dissented, stating that under the terms of the agreement the only inevitable conclusion was that defendant was not plaintiff’s employer and was not entitled to the exclusive remedy provision of the wdca. Judge Neff distinguished this case from Farrell and its progeny on the basis of explicit language of the contract, which she noted was never repudiated by either party. Because the majority’s opinion “turns the agreement between defendant and CLS on its head *30and effectively nullifies an otherwise valid contractual agreement between the parties,” Judge Neff would have reversed and remanded this case for trial on the liability issue. Id.2
B
Plaintiff Kidder, an iron worker also leased to Miller-Davis through cls, was working on the renovation of the Kalamazoo Center when he plummeted ten feet from some roof decking through an open skylight, shattering both his elbows. Plaintiff Kidder filed suit against Miller-Davis for negligently failing to *31place covers on or guardrails around the open skylights, negligently failing to supervise subcontractors, and negligently failing to provide a safe workplace. Miller-Davis moved for summary disposition, claiming that the plaintiff was an employee of both cls and Miller-Davis and therefore his only remedy was under the WDCA. MCL 418.131; MSA 17.237(131). The trial court agreed in spite of express language in the CLS-Miller-Davis contract, stating “[s]uch leased personnel shall be employees of cls for all purposes. Such leased personnel shall not be employees of [Miller-Davis] for any reason.”
The Court of Appeals affirmed, relying on Farrell v Dearborn Mfg Co, supra, and held that a labor broker relationship existed between cls and Miller-Davis. The Court stated: “Accordingly, the economic reality being that both CLS and defendant were employers within the meaning of the WDCA, defendant was entitled to the enforcement of the wdca exclusive remedy provision.” Unpublished opinion per curiam, issued January 12, 1996 (Docket No. 166059). Judge Neff again dissented, reiterating her dissent in Wolthuis.
n
A
The determination of who is an employer for purposes of the exclusive remedy provision of the wdca, as well as other statutory schemes, has had a rich and voluminous history. Early cases that addressed this question, first in the arena of unemployment compensation, evaluated the existence of the relationship under a “control” test.
The control theory is the traditional common-law test used to delineate the master-servant relationship. The the*32oiy, in its delineation of the servant concept, has for its purpose the definition and delimitation of the scope of the master’s liability under the doctrine of respondeat superior. Because most compensation acts contain no specific definition of the term “employee,” it was generally taken for granted that the common-law definition of employee, or servant, used for purposes of vicarious tort liability was to be used for purposes of workmen’s compensation laws. [Nichol v Billot, 406 Mich 284, 293-294; 279 NW2d 761 (1979).]
This test attempted to separate those persons who were employees from those who were independent contractors. This assessment examined who controlled the work, the horns, the process, and the methods of the work involved. The initial rule was “where the person employed is in the exercise of an independent and distinct employment, and not under the immediate control, direction, or supervision of the employer, the latter is not responsible for the negligence or misdoings of the former.” Powell v Employment Security Comm, 345 Mich 455, 469; 75 NW2d 874 (1956) (Smith, J., dissenting) (citations omitted). “The control test is also used to determine when an independent contractor has become an employee for purposes of respondeat superior liability on the part of his employer who has exercised a degree of control inconsistent with independent contractor status.” Nichol v Billot, supra at 296 citing Powell at 471.
Justice Talbot Smith’s thoughtful dissent in Powell, which later became the prevailing law, strongly criticized the rigid control test and the absurd, chaotic decisions that resulted from it. He noted that the control test failed to achieve any uniformity or certainty and that this tort concept should be inapplicable to *33the administration of social legislation.3 Id. at 471. Justice Smith noted that accepting this rigid control test could lead to much mischief, in that employers would start hiring independent contractors to do everything to avoid their “statutory obligations under the national labor relations act, under the social security, unemployment compensation, and fair labor standards acts, and, in addition, many of the other Federal and State laws . . . .” Id. at 473.
Justice Smith was influenced by federal case law administering the National Labor Relations Act. In United States v Silk, 331 US 704, 713; 67 S Ct 1463; 91 L Ed 1757 (1947), the United States Supreme Court first articulated an economic-reality test for assessing the employer-employee relationship.
