Opinion by Judge SILVERMAN; Dissent by Judge CYNTHIA HOLCOMB HALL.
SILVERMAN, Circuit Judge:A district court approved a consent decree that settled a civil rights lawsuit between a group of bus passengers and the Los Angeles County Metropolitan Transportation Authority (“MTA”) concerning the quality of bus service in their community. Fourteen months later, after certain service improvement goals had not been met, the district court-first through a Special Master, then directly-entered detailed orders concerning the operation of the L.A. County transportation system, including an order that MTA immediately acquire 248 additional buses to reduce passenger overcrowding even if that meant diverting funds from other transportation services under MTA’s jurisdiction. MTA appealed. Holding that the Special Master and district court correctly interpreted and applied the Consent Decree, and that the Special Master and district court acted within their power, we affirm the district court.
I. Facts
On August 31, 1994, a plaintiff class of Los Angeles County bus riders filed suit against MTA, alleging that MTA’s transportation policies discriminated against minorities in violation of the Civil Rights Act of 1964 and the Fourteenth Amendment to the United States Constitution. MTA is the statutorily created regional transportation planning, construction, funding, and operating agency for Los Angeles County. The suit alleged that MTA was spending a disproportionately large portion of its budget on rail lines and suburban bus systems that would primarily benefit white suburban commuters, while intentionally neglecting inner-city and transit-dependent minority bus riders who relied on the city bus system. The lawsuit was triggered by MTA’s decision to spend several hundred million dollars on a new rail line, foregoing an opportunity to reduce overcrowding problems on city buses, while at the same time increasing bus fares and eliminating monthly discount passes. In October 1996, after over two years of discovery and just before a trial was scheduled to begin, the parties reached a settlement and submitted to the district court a proposed consent decree that set forth a detailed plan to improve bus service.
*1044A. The Consent Decree
Under the consent decree’s terms, MTA agreed to make service improvements in the bus fleet to alleviate overcrowding and agreed to a set fare structure and fare increase procedure. In order to reduce bus overcrowding, the decree set forth specific “load factor targets” (“LFTs”)1 that were to be met by specific dates:
Improved Performance Goal: Reduced Load Factor Targets. MTA’s performance in meeting this critical objective of responding to consumer demand for bus services efficiently shall be measured by the reduction in levels of crowding on board buses. MTA shall establish as a five-year goal to be reached by the end of the fifth complete fiscal year following the approval of this Consent Decree, the reduction of the maximum load factor ceiling for all bus routes from 1.45 to 1.2 in the following increments (“target load factors”):
December 31,1997,1.35
June 30, 2000, 1.25
June 30, 2002,1.2
Thereafter, MTA shall maintain the 1.2 load factor for the duration of this Consent Decree.
The decree also called for the formation of a Joint Working Group (“JWG”) made up of an equal number of representatives from the plaintiffs’ class and MTA. The decree outlined the process by which the load factor targets were to be met:
MTA Plans to Meet Targets. MTA will plan to make available sufficient ad- . ditional buses and other vehicles to meet these target load factors. While MTA will have the discretion in determining how the targets will be met, MTA will consult with the JWG in formulating and implementing this plan. MTA will prepare at least 90 days prior to the beginning of each fiscal year, and make publicly available, a report setting forth its plan to meet the targets as of the date the report is issued, recognizing that changes in ridership, fares, the economy and other factors may require modifications to the plan. In addition, when MTA makes its scheduled modifications to its long range plan it shall incorporate plans to insure the availability and operation of the additional buses and other vehicles required to meet these targets. If ridership shall increase by more than 15 percent on any bus line MTA shall nevertheless make its best efforts to meet the target for that line and the target for that line may be deferred one (and only one) year. In addition, the JWG will designate a list of bus lines which may be exempted from the load factor requirement, such as lines with low frequency service.
