William Earl JONES, Plaintiff-Appellant, v. UNIVERSITY OF CENTRAL OKLAHOMA, formerly known as Central State University, State of Oklahoma, formerly known as Central State University; William Lillard, President, University of Central Oklahoma; Kenneth Elsner, Dean of the College of Education; Emmet Osgood, individually and in his official capacity; Carl Breazeale, individually and in his official capacity; Lonnie Gilliland, Jr., Director of Safety Institute, University of Central Oklahoma, Defendants-Appellees, 13 F.3d 361

McKAY, Chief Judge. The parties have agreed that this case may be submitted for decision on the briefs. See Fed.RApp.P. 34(f); 10th Cir.R. 34.1.2. The case is therefore ordered submitted without oral argument. Plaintiff-Appellant (Plaintiff) appeals the federal district court’s grant of

summary judgment on all counts to the Defendants-Appellees, the University of Central Oklahoma and various officials thereof (the “University”). Plaintiff, a former instructor at the University, argues that in rejecting his application for tenure, the University violated his due process rights giving rise to a claim under 42 U.S.C. § 1983, discriminated against him on the basis of race giving rise to a Title VII claim, and intentionally inflicted him with emotional distress. We first address Plaintiffs § 1983 claim that the University violated his due process rights by rejecting his application for tenure.

Plaintiffs request for tenure was denied because he failed to satisfy the formalized nineteen-step process utilized by the University to evaluate his application. Plaintiff contends that, because of the University’s past practice and representations, he had a “legitimate claim of entitlement” — and thus a protected property interest — that required the University to evaluate his application under an informal and less strenuous “local tenure” process.

Under the unwritten “local tenure” guidelines, Plaintiff claims, the main factor considered is longevity, candidates are not required to submit a formal application, and candidates are awarded tenure as a matter of course. Plaintiff submits that he would have been awarded tenure had the informal procedure been used as he had been led to believe it would. The district court, without elaboration and relying solely on LaVerne v.

University of Texas System, 611 F.Supp. 66 (D.C.Tex.1985), held as a matter of law that Plaintiff did not have a legitimate claim of entitlement to a grant of tenure under the informal process because the University had a formal, written procedure. The court stated simply, “[A]n unwritten or local ténure policy such as relied on by Plaintiff cannot exist if there is in force and effect a formal written tenure policy, as is the situation with the University of Central Oklahoma.” Jones v. University of Central Oklahoma, et al., No. CIV-91-1788—D, slip op. at 1 (W.D.Okla. April 1, 1993).

Although they do not represent a clean split of authority, there appear to be two general schools of thought among the circuits on the precise issue presented to the court today — whether a university professor can have a legitimate claim of entitlement in continued employment pursuant to an informal, unwritten tenure policy when there is a contrary, formal policy already on the books. The first view is evident in the Sixth Circuit’s decision in Soni v.

Board of Trustees of Univ. of Tennessee, 513 F.2d 347 (6th Cir.1975), cert. denied, 426 U.S. 919, 96 S.Ct. 2623, 49 L.Ed.2d 372 (1976), which held that despite a university’s written tenure regulations, a professor may have a protectable property interest if the university’s conduct and representations concerning an informal procedure give him a “reasonable expectation” of continued employment. Id. 513 F.2d at 350-51. Apparently, the Sixth Circuit concluded that one has a “legitimate claim of entitlement” whenever he has a “reasonable expectation” that he will receive a certain benefit from the state.

Although the Sixth Circuit has subsequently gone to great lengths to distinguish other cases from Soni, this broad view of a protected property interest still prevails in that jurisdiction and has had a modicum of influence on other courts. See Hannon v. Turnage, 892 F.2d 653, 658 (7th Cir.1990) (citing Soni and noting that “a property interest can arise through a legitimate and reasonable reliance on a promise made from the government”); Tyler v. College of William and Mary, 429 F.Supp. 29, 31-32 (E.D.Va.1977) (discussing “reasonable expectation” test of Soni and distinguishing on facts).

On the other hand, the Ninth Circuit held in Haimowitz v. University of Nevada, 579 F.2d 526 (9th Cir.1978), that the existence of a written tenure policy precludes a professor from gaining a property interest in continued employment through any informal, unwritten policies absent extraordinary circumstances. Id. at 528-29. Haimowitz discredited the Sixth Circuit’s broad view in Soni and seemed compelled by policy considerations concerning a university’s need to enforce its written tenure guidelines. Id.

