that would seem to present a stronger case for application of the FLSA). 931 F.2d at 1326; id. at 1330-31 (Rymer, J., concurring); see also Hale, 967 F.2d at 1368 (Fletcher, J., dissenting). Finally, to the extent that Hale may rule that a prisoner working within the prison and for the prison is an “employee” of the prison under the FLSA, we respectfully disagree with its conclusion.

Under Illinois law, the DOC is authorized to assign work to prisoners, and to provide wages for the work. The legislature’s purpose in authorizing prisoner work assignments is to “equip such persons with marketable skills, promote habits of work and responsibility and contribute to the expense of the employment program and the committed person’s cost of incarceration.” Ill.Rev.Stat. ch. 38, ¶ 1003-12-1 (1991). It is strictly forbidden, however, for the DOC to sell, contract or hire out a prisoner’s labor except as permitted under the work release provision of the statute. Ill.Rev. Stat. ch. 38, ¶ 1003-12-2 (1991). There is no indication that the DOC has a pecuniary, in contrast to a rehabilitative or penological, interest in inmate labor. See Gilbreath, 931 F.2d at 1330 (Rymer, J., concurring). Moreover, the relationship between the DOC and a prisoner is far different from a traditional employer-employee relationship, because (certainly in these circumstances) inmate labor belongs to the institution. Id. at 1331. The Thirteenth Amendment excludes convicted criminals from the prohibition of involuntary servi tude, so prisoners may be required to work. Draper v. Rhay, 315 F.2d 193, 197 (9th Cir.), cert. denied, 375 U.S. 915, 84 S.Ct. 214, 11 L.Ed.2d 153 (1963). Further, there is no Constitutional right to compensation for such work; compensation for prison labor is “by grace of the state.” Sigler v. Lowrie, 404 F.2d 659, 661 (8th Cir.1968), cert. denied, 395 U.S. 940, 89 S.Ct. 2010, 23 L.Ed.2d 456 (1969). That there is no Constitutional right does not, however, foreclose the possibility of a statutory right to compensation.

What is the “economic reality” of the relationship between the DOC and Van-skike? As noted earlier, several other courts have applied the four-factor Bon-nette standard in determining the status of prisoners who work. We think, however, that that standard is not the most helpful guide in the situation presented here. The Bonnette factors, with their emphasis on control over the terms and structure of the employment relationship, are particularly appropriate where (as in Bonnette itself) it is clear that some entity is an “employer” and the question is which one. The dispute in this case is a more fundamental one: Can this prisoner plausibly be said to be “employed” in the relevant sense at all? Consider a literal application of the Bon-nette factors in the present context. The DOC might be said to have “had the power to hire and fire” Vanskike; it surely “supervised and controlled” his schedule and work conditions, determined his pay and (presumably) kept work records. But the Bonnette factors fail to capture the true nature of the relationship for essentially they presuppose a free labor situation. Put simply, the DOC’s “control” over Van-skike does not stem from any remunerative relationship or bargained-for exchange of labor for consideration, but from incarceration itself. The control that the DOC exercises over a prisoner is nearly total, and control over his work is merely incidental to that general control. Indeed, the Thirteenth Amendment’s specific exclusion of prisoner labor supports the idea that a prisoner performing required work for the prison is actually engaged in involuntary servitude, not employment. 4 The same ba *810 sic point can be made in a slightly different way. Bonnette’s emphasis on control harkens back to the common law distinction between an employer and an independent contractor; indeed, the cases cited for Bonnette’s four-factor standard involve precisely that distinction. See Bonnette, 704 F.2d at 1470 (citing Real v. Driscoll Strawberry Assoc., 603 F.2d 748, 756 (9th Cir.1979); Hodgson v. Griffin & Brand of McAllen, Inc., 471 F.2d 235, 237-38 (5th Cir.), cert. denied, 414 U.S. 819, 94 S.Ct. 43, 38 L.Ed.2d 51 (1973)). In those cases the question is essentially whether there is enough control over the individual to classify him as an employee. But here we are coming at the definition of “employee” from the opposite direction: there is obviously enough control over the prisoner; the problematic point is that there is too much control to classify the relationship as one of employment. The Bonnette factors thus primarily shed light on just one boundary of the definition of “employee,” and we are concerned with a different boundary. Prisoners are essentially taken out of the national economy upon incarceration. When they are assigned work within the prison for purposes of training and rehabilitation, they have not contracted with the government to become its employees. Rather, they are working as part of their sentences of incarceration. See, e.g., Harris v. Yeager, 291 F.Supp. 1015, 1017 (D.N.J.1968) (payments for prison work assignments “are not wages in a realistic economic employer-employee relationship”), aff'd, 410 F.2d 1376 (3d Cir.1969). Because Van-skike’s allegations reveal that he worked in the prison and for the DOC pursuant to penological work assignments, the economic reality is that he was not an “employee” under the FLSA. 5

