Opinion
PETERS, J.The Workers’ Compensation Act (act), General Statutes § 31-275 etseq., provides benefits only for those workers who have the status of “employees” at the time of their injury. The principal issue in this certified appeal is whether an owner-operator of a taxicab qualifies as an employee of the taxicab company with which he contracted to provide taxicab service.
The plaintiff, Janet Hanson, the surviving spouse of Allen Hanson (decedent), filed a workers’ compensation claim to recover survivor benefits for the death of the decedent while he was driving a taxicab in accordance with his written agreement with the defendants, Transportation General, Inc. doing business as Metro
*615Taxi Service (Metro), and Hartford Insurance Company.1 The compensation commissioner for the third district (commissioner) dismissed Hanson’s claim as a result of his finding that, on the date of the decedent’s death, he had been an independent contractor and not an employee of Metro. Hanson appealed to the compensation review board (board), which affirmed the decision of the commissioner. That decision was then affirmed by the Appellate Court. Hanson v. Transportation General, Inc., 45 Conn. App. 441, 447, 696 A.2d 1026 (1997).
We granted Hanson’s petition for certification to consider the merits of her claim for compensation, either under a proposed “relative nature of the work” test or under the existing “right to control” test. Hanson v. Transportation General, Inc., 243 Conn. 914, 914-15, 701 A.2d 329 (1997). We also agreed to determine Hanson’s ancillary evidentiary claim.2 We affirm the judgment of the Appellate Court.
The opinion of the Appellate Court outlines the basic facts. “[Metro] . . . operated a fleet of taxicabs through the use of both owner-operator and lease agreements. Drivers could either lease or purchase cabs from [Metro]. The decedent, under an owner-operator *616agreement, drove a Metro [cab] from 1987 until he was murdered while operating a Metro [taxicab] on April 16, 1990. Under the terms of [Metro’s] standard owner-operator agreement, [Metro] held legal title to the [taxicab] and the owner-operator maintained a beneficial ownership.3 Upon the termination of the relationship, which could be done by either party after forty-eight hours notice, legal title [to the taxicab] would be transferred back to the owner-operator. In return for the weekly payment by the owner-operator of ‘stand dues’4 in the approximate amount of $425, plus an additional $100 per week if a second driver was used, [Metro], under a certificate issued to it by the department of transportation, allowed the owner-operator the right to operate a taxicab. [Metro] required the owner-operator to be properly licensed, to use a taxicab meter and radio approved by the department of transportation and to obey its regulations and to adhere to all federal, state and local laws. The owner-operators also were required to maintain insurance in the minimum amount required by law at their own expense and to pay for the use and operation of the [taxicab] including repairs, maintenance, fines, tickets, towing and all taxes or other assessments imposed by governmental agencies. Under the agreement, the owner-operator could set the hours of operation, hire a driver for the taxicab, use the vehicle for personal use and keep all fares derived from the operation of the [taxicab]. [Metro] did not pay the owner-operator, provide benefits, request income records or collect payroll or social security taxes.” Hanson v. Transportation General, Inc., supra, 45 Conn. App. 442-43.
*617I
The first issue presented by Hanson’s appeal is whether, in deciding whether a worker is an employee under the act, we should adopt the “relative nature of the work” test as an alternative to the “right to control” test. The Appellate Court declined to address this issue on its merits because it felt bound by this court’s decision in Ross v. Post Publishing Co., 129 Conn. 564, 29 A.2d 768 (1943). In that case, we held that “[t]here is no dispute about the ultimate test. It is the right of general control of the means and methods used by the person whose status is involved.” Id., 567. Although the issue of the propriety of recourse to the “relative nature of the work” test raises serious questions that warrant plenary exploration by this court; Doe v. Stamford, 241 Conn. 692, 696-97, 699 A.2d 52 (1997); we are not persuaded that we should engraft this test onto our longstanding workers’ compensation law.
A
The issue that Hanson has raised must be decided in the context of two jurisprudential principles that limit our authority to afford her the relief that she seeks. First, because this is a workers’ compensation case, it is governed by the special rules of construction that we have long held applicable in such cases.5 Second, because this is a case in which Hanson asks us to *618overrule a line of judicial precedents that the legislature has left intact for eighty years, it must be reconciled with the principles of stare decisis.
