DECISION AND ORDER

LYNN ADELMAN, District Judge.

In the present case, two organizations representing residents of Milwaukee's inner city challenge a decision of the Federal Highway Administration ("FHWA") and the Wisconsin Department of Transportation ("WisDOT") to make improvements to the "Zoo Interchange," which is a part of the Interstate Highway System located in the Milwaukee Metropolitan Area. The plaintiffs contend that the agencies failed to prepare an adequate environmental impact statement before deciding to proceed with the project, in violation of the National Environmental Policy Act of 1969 ("NEPA"). Before me now is the plaintiffs' motion for a preliminary injunction. The plaintiffs seek an order prohibiting the agencies from taking further action in connection with the project pending a final decision on the merits of this case.

For the reasons stated below, I find that the plaintiffs have a likelihood of success on the merits and that they are likely to suffer irreparable harm in the absence of an injunction. However, an injunction could delay the project and increase its cost, and in deciding whether to issue an injunction I must balance this potential delay and potential increased cost against the harm that the plaintiffs would suffer in the absence of an injunction. As explained below, I find that before I can strike the proper balance I must hold an evidentiary hearing.

I. BACKGROUND

The Zoo Interchange (so named because of its proximity to the Milwaukee County Zoo) is located in western Milwaukee County, at the junction of Interstate 94, Interstate 894, and U.S. Highway 45. It is a busy interchange that was first built in 1963. In May 2008, FHWA and WisDOT began to seriously consider making long-term improvements to the interchange. Eventually, the agencies decided to make improvements to the interchange itself and its four approach legs, as depicted in the image below. The project area's total east-west distance is about 3.5 miles and its total north-south distance is about 5.5 miles.

In studying potential improvement projects, the agencies had four main goals in mind. The first was to address the deteriorating condition of the Zoo Interchange — specifically its deteriorating pavement and bridges. The second was to address the obsolete design of the interchange. The interchange currently has several left-hand exits and other design flaws that increase the risk of crashes and otherwise impede the flow of traffic. The proposed improvements were designed to eliminate those flaws. The third goal was to accommodate traffic volumes through the year 2035. The final goal was to address the high crash rate in the interchange.

After studying various options, the agencies issued a Record of Decision on February 10, 2012, in which they reported that they would implement a specific improvement project known as the "Reduced Impacts Alternative with Adjacent Arterials Component." This will involve rebuilding the entire interchange to address the deteriorating conditions and design flaws. In addition, the capacity of the interchange will be expanded: the highway will be expanded from six to eight lanes in the north-south direction in the vicinity of the interchange, and an additional lane will be added to each of four interchange ramps, making them two lanes wide instead of one. No lanes will be added in the east-west direction, but the shoulders will be widened to eighteen feet and could be converted to an additional lane in the future. The project will also make improvements to certain arterial streets in the vicinity of the interchange. Federal and state funds will be used to complete the project, which is expected to cost $1.71 billion.

The agencies intend to implement the Zoo Interchange project in twenty-six segments over the course of the next five years. Some preparatory work, such as the relocation of utilities, has already commenced. WisDOT has awarded three construction contracts in the amount of $42 million to bidders, and it intends to award more contracts in the near future.

The plaintiffs are two associations that represent minority residents of Milwaukee's inner city: the Milwaukee Inner-City Congregations Allied for Hope and the Black Health Coalition of Wisconsin. Both of these associations have monitored and participated in the decisionmaking process for the Zoo Interchange project, and they now oppose certain aspects of the project. Their principal objection is that the project does not incorporate any public transportation. In particular, they object to the failure to include some form of public transportation, such as rapid bus service, that would help transit-dependent inner city residents access jobs and other services in Waukesha County, which is located at the western end of the project area. The associations also believe that the project will have an adverse impact on air quality in the Milwaukee area, and that it will contribute to suburban sprawl.

The plaintiffs commenced this lawsuit on August 6, 2012 and filed the present motion for preliminary injunction on February 6, 2013.

II. DISCUSSION

The controlling statute at issue here, NEPA, "declares a broad national commitment to protecting and promoting environmental quality." Robertson v. Methow Valley Citizens Council, 490 U.S. 332, 348, 109 S.Ct. 1835, 104 L.Ed.2d 351 (1989). It has been described as a "procedural" or "action-forcing" statute that does not "mandate particular results" but instead requires agencies to study and describe the environmental consequences of their proposed actions. Id. at 348-51, 109 S.Ct. 1835; Vermont Yankee Nuclear Power Corp. v. Nat. Res. Defense Council, 435 U.S. 519, 558, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1978). Thus, under NEPA, if an agency has adequately identified and evaluated the environmental effects of its proposed action, it is permitted to take that action even if the environmental effects will be devastating. Robertson, 490 U.S. at 350, 109 S.Ct. 1835. Put differently, "NEPA merely prohibits uninformed — rather than unwise — agency action." Id. at 351, 109 S.Ct. 1835.

The key procedural element of NEPA is the requirement that the agency prepare an "environmental impact statement" ("EIS") for "major Federal actions significantly affecting the quality of the human environment." 42 U.S.C. § 4332(2)(C). The EIS is "a detailed analysis and study conducted to determine if, or the extent to which, a particular agency action will impact the environment." Highway J Citizens Group v. Mineta, 349 F.3d 938, 953 (7th Cir.2003). Requiring an agency to prepare an EIS serves NEPA's action-forcing purpose in two respects. Robertson, 490 U.S. at 349, 109 S.Ct. 1835. First, "[i]t ensures that the agency, in reaching its decision, will have available, and will carefully consider, detailed information concerning significant environmental impacts." Id. Second, it "guarantees that the relevant information will be made available to the larger audience that may also play a role in both the decisionmaking process and the implementation of that decision." Id. Thus, in the EIS, the agency must "articulate why [it has] settled upon a particular plan and what environmental harms (or benefits) [its] choice entails." Simmons v. U.S. Army Corps of Eng'rs, 120 F.3d 664, 666 (7th Cir.1997). The EIS must show that agency officials have "[thought] through the consequences of — and alternatives to — their contemplated acts," and must ensure that "citizens get a chance to hear and consider the rationales the officials offer." Id.