As to how the improvements in load factors would be paid for, the decree stated:
Consistent with MTA’s other statutory responsibilities and obligations, MTA’s first priority for the use of all bus-eligible revenue realized in excess of funds already specifically budgeted for other purposes shall be to improve bus service for the transit-dependent by implementing MTA’s obligations pursuant to this Consent Decree. If sufficient funding is not provided to meet the obligations set forth in this Consent Decree, the matter *1045shall be addressed in accordance with the procedures set forth in this Consent Decree.
The decree also outlined a procedure to be followed in the event that MTA failed to meet the required LFTs by the dates scheduled:
Failure to Meet Targets. If MTA fails to meet the target load factors for all bus lines by the dates specified ..., MTA shall meet the target as soon as possible and reallocate sufficient funds from other programs to meet the next lower target load factor as scheduled. The reprogrammed funds, which may include but not be limited to revenues from Propositions A and C discretionary funds, shall be used to meet the target load factors. Any dispute concerning whether the targets have been met; or if targets have not been met, whether sufficient funds have been reprogrammed to meet the next target will be reviewed by the JWG. If the JWG cannot resolve the matter it will be referred to the Special Master. The failure of MTA to meet the target load factors shall not be deemed a changed or unforeseen factual condition for purposes of seeking a modification of this Consent Decree by MTA.
The decree also provided for the appointment of a Special Master to facilitate the resolution of disputes. As to this dispute resolution process, the decree provided:
Any dispute arising under any provision of Sections I through IV [which included the schedule of target load factor deadlines and funding reallocation requirements] of this Consent Decree in which the JWG has a role shall initially be addressed by the JWG. If the JWG cannot resolve the matter, or if the JWG does not have a role in the disputed function, this dispute shall be referred to the attorneys to the parties. If the attorneys cannot resolve the matter informally, the attorneys shall refer the matter to the Special Master for resolution, pursuant to procedures set forth by the Special Master. Any matter resolved by or referred to the Special Master may be reviewed by the District Court, along with the recommendations of the Special Master, if any, upon motion by either of the parties.
On October 29, 1996, the district court approved the consent decree.
B. The Special Master
Fourteen months later, soon after the first LFT deadline passed on December 31, 1997, the Special Master was pressed into service. The plaintiffs alleged that MTA had failed to meet the first LFT. In addition, the parties disputed the meaning of the LFTs, the extent to which MTA had or had not met its obligations under the consent decree, and the proper remedy to achieve compliance with the decree. After going through the preliminary dispute resolution procedure involving the JWG outlined in the decree, the parties submitted the dispute to the Special Master.
The Special Master set out a bifurcated procedure for resolving the matter. First, the parties were to brief the issue of how to measure compliance with the decree.2 *1046Second, once the Special Master had determined a method for measuring compliance, the matter was to be referred back to the JWG to determine whether MTA had fulfilled its obligations, and to craft a remedial plan if it had not. If the JWG could not agree on a remedial plan, the parties would then submit their respective remedial proposals to the Special Master for resolution of the issue.
In September 1998, the JWG found that MTA had failed to meet the decree’s December 31, 1997, 1.35 LFT on 75 out of the 79 bus lines measured. The JWG could not agree on a remedial plan, so the parties again submitted briefs to the Special Master assessing the violations and proposing remedies.
The Special Master issued a “Memorandum Decision and Order” that attempted to steer a middle course between the competing remedial plans. In analyzing the submitted data, the Special Master found that the LFT violations were caused primarily by two factors: 1) “missing buses”' — buses that were already in MTA’s fleet and scheduled to run, but failed to show up on a given day for some reason (usually mechanical problems, but also lack of a driver, traffic problems, etc.); and 2) “insufficient capacity” — simply not having enough buses available in MTA’s fleet to schedule sufficient service to meet demand on a particular line.
On the question of missing buses, the plaintiffs argued that 333 new buses were needed to solve the problem. However, the Special Master concluded that MTA’s remedial plan to accelerate procurement of new buses and to convert unreliable alternative fuel vehicles to diesel power would adequately address the missing bus problem, and that no additional bus purchases were needed.