Connecticut first adopted a statutory scheme of workers’ compensation in 1913. Dodd v. Middlesex Mutual Assurance Co., 242 Conn. 375, 381, 698 A.2d 859 (1997). Although some statutes have common-law roots that may enlighten their construction, the act is not such a statute. “[T]he workers’ compensation system in Connecticut is derived exclusively from statute. ... A commissioner may exercise jurisdiction to hear a claim only under the precise circumstances and in the manner particularly prescribed by the enabling legislation.” (Citations omitted; internal quotation marks omitted.) Discuillo v. Stone & Webster, 242 Conn. 570, 576, 698 A.2d 873 (1997); see also Kinney v. State, 213 Conn. 54, 60, 566 A.2d 670 (1989); Gagnon v. United Aircraft Corp., 159 Conn. 302, 305, 268 A.2d 660 (1970). “Because of the statutory nature of our workers’ compensation system, policy determinations as to what injuries are compensable and what jurisdictional limitations apply thereto are for the legislature, not the judiciary or the board, to make.” Discuillo v. Stone & Webster, supra, 577. Whatever the policy implications of this case may be, “the issue presented is, at bottom, a matter of statutory construction.” Doe v. Stamford, supra, 241 Conn. 697.
As a matter of statutory construction, to overrule our long-standing invocation of the “right to control” test, we would have to reconcile such a ruling with the presumption of legislative acquiescence in judicial interpretations that the legislature has not overturned. We have long acted on the hypothesis that the legislature is aware of the interpretation that the courts have placed upon one of its legislative enactments. Once an appropriate interval to permit legislative reconsideration has *619passed without corrective legislative action, the inference of legislative acquiescence limits judicial authority to reconsider the merits of its earlier decision. Hall v. Gilbert & Bennett Mfg. Co., 241 Conn. 282, 297-98, 695 A.2d 1051 (1997); see Jolly, Inc. v. Zoning Board of Appeals, 237 Conn. 184, 196, 676 A.2d 831 (1996); White v. Burns, 213 Conn. 307, 333-34, 567 A.2d 1195 (1990); Herald Publishing Co. v. Bill, 142 Conn. 53, 63, 111 A.2d 4 (1955). In this context, only “the most cogent reasons and inescapable logic”; Jolly, Inc. v. Zoning Board of Appeals, supra, 196; can justify judicial abandonment of judicial precedents dating back more than eighty years. See Thompson v. Twiss, 90 Conn. 444, 447, 97 A. 328 (1916).
We can perceive no such cogent reasons or inescapable logic to justify altering our long-standing judicial interpretation of one of the fundamental premises of our workers’ compensation statute. This is particularly true for this statute, which has been the subject of extensive legislative amendment over its entire life, without any legislative indication of disagreement with the right to control test. Furthermore, it is fair to assume that employers and insurers have relied justifiably on that interpretation in ordering their business affairs. For us to abandon that interpretation would upset their legitimate expectations.
B
Recognizing that these principles create obstacles to judicial reconsideration of our long-standing “right to control” test, Hanson attempts a bypass, arguing that the “relative nature of the work” test can coexist with the “right to control” test. We disagree.
The “relative nature of the work” test determines the relationship between a worker and a putative employer by asking whether the worker’s performance is an integral part of the regular business of the putative *620employer. 3 A. Larson & L. Larson, Workmen’s Compensation (1997) § 45.00, p. 8-193. The “right to control” test determines the same relationship by asking whether the putative employer has “the right to control the means and methods” used by the worker in the performance of his or her job. Hunte v. Blumenthal, 238 Conn. 146, 154, 680 A.2d 1231 (1996); Silverberg v. Great Southwest Fire Ins. Co., 214 Conn. 632, 639, 573 A.2d 724 (1990); Ross v. Post Publishing Co., supra, 129 Conn. 567.