The problem of differentiating between employee and an independent contractor or between an agent and an independent contractor has given difficulty through the years before social legislation multiplied its importance. When the matter arose in the administration of the National Labor Relations Act, we pointed out that the legal standards to fix responsibility for acts of servants, employees or agents had not been reduced to such certainty that it could be said there was “some simple, uniform and easily applicable test.” The word “employee,” we said, was not there used as a word of art, and its content in its context was a federal problem to be construed “in the light of the mischief to be corrected and the end to be attained.” We concluded that, since that end was the elimination of labor disputes and industrial strife, “employees” included workers who were such as a matter of economic reality. The aim of the Act *34was to remedy the inequality of bargaining power in controversies over wages, hours and working conditions. We rejected the test of the “technical concepts pertinent to an employer’s legal responsibility to third persons for acts of his servants.” This is often referred to as power of control, whether exercised or not, over the manner of performing service to the industry. [Id. (Citations omitted.)]
Persuaded by the reasoning in Silk, Justice Smith adopted the economic-reality test in his frequently cited dissent in Powell. He determined that a court must examine the work performed, whether it is part of a common objective integral to the employer’s business, and whether this work would normally follow the usual path of an employee. Powell at 479, citing Rutherford Food Corp v McComb, 331 US 722; 67 S Ct 1473; 91 L Ed 1772 (1947). Accordingly, control is only one factor to be considered.
In Tata v Muskovitz, 354 Mich 695; 94 NW2d 71 (1959), this Court overwhelmingly abandoned the control test in favor of the economic-reality test to determine who is an employer for purposes of the Worker’s Disability Compensation Act. In Schulte v American Box Board Co, 358 Mich 21; 99 NW2d 367 (1959), Justice Smith defined more fully the parameters of the economic-reality test:
The earlier cases tested this relationship through application of the “control” factor, originally a test for tortious liability, having its roots in the relationship of the apprentice to his master in early English industrial society. As applied to today’s complex economy of the assembly line, of dispersed industrial operations, of concentrated operations but with semi-autonomous “departments” or branches, and of general contractors who, in turn, employ subcontractors and sub-subcontractors, the “control” test is often meaningless, usually ambiguous, and always susceptible of paper-*35writing evasions. Consequently we have abandoned it. The test is now one of economic reality. This is not a matter of terminology, oral or written, but of the realities of the work performed. Control is a factor, as is payment of wages, hiring and firing, and the responsibility for the maintenance of discipline, but the test of economic reality views these elements as a whole, assigning primacy to no single one. [Id. at 32-33 (Smith, J., concurring) (citations omitted).]
The economic-reality test was embraced by this Court as a more realistic attempt to define the employer-employee relationship through a “balancing of all the relevant factors in each case,” than the rigid control test. Renfroe, supra at 265. Given the increasingly complicated relationships developing in today’s business and economic marketplaces anything other than a totality of the circumstances test would be an insufficient guide by which to evaluate the employee-employer relationship.
B
A labor broker-customer arrangement presents a unique employment relationship and adds a further dimension to the analysis of who is an employer for purposes of the wdca. Renfroe, supra, was one of the first cases to examine the labor broker relationship within the context of the Worker’s Disability Compensation Act. The issue in Renfroe was whether the defendant, Higgins Co., a customer of a labor broker, was an employer, for purposes of the exclusive remedy provision of the worker’s compensation act, thus precluding separate recovery. Id. at 261. The plaintiff in Renfroe was employed by Employers Temporary Service (ETS) whose sole business was supplying temporary labor to industrial and commercial customers. Customers would contact ETS on a daily basis, and ETS *36would dispatch the required numbers of workers. Ets paid the workers and withheld their taxes. The customer, however, instructed the workers and had the authority to dismiss them.
Higgins Co. frequently used ETS to supply it with temporary help. The plaintiff, supplied by ets to Higgins, injured his right hand in a punch press and received medical and disability compensation for his injury from ets’s worker’s compensation insurer. He then sought additional recovery from Higgins, arguing that a restrictive definition should be used to define “employer” because it was in his best interests to be able to sue Higgins as a third-party tortfeasor. The Court rejected the plaintiff’s argument, acknowledging the preference for a more liberal approach to defining the term “employer,” which necessarily followed the purposes behind “providing social remedial legislation in an attempt to provide benefits for injured employees under the workmen’s compensation act.” Id. at 263. That approach favors the economic-reality test over the rigid control test.
In evaluating this triangular relationship, the Court of Appeals noted that ets workers agreed to work for ets, and then agreed to work for ets customers. The Court of Appeals reasoned that, regarding labor brokers, the
economic reality is based on the fact that a profit can be made by efficiently matching workers with temporary work needs. Ets maintained control of the workers by its practice of daily reassignment and daily payment at its offices. It also maintained the formalities of employment by handling all paper work and payments incident to the employment. [Id. at 266.]