As to insufficient capacity, MTA proposed purchasing 130 new buses by June 2000, in addition to 53 that were already scheduled to be procured for fleet expansion. MTA also claimed that by more effectively managing its existing bus fleet it could increase the operating capacity of the fleet, in effect getting more work out of the buses it already had rather than buying new buses to do that work. On the other hand, the plaintiffs proposed the purchase of 553 additional buses to meet the insufficient capacity problem. They argued that any reliance on MTA increasing its capacity through improved management would have been misplaced, given that MTA could have implemented the proposals earlier if they were really workable.
The Special Master ruled that MTA had overestimated the extent to which better management could resolve insufficient capacity problems and concluded that MTA’s remedy would not resolve adequately the *1047LFT violations. However, the Special Master also held that the plaintiffs had overestimated the number of buses required to fix the problem. Accordingly, the Special Master found that:
[T]o remedy the load factor violations caused by ‘insufficient capacity’ the MTA would need to add a total of 430 new buses to the fleet (277 plus 55 spares to meet the [missed December 31, 1997] 1.35 LFT and 126 plus 25 spares to meet the [then-upcoming June 30, 2000] 1.25 LFT, minus 53 buses already scheduled by MTA for purchase).
Because of the inherent delay in the procurement process, the Special Master also held that further action was necessary:
I have concluded that [the plaintiffs are] correct in [their] assertion that the MTA likely cannot meet the 1.35 load factor target “as soon as possible” (and the 1.25 load factor target by June 2000) unless it acts immediately to obtain additional buses on a temporary basis to remedy the load factor violations caused by insufficient capacity. In order to meet the requirements of Section II.A of the Consent Decree [setting forth the LFT deadlines], the MTA therefore should use its best efforts to lease or obtain, by whatever means available, at least 277 additional buses of any type appropriate for service on the MTA system. These buses should be scheduled for delivery on or before December 31, 1999 and should remain in service until the new buses required by the remedial plan are delivered. (Footnote omitted.) 3
C. District Court
In accordance with the appeal provision of the consent decree, MTA challenged the Special Master’s ruling in district court, claiming: 1) that the remedy imposed was based on an erroneous interpretation of the consent decree; and 2) the remedy ordered by the Special Master exceeded both his power and the power of the district court. The district court affirmed the authority of the Special Master and the district court to order a remedy under the decree. The court also found that “the Special Master’s findings regarding compliance with the consent decree were not clearly erroneous.” As to the Special Master’s remedial plan, the district court stated its ruling as follows:
1. The Special Master’s determination that 2484 additional buses are needed, and must be purchased immediately, to resolve “insufficient capacity” violations so as to meet the 1.35 target load factor that should have been met be December 31, 1997 is affirmed. The MTA shall, within thirty days, through lease or by other means, obtain 248 buses on a temporary basis until the 248 purchased buses arrive.5
2. Given the apparent increased reliability of the MTA’s current fleet, the Special Master shall reconsider whether the additional 496 buses he ordered the *1048MTA to purchase for spares are still needed.
3. The Court believes that it is too early to determine whether MTA is incapable of meeting the 1.25 target load factor by June 30, 2000, given the current progress apparently made by the MTA and the age of the point check data presented to the Special Master and the Court. Therefore, the Special Master shall re-evaluate the likelihood of the MTA meeting the 1.25 target load factor after he is presented with more up-to-date point check date [sic]. (Footnotes added.)
MTA appealed the district court’s ruling, and obtained a stay pending the appeal.
