Am. Trucking Ass'ns v. Alviti

Full title: AMERICAN TRUCKING ASSOCIATIONS, INC.; CUMBERLAND FARMS, INC.; M&M…

Court: United States District Court, D. Rhode Island

Date published: Sep 21, 2022

630 F. Supp. 3d 357 (D.R.I. 2022)

Opinion

C.A. No. 18-378 WES

2022-09-21

AMERICAN TRUCKING ASSOCIATIONS, INC.; Cumberland Farms, Inc.; M&M Transport Services, Inc.; New England Motor Freight, Inc., Plaintiffs, v. Peter ALVITI, Jr., in his Official Capacity as Director of The Rhode Island Department of Transportation; Rhode Island Turnpike and Bridge Authority, Defendants.

Gerald C. DeMaria, James A. Ruggieri, Higgins Cavanagh & Cooney, LLP, Providence, RI, Charles Alan Rothfeld, Pro Hac Vice, Colleen M. Campbell, Pro Hac Vice, Evan Tager, Pro Hac Vice, James A. Fussell, Pro Hac Vice, Michelle Webster, Pro Hac Vice, Reginald R. Goeke, Pro Hac Vice, Sean P. McDonnell, Pro Hac Vice, Mayer Brown LLP, Washington, DC, Richard Pianka, Pro Hac Vice, ATA Litigation Center, Arlington, VA, for Plaintiff American Trucking Associations, Inc. Gerald C. DeMaria, James A. Ruggieri, Higgins Cavanagh & Cooney, LLP, Providence, RI, Charles Alan Rothfeld, Pro Hac Vice, Colleen M. Campbell, Pro Hac Vice, Evan Tager, Pro Hac Vice, James A. Fussell, Pro Hac Vice, Michelle Webster, Pro Hac Vice, Reginald R. Goeke, Pro Hac Vice, Sean P. McDonnell, Pro Hac Vice, Mayer Brown LLP, Washington, DC, for Plaintiffs Cumberland Farms, Inc., M&M Transport Services, Inc., New England Motor Freight, Inc. Keith David Hoffmann, Rhode Island Office of the Attorney General, Providence, RI, Michael W. Field, Department of Attorney General, 150 South Main Street — Civil Division, Providence, RI, for Defendants.

WILLIAM E. SMITH, District Judge

Gerald C. DeMaria, James A. Ruggieri, Higgins Cavanagh & Cooney, LLP, Providence, RI, Charles Alan Rothfeld, Pro Hac

Vice, Colleen M. Campbell, Pro Hac Vice, Evan Tager, Pro Hac Vice, James A. Fussell, Pro Hac Vice, Michelle Webster, Pro Hac Vice, Reginald R. Goeke, Pro Hac Vice, Sean P. McDonnell, Pro Hac Vice, Mayer Brown LLP, Washington, DC, Richard Pianka, Pro Hac Vice, ATA Litigation Center, Arlington, VA, for Plaintiff American Trucking Associations, Inc.

Gerald C. DeMaria, James A. Ruggieri, Higgins Cavanagh & Cooney, LLP, Providence, RI, Charles Alan Rothfeld, Pro Hac Vice, Colleen M. Campbell, Pro Hac Vice, Evan Tager, Pro Hac Vice, James A. Fussell, Pro Hac Vice, Michelle Webster, Pro Hac Vice, Reginald R. Goeke, Pro Hac Vice, Sean P. McDonnell, Pro Hac Vice, Mayer Brown LLP, Washington, DC, for Plaintiffs Cumberland Farms, Inc., M&M Transport Services, Inc., New England Motor Freight, Inc.

Keith David Hoffmann, Rhode Island Office of the Attorney General, Providence, RI, Michael W. Field, Department of Attorney General, 150 South Main Street — Civil Division, Providence, RI, for Defendants.

FINDINGS OF FACT AND CONCLUSIONS OF LAW

WILLIAM E. SMITH, District Judge.

In the early 1980s, an article appeared in the Wall Street Journal that described the State of Rhode Island as "little more than a smudge on the fast lane to Cape Cod." Leon Daniel, Poor Little Rhode Island, Maligned but Fighting Back, UPI, Sept. 11, 1983. That fast lane runs from the state's southern border with Connecticut on I-95 North, to Providence about forty-five miles distant, then to I-195 into Massachusetts and on to the Cape. (I-95 continues north into Massachusetts as well.) The governor at the time, J. Joseph Garrahy, was less than amused with the tongue-in-cheek comment. "With friends like that, who needs enemies?" he was quoted as saying about the Rhode Islander who penned the piece. Milly McLean, Rhode Island: Even the Natives Poke Fun at It, UPI, Dec. 10, 1984. The remark struck a nerve: Rhode Islanders have always been a little sensitive that their state's diminutive size garners it less respect than it deserves for its history, culture, beautiful beaches, and diversity.¹ That lack of respect is exacerbated by the fact that the busiest highway in the United States — the I-95 North/South corridor — effectively bisects the state. More than 250,000 cars and trucks speed through the state on I-95 every day, many never stopping to appreciate the charms of the Ocean State.

1 Articles like "Good-Bye, Rhode Island," advocating that Rhode Island should not even be a state, but rather subsumed into one of its neighbors, don't help. See Donald Dale Jackson, GoodBye, Rhode Island, Smithsonian Magazine, Jan. 2000.

While governors and business owners and tourism officials have grumbled for decades about this reality, in 2015 then-governor Gina Raimondo and her compatriots in the General Assembly figured out a way finally to monetize the "fast lane to Cape Cod" — by tolling the bridges along the major interstate and state highway corridors that connect "little Rhody" to its larger neighbors and states beyond.

But tolling highways is tricky and controversial business. The need to raise money to fund construction and repairs of bridges must be balanced against the political reality that many local residents and business owners use those same corridors daily to get around the state. The solution they contrived — dubbed "RhodeWorks," the first and only of its kind in the United States — was to toll only large commercial

trucks (or tractor trailers) at various bridge locations along these major corridors. This plan had the obvious appeal of raising tens of millions of needed dollars from tractor trailers while leaving locals largely unaffected.

Until now. In 2018, the trucking industry represented by its association, the American Trucking Associations, Inc. ("ATA"), and two companies,² M&M Transport Services, Inc. and Cumberland Farms, Inc., filed suit to block the RhodeWorks tolls. This Court denied the requested injunction, and litigation proceeded. A couple trips to the Court of Appeals stalled the process,³ but in May and June of 2022, the Court conducted a twelve-day bench trial featuring fine lawyers and numerous national experts in transportation economics and highway engineering. This decision decides the fate of the RhodeWorks tolling program.

2 A third is no longer an active party.

3 For a full procedural history, those interested can review both this Court's and the First Circuit's previous decisions.

As the Court describes in detail below, the RhodeWorks tolling program violates the Commerce Clause of the United States Constitution. For this reason, the State is permanently enjoined from further tolling under the law.

This decision, consistent with Federal Rule of Civil Procedure 52, sets forth the Court's findings of fact and conclusions of law. But this decision is one largely grounded in legal analysis, turning on data analytics and expert testimony; for that reason, and to make this decision more readable, the Court's initial findings of fact section is sparse relative to its conclusions of law, and is best understood as an overview. Within the legal analysis, the Court finds many additional facts, and these are specifically incorporated as additional findings of fact.

The Court begins its conclusions of law — where it also rules on several lingering evidentiary motions — with three issues antecedent to the merits. First, a surprise trial motion necessitates a lengthy standing analysis, and the Court finds standing does exist for all Plaintiffs. Next, the familiar issue of congressional authorization is addressed, followed by a discussion of the burden of proof. Moving then to the heart of the case, the Court considers the trial evidence under dormant Commerce Clause principles. That analysis starts with fair approximation, holding that the RhodeWorks tolling program does not fairly approximate use of the facilities under any relevant measurement. Next, the Court analyzes the discrimination issue. It concludes that the General Assembly enacted the program with a discriminatory purpose and that the program indeed does discriminate in effect. With that, the Court applies strict scrutiny, and finds the State fails to meet this high bar.

I. Findings of Fact

A. RhodeWorks Background

Rhode Island is home to more than 700 bridges listed on the National Bridge Inventory System. May 25 Tr. 58:18-24, ECF No. 235 (analyzing PX30, and indicating the number is 772); June 1 Tr. 198:12-22, ECF No. 238 (testifying that the number is 779); PX419 at 1 (reflecting the number is 766). For decades, its bridges ranked among the worst in the country, with more than 100 rated as being in poor condition. PX419 at 1. This deterioration was in part due to historic underfunding of the state's infrastructure. Id.; PX30 at 7. Of the 611 state-owned bridges on the National

Bridge Inventory, 142 were structurally deficient as of 2015.⁴ June 1 Tr. 87:1-9.

4 Many bridges were also designated as functionally obsolete. June 1 Tr. 87:10-13, ECF No. 238.

Seeking a cure, the Rhode Island Department of Transportation ("RIDOT") began studying options for investing in its transportation assets, including bridges, as early as 2008. See PX30. By 2015, RIDOT had begun to lay the foundation for the legislation that became known as RhodeWorks. See June 1 Tr. 72:8-15. With RhodeWorks, RIDOT hoped to address funding deficiencies, reorganize financing, and better execute project planning. Id. 74:12-75:5. But its primary goal was to fix Rhode Island's deficient bridges. Id. 74:19-23.