The purposes underlying the FLSA bolster our conclusion. “The central aim of the Act was to achieve, in those industries within its scope, certain minimum labor standards.” Mitchell v. Robert DeMario Jewelry, Inc., 361 U.S. 288, 292, 80 S.Ct. 332, 335, 4 L.Ed.2d 323 (1960). Congress sought to correct labor conditions that are “detrimental to the minimum standard of living necessary for health, efficiency, and general well-being of workers.” 29 U.S.C. § 202(a). In addition, the FLSA was intended to prevent unfair competition in commerce from the use of underpaid labor. 29 U.S.C. § 202(a)(3).

The first purpose of the FLSA has little or no application in the context presented here. Prisoners’ basic needs are met in prison, irrespective of their ability to pay. Requiring the payment of minimum wage for a prisoner’s work in prison would not further the policy of ensuring a “minimum standard of living,” because a prisoner’s minimum standard of living is established by state policy; it is not substantially affected by wages received by the prisoner. It is true, as Vanskike points out, that some cases have characterized the FLSA’s primary purpose more specifically, as aimed at “substandard wages and oppressive working hours.” Barrentine v. Arkansas-Best Freight Sys., Inc., 450 U.S. *811 728, 739, 101 S.Ct. 1437, 1444, 67 L.Ed.2d 641 (1981). The evil of substandard wages, however, as just noted, does not apply where worker welfare is not a function of wages. As for oppressive working hours, Vanskike alleges only that he was underpaid, not that he was overworked, so only the minimum wage provision of 29 U.S.C. § 206, and not the separate working-hours provision of § 207, is directly at issue here. A prisoner may, of course, challenge his conditions of incarceration under applicable statutory and Constitutional provisions such as the Eighth Amendment. But the fundamental goal of ensuring workers’ welfare and standard of living does not call for the application of the minimum wage in these circumstances.

The second purpose of the Act — preventing unfair competition — presents a closer and more difficult question. Cheap labor can give an unfair advantage to an enterprise that competes in the marketplace. Vanskike argues that low- or no-wage prison labor surely raises the specter of such unfair competition. The Ninth Circuit’s recent decision in Hale relied heavily on this second purpose of the FLSA to find working prisoners covered by the Act. The court noted that the prisoners produced goods that were ultimately sold in the private sector as well as goods that were used by state and local governments (such as license plates). The use of cheap prison labor to make such products, the court concluded, poses a risk of unfair competition.

In the case before us, it is not at all clear whether Vanskike works to produce goods that are distributed outside the prison. Assuming, for example, that he works to manufacture license plates, then the state (as producer) has an advantage over other potential producers of license plates in the economy, because it is able to produce that item at low cost. But this version of the unfair competition rationale is not limited to the production of goods. It would also seem to extend to simple service work, such as that suggested by Vanskike’s “janitor” and “kitchen worker” assignments. For every prisoner who is assigned to sweep a floor or wash dishes for little or no pay, there is presumably someone in the outside world who could be hired to do the job — someone who would have to be paid at least $4.25 an hour. This approach to the FLSA’s second purpose thus cuts a broad swath: carried to its logical conclusion, prisoners must be paid minimum wage for anything they do in prison that can be considered “work.”

