E. GRADY JOLLY, Circuit Judge:

This appeal arises out of Darrel Rundus's insistence on distributing free Bible tracts at the annual Texas State Fair (the "Fair"). He has twice had his efforts thwarted by the private corporation, the State Fair of Texas ("SFOT"), that runs the Fair. Shortly after his second failed attempt, Rundus filed this suit under 42 U.S.C. § 1983 against the City of Dallas (the "City") and SFOT, alleging that they had violated his First Amendment rights. The trial court held that no state action was involved in preventing Rundus from distributing the tracts, and therefore it did not reach the First Amendment question. The trial court also assessed against Rundus the costs incurred by Appellees in responding to his voluminous discovery requests. Rundus appeals.

I.

The Fair is held annually at Fair Park, a parcel of land that the City has owned since 1904. As we have said, Rundus has twice attempted to distribute Bible tracts at the Fair. Standing in his way is SFOT, the private corporation that runs the Fair, and its Exhibitor Rule 9 ("the restriction").1 SFOT is governed by an Executive Committee; no government employees, officials, or appointees serve on the Committee,2 SFOT does not receive any payments from the City, and SFOT pays the City rent and a marketing fee. SFOT uses its Fair revenues to improve Fair Park, but it improves only the areas it utilizes during the Fair, and cannot make any improvements without prior written consent from the City.3 SFOT is also required to maintain a reserve fund of at least $4.5 million, in order to ensure that the Fair is held during times of financial instability.

Since 1904, the Fair has been run by private organizations. SFOT is responsible for running the Fair pursuant to a contract it has with the City—the Fair Park Contract ("FPC"). Under the FPC, in non-Fair time, the City has primary control over Fair Park; under an earlier version of the FPC, SFOT had year round control. During the Fair, however, SFOT has primary control over the grounds, and it also decides who to admit into Fair Park in the days immediately preceding and following the Fair. A ticket is required for admission to the Fair, and ticket prices are within SFOT's sole discretion. SFOT also enacts its own rules and regulations, including the restriction—which prohibits the distribution of literature without a booth rental. Rundus has declined to rent a booth.

The City assigns around 160 police officers to work the Fair. They enforce applicable laws, including criminal trespass, but do not enforce SFOT's rules and regulations. SFOT attempts to remove unwelcome individuals without police intervention, involving the police only if the individual refuses to leave.

After his most recent rebuffed attempt to distribute tracts, Rundus brought suit against the City and SFOT under 42 U.S.C. § 1983, alleging that the restriction violates his First Amendment rights. During discovery, Rundus requested voluminous documents, ostensibly so he could examine Appellees' historical relationship. In response to these requests, SFOT and the City produced copies of the requested documents.

The case was tried to the bench on stipulated facts. The trial court held that SFOT is not a state actor. The trial court also assessed certain costs against Rundus, including the costs Appellees incurred in making copies to respond to his discovery requests. Rundus appeals.

II.

Rundus argues that his First Amendment rights have been violated. There is a hurdle to overcome first, however. Thus, Rundus first argues that the trial court erred in finding that no state action was involved in the prevention of his distribution of religious tracts. Consequently, before turning to the constitutional question, we will address whether state action is involved in the claims Rundus presents, for state action is a prerequisite for bringing an action under Section 1983. See, e.g., Miss. Women's Med. Clinic v. McMillan, 866 F.2d 788, 792-93 (5th Cir. 1989). We review de novo the trial court's determination that no state action was involved in enforcement of the restriction. See Eugene v. Alief Indep. Sch. Dist., 65 F.3d 1299, 1303 (5th Cir.1995).