*37The Court of Appeals held that under the economic-reality test both ETS and Higgins were the plaintiff’s employers, even though each employment situation was different. In holding that the plaintiff could not proceed against Higgins, and that the suit was barred by the exclusive remedy provision of the worker’s compensation act, the Court stated: “It is not necessary to make fine semantic distinctions as to types of degrees of control, et cetera. It is enough to say that either could be liable under the workmen’s compensation act, therefore, both are protected by it.” Id. at 266-267.
In Nichol v Billot, supra at 299, this Court examined the employer-independent contractor relationship and held that the economic-reality test “in its consideration of factors beyond the control element, achieves a more realistic determination” of who is entitled to the exclusive remedy defense under the worker’s compensation act. Id. In addition, and more importantly, we held that the question of the defendant’s status is one of law for the court and not one of fact for the jury when the evidence and facts are susceptible of one inference only. Id. at 302, quoting Flick v Crouch, 434 P2d 256 (Okla, 1967).4
In Farrell v Dearborn Mfg Co, supra, this Court had yet another chance to evaluate the application of the exclusive remedy provision of the wdca in four cases, three of which involved labor brokers.5 In each *38of these cases, labor brokers State Labor, Inc., Kelly Services, Inc., and Employers Temporary Service furnished employees to others. In each, a worker was injured, received worker’s compensation benefits, and brought suit against the customer of the labor broker. We held that “the exclusive remedy available to the employee in a labor broker situation is provided by the worker’s compensation statute and that a separate tort action against the customer of the labor broker may not be maintained.” Id. at 278.
In arriving at this conclusion, we first discussed the purpose and meaning behind the wdca and its exclusive remedy provision. We noted that a fundamental premise of the act is that if the employee’s injury falls within its provisions, then worker’s compensation will be the only remedy against the employer and the employer’s insurance carrier. In assuming responsibility for work-related injuries, the employer is thus protected from excessive damage awards, and injured employees are automatically compensated for their injuries.6 Id. at 274.
In Farrell, the plaintiffs argued that this separate cause of action should not be barred under the exclusive remedy provision because the conditions of liability as stated in the statute did not exist. In rejecting this argument, we examined the roles of the labor *39broker and its customers to assess who should be considered an employer for purposes of the statute. We noted that in each of the cases the labor broker’s business was to supply the customer with temporary employees pursuant to the customer’s expressed needs. In addition, we noted that, once the employee reported to the customer, the employee was under the control and supervision of the customer’s supervisory personnel. The customer could discharge an employee, or refuse to accept an employee. We noted further that the customer does not pay the employee directly. The employee is paid by the labor broker who charges the customer a fee for the services provided. Id. at 275.
We also addressed whether an employment relationship existed using the economic-reality test as our guide and adopted the standard enunciated in Renfroe. See supra. In Farrell, we stated that “[t]he economic reality test looks to the employment situation in relation to the statutory scheme of workers’ compensation law with the goal of preserving and securing the rights and privileges of all parties. No one factor is controlling.” Id. at 276 (emphasis added). Embracing the reasoning in Farrell we stated:
The customer received a worker each day who was subject to its authority. By engaging the services of the labor broker, the customer knew that, in exchange for a set fee, the broker would pay the employees, handle all paperwork, and provide compensation coverage. The Court concluded that the economic reality was that both the labor broker and its customer were employers within the meaning of the workers’ compensation statutes. We agree with the reasoning and result of this decision.
The labor broker providing personnel for temporary employment is a common business practice. The roles of *40the broker and its customer are defined and structured to fulfill short-term needs of many types of industries. When viewed in terms of control, payment of wages, allocation of responsibilities, maintenance of discipline, etc., it is clear that the two are so integrally related that their common objectives are only realized by a combined business effort. The broker supplies as the customers demand.
To conclude that an individual so employed is outside the scope of the exclusive remedy provision would clearly disregard the overall objectives of the statutory scheme. [Id. at 277.]
Accordingly, we held that a separate action against the customer could not be maintained.
in
Turning our focus to the cases at hand, we find that a labor broker relationship existed between cls and Miller-Davis to such an extent that a dual or coemployer relationship was formed, precluding a separate tort action against the customer of the labor broker.7 This conclusion is further supported by our analysis of the economic realities of the employment situation which compels us to hold that Miller-Davis was the plaintiffs’ employer for purposes of the exclu*41sive remedy provision of the worker’s compensation statute.