Judicial review of an agency's compliance with NEPA occurs under the Administrative Procedure Act, which instructs courts to set aside agency action only if it is "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law." 5 U.S.C. § 706(2)(A); Highway J Citizens Group, 349 F.3d at 952. In a NEPA case, the court's role is to ensure that the agency has adequately considered and disclosed the environmental impact of its actions. Baltimore Gas & Elec. Co. v. Nat. Res. Defense Council, 462 U.S. 87, 97-98, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983).

In the present case, the plaintiffs contend that the EIS for the Zoo Interchange project is deficient in two general respects. First, they contend that the EIS is deficient because the agencies failed to consider all reasonable alternatives to the chosen action, as required by NEPA and its associated regulations (which are promulgated by the Council on Environmental Quality ("CEQ")). See 42 U.S.C. § 4332(2)(C)(iii); 40 C.F.R. § 1502.14. Specifically, they contend that the agencies should have considered an alternative to the project that included public transit elements in addition to or in lieu of freeway reconstruction and expansion elements. Second, the plaintiffs contend that the agencies did not adequately analyze several kinds of environmental impacts likely to be caused by the project, including the project's impact on air quality and the extent to which expanding the interchange's capacity will contribute to suburban sprawl and higher traffic volumes. Another impact which plaintiffs believe the agencies failed to adequately address is the project's social and economic impact on communities in the area, particularly in Milwaukee's inner city.

As noted, the matter before the court is the plaintiffs' motion for a preliminary injunction. When confronted with a motion for a preliminary injunction, a district court proceeds in two distinct phases: a threshold phase and, if necessary, a balancing phase. Girl Scouts of Manitou Council, Inc. v. Girl Scouts of the United States of America, Inc., 549 F.3d 1079, 1085-86 (7th Cir.2008). To survive the threshold phase, a party seeking a preliminary injunction must establish three elements: First, that absent a preliminary injunction, the party will suffer irreparable harm in the interim period prior to final resolution of the case. Second, that traditional legal remedies would be inadequate. And third, that the party's claim has some likelihood of succeeding on the merits. Id. at 1086. If the court determines that the movant fails to establish any of these elements, it must deny the injunction. Id. If, however, the court finds that the movant passes the initial threshold, it then proceeds to the balancing phase of the analysis. Id.

In the balancing phase, the court weighs the irreparable harm that the movant would endure without the protection of the preliminary injunction against any irreparable harm the nonmoving party would suffer if the court granted the requested relief. Id. In doing so, the court employs a sliding-scale approach: "[t]he more likely the plaintiff is to win, the less heavily need the balance of harms weigh in his favor; the less likely he is to win, the more need it weigh in his favor." Id. (internal quotation marks omitted). Where appropriate, this balancing process should also encompass any effects that granting or denying the preliminary injunction would have on nonparties (something courts have termed the "public interest"). Id. The court's goal in conducting the balancing phase of the analysis is to minimize the cost of potential error: "the error of denying an injunction to one who will in fact (though no one can know this for sure) go on to win the case on the merits, and the error of granting an injunction to one who will go on to lose." Roland Mach. Co. v. Dresser Indus., Inc., 749 F.2d 380, 388 (7th Cir.1984). The court must "try to avoid the error that is more costly in the circumstances." Id.

Below, I examine whether the plaintiffs have established these elements.

A. Irreparable Harm

At first blush, it may appear that the plaintiffs are not likely to suffer irreparable harm absent an injunction between now and the time that a final decision on the merits of this case is rendered. This is for two interrelated reasons. First, it appears that most of the work on the Zoo Interchange will not take place for some time. The agencies intend to complete some preliminary work on the project this year, but the bulk of the work on the project is not expected to begin until 2014 or 2015. A final decision on the merits of this case will likely be reached before then. Second, unlike in many NEPA cases, the plaintiffs are not concerned about the imminent destruction of environmental resources. They are not, for example, concerned about specific woodlands being cut down in the coming months to make way for new components of the interchange. Rather, they are worried about the long-term effects of the completed project on the environment, such as the effect of the interchange's expanded capacity on air quality and its social and economic effects. Those effects will not begin to materialize until at least five years from now, well after a final decision on the merits of this case had been rendered.

On closer inspection, however, it appears that the plaintiffs will suffer a form of irreparable harm in the absence of a preliminary injunction. WisDOT has made plans to implement the project and is in the process of awarding contracts to bidders and making other commitments that would be difficult to break. See generally Aff. of Roberto Gutierrez, ECF No. 50. This means that the longer it takes to reach a decision on the merits of this case, the more committed WisDOT and the FHWA will be to this particular version of the project. If the plaintiffs win this case, the remedy will involve vacating the agencies' decision to implement the project and requiring them to make a fresh decision after correcting the deficiencies in the EIS. However, in the absence of an injunction, by the time the agencies revise the EIS and make a new decision, they may have made so many commitments to the preset version of the project that they will have no choice but to select that version once again, even if the corrected EIS reveals that a different version of the project would be preferable. If that happens, NEPA's action-forcing purpose will have been defeated. The agency will have made up its mind based on deficient environmental information, and when the agency corrects the deficiency and assembles an accurate EIS, the agency will feel compelled to ignore the accurate information and simply choose the project to which it had already committed itself. For this reason, courts have recognized that NEPA plaintiffs are likely to suffer irreparable harm when an agency is allowed to commit itself to a project before it has fully complied with NEPA. See Sierra Club v. Marsh, 872 F.2d 497, 499-501 (1st Cir.1989); Wisconsin v. Weinberger, 745 F.2d 412, 426-27 (7th Cir.1984); Massachusetts v. Watt, 716 F.2d 946, 951-53 (1st Cir.1983); Scherr v. Volpe, 466 F.2d 1027, 1034 (7th Cir.1972). Accordingly, I conclude that the plaintiffs are likely to suffer irreparable harm in the absence of a preliminary injunction.1

B. Likelihood of Success on the Merits

In this section, I examine the alleged deficiencies in the EIS to determine whether the plaintiffs are likely to succeed on the merits. But before proceeding, I note that it is somewhat awkward to ask whether the plaintiffs are "likely to succeed" in showing that the EIS is deficient in certain respects. The EIS says what it says, and if right now it appears to be deficient, then it will almost certainly appear to be deficient at the end of the case as well — the defendants will not be revising the EIS between now and then, and no other evidence is expected to be introduced concerning its adequacy. Thus, because I am currently in a position to review the adequacy of the EIS, I could likely render a final decision on the merits right now. However, the parties have not asked me to do that, and it is conceivable that in further proceedings they will raise new issues or arguments or point to new evidence. The parties' preliminary-injunction briefs were filed before the administrative record was compiled, and it is conceivable (though unlikely) that there is information in that record that will support new arguments on the issues discussed below. Accordingly, I will continue to ask whether the plaintiffs have shown a "likelihood of success" on the merits of their claims, even though it might seem as if I could simply decide once and for all whether the plaintiffs have succeeded on the merits of those claims.