We do not believe that Congress intended the FLSA to dictate such a result, even given its goal of preventing unfair competition. As noted earlier, the State of Illinois strictly regulates the use of prison labor. More important, Congress has addressed the problem of unfair competition by regulating prison-made goods. The Ashurst-Sumners Act, 18 U.S.C. §§ 1761-62, penalizes the knowing transportation of prison-made goods in commerce and was specifically intended to combat unfair competition. Kentucky Whip & Collar Co. v. Illinois Central R.R. Co., 299 U.S. 334, 351, 57 S.Ct. 277, 282, 81 L.Ed. 270 (1937). The Ashurst-Sumners Act undermines Vanskike’s argument in two ways. First, it is noteworthy that the Act exempts commodities manufactured for use by federal, state and local governments. Thus, Congress has already struck the balance by precluding a wide range of inmate-labor competition while permitting governments to use the fruits of such labor. That Congress drew the line where it did suggests that it considered a certain range of prison labor for the benefit of government outside the boundaries of the targeted evil. The second purpose of the FLSA coincides with the single purpose of the Ashurst-Sumners Act — preventing unfair competition — and the latter statute, by its exception for goods used by government, belies the notion that any and all uses of prison labor by the government unduly obstruct fair competition. Moreover, the balance struck by Congress makes sense. For the government, competition in the marketplace is not a dominant mode and profits are not the ultimate goal. A governmental advantage from the use of prisoner labor is not the same as a similar low-wage advantage on the part of a private entity: while the latter *812 amounts to an unfair windfall, the former may be seen as simply paying the costs of public goods — including the costs of incarceration (as the Illinois statute expressly provides).

Second, the Ashurst-Sumners Act supports the conclusion that Congress did not intend to extend the FLSA’s definition of “employee” to prisoners working in prison. The Ashurst-Sumners Act was enacted in 1935 — just three years before the enactment of the FLSA. It is difficult to imagine that Congress would have enacted legislation in 1938 that rendered its recently passed prison-goods law essentially superfluous (for the FLSA, so construed, would have addressed the problem of unfair competition from cheaply made prison goods by eliminating the low-labor-cost advantage). Even if one could plausibly accept this scenario, it is harder to hypothesize why, assuming the FLSA was intended to cover prisoners, Congress continued to make several minor amendments to and to recodify the Ashurst-Sumners Act over the years. Given Congress’s specific treatment of unfair competition in the prison labor context, we find it difficult to conclude that the federal minimum wage was intended to address unfair competition arising from prison labor performed for the prison.

Vanskike points out that the Ashurst-Sumners Act applies only to goods, not services. This distinction might carry some weight if there were a suggestion in the present case of a private, outside employer using the labor of prisoners. It is in such cases, like Watson and Carter, that the services of prisoners might be used to confer a prohibited unfair advantage on a competing business. We do not think, however, that the DOC’s use of prison labor presents such a case. Moreover, while it is true that the Ashurst-Sumners Act does not directly regulate inmate labor, the Act clearly presupposes that the labor of prisoners performed for the prison belongs to the institution; that is why the fruits of prison labor are assumed to be low-cost goods.

In sum, it cannot be denied, as the Ninth Circuit found in Hale, that the unfair competition rationale, broadly conceived, triggers some concerns in the context of prison labor. Nevertheless, we conclude that this second purpose of the FLSA does not call for application of the minimum wage provision in these circumstances, for several reasons. First, we emphasize that Vanskike was not in a true economic employer-employee relationship with the DOC, so the statutory language does not cover him. Second, as the court in Hale appears to concede, the first purpose of the FLSA is basically not implicated here. And third, Congress’s approach to prison labor in the Ashurst-Sumners Act suggests that the FLSA was not designed to encompass the present scenario, even given its general aim of preventing unfair competition. 6

Vanskike challenges the district court’s ruling that he could not bring his claim as a class action. The court rejected Vanskike’s request for class certification, noting that the FLSA “expressly disallows class actions.” It is not clear, however, that the district judge intended such a sweeping declaration, for he proceeded to state that “[o]nly prisoners who have filed written consents” are bound by such actions under the Act. In any event, class actions are expressly provided for under the FLSA, although they are governed by 29 U.S.C. § 216(b) rather than by Federal Rule of Civil Procedure 23. The difference is that under § 216(b) the class member must opt in to be bound, while under Rule *813 23 the class member must opt out in order not to be bound. Woods v. New York Life Ins. Co., 686 F.2d 578, 580 (7th Cir.1982).

Nevertheless, a class representative must have a cause of action in his own right in order to bring a class action. See, e.g., Tidwell v. Schweiker, 677 F.2d 560, 566 (7th Cir.1982), cert. denied, 461 U.S. 905, 103 S.Ct. 1874, 76 L.Ed.2d 806 (1983). Because we have concluded that Vanskike has failed to state a claim, we also conclude that he is not entitled to pursue a class action on the present claim.