In order to show that there is state action, Rundus must show that either: 1) The restriction represents an official City policy or custom, see Monell v. Dept. of Soc. Servs., 436 U.S. 658, 691, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978) or 2) SFOT's conduct in enacting and enforcing the restriction is "fairly attributable" to the City. See Lugar v. Edmondson Oil, 457 U.S. 922, 937, 102 S.Ct. 2744, 73 L.Ed.2d 482 (1982). Rundus focuses solely on the latter theory, so we must ask whether SFOT, although a private corporation, acts under color of state law, and is thus a state actor for Section 1983 purposes.4

The attribution test has not been precisely defined, as the Supreme Court has addressed it on a case by case basis. See, e.g., Brentwood Acad. v. Tenn. Secondary Sch. Athletic Ass'n., 531 U.S. 288, 295-96, 121 S.Ct. 924, 148 L.Ed.2d 807 (2001). In Brentwood, the Court held that, although in form the Athletic Association was a private corporation separated from the state and its agencies, it nevertheless was a state actor, because it was created to govern public school athletics; its members were mostly public schools; its employees were treated as state employees, and were eligible for state retirement benefits; and it was supported by gate receipts from games played between public schools and from membership fees paid by those schools. Id. at 299-300, 121 S.Ct. 924. Although Tennessee deleted its rule that had officially recognized the Association as its regulator in the field of interscholastic athletics, the Court determined that this change was merely formulaic. Id. at 300-01, 121 S.Ct. 924.

Rundus, relying heavily on Brentwood, contends that SFOT is a state actor when it runs the Fair, and that it therefore must meet constitutional standards when it restricts speech. He argues that the Appellees, by virtue of the FPC, are entwined in a joint venture, which is demonstrated by their financially interdependent relationship. He contends that this financial entwinement is established under the factual analysis endorsed in Brentwood, i.e., the Appellees mutually commit substantial financial sums to improve Fair Park; SFOT pays a portion of the police officers' wages earned during the Fair; and SFOT is required to maintain a reserve fund to ensure that the Fair will be held in times of financial distress. Rundus essentially argues that SFOT is a shell corporation, allowing the City to run the Fair and inject funds into Fair Park while avoiding liability. SFOT counters that Brentwood is not this case, because the overlapping leadership structure, financial dependence, and overall symbiosis, so essential to that holding, are not present. See 531 U.S. at 298-301, 121 S.Ct. 924. SFOT further argues that it is not financially dependent on the City, and, indeed, it receives no payments from the City. Moreover, SFOT says the FPC is representative of a long term commercial lease, a lease formalized only after extensive negotiations.5

Insisting that state action is present here, Rundus directs us to decisions in which other circuits, when presented with similar facts, have found state action present. See, e.g., Wickersham v. City of Columbia, 481 F.3d 591, 599 (8th Cir.2007); Parks v. City of Columbus, 395 F.3d 643, 652-53 (6th Cir.2005). Not to be outdone, SFOT diverts us to competing precedent indicating that running a public event like the Fair does not automatically amount to state action. Lansing v. City of Memphis, 202 F.3d 821, 833 (6th Cir.2000); Reinhart v. City of Brookings, 84 F.3d 1071, 1073 (8th Cir.1996). That each side attempts to confound us with colorable support for its position is not surprising, given that the Supreme Court has held that fact-intensive inquiries are necessary in Section 1983 cases. Brentwood, 531 U.S. at 298-99, 121 S.Ct. 924 (quoting Lugar, 457 U.S. at 939, 102 S.Ct. 2744). Although the cases from other circuits that we have cited do not bind us, we can draw from them for our decision today; indeed, the Supreme Court has said that "[a]midst such variety, examples may be the best teachers." Id. at 296, 102 S.Ct. 2744.

We turn first to the Sixth Circuit. In Lansing, a private corporation—with city officials on its board and with city funding—ran a festival on the streets of Memphis, Tennessee. 202 F.3d at 830-34. The Sixth Circuit held that the corporation was not a state actor, even though it cooperated with—and provided economic benefit to—the city. Id. at 831. Thus, faced with significantly more compelling evidence of state funding and involvement than Rundus has provided, the Sixth Circuit rejected the notion of state action.