1. Failure to consider reasonable alternatives

An EIS must discuss alternatives to a proposed action. 42 U.S.C. § 4332(2)(C)(iii). The NEPA regulations specify that an agency preparing an EIS must "[r]igorously explore and objectively evaluate all reasonable alternatives, and for alternatives which were eliminated from detailed study, briefly discuss the reasons for their having been eliminated." 40 C.F.R. § 1502.14(a). The plaintiffs contend that the defendants failed to comply with this requirement because they did not study an alternative that incorporated public transit.

In their discussion of reasonable alternatives in the EIS, the agencies noted that they briefly considered an alternative known as the "Transportation Demand Management Alternative," which would have involved "reduc[ing] the number of automobile trips through increased transit ridership and other strategies." Final EIS at 2-2. The EIS explains that this alternative would have included options such a bus system operating on the freeway and an expansion of existing local bus service in the area, as well as other measures designed to reduce trips during peak hours such as encouraging telecommuting and encouraging flexible work schedules. However, the agencies decided that they would not study this alternative in detail because it would not have satisfied the project's purposes. Id. at 2-4 to 2-5. They noted that it would not have addressed the obsolete design of the interchange or lowered the crash rate, and that it would have left the deteriorating pavement and bridges in place. The agencies also noted that increasing transit ridership would not have eliminated the need to expand the capacity of the interchange. Id.

The plaintiffs do not dispute that the agencies acted reasonably in eliminating the Transportation Demand Management Alternative from detailed study. They seem to concede that increasing transit ridership would not address the interchange's obsolete design, deteriorating infrastructure, and high crash rate, and that therefore at least the reconstruction and redesign elements of the project needed to be implemented. However, the plaintiffs contend that the agencies should have explored the possibility of incorporating transit elements into the project as an alternative to expanding the capacity of the interchange to accommodate future traffic volumes.

The agencies respond to this by arguing that one of their reconstruction alternatives — which involved reconstructing and modernizing the interchange but not adding lanes — implicitly addressed the question of whether transit was a viable substitute for expanding the capacity of the interchange. To understand this argument, I must provide some background. The Zoo Interchange project is itself a component of a larger transportation plan for Southeastern Wisconsin that was developed by the Southeastern Wisconsin Regional Planning Commission ("SEWRPC").2 See Final EIS at 1-5 to 1-11. This plan, which was formulated in 2006, provides overall guidance for developing transportation systems in the region through the year 2035. See SEWRPC, A Regional Transportation System Plan for Southeastern Wisconsin: 2035 (2006). In formulating the plan, SEWRPC studied traffic volumes in the region and also considered the extent to which the region's transportation needs could be addressed through public transit and related infrastructure (such as bicycle and pedestrian infrastructure and other measures designed to reduce automobile trips). Specifically, SEWRPC evaluated a transportation plan that included no street or highway capacity expansion and that relied exclusively on public transit and related infrastructure to meet future needs. Id. at 297-302. SEWRPC concluded that this plan was not viable. It found that failing to add street and highway capacity while relying on public transit and related infrastructure alone would cause traffic congestion to double by 2035 and would result in only a modest reduction in congestion as compared to doing nothing at all. Id. at 300-02. For that reason, SEWRPC developed and recommended a hybrid plan under which public transit and related infrastructure would be improved and expanded while capacity would be added to streets and highways in order to address the congestion that would remain even after full implementation of the transit and related elements. Id. at 302, 358. This plan recommends that the responsible agencies double the amount of public transit service in the region (as measured by the number of miles traveled by transit vehicles) by 2035, id. at 366-67, and that they expand certain highways and arterial streets, including those in the Zoo Interchange, id. at 387-96. SEWRPC explains that the highway and arterial expansion components of the plan are needed "to address the residual congestion which may not be expected to be alleviated by proposed land use, systems management, demand management, bicycle and pedestrian facilities, and public transit measures proposed in the recommended plan." Id. at 387.

In the EIS for the Zoo Interchange project, the agencies identified SEWRPC's recommendation to double the amount of public transit in the region by 2035 and explained that all of their alternatives were analyzed under the assumption that this recommendation would be implemented by the appropriate agencies or units of government in separate projects. See Final EIS at 1-7 to 1-8, 1-36, 2-5 & 2-6. In other words, in their assessment of the amount of additional freeway capacity needed in the Zoo Interchange, the agencies assumed that the task at hand was to identify the amount of additional capacity needed to address the "residual congestion" that would not be addressed by the transit and related elements of the regional plan. Thus, all of the studied alternatives were, in a sense, "hybrid" alternatives that involved both expanding transit and expanding the Zoo Interchange in order to meet future capacity needs. It is just that the agencies expected the transit components of the SEWRPC regional plan to be implemented in separate projects, and possibly by other agencies.3

With this background in mind, it becomes clear that when FHWA and WisDOT studied the alternative under which the Zoo Interchange would be redesigned and reconstructed to address deterioration and design flaws but no lanes would be added, they essentially addressed the plaintiffs' proposed alternative. The agencies' "six lane" alternative4 would have kept the interchange at its present size and relied on the public transit and related elements of the SEWRPC plan to provide the additional capacity needed through 2035. However, the agencies determined that this alternative would not have accommodated expected future traffic volumes at an acceptable level of service, even if all of SEWRPC's public transit recommendations were implemented. See Final EIS at 2-39 to 2-40. Thus, the agencies concluded that additional freeway capacity would be needed, and that public transit was not an acceptable substitute for that capacity.