Part of the RhodeWorks proposal was a tolling system designed to fund bridge repairs. From early on, the program aimed at tolling only trucks, not passenger vehicles. See Peter Garino Dep. 15:10-16:3; see June 1 Tr. 72:8-15. This limitation stemmed from the belief that trucks cause more damage to bridges than do other vehicles. Garino Dep. 15:22-16:3. In their research, RIDOT employees found support for this approach in a 1979 Government Accounting Office ("GAO") study concerning truck damage. Id. 39:1-9; May 24 Tr. 96:19-98:22, ECF No. 234.

In the leadup to the first draft of the legislation, RIDOT commissioned several studies to assess potential toll corridors and bridges. See, e.g., PX65. The process looked to several factors, some of which included volume of traffic and damage to bridges. June 1 Tr. 78:15-25. RIDOT was interested in determining where it would be most "cost effective" to toll, id. 89:15-25; Garino Dep. 26:15-25, and also how the toll system would impact in-state and out-of-state registered vehicles, see PX190; PX208 at 6; PX217.

RIDOT set a projected revenue goal from the RhodeWorks program at approximately $45 million per year. June 1 Tr. 82:16-84:17. The Department estimated that it would use RhodeWorks to fund about ten percent of the total cost of bridge repairs needed throughout the state. Id.

In 2015, the General Assembly considered two versions of the RhodeWorks bill. See PX83; PX101. The first would have tolled Class 6+ vehicles.⁵ See PX83 at 9-10.

5 Throughout this decision, various vehicle classifications are discussed. A "Large Commercial Truck," pursuant to Rhode Island General Laws § 42-13.1-3(3), "shall be defined pursuant to the Federal Highway Administration (FHWA) vehicle classification schedule as any vehicle within Class 8 — single trailer, three (3) or four (4) axles, up to and including Class 13 — seven (7) or more axle multi-trailer trucks, as such classifications may be revised from time to time by the FHWA." Class 8+ vehicles are known as "combination trucks," which have two or three separate parts. May 25 Tr. 106:13-19, ECF No. 235. Lower classes — 5-7 — are commonly known as straight trucks (the truck consists of a single unit). Id. 106:6-13. These vehicles may include dump trucks, cement mixers, and garbage trucks. Id. 106:7-11. Class 5 vehicles include tow trucks and may also include single-unit trucks. Id. 106:6-7; see also PX653 at 8-9. Class 4 vehicles are buses. May 25 Tr. 106:5-6. Finally, Classes 1-3 are typically passenger vehicles, including motorcycles, cars, pickup trucks, or ambulances. Id. at 105:24-106:5; see also PX653 at 8-9. The record inconsistently defines straight trucks as Classes 4-7 and Classes 5-7. May 26 Tr. 31:13-17 (defining straight trucks as Classes 4-7), 153:23-25 (defining straight trucks as Classes 5-7). The Court understands it to be category error to include Class 4 vehicles (buses) in this definition, and so it does not. Most record evidence focuses on Classes 6 and 7 as compared to Class 8+, and so the Court does, too; however, where warranted, it expands its analysis to Classes 4-7, and other times to Classes 5-7.

After objections from local businesses, the legislation was modified to toll only Class 8+ vehicles with a limitation on toll charges for multiple trips over the same bridge. See PX101 at 11-12.

On February 11, 2016, the Rhode Island General Assembly enacted "The Rhode Island Bridge Replacement, Reconstruction, and Maintenance Fund Act of 2016" ("RhodeWorks"). See R.I. Gen. Laws § 42-13.1-1. The General Assembly made the following legislative findings, included in the statute:

(1) The state of Rhode Island, through the Rhode Island department of transportation ("the department"), funds the reconstruction, replacement, and maintenance of all bridges in Rhode Island, except the Newport Bridge, the Mount Hope Bridge, the Jamestown-Verrazano Bridge, and the Sakonnet River Bridge.

(2) According to the Federal Highway Administration (FHWA) 2015 National Bridge Inventory (NBI) data, there are seven hundred sixty-four (764) bridges in Rhode Island greater than twenty feet (20') in length. Of these NBI bridges, one hundred seventy-seven (177) bridges, or twenty-three percent (23%), are classified as structurally deficient.

(3) For the past several decades, Rhode Island has depended on three (3) primary sources for funding all transportation infrastructure construction, maintenance, and operations: federal funds, state bond funds, and motor fuel tax revenue. Of these sources, two (2), federal funds and motor fuel tax revenue, are mutable.

(4) The 2008 governor's blue ribbon panel on transportation funding, the 2011 senate special commission on sustainable transportation funding, and the 2013 special legislative commission to study the funding for East Bay bridges determined that there is insufficient revenue available from all existing sources to fund the maintenance and improvement of Rhode Island transportation infrastructure.

(5) In 2011, the general assembly adopted a component of the recommended systemic change to transportation funding by dedicating increased resources from the Rhode Island capital plan fund and creating the Rhode Island highway maintenance account, to be funded by an increase in license and registration fees, beginning in FY2014.

(6) In 2014, the general assembly adopted changes to the Rhode Island highway maintenance account to provide additional state revenue for transportation infrastructure in future years.

(7) Although the state is shifting from long-term borrowing to reliance upon annual revenues to fund transportation infrastructure on a pay-as-you go basis, and although a recurring state source of capital funds has been established, there is still a funding gap between the revenue needed to maintain all bridges in structurally sound and good condition and the annual amounts generated by current dedicated revenue sources.

(8) According to the U.S. General Accounting Office, just one, fully-loaded five-axle (5) tractor trailer has the same impact on the interstate as nine thousand six hundred (9,600) automobiles. The department estimates that tractor trailers cause in excess of seventy percent (70%) of the damage to the state's transportation infrastructure, including Rhode Island bridges, on an annual basis. However, revenue contributions attributable to tractor trailers account for less than twenty percent (20%) of the state's total annual revenues to fund transportation infrastructure.

(9) The United States Congress, consistent with its power to regulate interstate commerce and pursuant to 23 U.S.C. § 129, has authorized states to implement reconstruction or replacement of a toll-free bridge and conversion of the bridge to a toll facility, provided that the state:

(i) Has in effect a law that permits tolling on a bridge prior to commencing any such activity; and

(ii) Otherwise complies with the requirements of 23 U.S.C. § 129.

Id. § 42-13.1-2.

In its final form, RhodeWorks directs RIDOT to set and collect tolls from large commercial trucks "for the privilege of traveling on Rhode Island bridges to provide for replacement, reconstruction, maintenance, and operation of Rhode Island bridges." Id. § 42-13.1-4(a). It expressly prohibits tolling vehicles defined as Federal Highway Administration ("FHWA") Classes 1-7. Id. § 42-13.1-5. The director of RIDOT is authorized to "designate any Rhode Island bridge on the National Highway System as a toll bridge" and to determine the toll amount charged at each location.⁶ Id. §§ 42-13.1-7, 8. The revenue generated from tolling is collected in a special fund, id. § 42-13.1-6, to be "used to pay the costs associated with the operation and maintenance of the toll facility, and the replacement, reconstruction, maintenance, and operation of Rhode Island bridges on the National Highway System or any other use permitted under 23 U.S.C. § 129," id. § 42-13.1-9.

6 The National Highway System includes major roads and encompasses approximately 13.8% of the roads in Rhode Island. May 25 Tr. 131:25-132:6.

Built in to the RhodeWorks program are several limitations on tolling, known as "toll caps." See id. §§ 42-13.1-4(b)-(d). The first tasks RIDOT with establishing a program that "limit[s] the assessment of the tolls upon the same individual large commercial truck using a [radio frequency identification transponder ("RFID")] to once per toll facility, per day in each direction, or an equivalent frequency use program based upon individual large commercial truck use." Id. § 42-13.1-4(b). The second sets the maximum total charge at $20 for large commercial trucks making a "border-to-border through trip on Route 95 Connecticut to Route 95 Massachusetts" using an RFID. Id. § 42-13.1-4(c). The final restriction caps the amount a single large commercial truck using an RFID can be charged daily at $40. Id. § 42-13.1-4(d).

Nearly all vehicles driving over the RhodeWorks bridges do not pay any tolls.⁷ In fact, large commercial vehicles only account for approximately 3% of the total vehicle traffic on those bridges. May 25 Tr. 114:12-13. Passenger vehicles, Classes 1-3, account for more than 90% of the traffic. Id. 147:6-12. Transaction data demonstrate that approximately 19% of tolled vehicles are registered in Rhode Island, compared to about 81% of tolled vehicles registered outside the state.⁸ May 26 Tr. 25:5-22.

7 The parties introduced extensive data and statistics at trial. The Kapsch data record vehicles passing through RhodeWorks toll gantries and sort those vehicles into tollable and un-tollable vehicle classes. The Emovis data capture transaction-level data. That is, those data show billing information for tolled vehicles under the RhodeWorks program. Several pre-implementation sets of data were also introduced, including from studies conducted by two state consultants, CDM Smith and Louis Berger.

8 These data are based on a one-month sample in July 2020. Id. 149:25-150:6.

By all accounts, the RhodeWorks tolling system is unique. See id. 105:19-106:7,

ECF No. 236 (Plaintiffs' expert, Dr. Jonathan Peters, testifying that RhodeWorks is "unprecedented and well outside the norm[]"); June 2 Tr. 76:2-77:13, ECF No. 239 (Defendants' expert, Dr. Kenneth A. Small, testifying that he is not aware of other tolling agencies that toll only Class 8+ trucks); June 1 Tr. 124:16-125:20 (Director Peter Alviti, Jr., Director of the Rhode Island Department of Transportation,⁹ testifying that he "underst[ood] that no one had done truck only [tolling], but... that there were a large number of tolling agencies that were differentiating large commercial vehicles from smaller vehicles and charging them [disproportionately] more than they were charging other vehicles.").