II. Standard of Review
A consent decree is enforceable as a judicial decree and “is subject to the rules generally applicable to other judgments and decrees.” Rufo v. Inmates of Suffolk County Jail, 502 U.S. 367, 378, 112 S.Ct. 748, 116 L.Ed.2d 867 (1992). This court reviews de novo the district court’s interpretation of the consent decree, but must defer to the district court’s factual findings underlying the interpretation unless they are clearly erroneous. Gates v. Gomez, 60 F.3d 525, 530 (9th Cir.1995); United States v. Gila Valley Irrigation Dist., 31 F.3d 1428, 1432 (9th Cir.1994). We must “give deference to the district court’s interpretation based on the court’s extensive oversight of the decree from the commencement of the litigation to the current appeal.” Gates, 60 F.3d at 530 (quoting Officers for Justice v. Civil Serv. Comm’n, 934 F.2d 1092, 1094 (9th Cir.1991)).
A district court order requiring modification of a defendant’s policy to corn-ply with the consent decree “is effectively an injunction and will be reversed ‘only where the district court abused its discretion or based its decision on an erroneous legal standard or clearly erroneous findings of fact.’ ” Id. A federal court enjoys broad equitable powers; its choice of equitable remedies is reviewed for an abuse of discretion. Stone v. City and County of San Francisco, 968 F.2d 850, 861 (9th Cir.1992); Kasper v. Board of Election Comm’rs, 814 F.2d 332, 339 (7th Cir.1987).
III. Discussion
MTA advances the same two arguments to this court that it advanced to the district court: 1) the remedial plan imposed by the district court and the Special Master was based on a misinterpretation of the consent decree; and 2) the district court and the Special Master did not have the power to order MTA to implement the remedial plan. MTA is wrong about the first; the remedial plan was based on a correct interpretation of the consent decree, and MTA’s claims to the contrary are without merit. Whether the court and the Special Master had the power to order MTA to immediately buy 248 new buses is a more difficult question.
A. Interpretation of the Consent Decree
1. Best Efforts
MTA’s primary argument on the interpretation of the consent decree is that the load factor targets were simply performance goals that MTA promised to use its “best efforts” to meet, but with which the decree only required “substantial compliance.” This argument is refuted by a reading of the decree as a whole. The decree set out a mathematically precise method of measuring bus overcrowding and a detailed schedule of load factor tar*1049gets that were to be met by specific dates. After the five year schedule of descending load factor targets, the decree provided that “MTA shall maintain the 1.2 load factor for the duration of this Consent Decree.” (Emphasis added.) The decree also provided that in the event MTA missed one of the scheduled LFTs, MTA “shall meet the target as soon as possible and reallocate sufficient funds from other programs to meet the next lower target load factor as scheduled.” (Emphasis added.) To say that MTA’s “best efforts” are enough for compliance would be to ignore the precise load factor schedule set out in the decree.
MTA cites “best efforts” language in the decree out of context as support for its position. In context, however, the language actually weighs against MTA’s argument. The decree provided that: “If ridership shall increase by more than 15 percent on any bus line MTA shall nevertheless make its best efforts to meet the target for that line and the target for that line may be deferred one (and only one) year.” The passage clearly indicates that MTA was required to meet the load factor targets as scheduled. In the event of a large unexpected ridership increase, the decree released MTA from that obligation for one and only one year. It is clear that MTA’s obligation was to meet the scheduled load factor targets, not simply to use its “best efforts.”
2. Statutory Obligations
MTA claims that because it does not have sufficient funds to purchase new buses under the Special Master’s remedial plan, the decree excuses compliance to the extent that the remedial plan would prevent MTA from meeting its other statutory obligations. In support of this claim, MTA cites to the following language in the consent decree:
Consistent with MTA’s other statutory responsibilities and obligations, MTA’s first priority for the use of all bus-eligible revenue realized in excess of funds already specifically budgeted for other purposes shall be to improve bus service for the transit-dependent by implementing MTA’s obligations pursuant to this Consent Decree.