The plaintiffs have not developed an argument in which they show that the agencies' consideration of the six-lane alternative did not satisfy their duty to consider whether additional interchange capacity could have been provided by transit elements rather than by adding lanes to the interchange. Instead, they contend that the agencies' rejection of this alternative was arbitrary and capricious because it was based on SEWRPC traffic forecasts that were outdated. See Pls.' Reply Br. at 3, ECF No. 79. Plaintiffs point out that, by 2008, there was evidence that higher gas prices had resulted in lower traffic volumes and increased transit ridership throughout the region, and that therefore there was reason to believe that the forecasts SEWRPC issued in 2006 were no longer reliable. However, in the Zoo Interchange EIS, the agencies acknowledge that traffic volumes in the region had been lower than expected and explain that they asked SEWRPC to determine if its 2035 forecast for the Zoo Interchange area remained accurate. See Final EIS at 1-33. The EIS states that SEWRPC reported that it reviewed its forecast in 2010 and determined that it remained valid for long-term transportation planning. Id.

The plaintiffs contend that this statement in the EIS about the continued validity of SEWRPC's traffic forecast is conclusory. And indeed it is, but that does not mean that the reported conclusion is not based on sound analysis. The SEWRPC analysis supporting this conclusion appears in the record, see Thrower Decl. Ex. 8, ECF No. 63-1, and the plaintiffs do not contend that this analysis was flawed in any respect. Thus, the plaintiffs have not shown that the agencies' conclusion, which was based on SEWRPC's analysis, was arbitrary or capricious. Moreover, NEPA does not require that the analysis supporting this conclusion appear in the EIS. Although it is true that an EIS must do more than set forth the agency's conclusions, that does not mean that all of the data and technical analysis on which the agency relied when preparing the EIS must appear in the EIS itself. In the context of SEWRPC's reevaluation of its 2035 traffic forecasts, all that a reader of the EIS needed to know was that the reevaluation took place and that the conclusion was that the forecasts remained valid. A reader who wanted to review the data and analysis supporting that conclusion could then consult the underlying SEWRPC documentation.

The plaintiffs also argue that FHWA and WisDOT were required to "verify" SEWRPC's conclusion that the 2035 traffic forecast remained valid. See Pls.' Reply Br. at 5. However, SEWRPC provided FHWA and WisDOT with the analysis that led it to conclude that its 2035 forecast remained valid, see Thrower Decl., Ex. 8, and, as noted, the plaintiffs have not suggested that this analysis was flawed in any respect. Thus, it is not clear what plaintiffs think FHWA and WisDOT needed to do to "verify" SEWRPC's conclusion. So again, there is no indication that FHWA and WisDOT's conclusion that SEWRPC's 2035 traffic forecast remained valid was arbitrary or capricious.

In light of the above, I conclude that plaintiffs are not likely to succeed on the merits of their claim that the EIS is deficient because the agencies did not consider all reasonable alternatives to the chosen project.

2. Failure to analyze the effects of the project

An EIS must analyze the environmental impact or effect of the proposed action. See 42 U.S.C. § 4332(2)(C)(i). The regulations use the terms "effect" and "impact" synonymously and explain that they "include[] ecological (such as the effects on natural resources and on the components, structures, and functioning of affected ecosystems), aesthetic, historic, cultural, economic, social, or health [effects], whether direct, indirect, or cumulative." 40 C.F.R. § 1508.8.

a. Failure to consider social and economic effects associated with expanding highway capacity while transit capacity declines

The plaintiffs first contend that the agencies' effects analysis was inadequate because they failed to consider the indirect and cumulative effects of the project on communities. Specifically, the plaintiffs contend that the agencies failed to consider the social and economic impacts of expanding highway capacity while transit capacity declines. The plaintiffs point out that state and local governments in the region have not implemented the transit recommendations in SEWRPC's 2035 transportation plan, and that, instead, there has been a decline in transit service in recent years. The plaintiffs contend that expanding highway capacity while transit languishes will have adverse social and economic impacts on communities in the region that the agencies failed to acknowledge.

Although the plaintiffs mention both indirect and cumulative impacts, they focus on cumulative impacts. "Cumulative impact" is:

40 C.F.R. § 1508.7. A proper cumulative impacts analysis will assess the proposed action in light of other activity that has affected or will affect the same environmental resources. The goal is to highlight any environmental degradation that might occur if the minor effects of multiple actions accumulate over time. For example, although a single highway-improvement project might have minimal environmental consequences, combining that project with those that preceded it and others that are anticipated might reveal a more serious overall impact. Placing the project into a broader context that includes these recent and anticipated projects helps prevent "the tyranny of small decisions." Council on Environmental Quality, Considering Cumulative Effects Under the National Environmental Policy Act 1 (January 1997), hereinafter cited as "CEQ Handbook," available at http://ceq.hss.doe.gov/nepa/ccenepa/ccenepa.htm (viewed May 7, 2013).5 In the present case, the issue is whether the EIS adequately discusses the cumulative impact of continuing to expand highway capacity in the Southeastern Wisconsin region while transit capacity declines.

The agencies do not dispute that, despite SEWRPC's recommendation to double the amount of public transit in the region by 2035, in fact the amount of such transit has declined. See Def. Resp. to PFOF ¶ 47, ECF No. 51. Moreover, in June 2010, SEWRPC published a document measuring compliance with the 2035 transportation plan and reported that the recommended transit expansions that were to take place between 2006 and 2008 had not occurred. See SEWRPC, Review, Update, and Reaffirmation of the Year 2035 Regional Transportation Plan at 81-84 (2010). In that document, SEWRPC also observed that there had been "reductions in transit service in the region between 2005 and 2010," and that "transit service was significantly reduced from 2000 to 2010." Id. at 130. SEWRPC also included a graph which appears to show that, by 2010, the amount of transit in the region had declined to below-1995 levels. Id. (Fig. 11). Another SEWRPC document reports that, between 2005 and June 2010, "transit service levels" had declined by approximately 4.5%. See SEWRPC, Memorandum Report No. 196, at 38 (2010).

In contrast, it appears that all of SEWRPC's recommendations concerning the expansion of highway capacity either have been implemented or are on track to be implemented. SEWRPC's 2010 review of compliance with the Year 2035 plan does not mention any failure to comply with the plan's highway-capacity recommendations. Moreover, in addition to the Zoo Interchange project, there are other large freeway projects in the region that either have been completed or are in the process of being completed. See SEWRPC Update at 97-99 (noting that reconstruction of Marquette Interchange had been completed, that Mitchell Interchange project is in process of being completed, and that 30 miles of I-94 are expected to be reconstructed by 2016).