A
The parties do not dispute that cls is a labor broker whose business is the provision of construction trades personnel to contractors. The labor broker-customer relationship is very different from an independent contractor-subcontractor relationship. Cls is not a subcontractor. No bids were ever solicited or received, and Miller-Davis did not relinquish its ultimate supervision or control of the work site to cls or any other subcontractor. Cls provided the workers “pursuant to the job description requests furnished to CLS by [Miller-Davis] from time to time.”8 Further, CLS designated supervisors on the work site also pursuant to job description requests furnished by Miller-Davis.
Cls billed Miller-Davis on a weekly basis for the workers’ wages, fringe benefits, payroll taxes, worker’s compensation and liability insurance premiums, state coiporate tax, trade travel pay, and FTCA. Cls also billed Miller-Davis an additional fee that covered the cost both of administering the contract and the handling of all other employment matters. The fact that Miller-Davis did not directly pay the workers is a distinction without a difference. Cls was a payment conduit for Miller-Davis. Miller-Davis reimbursed cls dollar for dollar, including an amount for worker’s disability compensation. Miller-Davis unequivocally assumed the responsibility for these payments. This alone is evidence that Miller-Davis *42was not giving up completely its role as employer. Clearly, the trades construction personnel worked for both cls and Miller-Davis. Miller-Davis paid a premium for this particular service. There is no question, under both our reasoning in Farrell and the Court of Appeals opinion in Renfroe that this was a coemployer situation. Accordingly, we reaffirm our decision in Farrell and hold that under these circumstances a labor broker and its customer both are subject to, and protected by, the exclusive remedy provision of the wdca.
B
Establishing a labor broker relationship does not, however, necessarily end the analysis. The relationship must still be evaluated under the economic reality test, which in this case yields the same conclusion: Both Miller-Davis and cls must be considered dual or coemployers for purposes of the exclusive remedy provision of the worker’s compensation statute. Again, this standard examines a number of criteria including control, payment of wages, hiring, firing, the maintenance of discipline, and common objective. These factors are viewed together in their entirety under a totality of the circumstances test. We repeat: No one factor is controlling.
First, both employers shared control of the workers. Miller-Davis requested the types of workers it needed to do its bidding at the job site. “[Miller-Davis] will make all directives deemed necessary for the efficient jobsite utilization of leased cls construction trades personnel through the assigned cls on-site supervisor (foreman), or the designated cls official.” Miller-Davis never relinquished its ultimate control or *43supervision over the work site.9 This decentralization of control is not tantamount to a lack of control or authority on the project. Miller-Davis also furnished the appropriate tools for the workers to perform their tasks (unless otherwise provided by union agreement *44or custom).10 Miller-Davis controlled the hours on the job, and daily told the workers exactly where they would be working. Miller-Davis did not have to stand over each worker’s shoulder, explaining what to do minute to minute or where to put each beam and rivet. “It is not necessary to make fíne semantic distinctions as to types of degrees of control, et cetera.” Renfroe at 266.
Second, the payment of wages indicates that both Miller-Davis and cls were employers. Miller-Davis reimbursed CLS, dollar for dollar, the actual amount of fringe benefits, withholding and other employment taxes, worker’s compensation, and life and health insurance premiums that were incurred by CLS. Again, who writes the check is not dispositive of the employer-employee relationship under Renfroe and Farrell. The economic reality is that Miller-Davis paid the wages and benefits for each plaintiff through CLS.
The third factor concerns who has the right to hire, fire, and discipline the employees. We find that, under the agreement, this was also a shared responsibility. Miller-Davis hired cls to provide workers, and although Miller-Davis could not directly hire, fire, or *45discipline it is undisputed that “cls agree[d] to remove any person leased to [Miller-Davis] as leased construction trades personnel which [Miller-Davis] deem[ed], at its sole discretion, as unsatisfactory . . . Thus it is clear that Miller-Davis could accomplish indirectly that which it could not do directly. Even though the agreement specifies that Miller-Davis could not evaluate cls employees, Miller-Davis had the right to order cls to remove unsatisfactory workers.