Several courts have followed the' Ninth Circuit’s approach in Haimowitz, and the rule has gradually solidified to the point where, as expressed by the district court below, the mere existence of a written tenure procedure in effect nullifies a professor’s claim to a property interest stemming from any contract implied through conduct or oral representations. See Lovelace v. Southeastern Massachusetts Univ., 793 F.2d 419, 423 (1st Cir.1986); Colburn v. Trustees of Indiana Univ., 739 F.Supp. 1268, 1293 (S.D.Ind.1990), aff'd, 973 F.2d 581 (7th Cir.1992); LaVerne v.

University of Texas System, 611 F.Supp. 66, 69 (S.D.Tex.1985). We do not believe that either of the approaches set forth above is faithful to the Supreme Court’s instructions in Perry v. Sindermann, 408 U.S. 593, 92 S.Ct. 2694, 33 L.Ed.2d 570 (1972), and Bishop v. Wood, 426 U.S. 341, 96 S.Ct. 2074, 48 L.Ed.2d 684 (1976), requiring federal courts to define pro-tectable property interests in the state-employment context solely through the application of state law.

In Perry, the Regents of the state university where the respondent was employed as a professor denied the respondent tenure without first offering him a hearing. Despite the lack of a contract or written tenure guidelines, the Court held that the professor could have a property interest in continued employment because of his reliance on the university’s conduct and representations which created an expectation that he would be given tenure pursuant to an informal practice or de facto tenure policy.

The Court stated, “A person’s interest in a benefit is a ‘property’ interest for due process purposes if there are such rules or mutually explicit understandings that support his claim of entitlement to the benefit....” Perry, 408 U.S. at 601, 92 S.Ct. at 2699 (emphasis added). The Court then cited Corbin on Contracts and discussed general theories of contract law, such as implied or quasi-contracts, through which a litigant can establish an entitlement or a contract right to certain property because of “mutually explicit understandings” inferred from the parties’ conduct, statements, or practices.

Id. at 601-602, 92 S.Ct. at 2699-2700. The Perry Court made clear, however, that it expressed no opinion as to whether the respondent had actually proved the existence of a quasi-contract with the university vesting him with a property right — that determination would have to be made by the district court on remand and solely in accordance with the appropriate state contract law. Id. at 603 n. 7, 92 S.Ct. at 2700 n. 7. The Court stated: We do not now hold that the respondent has any such legitimate claim of entitlement to job tenure. For “property interests ... are not created by the Constitution.

Rather, they are created and their dimensions are defined by existing rules or understandings that stem from an independent source such as state law.” Id. (citation omitted). Continuing its discussion, the Court held that the issue of whether the respondent had a protectable property interest must be decided pursuant to the law of Texas, the state where the cause of action arose.

The Court’s analysis indicated that if, upon remand, a determination were made that a teacher in the respondent’s position could bind the university to an implied contract under Texas contract law, then the respondent would have a “legitimate claim of entitlement,” and thus a protectable property interest. On the other hand, “[i]f it is the law of Texas that a teacher in the respondent’s position has no contractual or other claim to job tenure, the respondent’s claim would be defeated.” Id. Three years later, the Supreme Court reaffirmed its Perry analysis in Bishop v. Wood.

In Bishop, the Court stated, “A property interest in employment can, of course, be created by ordinance, or by an implied contract. In either ease, however, the sufficiency of the claim of entitlement must be decided by reference to state law.” Bishop, 426 U.S. at 344, 96 S.Ct. at 2077 (emphasis added). Both the Sixth Circuit’s analysis in Soni and the Ninth Circuit’s view in Haimowitz fail to use. substantive state contract law to determine whether a property interest has been created by a state university’s conduct.

Instead, the courts seem to rely on their own views of how narrowly or broadly the due process clause should be construed and on other extraneous policy considerations. For example, in Soni, the Sixth Circuit held that a professor had a legitimate claim of entitlement to continued employment because the University of Tennessee’s representations created in him a “reasonable expectation” of indefinite employment.