In the EIS, the agencies do not acknowledge this disparity between highway-capacity expansion and transit-capacity expansion or discuss any cumulative environmental impacts that might materialize if this disparity continues. Instead, the agencies explain that they do not have authority to implement SEWRPC's transit recommendations and that they cannot use the funds allocated for the Zoo Interchange project on transit projects. See Final EIS at 3-106 to 3-107. Likewise, in their briefs, the agencies emphasize that NEPA does not allow a court to "second-guess how agencies distribute their resources among competing policy priorities." Fed. Defs.' Br. in Opp. at 19, ECF No. 52; see also Gottlieb Sur-reply at 3-4, ECF No. 90. However, even if the defendants do not have authority to expand transit capacity in the region, and even if NEPA does not require that they prioritize transit over highway projects, NEPA still requires them to examine and identify any cumulative environmental impacts that might result if the responsible decisionmakers continue to implement SEWRPC's highway-capacity recommendations but not its transit-capacity recommendations. Remember that the goal of NEPA is to improve decisionmaking, and one of the ways it seeks to do that is to require agencies to identify any environmental degradation that may occur if the effects of multiple actions accumulate over time. That goal would be undermined if agencies could ignore a cumulative impact to which their actions contribute simply because they do not have authority to fully prevent it. Thus, although the agencies might not be able to implement SEWRPC's transit recommendations, they still have an obligation to acknowledge that the transit components of the Year 2035 plan are not being implemented and to identify the environmental harms that might materialize if transit continues to languish but the agencies continue to expand highway capacity in the region. See 40 C.F.R. § 1508.7 (requiring agency to analyze cumulative effect of agency action when combined with actions beyond the agency's control). Of course, even after the agencies identify those harms, the responsible decisionmakers will remain free to prioritize highway expansion over transit expansion, but at least the decision to do so will have been based on a full consideration of the environmental costs.

Accordingly, I conclude that the EIS is likely deficient because it does not address the cumulative impact of continuing to expand highway capacity in the region while transit capacity declines. Because it seems likely that this conclusion will hold at the end of the case, I will also identify what the agencies need to do to correct this deficiency. They must examine the potential social and economic impact on the transit-dependent of continuing to expand highway capacity in the region while transit capacity declines. If after conducting this examination the agencies determine that their continuing to expand highway capacity while transit capacity declines will have negative effects, the agencies must consider identifying and assessing an alternative to the project that might avoid, minimize, or mitigate those negative effects. See 40 C.F.R. §§ 1500.2(e); 1502.14(f). Such an alternative might include incorporating some form of transit into the project, such as rapid bus service between the City of Milwaukee and Waukesha County. Such bus service might offset the social and economic harm to inner city communities that might result if the continued expansion of highway capacity facilitates the movement of jobs and other services away from those in the inner city who do not have access to automobiles.

Although establishing rapid bus service or another form of transit might be outside the jurisdiction of FHWA and WisDOT, that should not prevent those agencies from proposing it or attempting to persuade the appropriate agencies or units of government to get involved in the project and establish such service. See 40 C.F.R. § 1502.14(c) (EIS must "[i]nclude reasonable alternatives not within the jurisdiction of the lead agency"). If the agencies propose an alternative involving transit and it becomes clear that the will to implement it is lacking, then that may be a reason to eliminate the alternative from detailed study. However, the agencies cannot simply assume that incorporating some form of transit into the project to avoid or minimize adverse social and economic harm is out of the question. Moreover, the agencies should not assume that adding transit entails eliminating other necessary elements of the project, such as elements that are designed to improve safety. Indeed, I see no reason why the agencies could not propose the same action that they have already selected (which involves rebuilding, redesigning, and expanding the vehicular capacity of the interchange) along with a supplemental transit component that is designed to minimize the project's negative effects on the transit-dependent. Cf. Gottlieb Sur-reply at 9 (stating that implementing the chosen alternative would not prevent WisDOT from implementing the plaintiffs' transit recommendations).

b. Failure to consider effects on air quality

The effect of a project on air quality must be assessed in an EIS. See, e.g., 40 C.F.R. § 1508.8(b). The plaintiffs argue that the agencies did not properly analyze the effect of the Zoo Interchange project on air quality because their analysis assumed that SEWRPC's recommendation to double the amount of transit service in the region by 2035 would be implemented, when in fact the agencies had no reasonable basis to conclude that those recommendations would be implemented. The agencies do not dispute that their analysis of the project's impact on air quality depended on the assumption that the amount of transit in the region would double by 2035, or that if this assumption proved to be unfounded their analysis of the effect of the project on air quality would be deficient. See Fed. Defs.' Br. in Opp. at 22-23. However, they argue that their analysis was not deficient because the assumption was supported by a "conformity assessment" prepared by SEWRPC in June 2010.

SEWRPC's purpose in preparing this conformity assessment was to document compliance with various air-quality requirements. The conformity assessment noted that, in SEWRPC's 2035 transportation plan, SEWRPC recommended that the amount of transit service in the region be doubled by the year 2035. See SEWRPC, Memorandum Report No. 196, at 38 (2010). The conformity assessment goes on to note that there has been a lag in implementing the recommended transit-service expansion, and that, because of this lag, the conformity assessment would assume that progress toward the proposed 100% increase in transit service would begin in the year 2012 and be implemented at a pace of 3.1% per year. Id. However, the conformity assessment does not explain how SEWRPC arrived at the conclusion that transit service would increase by 3.1% per year starting in 2012 or the conclusion that the 100% increase in transit service by 2035 was still attainable, and the defendants have not pointed to any other document that supports these conclusions. Thus, on the present record, it appears that SEWRPC simply adopted these conclusions with no supporting analysis. There is nothing in the record indicating that the expectation of doubling the amount of transit in the region by the year 2035 through 3.1% annual increases beginning in 2012 was anything other than a pipe dream.