This argument overlooks other language in the “Failure to Meet Targets ” section of the decree that addressed this specific situation and required MTA to “reallocate sufficient funds from other programs to meet the next lower target load factor as scheduled.” Any dispute about the fund reallocation was to be settled by the JWG, or if necessary, the Special Master. The Special Master pointed out that MTA had failed to demonstrate that it lacked sufficient funds to otherwise meet its statutory obligations:
In the MTA’s May 4, 1998 draft Restructuring Plan, the MTA identifies the many funding sources for which bus capital and/or operating expenses are eligible.... For many of these bus-eligible funding categories, no funds at all have been allocated to buses.... Thus, the fact that the MTA apparently has not applied for, allocated or received these bus-eligible funds somewhat undercuts the MTA’s argument that it will be forced to tap already-committed funds, and therefore violate its other statutory obligations, to comply with the Decree.
The Special Master’s factual findings are entitled to deference and are reviewed for clear error. Fed.R.Civ.P. 53(e)(2). MTA has not pointed to any evidence suggesting that the Special Master’s factual finding that MTA had not exhausted all sources of funding was clearly erroneous.
B. Power to Require Compliance with the Decree
1. Special Master as Mediator
MTA argues that the consent decree reflected the parties’ contemplation *1050that the Special Master would mediate disputes, not resolve them. This argument is without merit.
In outlining the procedure for resolving disputes, the decree explicitly provided that if the lawyers could not resolve a problem referred from the JWG, they “shall refer the matter to the Special Master for resolution, pursuant to procedures set out by the Special Master.” (Emphasis added.) Not for mediation. Not for suggestions. For resolution. The fact that the decree allowed the parties to challenge the Special Master’s decisions in the district court is further evidence that the Special Master was intended to be a deci-sionmaker. For there to be something to appeal from, the Special Master must have had the power to make a decision.
MTA’s argument also is quite disingenuous. Previously, when it suited MTA’s purposes, it had claimed that the decree gave the Special Master vast powers to decide matters involving the parties. For example, when the plaintiffs began to organize a fare strike against MTA in retaliation for MTA’s noncompliance with the load factor targets, MTA was only too happy to petition the Special Master for a temporary restraining order to enjoin the strike. In arguing for the restraining order, MTA contended that even though the consent decree was silent about the Special Master’s power to issue a TRO to enjoin a strike, the Special Master had broad powers unless specifically limited by the decree. At that time, MTA argued:
[T]he Special Master was appointed to monitor the parties’ compliance with, and resolve any disputes arising under, the Consent Decree. The Consent Decree does not impose any limitations on the Special Master’s powers or authorities [sic] to effectuate these objectives. Accordingly, the Special Master has inherent authority to issue orders and resolve disputes arising under the Consent Decree.
(Emphasis added.) Under the decree’s plain language, the Special Master had the power to resolve disputes, subject to appeal, and that is exactly what he did.
2. Scope and Intrusiveness of the Remedial Order
Finally, MTA raises federalism concerns regarding the Special Master’s and district court’s remedial order. Specifically, MTA contends that the order requires MTA to violate state and federal environmental laws. We reject this contention and find that the district court did not abuse its discretion because (1) MTA consented to this dispute resolution, (2) MTA had the opportunity to comply with the Consent Decree but failed to do so, and (3) the remedial order does not require a violation of state or federal laws.
When imposing a remedial scheme on a state institution, a federal court must not unduly insert itself into the institution’s management. However, “federalism concerns in institutional reform litigation ... do not automatically trump the powers of federal courts to enforce the Constitution or a consent decree.” Stone, 968 F.2d at 861. “[S]everal courts have held that federalism concerns do not prevent a federal court from enforcing a consent decree to which state officials have consented.” Id. at 861 n. 20 (citing United States v. City of Yonkers, 866 F.2d 444, 454 (2d Cir.1988), rev’d in part on other grounds sub nom. Spallone v. United States, 493 U.S. 265, 110 S.Ct. 625, 107 L.Ed.2d 644 (1990); Allen v. Alabama State Bd. of Educ., 816 F.2d 575, 577 (11th Cir.1987); United States v. District of Columbia, 654 F.2d 802, 808 & n. 11 (D.C.Cir.1981)). MTA’s consent to this form of dispute resolution relieves many federalism concerns.