9 In his testimony, Director Alviti recalled being told that one other turnpike in New York tolled only large commercial vehicles. June 1 Tr. 125:6-20. Aside from RhodeWorks, no other record evidence reveals a tolling system targeting only large commercial trucks.

On cue, RIDOT implemented tolls at locations across the state. See DX250. Each reconstructed bridge or group of bridges is linked to a nearby toll gantry. See, e.g., DX41 § 1.1; DX63 § 1.1. Five of the tolls are located on the I-95 corridor, three on I-295 (a beltway around Providence), one on I-195 (which travels to Southeastern Massachusetts), and three on U.S. Route 6 or Rhode Island Route 146. See DX250; DX63 § 1.1. Toll locations have incrementally gone live (though at least one is not yet active).¹⁰ See DX250, DX44, DX67, DX86, DX107. The toll amounts range from $2.25 to $9.50,¹¹ and the five tolls on the I-95 corridor total $17.75. See DX250.

10 Evidence in the record shows that Location 5 is connected to the I-95 Viaduct in Providence. June 6 Tr. 22:19-22, ECF No. 241. This location has not yet gone live. See DX250.

11 RIDOT set the following toll rates for each location: Location 1-$3.25; Location 2-$3.50; Location 3-$6.25; Location 4-$2.25; Location 6-$2.50; Location 7-$6.50; Location 8-$8.50; Location 9-$7.50; Location 10-$9.50; Location 11-$3.50; Location 12-$6.75; Location 13-$5.00. DX250. (This skips Location 5.)

RIDOT created separate accounts for each location. June 10 Tr. 108:4-25, ECF No. 242. The revenue collected from each location is first used to reimburse funds to reconstruct the tolled bridge, id. 110:6-10, and is allocated to reimburse these at a rate that corresponds to a damage calculation done by RIDOT engineers specific to each bridge, id. 109:1-111:23. The costs to rebuild the bridge above and beyond that damage allocation are generally paid according to an 80/20 split between federal and state funds.¹² Id. 109:25-111:3. Once the bridge reconstruction costs have been repaid, RIDOT will allocate a certain amount of toll revenue for yearly maintenance costs and then use the remaining revenue for other permissible projects.¹³ Id. 111:21-114:6.

12 Most of the bridges were paid for by federal funds (80%) and state funds (20%). June 10 Tr. 109:25-110:10, ECF No. 242. One bridge in the program will not be paid for using state funds, only toll revenue and federal funds. Id. 122:7-123:2.

13 RIDOT's Acting Chief Operating Officer testified that RIDOT intended to use the excess funds to repair other bridges but also acknowledged that excess revenue could be used for other projects permitted by federal legislation. Id. 97:19-21, 113:6-19.

B. ISTEA

The RhodeWorks tolling program would not be permissible absent authorization from Congress in the Intermodal Surface Transportation Efficiency Act of 1991 ("ISTEA").¹⁴

14 ISTEA was enacted to "develop a National Intermodal Transportation System that is economically efficient and environmentally sound[.]" Pub. L. 102-240.

This is because states may not toll interstate highways. 23 U.S.C. § 301. An exception to this general prohibition is 23 U.S.C. § 129, which permits "reconstruction or replacement of a toll-free bridge or tunnel and conversion of the bridge or tunnel to a toll facility." Id. § 129(a)(1)(E). A public authority may use revenues generated from such tolling for any purpose listed under § 129(a)(3), including repayment of debt and maintenance and improvement of "the facility." See id. § 129(a)(3). Additionally, "if the public authority certifies annually that the tolled facility is being adequately maintained," it may use excess revenue for "any other purpose for which Federal funds may be obligated by a State under this title." Id. § 129(a)(3)(A)(v).

Prior to implementation of the RhodeWorks tolls, RIDOT and FHWA executed separate Memoranda of Understanding ("MOU") for each proposed toll location. See DX34; DX99. Before the toll gantries went live, FHWA approval was required.¹⁵ See June 10 Tr. 159:23-160:2, ECF No. 242. These MOUs set forth RIDOT's intent to reconstruct specified bridges and toll them in accordance with 23 U.S.C. § 129(a). See DX34; DX99. The MOUs also stated that "RIDOT desires to implement tolls on large commercial vehicles or 'tractor trailers' using an open road tolling structure using one or more gantries to collect tolls" on those bridges. See DX34; DX99. Lastly, RIDOT and FHWA agreed that the "Toll Project[s] m[et] the toll eligibility requirements of 23 U.S.C. 129(a)(1)" and that RIDOT must comply with all other requirements set forth in that statute. See DX34; DX99.

15 Each location was also required to have an environmental assessment. June 10 Tr. 159:23-160:2, ECF No. 242.

II. Conclusions of Law

A. Preliminary Matters

1. Standing

Standing exists to keep courts within their constitutional limits, ensuring they exercise "the Judicial Power of the United States" only over "Cases" and "Controversies." Spokeo, Inc. v. Robins, 578 U.S. 330, 337, 136 S.Ct. 1540, 194 L.Ed.2d 635 (2016) (cleaned up). Thus, before a federal court can reach the merits of any dispute, the party asserting jurisdiction must meet the familiar, three-part "irreducible constitutional minimum of standing:" injury in fact, causation, and redressability. Lujan v. Defs. of Wildlife, 504 U.S. 555, 560, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992).

By its canonical definition, an injury in fact is the "invasion of a legally protected interest which is (a) concrete and particularized and (b) actual or imminent, not conjectural or hypothetical." Id. (internal citations and quotation marks omitted). "Financial loss," even in small amounts, "is a paradigmatic example of an injury in fact." Pincus v. Am. Traffic Sols., Inc., 986 F.3d 1305, 1310 n.7 (11th Cir. 2021) (concluding allegedly wrongful payment of $7.90 gave standing). The amount of pecuniary loss is generally irrelevant because money unlawfully taken is a clear legal injury that is actual, particularized, concrete, and redressable.¹⁶ See Bacchus

16 Some financial losses may be "too trifling of an injury to support constitutional standing." Santos v. TWC Admin. LLC, No. CV 13-04799 MMM, 2014 WL 12558009, at *18 C.D. Cal. Aug. 4, 2014 (quoting Skaff v. Meridien N. Am. Beverly Hills, LLC, 506 F.3d 832, 840 (9th Cir. 2007)). But the financial losses courts have found to be de minimis are typically smaller than even a single toll here. See id. at *19 ("[A]n error of less than one cent appears to be the type of de minimis injury that does not suffice to support Article III standing."); Epstein v. JPMorgan Chase & Co., No. 13 Civ. 4744(KPF), 2014 WL 1133567, *2 (S.D.N.Y. Mar. 21, 2014) (finding loss of interest accruing on $0.67 over seven weeks to be de minimis injury insufficient for standing). The losses alleged and proven at trial are not de minimis.

Imps., Ltd. v. Dias, 468 U.S. 263, 267, 104 S.Ct. 3049, 82 L.Ed.2d 200 (1984) (liability for a tax alone gave standing in dormant Commerce Clause challenge, even if that cost was passed on to customers); All. of Auto. Mfrs. v. Gwadosky, 430 F.3d 30, 37 (1st Cir. 2005) (recognizing "concrete pecuniary injury" sufficient to give standing without analyzing amount).

As a case progresses, all three elements of standing "must be supported in the same way as any other matter on which the plaintiff bears the burden of proof, i.e., with the manner and degree of evidence required at the successive stages of the litigation." Lujan, 504 U.S. at 561, 112 S.Ct. 2130. "[A]t the final stage, those facts [that establish jurisdiction] (if controverted) must be 'supported adequately by the evidence adduced at trial.'" Id. (quoting Gladstone Realtors v. Vill. of Bellwood, 441 U.S. 91, 115 n.31, 99 S.Ct. 1601, 60 L.Ed.2d 66 (1979)). Jurisdiction is so foundational that it is a question "the court is bound to ask and answer for itself, even when not otherwise suggested, and without respect to the relation of the parties to it." Steel Co. v. Citizens for a Better Env't, 523 U.S. 83, 94, 118 S.Ct. 1003, 140 L.Ed.2d 210 (1998) (internal quotation marks omitted).

Because Plaintiffs' claimed loss is financial, the individual Plaintiffs must prove that one of their interstate trucks paid an allegedly unconstitutional RhodeWorks toll. And ATA can show standing on associational grounds if it proves one of its members paid an unconstitutional toll.¹⁷ What evidence the Court may look to in evaluating those showings here, however, is more fraught.

17 For associational standing, ATA must prove that one of its "members would otherwise have standing to sue in their own right, the interests at stake are germane to the organization's purpose, and neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit." Friends of the Earth, Inc. v. Laidlaw Env't Servs. (TOC), Inc., 528 U.S. 167, 181, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000). There is little question the latter two requirements are met.

At the close of Plaintiffs' case, they hit a speed bump. Defendants surprised them by moving for Judgment on Partial Findings under Rule 52(c), ECF No. 222, arguing Plaintiffs failed to prove an injury in fact during their case-in-chief. Defendants pointed out that none of Plaintiffs' witnesses mentioned M&M Transport or Cumberland Farms, nor did Plaintiffs offer evidence that any member of ATA drove a Class 8+ truck through a Rhode Island toll gantry. Because of these failures, Defendants contend that when Plaintiffs rested their case, the Court lost jurisdiction and was powerless to reach the merits or reopen the case.

Plaintiffs' response is twofold: ATA has associational standing, which suffices on its own; and, in any event, evidence of standing existed in the trial record — both in a deposition designation they believed Defendants submitted and in objected-to testimony solicited from a witness in Defendants' case. To top it off, Plaintiffs cry sandbagging because Defendants never questioned standing up to this point.