Because the defendants relied on SEWRPC's unsupported conclusions as to expected transit expansion in their air-quality analysis, it follows that their analysis of the effect of the Zoo Interchange project on air quality was deficient. Accordingly, the plaintiffs are likely to succeed on this issue.

c. Failure to consider growth-inducing effects

The plaintiffs next contend that the agencies did not adequately consider the "growth inducing effects" of the Zoo Interchange project. 40 C.F.R. § 1508.8(b). Agencies are required to consider such effects as part of their analysis of indirect effects, "which are caused by the action and are later in time or farther removed in distance, but are still reasonably foreseeable." Id. The EIS contains an extensive discussion of the indirect effects of the project on growth in the immediate vicinity of the Zoo Interchange, see Final EIS at 3-9 to 3-16, and the plaintiffs do not point to any deficiencies in that discussion. What the plaintiffs appear to be arguing is that the agencies did not adequately discuss the cumulative effect of the Zoo Interchange project and all other freeway projects in Southeastern Wisconsin on growth in the region. That is, they appear to be arguing that the agencies did not examine the extent to which the combined effect of implementing all of SEWRPC's highway-capacity recommendations could be expected to contribute to suburban sprawl and its related environmental effects across Southeastern Wisconsin.

In the EIS, the agencies list various past, present, and reasonably foreseeable future highway projects in Southeastern Wisconsin as projects that, when considered together with the Zoo Interchange project, could have cumulative effects on the environment. See Final EIS at 3-17 to 3-18 & Table 3-3.6 However, in the discussion of cumulative effects that follows this list, the agencies mostly discuss only those effects that are expected to occur within the immediate vicinity of the Zoo Interchange project itself. Id. at 3-16 to 3-27. The agencies do not fully discuss the extent to which the listed highway projects can be expected to generate adverse environmental affects across the entire region.7 Moreover, the agencies do not give any reasons for, on the one hand, identifying all regional highway projects as projects that can be expected to have a cumulative impact on the environment when combined with the Zoo Interchange project and, on the other hand, limiting the scope of their analysis of the cumulative impact to the immediate vicinity of the Zoo Interchange. In the absence of any reasons, this decision seems arbitrary. If the Zoo Interchange and all other regional highway projects are expected to have a combined cumulative impact, wouldn't that impact be on the entire region rather than on just the area around the Zoo Interchange? And it seems that one effect of implementing SEWRPC's highway-expansion recommendations across the region would be to facilitate suburban sprawl and its associated environmental effects, such as the destruction of natural areas. See Swain v. Brinegar, 517 F.2d 766, 777 (7th Cir. 1975) (explaining that highway construction has a profound influence on population growth, high-density urbanization, industrial expansion, and resource exploitation); cf. CEQ Handbook at 2, Table 1-1 (listing "cumulative commercial and residential development and highway construction associated with suburban sprawl" as a potential cumulative effect of FHWA action). Yet the EIS contains no discussion of the potential sprawl-inducing effect of highway construction on the region as a whole.

Indeed, it appears that no one has studied the overall environmental effect on the region of implementing all of the highway-expansion projects in SEWRPC's 2035 plan. Rather, it seems that the agencies which implement specific projects study only the effects that could be expected to occur in the immediate vicinity of the project under consideration. By proceeding in this fashion, the agencies run the risk of overlooking an environmental effect that emerges on the regional level. For example, in the present case, by focusing solely on natural resources in the immediate vicinity of the Zoo Interchange, the agencies may overlook the fact that by facilitating automobile travel to suburban areas (i.e., by facilitating suburban sprawl), the Zoo Interchange project and other highway projects may be contributing to the destruction of natural resources in various other parts of the region. Again, the Zoo Interchange project by itself might not have an appreciable effect on the region as a whole, but when that project is combined with SEWRPC's other recommended highway projects, a regional environmental impact may emerge. See, e.g., Klamath-Siskiyou Wildlands Ctr. v. Bureau of Land Mgmt., 387 F.3d 989, 994 (9th Cir. 2004) ("Sometimes the total impact from a set of actions may be greater than the sum of the parts."); Swain, 517 F.2d at 775 (emphasizing that NEPA is intended to focus concern on the "big picture" created when the effects of "limited" projects combine to form a large mosaic). The agencies must consider this potential cumulative impact on the region rather than simply focus on environmental impacts that will be felt only in the vicinity of their specific project. Cf. CEQ Handbook at 12 (advising agencies that they should expand the geographic scope of their analysis of cumulative impacts beyond the immediate area of the proposed action).

Accordingly, the plaintiffs are likely to succeed on their claim that the EIS is deficient because it does not address the potential growth-inducing cumulative effect of highway expansion on the Southeastern Wisconsin region.

d. Failure to consider "induced travel"

The plaintiffs' remaining claim is that the EIS is deficient because the agencies did not consider "induced travel," which is a term used to describe an observed increase in traffic volume that occurs after capacity is added to a previously congested highway. See Highway J Citizens Group v. United States Dept. of Transp., 656 F.Supp.2d 868, 887 & n. 10 (E.D.Wis.2009). However, the agencies explain that they used traffic models to predict the amount of traffic that could be expected to be generated by each of the studied alternatives, and that those models show that expanding the capacity of the Zoo Interchange will cause traffic volumes to increase by 0.8 percent. See Final EIS at 5-55 to 5-56. The plaintiffs contend that this predicted increase in traffic volumes can be explained by traffic being diverted from arterial streets to the Zoo Interchange, and that the models did not in fact account for the possibility that lessening congestion on the Zoo Interchange might cause motorists to make entirely new trips. However, the plaintiffs have directed me to no evidence showing that the agencies' traffic models did not account for the possibility that expanding the capacity of the Zoo Interchange might cause motorists to make new trips. Although the plaintiffs point out that the EIS mentions traffic being diverted from arterial roads to the Zoo Interchange, that does not imply that the traffic models did not also account for entirely new trips. Accordingly, the plaintiffs have not shown that they are likely to succeed on this issue.

C. Balancing Phase

Having found that the plaintiffs will suffer irreparable harm in the absence of an injunction and that they have a likelihood of success on certain issues, I must proceed to the balancing phase, during which I balance the harm the plaintiffs would suffer in the absence of an injunction against the harm that the agencies would suffer if an injunction were granted. I must also consider where the public interest lies. During this balancing phase, I employ a sliding scale: the more likely the plaintiffs are to win, the less heavily need the balance of harms weigh in their favor; the less likely they are to win, the more need it weigh in their favor.