Finally, Vanskike argues that the district court abused its discretion in denying his motion for appointment of counsel. Although counsel might have been a big help here, we conclude reluctantly that there was no abuse of discretion in the district court’s application of the five-factor standard for the appointment of counsel for an indigent. See Merritt v. Faulkner, 697 F.2d 761, 764 (7th Cir.), cert. denied, 464 U.S. 986, 104 S.Ct. 434, 78 L.Ed.2d 366 (1983); Maclin v. Freake, 650 F.2d 885, 887-89 (7th Cir.1981). 7 We have determined that Vanskike’s complaint fails, as a matter of law, to state a valid claim upon which relief may be granted. In addition, we note that Vanskike’s appointed counsel on appeal have served him vigorously and effectively, and we are confident that Van-skike’s present claim has received a full and fair hearing.

III.

For the foregoing reasons, the judgment of the district court is

Affirmed.

. The district court construed the complaint’s references to "forced labor” as a claim under 42 U.S.C. § 1983 for violation of the Thirteenth Amendment and dismissed that claim as well. Vanskike does not pursue that issue on appeal.

. Vanskike also points out that the FLSA lists specific exceptions to its coverage of "employees” but does not list prisoners as an exception. This framework does suggest that all individuals within the general category of “employees,” if not specifically excluded, come within the statute’s scope. See Powell v. United States Cartridge Co., 339 U.S. 497, 516-17, 70 S.Ct. 755, 765-66, 94 L.Ed. 1017 (1950); Patel v. Quality Inn South, 846 F.2d 700, 702 (11th Cir.1988), cert. denied, 489 U.S. 1011, 109 S.Ct. 1120, 103 L.Ed.2d 182 (1989). The argument does not take us anywhere, however, because it assumes that prisoners plainly come within the meaning of the term "employees.”

. Thus the court in Carter carefully limited its holding to prisoners working for "outside employer[s],” and distinguished cases involving work within the prison and where "sole control was with prison officials.” 735 F.2d at 14, 15. Similarly, the court in Watson distinguished two classes of FLSA prisoner claims — work for outside employers performed within the prison, and work for outside employers performed outside the prison — apparently assuming that work performed for the prison itself would clearly not be within the reach of the FLSA. 909 F.2d at 1553.

. Section 1 of the Thirteenth Amendment provides in full: "Neither slavery nor involuntary servitude, except as a punishment for crime whereof the party shall have been duly convict *810 ed, shall exist within the United States, or any place subject to their jurisdiction.”

. The Ninth Circuit’s decision in Baker v. McNeil Island Corrections Ctr., 859 F.2d 124 (9th Cir.1988), is distinguishable. In Baker, a prisoner had applied for a voluntary position as an aide in the prison library and was rejected, allegedly on the basis of his race. The court reversed the dismissal of his claim under Title VII of the Civil Rights Act of 1964, focusing on the economic reality of the relationship and concluding that he could be an "employee” of the prison under Title VII. One reason Baker is different from the case at hand is that it involved a voluntary position rather than a mandatory work assignment. Perhaps more important, however, are the differences in purpose between Title VII and the FLSA. It is true that cases interpreting Title VII are often helpful and persuasive in construing the FLSA, and vice versa. See, e.g., Hyland v. New Haven Radiology Assoc., 794 F.2d 793, 796 (2d Cir.1986). But we think that this case highlights an area in which the FLSA does not track Title VII. Prison is in many ways a society separate from the outside world. Discrimination, however, maintains the same invidious character within the world of the prison and outside it. Given the broad policies behind Title VII, there would appear to be no reason to withhold Title VII’s protections from extending inside the prison walls. The policies underlying the FLSA, in contrast, are tied to the national economy, and those policies have limited application in the separate world of the prison.

. We note that, as Vanskike's counsel conceded at oral argument, if we were to hold that Van-skike is an "employee" of the DOC for purposes of the FLSA, the cost of his board and lodging would presumably have to be deducted as wages. 29 U.S.C. § 203(m) provides that ‘"[w]age’ paid to any employee includes the reasonable cost ... to the employer of furnishing such employee with board, lodging, or other facilities, if such board, lodging, or other facilities are customarily furnished by such employer to his employees.” Of course we need not address here the potentially difficult question of the "reasonable cost” of a prisoner’s board and lodging — and whether such cost might itself equal or exceed income attributable to the minimum wage.

. To this five-factor standard we have recently added a threshold inquiry into whether the indigent made reasonable efforts to obtain counsel or was effectively precluded from making such efforts. Jackson v. County of McLean, 953 F.2d 1070, 1072-73 (7th Cir.1992). This added factor does not apply retroactively, however, and is therefore irrelevant here.