Dual standards for determining whether a worker is an employee or an independent contractor would complicate the administration of our workers’ compensation system. Although there well may be circumstances under which the results reached by application of the two tests will be identical, more often than not the results will differ. In all probability, a workers’ compensation commissioner applying the “relative nature of the work” test would be required to broaden the class of “employees” to include workers who currently are characterized as independent contractors. See, e.g., Harger v. Structural Services, Inc., 121 N.M. 657, 916 P.2d 1324, 1329-31 (1996). Indeed, unless differing results were contemplated, there would be little to gain from adopting a new test. Hanson has proffered no criteria to define the underlying circumstances under which one or the other test would be the more appropriate. If a workers’ compensation commissioner had undefined and hence unlimited authority to invoke either test, the scope of the term “employee” would be left in confusion. We are persuaded, therefore, that we must evaluate Hanson’s argument in favor of the “relative nature of the work” test as an outright alternative to the “right to control” test.6
*621C
We turn now to the merits of replacing the “right to control” test with the “relative nature of the work” test. As advocated by Hanson, the “relative nature of the work” test would require a two part analysis, each part requiring consideration of three separate factors. With respect to the character of the work performed, the claimant would be required to show: (1) the degree of skill involved; (2) the degree to which the work is a separate calling or business; and (3) the extent to which a worker so situated reasonably can be expected to carry the burden of accident. With respect to the relationship of the work to the putative employer’s business, the claimant would have to establish: (1) the extent to which the work is a regular part of the putative employer’s regular business; (2) the extent to which the work is being performed continuously or intermittently; and (3) the extent to which the work is of sufficient duration to constitute continuing services rather than a particular assignment. See 3 A. Larson & L. Larson, supra, § 43.51, p. 8-23, and § 45.10, p. 8-193.7
*622Hanson advances four arguments in favor of adoption of the “relative nature of the work” test. She derives three of these arguments from the unchallenged proposition that workers’ compensation is social legislation. In that light, she argues that, in preference to the “right to control” test, the proposed test: (1) more effectively implements the objectives of this social legislation; (2) more accurately depicts the employer-employee relationship by precluding employer reliance on contracts drafted to avoid the costs of this social legislation; (3) more directly advances the remedial purposes of social legislation. In addition, she argues that the “right to control” test is too vague to be applied fairly. We are not persuaded. Evaluation of Hanson’s first three arguments cannot be undertaken by this forum; their resolution is grist only for the legislative mill. Hanson’s final argument is unpersuasive on its merits.
In the absence of any claim of constitutional vagueness in the “right to control” test, Hanson must demonstrate that fact-finding under this test exceeds the discretionary authority normally afforded to any trier of fact. Undeniably, whenever the parties disagree, a determination of fact has elements of uncertainty. Hanson has not shown that, in application, the “right to control” test has created more uncertainty than factual disputes about, for example, contractual intent or tor-tious misbehavior. Indeed, as is illustrated by the Alaska case upon which Hanson relies, close questions of fact are also likely to arise in the application of the “relative nature of the work” test. Grothe v. Olafson, 659 P.2d 602, 606 (Alaska 1983).8
*623II
The second issue raised by Hanson’s appeal is whether, if the “right to control” test applies, the trier of fact improperly determined that the decedent was not an employee at the time of his death. She claims that Metro retained enough authority over the taxicab fleet bearing its name to demonstrate a right to general control that made the decedent its employee rather than an independent contractor. We disagree.
“The determination of whether an injury arose out of and in the course of employment is a question of fact for the commissioner.” Spatafore v. Yale University, 239 Conn. 408, 418, 684 A.2d 1155 (1996). Accordingly, we apply the same deferential standard of review that we accord to findings of fact by a trial judge or jury. Id., 419. Keeping these principles in mind, we turn to Hanson’s claim that Metro retained sufficient control to render her decedent an employee.
For Connecticut authority, Hanson relies principally on Lassen v. Stamford Transit Co., 102 Conn. 76, 128 A. 117 (1925), a case which, like this one, involved the “right to control” test for taxicab drivers, and which held that the drivers were employees. That case, however, is distinguishable on its facts. For example, the taxicab company in that case required drivers to turn over their fares to the company and paid the drivers 75 percent of that amount. Id., 79. Furthermore, company dispatchers designated the driver who was to transport any particular passenger. Id., 78-79. In this case, Metro took care to avoid such indicia of control, leaving it to taxicab drivers like the decedent to decide, in the *624exercise of their own discretion, when and where they would take the taxicabs on the road.