The court made no reference to Tennessee contract law as to whether an implied contract is created in Tennessee when one engages in conduct that leads another to a “reasonable expectation” that certain representations will be followed. Likewise, in Haimowitz, the Ninth Circuit did not consider Nevada contract law when it decided that an implied contract cannot arise if a contradictory written document already exists. By ignoring state law, the analytical framework developed by these cases fails to discipline the definition of “legitimate claim of entitlement.” Consider the Ninth Circuit’s approach in Haimowitz.

Based apparently on policy considerations, the court held that the university’s conduct could not form an implied contract or modify the written tenure guidelines. However, if under Nevada law, Professor Haimowitz could have established an implied contract binding the university to its representations concerning a less strenuous, “local tenure” process, how can one reasonably conclude that his “claim” to the local tenure process was not “legitimate,” or that, under Nevada law, he was not “entitled” to have his application judged under the less strenuous procedure?

Likewise, the Sixth Circuit found that Professor Soni had a property interest because he had a “reasonable expectation” of continued employment. If Tennessee law holds that no implied contract is created when a party has a mere “reasonable expectation” that a promise will be kept, how can one conclude that Soni was “entitled” to the less strenuous tenure procedure under Tennessee law?

By departing from the Supreme Court’s instructions that “property interests ... are not created by the Constitution.... they are created and their dimensions are defined by existing rules or understandings that stem from an independent source such as state law,” Perry, 408 U.S. at 602 n. 7, 92 S.Ct. at 2700 n. 7, the federal courts have left the states without a principled or predictable way to define property rights or organize their internal institutions.

For example, if a state wanted its universities to be able to enforce written tenure guidelines in every case, it could simply pass a law that makes implied contracts void in the tenure setting, or a law that would divest university officials of all authority to enter into implied contracts with their employees, thereby making the implied contracts ultra vires. Assuming the law were upheld, the state, instead of the federal courts, would continue to have control over grants of property interests within its borders.

If a state has not taken such actions, and its contract law allows a professor to bind a university to its oral representations despite a contrary written contract, then the state has granted professors property rights in continued employment in certain situations. This result is more consistent with the language and spirit of Perry and Bishop. Thus, we hold that “legitimate claim of entitlement” in the state university tenure setting is defined solely through the application of state contract and employment law.

If, under state law, the employee can establish a contract right to continued employment through an alleged unwritten, “local tenure” procedure — even when there exists a written procedure already on the books — then his claim is legitimate and the state has granted him a property interest protected by the due process clause. If he cannot establish a contract right under any of the contract theories recognized by the law of that state, then his claim that he has been granted an entitlement from that state is not legitimate. See, e.g., Woolsey v.

Hunt, 932 F.2d 555 (6th Cir.), cert. denied, — U.S. -, 112 S.Ct. 195, 116 L.Ed.2d 155 (1991). The district court erred in holding that an implied contract cannot exist simultaneously with a written contract without first referring to Oklahoma law on the issue. Just as property interests are not created or defined by the Constitution, in the state employment context property rights cannot be created or taken away by federal courts. We have made no attempt at a definitive determination of what Oklahoma law would hold on the issue of implied contracts in this case.

It may be that the trial court may wish to utilize the available processes for certifying this question to the Oklahoma Supreme Court. That, of course, is a matter for the trial court’s discretion. Because it appears from our initial research that Oklahoma law may allow implied contracts to form in these situations, and because Plaintiff has alleged facts and circumstances that could potentially give rise to an implied contract under general principles of contract law, the district court should not have granted

summary judgment for the University without a careful review of Oklahoma law. Thus, a reversal is necessary. The district court also granted judgment to the University on the Title VII discrimination issue partially because of its view that only one evaluation procedure — the written guidelines — existed at the University. Because our disposition sheds doubt on that view and may allow Plaintiff a chance to prove that a second evaluation procedure existed, the district court must reconsider the Title VII suit in light of this opinion and the subsequent proceedings on remand.

Finally, we reverse the district court’s dismissal on the pendant state law claim as premature now that the federal claims have been reinstituted. The judgment of the district court is reversed in all respects and is remanded for further proceedings consistent with this opinion. REVERSED and REMANDED. . See, e.g., 15 Okla.Stat.Ann. § 2-133 (West 1983); Hinson v. Cameron, 742 P.2d 549, 554-557 (Okl.1987). We do not intend our statements as to Oklahoma law to be binding on the district court. The parties on appeal have not briefed the issue of Oklahoma law, and our research was not exhaustive.

We also express no opinion concerning the immunity issues not reached by the district court.