Before going further, it is important to recognize that, in employing the sliding scale, I must take into consideration the fact that even if the plaintiffs ultimately succeed in showing that the agencies violated NEPA by preparing a deficient EIS, it would not follow that an injunction against the Zoo Interchange project should issue. Rather, if at the end of this case I enter a declaratory judgment stating that the EIS is deficient, I would then have to consider whether to also issue a permanent injunction. See Wisconsin v. Weinberger, 745 F.2d 412, 424-28 (7th Cir.1984). And in considering whether to issue a permanent injunction, I would have to balance the concerns of NEPA against other concerns, such as the harm that the agencies and the public would suffer if the Zoo Interchange project were enjoined pending completion of a NEPA-compliant EIS. See id. Thus, even though my analysis of the merits of this case, above, reveals that the plaintiffs are almost certain to succeed on the merits of their claim that the EIS is deficient in certain respects, it does not follow that, by virtue of the sliding scale, the balance of harms tips in their favor and requires issuance of a preliminary injunction. Instead, whether the balance of harms tips in their favor depends on the propriety of an injunction as a remedy for the NEPA violations I have found.

In assessing the propriety of an injunction as a remedy, I start by attempting to quantify the harm that the plaintiffs would suffer in the absence of an injunction, the harm that the defendants would suffer if an injunction were issued, and the harm to the public interest that would be caused by either course of action. Recall that the plaintiffs' harm is based on the need to stop the "bureaucratic steam roller." See Sierra Club v. Marsh, 872 F.2d 497, 504 (1st Cir.1989). As discussed, the defendants are in the process of committing themselves to the present version of the Zoo Interchange project, even though the selection of that version of the project was made on the basis of a deficient environmental impact statement. In the absence of an injunction, the defendants will commit themselves further to that version, and thus by the time the deficiencies in the impact statement are corrected it may, as a practical matter, be too late to make a truly objective choice among the available alternatives. So, in the absence of an injunction, the plaintiffs (and other members of the public) are likely to lose their right to an objective decision made on the basis of a proper environmental impact statement.

On the other side of the balance we find the harms that the agencies (and members of the public) will suffer if the Zoo Interchange project is delayed. First, delaying the project could cause costs to increase substantially. The defendants estimate that the effects of inflation by itself would add $36 million to $39 million to the project per year of delay. See Gutierrez Aff. ¶¶ 11, 21.8 The defendants also estimate that a delay of three months would add $20 million to the cost of the project because such a delay would make "major contractors and materials" unavailable and otherwise affect the bidding process for labor and materials. Id. ¶¶ 13-19. The defendants further estimate that they would incur between $2 million and $4 million in "maintenance costs" per year of delay. Id. ¶ 20. The defendants also state that any delay might require WisDOT to pay damages to the contractors who have already been awarded contracts. Id. ¶ 22. Finally, the defendants state that any delay will disrupt the traffic-management plan they have developed for the project. Id. ¶¶ 23-24. The defendants have already spent $9 million developing this plan and expect to spend another $10 million on the plan.

Another form of delay-related harm is harm to the members of the public who use the Zoo Interchange. As noted, the Zoo Interchange has a high crash rate, and one of the purposes of the project is to improve safety. WisDOT estimates that for every year that completion of the project is delayed, 400 additional crashes will occur. Gutierrez Aff. ¶ 27. FHWA estimates that failing to prevent these crashes will result in an average of two fatalities per year. Br. in Opp. at 32. In addition to these safety issues, an injunction will deprive those who use the interchange of the benefits of improved traffic flow and thus cause other forms of harm, such as increased travel times. Finally, the defendants note that several institutions are expanding facilities in the vicinity of the interchange in reliance on the completion of the project, including the Milwaukee Regional Medical Center and the University of Wisconsin-Milwaukee. See Gutierrez Aff. ¶¶ 25-26. These institutions would suffer at least some harm if the project were delayed, although the precise nature of this harm is difficult to identify.

Now that the respective harms have been identified, it may seem that the balance tips decidedly against the issuance of an injunction. That is, it may seem clear that the need to make revisions to an environmental impact statement cannot outweigh millions of dollars in costs and the safety risks associated with delaying the project. However, the action-forcing purpose of NEPA is itself extremely valuable, in that an agency's failure to fully consider the environmental effects of a project before committing itself to a course of action can impact an entire region for generations to come. Cf. Amoco Production Co. v. Vill. of Gambell, 480 U.S. 531, 545, 107 S.Ct. 1396, 94 L.Ed.2d 542 (1987) (explaining that environmental injury is often permanent or at least of long duration and that the balance of harms will usually favor the issuance of an injunction to protect the environment). Moreover, in almost any major federal action involving highways, the cost of delay will be significant both in terms of dollars and public safety. If these delay-related costs were always allowed to trump NEPA's action-forcing purposes, then agencies involved in highway projects could routinely fail to comply with NEPA without fear of a court's interfering with their plans. Thus, allowing delay-related costs to automatically tip the balance of harms in the agencies' favor could render NEPA toothless, at least in the context of major highway projects. For this reason, it is not so obvious that the balance of harms tips in the defendants' favor.

Moreover, it is possible that an injunction could be crafted in a way that minimizes any delay-related harm. As indicated, the plaintiffs are not opposed to the Zoo Interchange project in its entirety; they concede that the interchange must be rebuilt and redesigned to address the deteriorating conditions, design flaws, and high crash rate. Rather, the environmental harm that concerns the plaintiffs is associated with expanding the capacity of the interchange. Only ten percent of the overall cost of the Zoo Interchange project is related to capacity expansion, see Aff. of Mark J. Wolfgram ¶ 24, ECF No. 91, and it does not appear that the agencies will be doing any work related to capacity-expansion in the near future. Thus, the agencies might be able to continue their current work without making any further commitments to expanding the capacity of the interchange. Moreover, although the agencies contend that they cannot segregate work involving capacity expansion from other work on the interchange without "redesigning all current construction plans," Gutierrez Aff. ¶ 6, they do not contend that it would be impossible to redesign those plans. Rather, it appears that they could redesign those plans at a cost of approximately $5 million. Id. ¶ 9. And the agencies have not shown that such a redesign would likely delay the ultimate completion of the project. Thus, an injunction against capacity-expansion activities alone might cost as little as $5 million and might not delay the safety-improvement components of the project at all.

The agencies stress that they have already signed construction contracts in the amount of $42 million, and that they have already committed to $75 million in costs in connection with utility work for the project. See Gutierrez Aff. ¶¶ 10, 22. However, the defendants have not shown that an injunction limited to capacity-expansion activities would disrupt either the construction contracts or the utility work. Perhaps the agencies would have to complete this work even if they ultimately decided that they would only rebuild and redesign the interchange.