As a more general matter, Hanson cites the fact that Metro required cabdrivers to conform to department of transportation regulations with respect to such matters as licenses, meters, fares, records, dispatch and safety. The argument that such state regulatory requirements automatically satisfy the right to control test would make irrelevant the contractual terms of the relationship between the parties. It is axiomatic that every taxicab company must comply with state regulations. We are not persuaded, therefore, that Metro’s requirement that owner-operators comply with government regulations definitively demonstrates an employer-employee relationship. See, e.g., La Grande v. B & L Services, Inc., 432 So. 2d 1364, 1366 (Fla. App. 1983).
Contrary to Hanson’s argument, we agree with the Appellate Court, the board and the commissioner that the totality of the evidence did not demonstrate that Metro had retained sufficient control to require a finding of an employer-employee relationship between Metro and the decedent. True, the commissioner made some subordinate factual findings that, if viewed in isolation, might have supported a different determination. For example, Metro required its drivers to: (1) report all traffic violations to Metro; and (2) obtain Metro’s approval of any outside insurance. Metro also held legal title to the vehicles, and painted and marked the cars as Metro cabs. Other subordinate findings by the commissioner, however, paint a clearer picture of the contractual relationship between Metro and its drivers. In the daily use of their cabs, drivers could set their own hours, work anywhere in the Metro service area, refuse to accept dispatch calls,9 and hire a second driver for *625the cab. Drivers had sole responsibility for all expenses related to operation of their cabs and had the right to regain total ownership rights to their cabs upon the termination of their relationship with Metro. Administratively, moreover, Metro did not pay any salary or fringe benefits to the drivers and did not require the drivers to report their fares to Metro. We conclude that the commissioner reasonably found that the totality of factors in this case indicates that Metro taxi drivers were not employees as that term is used under the act.10
In sum, the Appellate Court properly reviewed the commissioner’s determination under the terms of the “right to control” test. We concur in the Appellate Court’s conclusion that, under the circumstances of this case, Hanson’s decedent was not an employee, but, rather, an independent contractor.
Ill
The third and final issue raised by Hanson’s appeal is whether the commissioner improperly refused to amend its findings to include certain requested additions of fact. We agree with the Appellate Court’s resolution of this issue.11 Once it has been decided that the appropriate test is the “right to control” test, “even if *626the findings were corrected as requested, the ultimate conclusion would not change.” Hanson v. Transportation General, Inc., supra, 45 Conn. App. 447.
The judgment of the Appellate Court is affirmed.
In this opinion BORDEN, NORCOTT and KATZ, Js., concurred.
Because the defendants apparently present the same arguments to this court, having filed a joint brief that makes no distinctions between them in this respect, we will refer to them as one party, Metro. They were so denominated in the proceedings before the commissioner and the board.
Our grant of certification was limited to the following issues:
“Should the relative nature of the work test be adopted as an alternative to the right to control test in concluding whether the plaintiffs decedent, a cab driver, was an independent contractor and not an employee of the defendant taxi cab company?
“Did the Appellate Court properly affirm the finding of dismissal pursuant to the right to control test in light of the department of transportation regulations?
“Did the Appellate Court properly affirm the compensation review board’s refusal to correct the finding of dismissal as requested by the plaintiff?” Hanson v. Transportation General, Inc., supra, 243 Conn. 914-15.
“[Metro] was required to hold legal title to the vehicle because, in accordance with Connecticut law, all taxicabs are required to be titled in the name of the certificate holder, in this case, [Metro].” Hanson v. Transportation General, Inc., supra, 45 Conn. App. 442 n.2.
“ ‘Stand dues’ are a combination of leasing, taxes and insurance costs.” Hanson v. Transportation General, Inc., supra, 45 Conn. App. 442 n.3.