Adhering to caution, Plaintiffs moved for the Court to reopen the record to receive additional evidence of standing, Mot. to

Reopen Case, ECF No. 221. They argued reopening was both permissible and proper because standing had not been controverted at any time before Defendants' Motion, save for some pro forma language in Defendants' Answer nearly four years ago. Along with their contention that the Court has jurisdiction to reopen, Plaintiffs argue that refusal to reopen would unfairly reward Defendants' trial-by-ambush tactics and waste the copious resources poured into this litigation from all sides. Nothing would prevent Plaintiffs from walking downstairs and refiling their suit the same day it was dismissed.¹⁸ The Court conditionally denied the 52(c) Motion and conditionally reopened the record to allow Plaintiffs to present additional evidence of standing. June 13 Tr. 2:12-15, ECF No. 243.

18 Defendants make the peculiar claim that the Court is powerless to render any judgment or to reopen the record because it lacks jurisdiction and that a dismissal would act as a judgment on the merits worthy of res judicata effect in their favor. This argument confuses cases in which damages are a necessary but unproven element at trial and jurisdictional challenges, like that mounted here. See Am. C.L. Union of Ill. v. City of St. Charles, 794 F.2d 265, 269 (7th Cir. 1986) (distinguishing between injury that "goes to damages (or in an equity case like this, to the right to obtain an injunction)" and "a dismissal on jurisdictional grounds that might allow the plaintiff to start the suit over again").

The Court must answer three distinct questions: (1) May the Court reopen the record?; (2) If yes, should it do so as a matter of discretion?; and (3) Have Plaintiffs proven standing on whatever part of the record is appropriate to consider? The answer to each is yes. Plaintiffs' Motion to Reopen is GRANTED and Defendants' Motion for Judgment on Partial Findings is DENIED.

a. May the Court reopen the record?

Defendants' argument that the Court lost jurisdiction at the close of Plaintiffs' case rests on two venerable principles: a plaintiff bears the burden of proving jurisdiction at trial, see Lujan, 504 U.S. at 561, 112 S.Ct. 2130, and a court may not adjudicate a dispute without jurisdiction, Steel Co., 523 U.S. at 94, 118 S.Ct. 1003. Defendants argue that even though jurisdiction was never seriously questioned, Plaintiffs' evidentiary failure at trial suddenly and permanently deprived the Court of jurisdiction, preventing the Court from reopening the record. This is plainly wrong.

The Court obviously cannot reach the merits or render any judgment without jurisdiction. But its evaluation of its jurisdiction is not confined to the evidence submitted by Plaintiffs in their case-in-chief, especially when standing was not in dispute. Instead, "it is familiar law that a federal court always has jurisdiction to determine its own jurisdiction." United States v. Ruiz, 536 U.S. 622, 628, 122 S.Ct. 2450, 153 L.Ed.2d 586 (2002). This hoary rule has always included the power to take in evidence of standing appropriate to the stage of the proceeding, both before and after trial. See Rivera-Flores v. P.R. Tel. Co., 64 F.3d 742, 748 (1st Cir. 1995) ("[E]ven where the claim is set for jury trial, the court has great latitude to direct limited discovery and to make such factual findings as are necessary to determine its subject matter jurisdiction."); see also Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 718, 127 S.Ct. 2738, 168 L.Ed.2d 508 (2007) (relying, in part, on facts asserted in new affidavit submitted under Supreme Court Rule 32.3 to assure itself of jurisdiction after new circumstances resulted in jurisdiction being questioned); Warth v. Seldin, 422 U.S. 490, 501, 95 S.Ct. 2197, 45 L.Ed.2d 343 (1975) (noting that, at the motion-to-dismiss stage, "it is within the trial court's

power to allow or to require the plaintiff to supply, by amendment to the complaint or by affidavits, further particularized allegations of fact deemed supportive of plaintiff's standing").

The Supreme Court's seminal standing cases place the burden on plaintiffs to prove standing at trial, but only "if controverted." Lujan, 504 U.S. at 561, 112 S.Ct. 2130. For example, in Alabama Legislative Black Caucus v. Alabama, after a bench trial, the district court found that one of the plaintiffs had failed to prove standing because the record did "not clearly identify the districts in which the individual members of the [Conference] reside." 989 F. Supp. 2d 1227, 1292 (M.D. Ala. 2013), vacated and remanded, 575 U.S. 254, 135 S.Ct. 1257, 191 L.Ed.2d 314 (2015). On review, the Supreme Court decided it was reasonable for the Conference to believe standing was not at issue, and thus "elementary principles of procedural fairness required that the District Court, rather than acting sua sponte, give the Conference an opportunity to provide evidence of member residence." Ala. Leg. Black Caucus v. Alabama, 575 U.S. 254, 271, 135 S.Ct. 1257, 191 L.Ed.2d 314 (2015) (emphasis added). Neither the close of the plaintiffs' case-in-chief nor the completion of trial altered the Supreme Court's conclusion that the district court could properly accept such evidence; indeed, the Court went further, directing it to do so. Id.

Similarly, in Gill v. Whitford, — U.S. —, 138 S. Ct. 1916, 201 L.Ed.2d 313 (2018), the Court found the district court lacked jurisdiction because the plaintiffs had failed to produce evidence at trial of sufficiently individualized harm. Id. at 1932-33.¹⁹ Despite this failing, the Court remanded to give the plaintiffs "an opportunity to prove concrete and particularized injuries." Id. at 1934. It is telling that Defendants do not point to a single case in which a court has held it was constitutionally powerless to reopen a case to hear jurisdictional evidence after a plaintiff failed to prove standing at trial (which, of course, is different from declining to do so under ordinary principles guiding a court's discretion to reopen), without having notice that it was at issue.²⁰ Instead, many

19 Gill v. Whitford, — U.S. —, 138 S. Ct. 1916, 201 L.Ed.2d 313 (2018) effectively guts Defendants' attempt to distinguish Alabama Legislative Black Caucus from this case on the grounds that in the latter case, the Court found standing for at least some members and therefore had not completely "lost" jurisdiction. See Defs.' Obj. Pl.'s Mot. Reopen 10-11, ECF No. 225. In Whitford, the Court found none of the plaintiffs had standing but still remanded so they could "attempt to demonstrate standing in accord with the analysis in this opinion." 138 S. Ct. at 1923.

20 Nor do the cases Defendants rely on support their conclusion. Although a court "lacks the power to create jurisdiction by embellishing a deficient allegation of injury," Elend v. Basham, 471 F.3d 1199, 1206 (11th Cir. 2006) (internal quotation marks omitted), this admonition does not apply to the question of whether it has the power to reopen the record. Reopening to receive new jurisdictional evidence when the Court's jurisdiction has suddenly been cast into doubt is different from "embellishing" the record by stretching inferences or assuming harms where they are insufficiently pleaded, "wholly inchoate," or "entirely conjectural." Id. at 1209.
Defendants also cite Steel Co. for the proposition that "[w]ithout jurisdiction the court cannot proceed at all in any cause. Jurisdiction is power to declare the law, and when it ceases to exist, the only function remaining to the court is that of announcing the fact and dismissing the cause." 523 U.S. at 94, 118 S.Ct. 1003 (quoting Ex parte McCardle, 7 Wall. 506, 514, 74 U.S. 506, 19 L.Ed. 264 (1868)). True enough. But applying this directive here requires the Court to beg the question by assuming that jurisdiction has "ceased to exist," rather than merely having been cast into doubt by Defendants' surprise motion. The broader point of Steel Co. was to rebuke a practice brewing in the lower courts of assuming hypothetical jurisdiction when the merits were clear, the jurisdictional question thorny, and the merits favored the party challenging jurisdiction, such that that party would win anyways. Id. at 93-94, 101, 118 S.Ct. 1003. The case says nothing about what to do with a surprise jurisdictional attack or what part of the trial record a court should use in determining jurisdiction (for good reason, Steel Co. was decided on appeal of a motion to dismiss). Rather, it holds that the jurisdictional question must be answered before a court addresses the merits. Id. at 101, 118 S.Ct. 1003.

courts have concluded they have this power and then received new standing evidence. See, e.g., Pub. Emps. for Envt'l Resp. v. Bright, No.: 3:18-CV-13-TAV-HBG, 2021 WL 1220928, at *4 (E.D. Tenn. Mar. 31, 2021) (reopening case to accept affidavits when the defendants challenged Article III standing after trial but standing was not controverted before trial); see also Cranpark, Inc. v. Rogers Grp., Inc., 821 F.3d 723, 733-35 (6th Cir. 2016) (holding district court erred in refusing to consider evidence of standing which was not "adduced at trial," where the plaintiff's mistake was reasonable).

Thus, the Court concludes it has constitutional power to resolve doubts about its jurisdiction and take evidence to determine those facts which would support the exercise of its jurisdiction, even after the close of Plaintiffs' case-in-chief. Thus, having concluded the Court may reopen the record, it turns to whether it should exercise its discretion to do so here.

b. Should the Court exercise its discretion to reopen the record?

"[T]he particular criteria that guide a trial court's decision to reopen are necessarily flexible and case-specific." Rivera-Flores, 64 F.3d at 746. That said, the First Circuit has directed courts to non-exclusively consider "whether: (1) the evidence sought to be introduced is especially important and probative; (2) the moving party's explanation for failing to introduce the evidence earlier is bona fide; and (3) reopening will cause no undue prejudice to the nonmoving party." Id.

Here, there is no question the evidence is important and probative. Without it, M&M Transport and Cumberland Farms would likely be dismissed.