The fourth criterion traditionally evaluated under the economic-reality test examines the work performed and whether it is part of a common objective, integral to the employer’s business. In both cases a common objective or goal was undertaken by both employers. Miller-Davis could not have completed its renovations without employing cls workers and the cls workers were working toward completing Miller-Davis’ project goals. In fact, in testimony taken during the Wolthuis trial a person stated that there were so many workers on the site and the duties were so intertwined that it could not readily be determined who worked for CLS and who worked for Miller-Davis. Clearly, the two were “so integrally related that their common objectives [were] only realized by a combined business effort.” Farrell at 277. A labor broker-customer relationship may very well presume a common objective. See Tolbert v U S Truck Co, 179 Mich App 471, 476; 446 NW2d 484 (1989) (“[a] labor broker relationship established ... [a] common objective in a business effort”).
Finally, this analysis would not be complete without discussing the meaning of the agreement between cls and Miller-Davis. We are cognizant that this con*46tract attempts to disavow any employer-employee relationship between the two parties. However, we cannot accept the plaintiffs’ characterization of the agreement and refuse to hold that it alone is dispositive of the status of the parties. Just as we have held that control is but one factor to consider under the economic-reality test, so is the contract but one factor. Parties enter into contracts to serve all kinds of purposes in delineating rights, responsibilities, and liabilities, whether for tax, insurance, or other reasons.11 We hold that the economic reality of these cases is that the rights and responsibilities over these workers were shared by both Miller-Davis and CLS and that the agreement is neither dispositive nor controlling. See Fitzgerald v Mobil Oil Corp, supra at 1303; White v Central Transport, Inc, 150 Mich App 128; 388 NW2d 274 (1986); Tolbert v U S Track Co, supra.12 We believe this conclusion best advances the principles embodied within the wdca.
*47IV
In sum, we find that a labor broker relationship existed between CLS and its customer, Miller-Davis, to an extent that each was a coemployer of the plaintiffs for the purposes of the wdca. Accordingly, each employer is protected by, and entitled to, the benefit of the exclusive remedy provision of the wdca. In addition, we further hold that under the economic-reality test, which examines the totality of the circumstances surrounding the employment relationship and work performed, Miller-Davis was the plaintiffs’ employer for purposes of the wdca and thus entitled to enforcement of the exclusive remedy provision.
Brickley, Riley, and Weaver, JJ., concurred with Mallett, C.J.In her dissent, Judge Neff highlighted the following provisions in the agreement between CLS and Miller-Davis, entitled “Agreement Between Construction Labor Services, Inc. and Miller-Davis Company”:
1. CLS provided construction trades personnel on an independent contractor basis.
2. Such leased personnel were not to be employees of Miller-Davis for any reason.
3. CLS was responsible for all employment matters. This included withholding taxes; payment of withheld taxes and employment-related taxes; providing worker’s compensation, disability, life and group health insurance; providing a pension plan. In addition CLS was responsible for any contract administration matters respecting any collective bargaining agreements covering the leased employees.
7. CLS had the sole responsibility for recruiting, hiring, training, evaluating, replacing, supervising, disciplining, and discharging construction workers leased to Miller-Davis. CLS also was to designate on-site supervisors to direct operational and administrative matters and directly supervise CLS trades personnel. CLS retained the. responsibility to establish procedures under which CLS employees conducted their job performance and times of employment. Further, Miller-Davis could not modify the contract between CLS and its leased personnel, nor could it recruit, hire, train, evaluate, replace, supervise, discipline, or discharge leased individuals. However, CLS agreed to remove any person leased to Miller-Davis that Miller-Davis in its sole discretion deemed unsatisfactory and find a suitable replacement.
8. CLS was to implement its policies and procedures relating to leased personnel.
9. CLS agreed to indemnify and hold harmless Miller-Davis as owner and architect.
In Powell, the majority found that a woman employed as an industrial homeworker retouching photographic negatives was an independent contractor and thus not an employee for purposes of unemployment compensation. Everything about her employment was controlled by the employer except that her work was completed at home. Id. at 457, 460.
In Nichol, after applying the economic-reality test, we held that the defendant was an independent contractor as a matter of law, and remanded the case for trial on the issue of negligence. Id. at 306.
The fourth case involved an International Hockey League linesman who was attacked by an irritated hockey player. Applying the economic-reality test to the relationship between the linesman, the league, and the team, we held that the hockey team was not the plaintiff’s employer; thus, *38the hockey player was not a fellow employee, and the linesman’s suit against the team was not barred.
“[T]ort litigation is an adversary contest to right a wrong between contestants; workers’ compensation is a system, not a contest, to supply security to injured workers and distribute the cost to the consumers of the product.” 1 Larson, Workmen’s Compensation, § 2.70, p 1-16. The notion of fault is eliminated, and the idea is compensation tied to earnings, the costs of which are ultimately passed on to the consumers. Id., §§ 3.00-3.30, pp 1-17 to 1-19.