In Parks, however, the Sixth Circuit held that there was state action in the operation of an arts festival on the riverfront in Columbus, Ohio. 395 F.3d at 653. It found particularly important that the city aided the private event organizers in enforcing their speech restrictive regulations. Id. It is worth noting that the event organizers held a permit, and were not the city's tenants, id. at 645; as such, they did not hold the same legal rights that a private tenant like SFOT does.

The Eighth Circuit cases present a similar dichotomy. In Reinhart, the court held that a privately run arts festival was not a state actor, even though the festival took place on public property. 84 F.3d at 1072-73. The private event organizers adopted and enforced a literature distribution rule that required anyone seeking to distribute literature to obtain a booth. Id. A candidate for political office who wished to distribute literature in support of his campaign was forced to leave the park, as he had no booth, and, indeed, could not have obtained one, as the private committee categorically refused to give booths to political candidates. Id. at 1072.

In Wickersham, the Eighth Circuit held that a Memorial Day air show, held at the city's airport and run by a nonprofit corporation, involved state action. 481 F.3d at 599. The court said: "[T]he city [of Columbia] not only provided critical assistance in planning and operating the show, but also played an active role in enforcing the particular speech restrictions . . . ." Id. at 598. In the court's judgment, these facts merited departure from Reinhart, because "[t]here, the city had no role in planning, advertising, or managing the festival." Id. This description of Reinhart cogently summarizes the Fair as well. Furthermore, the Eighth Circuit observed, "[t]he contributions of the Columbia police go beyond the kind of neutral assistance that would normally be offered to private citizens in enforcing the law of trespass." Id.

Rundus, referring to Parks and Wickersham, draws an analogy to these cases, pointing out that Dallas police officers provide security at the Fair. There is a difference, however. During the Fair, Dallas police enforce only criminal statutes and ordinances that provide "neutral assistance."6 SFOT is not a state actor merely because it takes advantage of law enforcement services provided to the public. See Tulsa Prof'l Collection Servs., Inc. v. Pope, 485 U.S. 478, 485, 108 S.Ct. 1340, 99 L.Ed.2d 565 (1988) ("Private use of state-sanctioned private remedies or procedures does not rise to the level of state action."); Moose Lodge No. 107 v. Irvis, 407 U.S. 163, 173, 92 S.Ct. 1965, 32 L.Ed.2d 627 (1972) (converting a private actor into a state actor merely because the private entity uses state services would destroy the distinction).

We hold that the facts here clearly indicate SFOT is not a state actor; it runs a private event on public property. The pervasive entwinement present in Brentwood is not presented in the facts before us. The City has no say in SFOT's internal decision making, and had no role in enacting or enforcing the restriction on distribution of literature. Nor are we convinced by Rundus's argument that Appellees' mutual commitment to improve Fair Park demonstrates state action, because SFOT improves only the portions of Fair Park that will attract more fairgoers. In short, the facts presented are not sufficiently analogous to Brentwood to conclude that SFOT is a state actor. We thus need not address whether the restriction meets constitutional muster, and will proceed directly to address whether the trial court erred in assessing costs against Rundus.

III.

Rundus argues that the trial court erred in ordering him, under FED. R. CIV. P. 54, to repay Appellees costs they incurred— $14,547.23 by SFOT and $6,410.10 by the City—in responding to his discovery requests. Specifically, Rundus objects to the assessment of costs Appellees incurred in copying documents that he requested them to produce. To succeed, he must overcome the provision of FED. R. CIV. P. 54(d)(1). "Unless a federal statute, these rules, or a court order provides otherwise, costs—other than attorney's fees—should be allowed to the prevailing party." FED. R. CIV. P. 54(d)(1).

A.