Similarly, Plaintiffs have shown a bona fide reason for their failure to introduce more evidence of standing in their case-in-chief. No dispute over standing was raised in the long history of this case outside of Defendants' now-dusty Answer. It was not mentioned in two appeals to the Court of Appeals, nor in the pretrial memoranda, nor in Defendants' opening statement. See Cranpark, 821 F.3d at 730, 733 (concluding standing was a foregone conclusion, despite boilerplate objection in the defendants' answer). In any event, Plaintiffs had a good-faith reason to believe as recently as two days before trial that Defendants would be submitting a deposition designation containing clear evidence of standing. Defendants did not submit that deposition. The Court need not decide whether this last-minute decision was entirely innocent, a wily litigation tactic, or something more nefarious; whatever the reason, this shift supports a finding that Plaintiffs' reasons for not submitting this deposition were bona fide. And given the obviously concrete and particularized harm associated with paying an allegedly unconstitutional toll, there was good reason to think standing was so patently obvious that it was not a genuine dispute. See Ala. Leg. Black Caucus, 575 U.S. at 270, 135 S.Ct. 1257 (noting "common sense inference" of standing, absent specific challenge, led the plaintiffs to reasonably believe their evidence of standing

was sufficient); Rivera-Flores, 64 F.3d at 747 ("[I]t may amount to an abuse of discretion for a trial court to decline to reopen in circumstances where the movant has demonstrated reasonably genuine surprise.") (internal quotation marks omitted). The Court finds Plaintiffs' surprise to be both genuine and reasonable and their explanation for the limited evidence of standing bona fide.

Finally, there is no undue prejudice to Defendants in reaching the merits of this dispute now. Litigation is not a game, and although some cases no doubt hinge on mistakes of counsel or technical errors, the orientation of the Court is "to secure the just, speedy, and inexpensive determination of every action and proceeding." Fed. R. Civ. P. 1. Refusal to sanction a procedural ambush is not prejudicial, especially if it merely allows Defendants to defend the law on the merits.²¹

21 One might even go so far as to say that the State should want a ruling on the constitutionality of RhodeWorks and ought to be reluctant to use procedural sleights of hand to deflect or delay such a ruling. After all, if the RhodeWorks scheme violates the Commerce Clause, as it does, the State should want to stop the improper tolling and fix the problem.

For all these reasons the Court holds it both can and should accept the evidence offered conditionally in reopening. Plaintiffs' Motion to Reopen, ECF No. 221, is GRANTED.

c. Is Plaintiffs' evidence of standing sufficient?

Having granted Plaintiffs' Motion to Reopen, the Court next considers whether the evidence introduced on reopening is enough to prove standing. It is. Then, in order to produce a thorough record, the Court makes additional holdings on the independent adequacy of two other potential bodies of standing evidence. It considers the propriety of standing evidence adduced in Defendants' case and then decides whether the evidence admitted in Plaintiffs' case-in-chief is itself sufficient to prove standing.

i. Plaintiffs' Evidence on Reopening

Defendants argue that even considering the evidence on reopening, Plaintiffs still failed to prove standing. Some of these arguments border on the frivolous.²² Considering the evidence on reopening, each Plaintiff has easily proven an injury in fact sufficient for standing.

22 See, e.g., Defs.' Post-Trial Br. ("Defs.' Br.") 15 n.14, ECF No. 229 (arguing $100,000 in tolls is a de minimis injury for standing purposes because Cumberland Farms is a large company).

First, the evidence showed that Plaintiff Cumberland Farms maintains a fleet of 120 trucks Classes 8 or above, which service gas stations and convenience stores throughout New England, including Rhode Island. June 13 Tr. 11:15-12:7, 13:5-13:17. Since RhodeWorks went into effect, the company's operating costs have increased by about $100,000 because of the tolls. Id. 16:4-16:20. Given testimony that its headquarters is in Massachusetts, id. 11:21-22, its business is spread across New England, and its trucks regularly use bridges tolled by RhodeWorks, the Court has no trouble finding by a preponderance of the evidence that Cumberland Farms has paid significant RhodeWorks tolls while in interstate commerce. This evidence proves injury in fact. And because Cumberland Farms continues to operate in Rhode Island, the Court finds that future harm is more than speculative, justifying prospective injunctive relief.

Second, the evidence showed that Plaintiff M&M Transport is a trucking company headquartered in Massachusetts

and uses a fleet of roughly 500 Class 8+ trucks to ship goods across the country, from warehouses to retail locations, including into Rhode Island and Connecticut. Id. 27:16-30:8. It has been subject to RhodeWorks tolls each day since the program began, increasing its operating costs. Id. 30:15-31:7. Thus, for M&M Transport, the Court finds by a preponderance of the evidence that the company has suffered pecuniary loss from paying RhodeWorks tolls in interstate commerce and will continue to do so in the future. It therefore has standing to sue based on the evidence admitted on reopening.

Finally, for associational standing, ATA must prove that its "members would otherwise have standing to sue in their own right, the interests at stake are germane to the organization's purpose, and neither the claim asserted nor the relief requested requires the participation of individual members in the lawsuit." Friends of the Earth, Inc. v. Laidlaw Env't Servs. (TOC), Inc., 528 U.S. 167, 181, 120 S.Ct. 693, 145 L.Ed.2d 610 (2000). The constitutionality of RhodeWorks, the first truck-only tolling system in the country, is an interest germane to ATA's purpose, and nothing about this suit requires participation of individual members. And because Cumberland Farms has standing and is a member of ATA, ATA too has standing. See June 13 Tr. 12:18-22; PX759. Separate from this, ATA's general counsel testified to personally witnessing various ATA members' trucks (UPS, Walmart, C.R. England, NFI Industries, Signature Transport, and DJ Cronin) driving on the RhodeWorks toll corridors, including moments before and after they passed through gantries. June 13 Tr. 46:1-47:9. The Court therefore finds it more likely than not these members paid a RhodeWorks toll in interstate commerce and this testimony independently supports ATA's standing.

ii. Evidence from Defendants' Case

On cross-examination, Plaintiffs elicited testimony from Dr. Vijay K. Saraf, a defense expert, showing that some data he used to form his expert opinions also proved that M&M Transport and Cumberland Farms paid RhodeWorks tolls. June 6 Tr. 68:17-72:19, 73:15-75:4, ECF No. 241. Defendants objected to this testimony and moved to strike, arguing that they had carefully avoided soliciting any opinion that relied on a Plaintiff-specific analysis of the underlying data and thus this testimony exceeded the proper scope of cross-examination by asking about the basis for an opinion Dr. Saraf never gave. Id. 64:10-65:21, 69:24-70:1, 72:22-23. The Court took the objection under advisement.

Now, after careful review of the record, the Court agrees with Defendants. Although opinions in Dr. Saraf's direct testimony relied on Emovis and Kapcha data, none of his direct testimony relied on searching or collating these data by E-Z Pass transponders known to belong to Plaintiffs M&M Transport or Cumberland Farms. Questions about a Plaintiff-specific analysis were therefore questions about an analysis that did not form the basis of any opinion given in his direct testimony and thus were beyond the permissible scope of cross-examination. Defendants' objection is sustained, and Dr. Saraf's testimony about M&M Transport and Cumberland Farms, id. 68:17-72:19, 73:15-75:4, is stricken from the trial record.²³ For this reason, the Court need not decide whether a deferred Rule 52(c) motion should be decided based

23 For the same reason, the Court sustains Defendants' objection to the admission of Exhibit DX236. See June 6 Tr. 72:20-23. The clerk is directed to revise the Exhibit and Witness List, ECF No. 227, accordingly.

on the entire trial record or the record at the time the motion is made — an issue the parties debate but that is rendered academic by the exclusion of this portion of Dr. Saraf's testimony.

iii. Evidence in Plaintiffs' Case-in-Chief

If limited only to their case-in-chief, Plaintiffs argue that a deposition designation they believed Defendants submitted shows M&M Transport has standing and that ATA has both associational and organizational standing. But the deposition designation Plaintiffs believed was submitted, was not. Whether this was a foul trick or a benign trial decision, the relevant portions of Shawn Sanford's deposition may not be considered as part of Plaintiffs' case-in-chief, and Plaintiffs cannot rely on that deposition to prove M&M Transport's standing.

The question of ATA's associational standing based on evidence in Plaintiffs' case-in-chief is a close call. As noted earlier, associational standing here depends on whether ATA has proven that one of its members has standing. To this end, Plaintiffs rely on the testimony of John Lynch, an ATA executive, stating that ATA has member companies in all fifty states. May 24 Tr. 11:4-5. Their argument proceeds again from Alabama Legislative Black Caucus, in which the trial court faulted the plaintiffs for failing to identify the legislative districts where individual members resided. 575 U.S. at 269, 135 S.Ct. 1257. The Supreme Court reversed, holding most plaintiffs' proof on standing was adequate because they had members in every county, their purpose was to help those less fortunate, and these facts together supported an inference that it was "highly likely" it had members in each majority-minority district. Id. at 269-270, 135 S.Ct. 1257. Plaintiffs argue that because ATA, an organization dedicated to advancing national trucking interests, has 2,000 members spread across the fifty states (including Rhode Island and every surrounding state) it is extremely likely that its members have paid a RhodeWorks toll.

The Court agrees. Following the logic of Alabama Legislative Black Caucus, the Court finds that it is more likely than not that at least one ATA member paid a RhodeWorks toll in interstate travel. Indeed, this is a near certainty. As the Court did in that case, this Court finds it is so highly likely that an ATA member paid a toll in interstate travel that ATA has standing based solely on the evidence in Plaintiffs' case-in-chief.