Another factor to consider is how long an injunction would need to remain in force. The injunction would need to remain in force at least until I rendered a final decision on the adequacy of the EIS, but as noted above, I have virtually made that decision already. Thus, rather than waiting to the end of the case to begin revising the EIS, the agencies could decide to begin that process immediately. Alternatively, I could decide that there is no need for further proceedings and enter a final decision now. The question would then become how long it would take the agencies to correct the deficiencies in the EIS and reconsider their decision to expand the capacity of the interchange. Currently, there is no evidence in the record pertaining to this question, but perhaps the agencies could act expeditiously and render a new decision before they incur any substantial delay-related harm.

Because there are a number of unanswered questions concerning the cost of an injunction limited to capacity-expansion activities, I conclude that the best course of action is to hold an evidentiary hearing. At the hearing, the parties should be prepared to introduce evidence on the balance-of-harm issues identified above and on any other issues that relate to the cost of an injunction limited to capacity-expansion activities.9 In addition, the parties should be prepared to address the question of whether the purpose of NEPA could be substantially vindicated (even if not completely vindicated) without issuing an injunction. I say this because it may be possible for the defendants to substantially redress the NEPA violations I have identified even if they further commit themselves to the present version of the Zoo Interchange project. For example, one of the deficiencies I identified is the failure to consider the social and economic effect of expanding the capacity of the interchange while transit service declines. If after studying this issue the agencies identify a serious negative impact, it might not be too late to do something about it even if the agencies have already committed themselves to expanding the capacity of the interchange, since the agencies could still propose adding a transit component to the project. See Gottlieb Surreply at 9 (stating that WisDOT could implement plaintiffs' transit recommendations as a separate project even if they fully implement the chosen alternative). Moreover, most of the deficiencies I have identified involve the failure to study the cumulative effect of continuing to implement SEWRPC's highway-expansion recommendations, rather than the isolated effect of the Zoo Interchange project itself. Even if the agencies irrevocably commit themselves to the present version of the project, it may be that studying this cumulative effect will vindicate the purpose of NEPA because it will influence decisions with respect to future potential projects, such as whether to further expand the I-94 corridor. See 77 Fed.Reg. 29750-51 (May 18, 2012) (notice of intent to prepare environmental impact statement for further improvements to I-94 corridor in Milwaukee area). If it turns out that the defendants can substantially redress the NEPA violations I have identified in this opinion even if they continue to make commitments to the present version of the project, then perhaps the irreparable harm the plaintiffs will suffer in the absence of an injunction will be minimal. The parties should be prepared to address this possibility at the hearing.

D. Bond

Another matter that the parties should be prepared to address at the hearing is the amount of any bond that the plaintiffs might be required to post in the event I decide to issue a preliminary injunction. Under Federal Rule of Civil Procedure 65(c), a court may issue a preliminary injunction "only if the movant gives security [i.e., posts a bond] in an amount that the court considers proper to pay the costs and damages sustained by any party found to have been wrongfully enjoined." The plaintiffs contend that they should not be required to post any bond at all on the ground that they are nonprofit associations that are litigating on behalf of the environment and the public interest. However, the Seventh Circuit has held that nonprofit entities litigating on behalf of the environment or another public cause are not automatically exempt from Rule 65(c)'s bond requirement. See Habitat Educ. Ctr. v. United States Forest Serv., 607 F.3d 453, 459 (7th Cir.2010).

Still, there may be reasons for exempting the plaintiffs from having to post a bond. First, a line of cases recognizes that a plaintiff may be excused from posting a bond when there is no danger that the opposing party will incur any damages from the injunction. See id. at 458 (collecting cases). Here, of course, there is danger that the defendants will be harmed by an injunction, but that harm would not stem from the preliminary nature of the injunction. The purpose of an injunction bond is to compensate the defendants, in the event they prevail on the merits, for the harm that an injunction entered before the final decision caused them, and so it is required only for a temporary restraining order or a preliminary injunction, not for a permanent injunction. Ty, Inc. v. Publ'ns Int'l Ltd., 292 F.3d 512, 516 (7th Cir.2002). As I have explained, due to the nature of the NEPA violations at issue in this case, the plaintiffs are virtually certain to prevail on the merits (at least at the district-court level), and so any injunction I issue will resemble a permanent injunction more than a preliminary injunction — there is virtually no chance that I will later in this case decide that the injunction was wrongfully issued. Thus, the defendants might not be permitted to recover against the bond for any delay-related harm that they incur, and so there may be no point in making the plaintiffs post such a bond in the first place. Another possibility is that, after the hearing, I will simply render final judgment and issue a permanent injunction since, as discussed, further proceedings on the merits are unlikely to affect the outcome of this case. If I do that, then the plaintiffs would, of course, not be required to post a bond, as bonds are not required for permanent injunctions.

Second, a line of cases recognizes that when a bond that would give the opposing party absolute security against incurring any loss from the injunction would exceed the applicants' ability to pay, the district court may balance the relative cost to the opponent of a smaller bond against the cost to the applicants of having to do without a preliminary injunction that they may need desperately. Habitat Educ. Ctr., 607 F.3d at 458 (collecting cases). In the present case, the plaintiffs may be able to show that their financial resources are so scarce that they could not afford to post a bond in an amount that would be adequate to compensate the defendants for any delay-related harm they may suffer. If that turns out to be the case, then I might allow the plaintiffs to post a smaller bond or, perhaps, no bond at all. Thus, at the hearing, the plaintiffs may wish to present evidence tending to show that they cannot afford to post a bond in the full amount needed to protect the defendants from a wrongfully issued preliminary injunction.

In any event, at the evidentiary hearing, the defendants should be prepared to present whatever evidence they wish as to the amount of the bond plaintiffs should be required to post in the event that their motion for a preliminary injunction is granted and they are not exempted from the bond requirement.

III. CONCLUSION

I find that the plaintiffs are likely to succeed on the merits of their claims and that they are likely to suffer irreparable harm if a preliminary injunction is not issued. However, a hearing must be held before I can determine whether and on what terms an injunction should be granted. A telephonic status conference will be held on May 28, 2013 at 10:30 a.m. for the purpose of scheduling that hearing. The court will initiate the call.