*1051Moreover, the remedial order does not violate the general principle that “federal courts in devising a remedy must take into account the interests of state and local authorities in managing their own affairs.” Milliken v. Bradley, 433 U.S. 267, 280-81, 97 S.Ct. 2749, 53 L.Ed.2d 745 (1977). MTA failed to comply with the Consent Decree before there ever was a Special Master’s order. After being found in violation of the Consent Decree, MTA had the opportunity to submit its own effective remedial scheme. The Special Master made detailed findings, reasonably concluded that MTA had submitted an inadequate remedial scheme, and meticulously fashioned the remedial order.
Finally, the federal and state laws to which MTA refers are simply funding provisions with which MTA may choose to comply. The reward for compliance is state and federal funding. However, failure to follow the requirements of the funding mandates is not a violation of the law. At most, it might disqualify MTA from entitlement to receive funds. MTA’s compliance with the district court’s order would no more violate federal law than would a state legislature’s decision to lower the statutory drinking age below twenty-one in spite of the federal law that conditions receipt of federal funds on a drinking age of twenty-one or older, 23 U.S.C. § 158. See South Dakota v. Dole, 483 U.S. 203, 107 S.Ct. 2793, 97 L.Ed.2d 171 (1987). The state might lose federal highway funds under federal law, but it would not be in violation of that law. Thus, the district court did not abuse its discretion.
The district court’s order is AFFIRMED.
. A “load factor” is a numerical representation of the number of people standing on a bus in relation to the number of seats. It is calculated by dividing the total number of bus passengers riding a bus during a specific period of time (20 minutes in this case) by the total number of seats on the bus. A load factor of 1.35, for example, means that the average number of standees on a given bus during a given 20-minute period exceeds the number of seats by a factor of .35.
. In measuring compliance, MTA argued that the load factors for all bus lines during a given period of time should be compiled and averaged, then reduced by a certain factor to account for a standard number of mechanical failures. The Special Master rejected MTA's proposal, ruling that compliance with the LFTs had to be measured by looking at each individual bus line. MTA had to show that during any 20-minute weekday period in the peak direction of travel on the line, the average ratio of passengers to seats available did not exceed the particular LFT. Under the Special Master’s interpretation, a single bus line would be tested at a number of different points during the same 20-minute period. If the average of those test results exceeded the *1046relevant target load factor, MTA would have failed to comply with the decree for that bus line. The Special Master based this interpretation on the language of the decree itself, such as the requirement that MTA reduce the maximum load factor ceiling “for all bus routes,” and the definition of the peak load factor as the total number of passengers divided by the total number of seats "during any 20 minute weekday period in the peak direction of travel on each bus line.”
The Special Master also resolved a dispute regarding the method of gathering load factor data. MTA argued that compliance should be measured by gathering data only during fixed time periods traditionally used by MTA, called the "fixed window” approach. The Special Master rejected this interpretation, holding that data could be gathered during any 20-minute period as long as that 20-minute window did not overlap with another 20-minute period. This was known as the "sliding window” approach. The Special Master based his interpretation on the language in the decree requiring the load ceiling to be achieved during "any” 20-minute period, and because he determined that the sliding window approach would provide a more accurate assessment of the load factor conditions on the buses.
. Although the Special Master initially ordered MTA to acquire 277 new buses, plus 55 spares, he reduced that number to 248 plus spares on May 14, 1999, following MTA's motion for clarification and modification of the March 6, 1999 order.
. The district court’s bus count mirrors the Special Master’s final order.
. The district court amended this portion of the order on October 6, 1999. The amended order required that, “MTA shall, by November 5, 1999, contract to obtain 248 buses on a temporary basis until the 248 purchased buses arrive. The 248 temporary buses shall be placed into operation no later than January 3, 2000."
.The 49 spare buses number appears to reflect the application of a standard 20% reserve replacement figure that the Special Master had used to calculate the number of new buses needed to the revised number of *1048temporary buses ordered by the district court. Twenty percent of the 248 buses the district court ordered procured is approximately 49.