But the Court rejects ATA's argument that it has organizational standing to sue. Organizational standing exists on proof of "concrete and demonstrable injury to the organization's activities — with the consequent drain on the organization's resources — [which] constitutes far more than simply a setback to the organization's abstract social interests." Havens Realty Corp. v. Coleman, 455 U.S. 363, 379, 102 S.Ct. 1114, 71 L.Ed.2d 214 (1982). ATA may have an interest in preventing truck-only tolling systems from spreading beyond Rhode Island, but expenditures solely for litigation of the matter at hand and lobbying cannot support organizational standing on a diversion of resources theory. See Equal Means Equal v. Ferriero, 3 F.4th 24, 30 (1st Cir. 2021).

In sum, the Court determines that it both may and should reopen the record to admit additional evidence of Plaintiffs' standing. The evidence produced in the reopened record proves all three Plaintiffs have a real stake in this litigation and would be injured by a constitutional infirmity in RhodeWorks. The Court therefore has jurisdiction. Although the evidence solicited in Defendants' case is excluded, the Court also finds that ATA successfully

demonstrated associational standing based solely on the evidence in its case-in-chief.

2. Congressional Authorization

Another of Defendants' threshold arguments involves congressional authorization; this, the Court has addressed at an earlier stage of this litigation. See Am. Trucking Ass'n, Inc. v. Alviti ("ATA I"), No. 18-378, 2020 WL 4050237, at *2 (D.R.I. July 20, 2020). Defendants' argument comes in two parts: first, Plaintiffs' dormant Commerce Clause challenge fails because Congress authorized the State to take the exact action at issue in this case; and second, Plaintiffs' fair approximation argument fails because of Congress's prior authorization of the tolling regime.

"Congress may 'redefine the distribution of power over interstate commerce' by 'permitting the states to regulate the commerce in a manner which would otherwise not be permissible.'" S.-Cent. Timber Dev., Inc. v. Wunnicke, 467 U.S. 82, 87-88, 104 S.Ct. 2237, 81 L.Ed.2d 71 (1984) (quoting S. Pac. Co. v. Arizona, 325 U.S. 761, 769, 65 S.Ct. 1515, 89 L.Ed. 1915 (1945)) (cleaned up). But congressional authorization must be "expressly stated" or "made unmistakably clear." N.Y. State Dairy Foods, Inc. v. Ne. Dairy Compact Comm., 198 F.3d 1, 9 (1st Cir. 1999) (internal citations and quotation marks omitted). This is because "the democratic process at the state level does not in and of itself function as an effective restraint against protectionist state laws because the burdens that such laws impose will fall on actors who are unrepresented in state legislatures." Ne. Patients Grp. v. United Cannabis Patients & Caregivers of Me., 45 F.4th 542, 552 (1st Cir. 2022). Therefore, this rule "reduces significantly the risk that unrepresented interests will be adversely affected." Id. (quoting S.-Cent. Timber Dev., 467 U.S. at 92, 104 S.Ct. 2237).

Earlier in this case, the Court ruled that it was "unpersuaded by Defendants' broad interpretation of congressional authorization in ISTEA," finding that the statutory scheme does not "evince an unmistakably clear intent by Congress to completely shield state highway tolling activity from Commerce Clause challenges." ATA I, 2020 WL 4050237, at *2 (internal quotation marks omitted). The Court sees no reason to disturb this ruling.²⁴ Through ISTEA, Congress authorized, subject to certain requirements, an activity it once barred: tolling on the interstate highways. See 23 U.S.C. 129(a). But this allowance does not grant states carte blanche to toll in ways that violate the dormant Commerce Clause.

24 Defendants suggest that the toll system here is "supported by ISTEA and by the [MOUs] entered into between the State and the FHWA, Congress's authorized agent." Defs.' Br. 24 n.24. But the MOUs address compliance with ISTEA and have no bearing on whether Congress, through ISTEA, authorized the challenged activity here. See DX34; DX99.

Defendants' second argument fares no better. It is true that ISTEA permits "a public authority to use toll revenues for non-toll road projects." Owner Operator Indep. Drivers Ass'n, Inc. v. Penn. Tpk. Comm'n, 934 F.3d 283, 292 (3d Cir. 2019). At the judgment on the pleadings stage, Plaintiffs conceded, and the Court agreed, this effectively displaced the "excessiveness"²⁵ part of the Evansville/Northwest

25 This asks whether a fee is "excessive in comparison with the governmental benefit conferred." Evansville-Vanderburgh Airport Auth. Dist. v. Delta Airlines, Inc., 405 U.S. 707, 717, 92 S.Ct. 1349, 31 L.Ed.2d 620 (1972).

Airlines ²⁶ analysis. ATA I, 2020 WL 4050237, at *2; see Owner Operator Indep., 934 F.3d at 293 ("ISTEA contemplated that tolls exceeding the amount needed to fund a toll road would be collected and spent on non-toll road projects."). Not satisfied with this congressional gift, Defendants take aim at Plaintiffs' reliance on highway cost allocation studies ("HCASs") to show that RhodeWorks' tolling of only Class 8+ trucks does not fairly approximate usage by all vehicles. Defendants say if allocation is required, "there could never be any excess revenue," so Plaintiffs' argument is nothing more than a "camouflaged attempt to challenge the excessiveness of the tolls."²⁷ Defs.' Post-Trial Br. ("Defs.' Br.") 27, ECF No. 229. As Plaintiffs note, ISTEA "says nothing at all about who may be required to pay the toll or the allocation of the toll burden." See Pls.' Reply Supp. Post-Trial Br. ("Pls.' Reply") 9, ECF No. 231. The fact that excess revenues are used on other projects does not absolve a state of its Commerce Clause duty to toll users fairly.

26 Id. at 716-17, 92 S.Ct. 1349; N.W. Airlines, Inc. v. Cty. of Kent, Mich., 510 U.S. 355, 369, 114 S.Ct. 855, 127 L.Ed.2d 183 (1994). The so-called Evansville test is discussed in detail below.

27 Foreshadowing their fair approximation argument, Defendants throw in that ISTEA precludes an argument that fair approximation must be measured only by the actual costs associated with the individual tolled facilities and that Congress recognized in ISTEA the importance of an interconnected transportation system. These issues are discussed in more detail below in the context of the Court's fair approximation analysis. See infra.

The excessiveness and fair approximation inquiries of the Evansville test serve distinct purposes. Fair approximation zeros in on the fairness of the method by which the fees are charged, not the reasonableness of the amount of money generated from them. Plaintiffs' reliance on HCAS analyses does not implicitly endorse the State's methodology; it merely attempts to show that a toll system should account in some fair way for all users of the tolled facilities.

3. Burden of Proof

And last, the parties disagree about Plaintiffs' burden of proof: Plaintiffs say the burden is the usual preponderance standard, while Defendants claim it's beyond a reasonable doubt in cases concerning constitutionality.

Superficially, at least two federal district court cases support Defendants' position. See Donahue v. City of Bos., 264 F. Supp. 2d 74 (D. Mass. 2003); Devaney v. Kilmartin, 88 F. Supp. 3d 34 (D.R.I. 2015). But delving deeper, both are distinguishable: in stating that the party contesting constitutionality bears the burden of a beyond-a-reasonable-doubt standard, these cases did so in reliance on state — not federal — law. In Donahue, the court cited to a Massachusetts Supreme Judicial Court case for the principle that "the party challenging the statute's constitutionality must demonstrate beyond a reasonable doubt that there are no conceivable grounds which could support its validity." 264 F. Supp. 2d at 82-83 (cleaned up). This principle, which the court considered in analyzing the constitutionality of a Massachusetts statute, traces back to 1917 in Massachusetts common law. See Perkins v. Inhabitants of Town of Westwood, 226 Mass. 268, 115 N.E. 411, 411-12 (1917) ("All rational presumptions are made in favor of the validity of every act of the legislative department of government, and the court will not refuse to enforce it unless its conflict with the Constitution is established beyond reasonable doubt."). And in Devaney, a Magistrate Judge of this Court analyzing alleged

violations of the federal and Rhode Island constitutions "pause[d] to assemble the state law principles applicable to interpreting ordinances." 88 F. Supp. 3d at 44. Citing only Rhode Island law, she said this: "Finally, the party contesting constitutionality... bears the burden of proving beyond a reasonable doubt that the challenged enactment is unconstitutional." Id. at 45 (cleaned up).

These principles mimic the burden that state courts impose on themselves when analyzing constitutionality. See Gomes v. Bristol Mfg. Corp., 95 R.I. 126, 184 A.2d 787, 790 (1962) ("In considering the question of the constitutionality of a statute this court is bound to uphold it unless its unconstitutionality appears beyond a reasonable doubt."); see also State v. Kofines, 33 R.I. 211, 80 A. 432, 435-36 (1911) ("'It is but a decent respect due to the wisdom, the integrity, and the patriotism of the legislative body by which any law is passed,' says Justice Washington, 'to presume in favor of its validity, until its violation of the Constitution is proved beyond all reasonable doubt.'") (quoting Ogden v. Saunders, 25 U.S. 12 Wheat. 213, 270, 6 L.Ed. 606 (1827) (opinion of Washington, J.)). So federal courts follow suit and apply this standard when analyzing violations of those states' constitutions. See R.I. Fishermen's All., Inc. ex rel. Fuka v. Dep't of Env't Mgmt., No. CIV.A.07-230ML, 2008 WL 4467186, at *5 (D.R.I. Oct. 3, 2008), aff'd sub nom. R.I. Fishermen's All., Inc. v. R.I. Dep't of Env't Mgmt., 585 F.3d 42 (1st Cir. 2009) ("Plaintiffs carry the burden of proving beyond a reasonable doubt that the regulations violate the Rhode Island Constitution.") (emphasis added).