This appeal does not, involve the rights of a third party tort victim to recover against Metro for the misconduct of one of its drivers on the basis of the principles of actual, implied, or apparent authority. See, e.g., Beckenstein v. Potter & Carrier, Inc., 191 Conn. 120, 132-41, 464 A.2d 6 (1983) (discussing agency principles in contract case); Fireman’s Fund Indemnity Co. v. Longshore Beach & Country Club, Inc., 127 Conn. 493, 496-97, 18 A.2d 347 (1941) (applying similar agency principles in tort action). The issue, rather, is what the decedent driver knew or should have known of his own status under the act.
It is true that the right to control test, in its present form, was embedded into the act by early cases relying on similar rules in tort cases. Nonetheless, that construction of the act has been ratified by legislative acquiescence.
Hanson’s argument relies on four out-of-state cases that we find unpersuasive. See Stampados v. Colorado D. & S. Enterprises, 833 P.2d 815, 817 (Colo. App. 1992); Kertesz v. Korsh, 296 N.J. Super. 146, 152-55, 686 A.2d 368 (App. Div. 1996); Claim of Griffin, 466 N.W.2d 148, 150 (N.D. 1991); *621Kaiel v. Cultural Homestay Institute, 129 Or. App. 471, 477, 879 P.2d 1319, rev. denied, 320 Or. 453, 887 P.2d 792 (1994).
As Metro observes, although Professor Larson’s treatise describes the “relative nature of the work” test as a new test currently superseding the “right to control” test, that statement no longer reflects the totality of the relevant case law. Earlier editions of his celebrated work demonstrate that, for almost thirty years, Professor Larson has been a staunch advocate of the “relative nature of the work” test. During this time period, however, only a few state courts have adopted this test in substitution for the “right to control” test. See, e.g., cases cited in footnote 6 of this opinion. Connecticut has not done so, either by way of legislation or by way of judicial opinion. See DaSilva v. Danbury Publishing Co., 39 Conn. App. 653, 656, 666 A.2d 440, cert. denied, 235 Conn. 936, 668 A.2d 374 (1995). Connecticut is not alone. See, e.g., Kirkwood v. Industrial Commission, 84 Ill. 2d 14, 23-24, 416 N.E.2d 1078 (1981); Youngblood v. North State Ford Truck Sales, 87 N.C. App. 35, 37-38, 359 S.E.2d 256 (1987), aff'd, 321 N.C. 380, 364 S.E.2d 433 (1988). A contrary decision by the Supreme Court of Maine; Timberlake v. Frigon & Frigon, 438 A.2d 1294, 1297-98 (Me. 1982); was overturned five years later by that state’s legislature. See West v. C.A.M. Logging, 670 A.2d 934, 937 n.2 (Me. 1996).
The relationship between the parties in this case illustrates the factual uncertainties that may arise in the application of the “relative nature of the work” test. Concededly, the decedent was free to use his taxicab for private engagements entirely unrelated to the business of Metro. Would the decedent’s status depend on whether, when he was killed, he was engaged in business obtained through Metro’s dispatcher or through his own enterprise? Would the status of other taxi drivers who were accidently injured depend *623solely, or principally, on the percentage of the time that they used the taxicab for Metro business? We do not suggest that these questions are unanswerable. They illustrate, however, that there will be close issues of fact, whichever test is applied.
Contrary to the implied assertion of Justice Berdon’s dissenting opinion, Metro could not order a driver to take radio dispatch calls. Metro dispatched *625drivers who agreed to be available ad hoc, it could not prevent a driver from sharing a car with other drivers.
The principal issue was the propriety of a motion to correct to add a finding that, on the date of his death, the decedent was on the road in response to a call that he had received from Metro’s radio dispatcher. There was, however, neither a finding, nor a request to add a finding, to the effect that the decedent had used his taxicab exclusively or even predominantly in response to calls from the dispatcher. Significantly, the commissioner found that the decedent had used his own cellular phone for servicing customers and that he had shared the use of his cab with a second driver. Furthermore, in one of the requests for correction of the finding, Hanson herself asked for inclusion of a finding that: “The work performed by [the decedent] was not for the completion of a particular job but consisted of an ongoing relationship between [the decedent] and [Metro].”
In reaching its conclusion, the Appellate Court concurred in the reasoning of the board.