We stop short of holding that a labor broker-customer relationship will always establish dual employer status as a matter of law. In this regard, we note that Farrell may be seen as ambiguous. We held in Farrell that the exclusive remedy provision bars a separate tort action against the employer in a labor broker situation. Yet, additional language in that opinion indicates that the inquiry does not end there because we recognized that the labor broker relationship must still be “viewed in terms of control, payment of wages, allocation of responsibilities, maintenance of discipline, etc. . . .” Id. at 277. We realize that this may appear to be inconsistent. However, each case turns on its own facts and, as hinted at in Farrell and reemphasized in our decision today, we note that the existence of a labor broker relationship in and of itself may not always be dispositive.
The fact that the contract was for twelve months, renewable yearly, does not change the fact that CLS was a labor broker providing temporary workers.
Excerpts from the depositions of CLS employees are illustrative: The first excerpt is from the deposition of Frank Lombardo, a CLS foreman on the site where plaintiff Kidder was injured. Benny Komdyke “was the big boss” from Miller-Davis:
Q. Was there a particular company or person from whom you received your daily supervision? In other words, who told you what to do when you got to the job every day?
A. Well the job supervisor was Benny Komdyke.
Q. Komdyke, K-o-r-n-d-y-k-e?
A. Yes.
Q. Would Benny Komdyke have the power and authority to tell you, hey, we are not going to start until 10:00 o’clock today, something like that?
A. Oh, yes.
Q. That would be one of his jobs?
A. Right.
Q. If Ben Komdyke saw something that you were doing that he believed should have been done in a different area or done differently, could he direct you to direct your people to do it differently?
A. Oh, yes. Yes.
Q. If Ben Komdyke was not happy with one of the people who was working under you, you have given me these three journeymen iron workers, could he have told you to remove them from the job?
A. Yes. He could have.
The following excerpts are taken from the depositions of James Lash, a Miller-Davis superintendent.
Q. What if CLS had too many people on the job, how would that affect you as a Miller-Davis superintendent?
A. If CLS had too many people on the job, then I have the authority to dismiss the ones that I don’t feel are necessary.
Gerald Lash testified:
Q. Would it be unusual for CLS, the company, to provide tools for its employees?
A. Yes, it would.
Still other excerpts are illustrative:
Q. When you were using the acetylene torch are there other pieces of equipment or clothing or safety things that are required for you to wear?
A. Yes.
Q. What are those things?
A. Goggles and you have to wear a hard hat all the time.
Q. Are gloves required?
A. No, not necessarily.
Q. Is that part of your standard equipment that you bring to a job or is that something you get from the Miller-Davis people?
A. Get that from Miller-Davis.
Miller-Davis did not want day-to-day administrative responsibility for paperwork and general employment matters. This is specifically why it hired a labor broker. However, this in no way lessens the fact that control of the work site and project belonged unequivocally to Miller-Davis. Further, the record indicates that the contract was necessary for the protection of the union workers hired by CLS and keeping their business affairs separate from nonunion Miller-Davis employees.
We are cognizant that other jurisdictions may decide differently just how much deference to give the employment agreement. Again, we are persuaded that our view, to treat the agreement as one of many factors, not dispositive in and of itself, best advances the policies and purposes embodied in the worker’s compensation act. See O’Loughlin v Servicemaster Co Ltd Partnership, 216 Ill App 3d 27; 576 NE2d 196 (1991) (contract is to be considered, but is not necessarily conclusive); Red Line Express Co v Workmen’s Compensation Appeal Bd, 138 Pa Commw 375, 384; 588 A2d 90 (1991) (lease providing for “exclusive possession and use” is an important factor in determining who is employer, but is not conclusive). But see Kirby v Union Carbide, 373 F2d 590, 595 (CA 4, 1967) (the contract is controlling); Dugas v Pelican Construction Co, 481 F2d 773, 778 (CA 5, 1973) (the contract is controlling); Thornton v *47Paktank Florida, Inc, 409 So 2d 31, 33 (Fla App, 1981) (the contract is controlling); Hill v Erdle Perforating Co, 53 AD2d 1008; 386 NYS2d 265 (1976) (the contract is a factor in creating the jury question with regard to an employer-employee relationship); see O H Jackson v Weaver, 516 So 2d 702, 705 (Ala App, 1987) (the lease agreement is a factor to be considered in assessing the employer-employee relationship).