But here Plaintiffs allege only a federal constitutional violation. Nothing in dormant Commerce Clause jurisprudence instructs that a heavier burden applies in testing a law against the federal Constitution, even when the relevant state law does impose such a standard. Compare N.H. Motor Transp. Ass'n v. Flynn, 751 F.2d 43, 47 (1st Cir. 1984) (never specifying that the burden is beyond a reasonable doubt) with Pet. of Bos. & Me. Corp., 109 N.H. 324, 251 A.2d 332, 335 (1969) (requiring statutes to be proven unconstitutional beyond all reasonable doubt); compare Cherry Hill Vineyard, LLC v. Baldacci, 505 F.3d 28, 33-34 (1st Cir. 2007) (making no mention of heightened standard of proof) with Op. of the Justs., 162 A.3d 188, 209 (Me. 2017), as revised (Sept. 19, 2017) ("[A] statute enjoys a 'heavy presumption' of constitutionality, it is the burden of the party challenging the statute to establish that it is unconstitutional, and the challenging party must meet that burden beyond a reasonable doubt.") (quoting Op. of the Justs., 850 A.2d 1145, 1149 (Me. 2004)). Proving the point, those principal dormant Commerce Clause cases the Court relies on today — Scheiner, Trailer Marine, Doran, Trailer Bridge, to name a few — make no mention of the beyond a reasonable doubt standard.²⁸

28 Even if the beyond a reasonable doubt standard applied, Plaintiffs would meet it, at the very least with respect to fair approximation and likely with respect to discrimination as well.

Seeing no reason to take a different route here, the Court gets to the merits.

B. Merits

1. Dormant Commerce Clause Principles

Under Article I, Section 8 of the United States Constitution, Congress shall have power to regulate commerce among the several states. The Supreme Court has "long held that this Clause also prohibits state laws that unduly restrict interstate commerce." Tenn. Wine & Spirits

Retailers Ass'n v. Thomas, — U.S. —, 139 S. Ct. 2449, 2459, 204 L.Ed.2d 801 (2019). "'This negative aspect of the Commerce Clause' prevents the States from adopting protectionist measures and thus preserves a national market for goods and services."²⁹ Id. (quoting New Energy Co. of Ind. v. Limbach, 486 U.S. 269, 273, 108 S.Ct. 1803, 100 L.Ed.2d 302 (1988)).

29 Straight out of the gate, Defendants argue Plaintiffs' case should be tossed because there is a good chance that the Supreme Court will re-examine the viability of the dormant Commerce Clause and revoke it. They suggest that, presented with such an opportunity, the Court will likely "reject the dormant Commerce Clause" because it is "not tethered to any constitutional text." Defs.' Br. 18. Although it is true that the dormant Commerce Clause has a "long and complicated history," and the Supreme Court has aptly described its jurisprudence as a "quagmire," the Court has also very recently reinforced its validity. See Tenn. Wine & Spirits Retailers Ass'n v. Thomas, — U.S. —, 139 S. Ct. 2449, 2459, 204 L.Ed.2d 801 (2019); Quill Corp. v. N. Dakota By & Through Heitkamp, 504 U.S. 298, 315, 112 S.Ct. 1904, 119 L.Ed.2d 91 (1992), overruled on other grounds by S. Dakota v. Wayfair, Inc., — U.S. —, 138 S. Ct. 2080, 201 L.Ed.2d 403 (2018); see also Am. Trucking Ass'ns, Inc. v. State, Dep't of Transp., 339 Or. 554, 562, 124 P.3d 1210 (2005). In Tenn. Wine, the Court examined the history of the dormant Commerce Clause, noting that restricting state protectionism "is deeply rooted in our case law" and that without the Clause "we would be left with a constitutional scheme that those who framed and ratified the Constitution would surely find surprising." 139 S. Ct. at 2459. And even if Defendants' crystal ball is accurate and the Supreme Court decides at some future point to reexamine this line of jurisprudence, for now this Court must respect and apply it.

The Supreme Court has held that modern dormant Commerce Clause doctrine concerns two primary principles: (1) "state regulations may not discriminate against interstate commerce" and (2) states "may not impose undue burdens on interstate commerce." S. Dakota v. Wayfair, Inc., — U.S. —, 138 S. Ct. 2080, 2091, 201 L.Ed.2d 403 (2018). "[T]he Commerce Clause was designed to prevent States from engaging in economic discrimination so they would not divide into isolated, separable units." Id. at 2093-94. In cases involving user fees, the fee is reasonable "if it (1) is based on some fair approximation of use of the facilities, (2) is not excessive in relation to the benefits conferred, and (3) does not discriminate against interstate commerce." N.W. Airlines, Inc. v. Cty. of Kent, Mich., 510 U.S. 355, 369, 114 S.Ct. 855, 127 L.Ed.2d 183 (1994) (citing Evansville-Vanderburgh Airport Auth. Dist. v. Delta Airlines, Inc., 405 U.S. 707, 716-17, 92 S.Ct. 1349, 31 L.Ed.2d 620 (1972)). Both this Court and the First Circuit, along with other circuits, have applied the Evansville/Northwest Airlines test in highway tolling cases. Cohen v. R.I. Tpk. & Bridge Auth., 775 F. Supp. 2d 439, 445 (D.R.I. 2011); see Doran v. Mass. Tpk. Auth., 348 F.3d 315, 320-21 (1st Cir. 2003); Selevan v. N.Y. Thruway Auth., 584 F.3d 82, 98 (2d. 2009). (And Evansville itself relied on highway tolling cases, finding them "instructive." 405 U.S. at 715, 92 S.Ct. 1349.)

The test is disjunctive: failing any of the three is sufficient to invalidate a law under the Commerce Clause. As noted above and in the Court's prior rulings, congressional authorization takes excessiveness out of play. See ATA I, 2020 WL 4050237, at *2. So, for Plaintiffs' challenge to succeed, they must show either that RhodeWorks' tolling regime fails to fairly approximate each tolled users' use of the tolled facilities or that it discriminates against interstate commerce. Plaintiffs prove RhodeWorks does both.

2. Does RhodeWorks fairly approximate use?

The fair approximation test asks whether the tolls "reflect a fair, if imperfect,

approximation of the use of the facilities for whose benefit they are imposed" and whether the law has reasonably drawn the line between users it exempts and users it charges for use of a facility. See Evansville, 405 U.S. at 717, 92 S.Ct. 1349; Industria y Distribuction de Alimentos v. Trailer Bridge, 797 F.3d 141, 145 (1st Cir. 2015). Defendants attempt to shape the analysis in three ways: first, they seek to limit what can be considered by excluding evidence of the costs of operating, maintaining, and repairing the tolled facilities; second, if costs are included, they want to dilute the impact by expanding the "denominator" — the benefit side of the equation — to the tolled bridge group's functionally-related facilities; and third, they contend that other sources of RIDOT's revenue must be factored into the equation so that the tolled trucks' share is compared to the financial share paid by other users through various taxes and fees beyond tolls. In line with that attempt, Defendants offer a corresponding idiosyncratic definition of use. None of these arguments are remotely convincing but must be dispatched to set the stage for explaining why the RhodeWorks tolls do not meet the fair approximation test.

a. Parameters of the Test

i. Are the costs of operating, maintaining, and repairing the bridges relevant to assessing use of the facilities?

The parties disagree about whether use in the fair approximation calculation includes costs associated with the facilities. This costs versus use debate is not mere semantics: inclusion of costs changes the inquiry substantially because the focus becomes allocation of all costs of the facility (past and future) among all users, as opposed to a narrower allocation of fees against only those vehicles Defendants say consume the facility over time. Defendants hope to cabin the inquiry to fees approximating a user's use (as they define it) of a facility, excluding that user's fair portion of what it costs the State to operate, maintain, and repair the facility. And they submit that their position must be correct because consideration of costs is reserved for excessiveness, which congressional authorization removed from the calculation. And because HCASs concern such costs, Defendants argue the Court should grant their Motion in Limine seeking to exclude Plaintiffs' HCAS models, ECF No. 166. But the Court sides with Plaintiffs and finds that the costs of operating, maintaining, and repairing the tolled facilities are the proper metric.

The distinction Defendants attempt to draw between costs and use is an illusory one. Incurred costs are often the basis for imposing a fee in the first place. See Evansville, 405 U.S. at 714, 92 S.Ct. 1349 ("[A] charge designed only to make the user of state-provided facilities pay a reasonable fee to help defray the costs of their construction and maintenance may constitutionally be imposed on interstate and domestic users alike."). Indeed, Evansville's "reasonable fee" asks whether the fee structure roughly apportions the costs of building and maintaining the facility among the various users who benefit from using it. See Nw. Airlines, 510 U.S. at 369, 114 S.Ct. 855 ("The Airport's decision to allocate costs according to a formula that accounts for [a] distinction appears to 'reflect a fair, if imperfect, approximation of the use of facilities for whose benefit they are imposed.'" (quoting Evansville, 405 U.S. at 716-17, 92 S.Ct. 1349)) (emphasis added); see Selevan, 584 F.3d at 98 (directing the district court to examine fair approximation by "consider[ing] whether the ... policy at issue reflects rational distinctions among different classes of motorists using the Bridge, so that each user,

on the whole, pay some approximation of his or her fair share of the state's cost for maintaining the Bridge") (internal quotation marks and citation omitted). Evansville's language further proves that both retrospective and prospective costs may be considered. See 405 U.S. at 714, 92 S.Ct. 1349.