Rundus's primary argument is that the Supreme Court has held that, under the Federal Rules of Civil Procedure, costs incurred in preparing discovery responses cannot be assessed. Oppenheimer Fund, Inc. v. Sanders, 437 U.S. 340, 358, 98 S.Ct. 2380, 57 L.Ed.2d 253 (1978). Rundus does not identify any specific rule of procedure that, by its text, establishes this point.7 Instead, he says "Article V of the Federal Rules of Civil Procedure," which is the Article containing all the discovery rules, "creates an obligation to comply with discovery requests made under [that] article and bear the costs of compliance." Rundus's argument that the Federal Rules of Civil Procedure forbid a trial court from awarding discovery costs is subject to de novo review. See Romaguera v. Gegenheimer, 162 F.3d 893, 895 (5th Cir.1998) (interpreting the Rules is a legal question).

The Supreme Court has held that "the responding party must bear the burden of responding to discovery requests." Oppenheimer Fund at 358, 98 S.Ct. 2380 (emphasis added). Appellees concede that parties must bear their own costs initially when actually responding to discovery, but they argue the trial court clearly has the authority to shift such costs at the litigation's conclusion.

In Oppenheimer Fund, the Court was faced with a dispute over which party should bear the costs of providing notice to potential class members in a class action suit. Id. To answer that question, the Court drew an analogy to the discovery rules. Id. Nowhere in the opinion does the Court mention, much less analyze, Rule 54. Id. Oppenheimer Fund thus does not purport to address the question here—whether, at the conclusion of the trial stage of litigation, discovery costs can be assessed against the losing party. In a case that does address that precise question, we have expressly held that the trial court can assess such costs. See Harrington v. Texaco, Inc., 339 F.2d 814, 822 (5th Cir.1964). In Harrington, we said that "the authority of the trial court to assess `necessary and reasonable' costs incurred during discovery `can hardly be doubted.'" Id. Rundus has presented no authority from this court that suggests that this holding was in error, and his argument is therefore rejected.

B.

Rundus advances an alternative argument. He says that the trial court abused its discretion in assessing the copying costs because the documents were not "necessarily obtained for use in the case." We have previously held that "whether a deposition or copy was necessarily obtained for use in the case is a factual determination to be made by the district court. We accord great latitude to this determination." Coats v. Penrod Drilling Corp., 5 F.3d 877, 891 (5th Cir.1993).

Rundus predicates his argument on the Supreme Court holding that the only costs that are recoverable under Rule 54 are those listed in 28 U.S.C. § 1920. See Crawford Fitting Co. v. J.T. Gibbons, Inc., 482 U.S. 437, 441, 107 S.Ct. 2494, 96 L.Ed.2d 385 (1987). In turn, Section 1920(4) authorizes the assessment of "the costs of making copies of any materials where the copies are necessarily obtained for use in the case."

Rundus then urges us to interpret Section 1920(4) in the same way that we have interpreted Section 1920(2). Section 1920(2), in relevant part, contains language identical to that of Section 1920(4), and allows the assessment of "[f]ees for printed or electronically recorded transcripts necessarily obtained for use in the case." In interpreting Section 1920(2), and, more specifically, the operative phrase—"necessarily obtained for use in the case"—we have held that costs incurred "merely for discovery" do not meet that standard. See Fogleman v. ARAMCO, 920 F.2d 278, 285-86 (5th Cir.1991) (internal citations and quotation marks omitted). We further held, however, that such costs are recoverable if the party making the copies has a reasonable belief that the documents will be used "during trial or for trial preparation." 920 F.2d at 285. Finally, as we have indicated, the determination of whether such copies are reasonably necessary is best made by the district court, and we give great deference to its decision. The trial court did not abuse this broad discretion by assessing copying costs against Rundus.

IV.

We have held that SFOT, as a private corporation performing the operation of the Fair, is not a state actor. We have further held that the trial court did not commit error by assessing copying costs against Rundus. The trial court's judgment is, therefore, in all respects,

AFFIRMED.