Defendants' reliance on Jorling v. United States Department of Energy, 218 F.3d 96 (2d Cir. 2000), for the principle that there is a meaningful difference between fair approximation of use of a facility and fair approximation of the costs of that use only gets them so far. In Jorling,³⁰ the Second Circuit observed that frequently the distinction between use and costs makes no difference to the outcome, but there may be situations in which the two diverge. Id. at 101. To illustrate the point, Jorling gives an extended helpful hypothetical about planes of varying sizes requiring different degrees of runway staff to land. Id. at 101-02. A bigger plane requires more runway staff and thus costs the airport more for each landing than does a smaller plane; a per-landing user fee would not capture this difference in costs but would still likely be permissible as a rough "surrogate for an otherwise complicated and expensive attempt to allocate costs." Id. at 102. Because the aim of the fair approximation analysis is deciding "whether the challenged method for imposing charges fairly apportions the cost of providing a service," "a method for imposing charges based on each payer's approximate use will pass muster as an adequate apportionment of costs." Id. at 103. Allocation of costs is what the test is after, even if simpler measures of use can in some cases be substituted in as a proxy. See Evansville, 405 U.S. at 715, 92 S.Ct. 1349 ("[W]e have sustained numerous tolls based on a variety of measures of actual [road] use, including: horsepower, number and capacity of vehicles, mileage within the State, gross-ton mileage, carrying capacity, and manufacturer's rated capacity and weight of trailers.") (internal citations omitted). Jorling, therefore, does not compel the result Defendants seek. Use is not divorced from costs, and so the Court considers costs of use for allocation.

30 There, the Second Circuit considered the fair approximation question in the context of the intergovernmental tax doctrine, which borrowed this element from dormant Commerce Clause jurisprudence. Jorling v. U.S. Dep't of Energy, 218 F.3d 96, 101 (2d Cir. 2000). The analysis is nevertheless germane.

Which is why Defendants' Motion in Limine to exclude evidence of HCASs is denied. HCASs are tools used by governments (federal and state) to allocate the costs of operating, maintaining, and repairing highways and bridges among different classes of users (e.g., pavement thickness to account for vehicle load or lane additions to account for passenger car volume). They are studies aimed at answering the very question the fair approximation analysis gets after. Although the HCASs introduced by Plaintiffs are neither perfect analogues, nor necessary to assess fair approximation of use, they are helpful and relevant. Defendants' Motion in Limine, ECF No. 166, is DENIED.³¹ ii. Should the analysis focus on the tolled bridges and bridge groups, all Rhode Island bridges, or the entire Rhode Island transportation system?

31 To be clear, this ruling does not in any way hint that tolls not based on a HCAS methodology could not pass the fair approximation test. Neither does the Court suggest that one methodology — HCAS or something else — is superior. Rather, the Court only holds that HCASs are relevant to the question of whether RhodeWorks' tolling method fairly approximates the use of the tolled facilities. As discussed below, the RhodeWorks toll program fails the fair approximation test even without considering the HCAS evidence. This evidence merely reinforces why it does. And, in line with this ruling, what remains of Defendants' Motion to Exclude Evidence and Expert Testimony on the Issue of Fair Approximation on the Basis of Congressional Authorization, ECF No. 168, is now DENIED in full. For the same reason, this resolves the relevance of expert testimony about the HCASs (see, e.g., ECF No. 178 (Daubert Motion seeking to exclude testimony of Dr. Rose Ray)).

With costs in the calculus, the Court next concludes that the relevant costs are those of the tolled bridges only. After all, the Court must evaluate whether the fees imposed at the tolled facilities fairly apportion the costs of those facilities among the different users. Defendants try to dilute the Court's holding that costs are relevant by expanding the measure of costs to include the costs of all other bridges and even Rhode Island's entire highway system (roads and bridges alike) — all of which the State pays to maintain — because those roads and bridges provide benefits to tolled trucks, too. Defendants' reasoning is that even if trucks pay a significant portion of the costs for the tolled bridges, their share shrinks dramatically if the denominator is RIDOT's bridge budget or RIDOT's entire budget, especially given the fees that other (un-tolled) users pay in registration fees, gas taxes, and the like discussed below.

In pursuit of this futile end, Defendants lean on this Court's decision in Cohen and several out-of-circuit cases, including Angus Partners LLC v. Walder, 52 F. Supp. 3d 546 (S.D.N.Y. 2014), to support their contention that all Rhode Island bridges and the whole highway system are functionally related to the tolled bridges. But none of these cases support this theory and the Court easily concludes that the tolled bridges are the proper measurement.

In Cohen, this Court found there was a "spillover effect" between the Mount Hope Bridge and the Newport Bridge, which have a functional relationship in that they make up two of the three access points to Aquidneck Island and closure of one would inevitably cause congestion on the other. 775 F. Supp. 2d at 449-50. And in Angus Partners, the district court analyzed fair approximation and excessiveness together, finding the Metropolitan Transportation Authority ("MTA") was an integrated transportation system from which all users benefitted. 52 F. Supp. 3d at 566. It surmised, "[e]ven if the tolls charged create a surplus, all that is required is a 'functional relationship' between the toll and the facilities used by the toll's payers." Id. at 567 (quoting Bridgeport & Port Jefferson Steamboat Co. v. Bridgeport Port Auth., 567 F.3d 79, 87 (2d Cir. 2009)). Finding a functional relationship existed, the court said the "traffic flow on the [Triborough Bridge and Tunnel Authority] bridges and tunnels would suffer significantly without the MTA's mass transit and commuter railway services," so it was not unfair to allocate "surpluses to support these functionally related facilities." Id. at 568.

Defendants' approach would stretch the functional relationship doctrine beyond all recognition and effectively gut the fair approximation test, so the Court rejects it. A "tolled facility," for purposes of the fair approximation analysis, is the bridge or bridge group for which a specific toll is charged. This conclusion is well supported in the record and the language of ISTEA itself.³² iii. May the Court consider non-toll sources of revenue that fund the RIDOT budget?

32 The tolling authorization in ISTEA clearly applies to reconstruction or replacement of "a toll-free bridge." 23 U.S.C. § 129(a)(1)(E). Ample record evidence demonstrates that RIDOT considered the individual bridge groups as the tolled facilities. One, RIDOT entered a separate MOU with the FHWA for each tolled facility. DX34; DX99. Two, RIDOT separately calculated damage percentages for each bridge, which percentages correlate to the relative proportion of bridge costs to be repaid using toll revenue. June 10 Tr. 119:11-120:2; May 25 Tr. 32:24-33:11; PX314-B. And three, RIDOT set up separate revenue accounts for each tolled facility from which the reconstruction costs of each bridge will be paid. June 10 Tr. 108:21-25, 111:21-112:2. That said, not all evidence relating to the RIDOT budget is irrelevant to the fair approximation analysis. Evidence demonstrating how RIDOT apportions the costs of rebuilding and replacing each bridge is helpful to understand the extent to which tolled and un-tolled users are paying their "fair share," particularly using Plaintiffs' methodology. Thus, the Court DENIES Plaintiffs' Motion to Exclude Evidence Related to the Rhode Island Department of Transportation's Road System Budget, ECF No. 181.

Last, considering non-toll funding sources of the RIDOT budget would be inappropriate. Defendants ask the Court to consider other ways users — both tolled and un-tolled — contribute to the costs of maintaining the tolled bridges. This would include, for example, other funding sources for the RIDOT budget like Rhode Island Capital Plan ("RICAP") funds (excess funds in the state's general revenue rainy day fund), the Rhode Island Highway Maintenance Account ("HMA") funded by DMV revenues (license fees, registration fees, etc.), RIDOT gas tax, and general bond obligations. June 10 Tr. 105:12-108:1. To the extent state funds pay for a part of the cost to reconstruct a RhodeWorks bridge, it is paid for using RICAP or HMA funds. Id. 109:25-110:10. Plaintiffs rely on American Trucking Ass'ns, Inc. v. Scheiner, 483 U.S. 266, 107 S.Ct. 2829, 97 L.Ed.2d 226 (1987), and Fulton Corp. v. Faulkner, 516 U.S. 325, 335, 116 S.Ct. 848, 133 L.Ed.2d 796 (1996), to argue that evidence regarding these contributions is irrelevant. The Court agrees with Plaintiffs.

In Scheiner, the Court refused to weigh the challenged flat tax against other taxes and fees paid by Pennsylvanians. It found that other taxes and fees, such as registration fees, were not paid solely for the use of Pennsylvania's highways and, "even if the relative amounts of the States' registration fees confer a competitive advantage on trucks based in other States, the Commerce Clause does not permit compensatory measures for the disparities that result from each State's choice of tax levels." 483 U.S. at 288, 107 S.Ct. 2829. Thus, the Scheiner Court held that the "flat taxes must stand or fall on their own." Id. at 289, 107 S.Ct. 2829. In Fulton, the Supreme Court likewise rejected this argument, noting the "danger of treating general revenue measures as relevant intrastate burdens for purposes of the compensatory tax doctrine" because it would "allow a state to tax interstate commerce more heavily than in-state commerce anytime the entities involved in interstate commerce happened to use facilities supported by general state tax funds." 516 U.S. at 335, 116 S.Ct. 848 (internal quotation marks omitted).

It is true, as Defendants point out, that Evansville and Northwest Airlines both considered other sources of revenue contributing to the costs of the facilities when assessing fair approximation. See Evansville, 405 U.S. at 718, 92 S.Ct. 1349 ("Furthermore, business users, like shops, restaurants, and private parking concessions, do contribute to airport upkeep through rent, a cost that is passed on in part at least to their patrons."); Nw. Airlines, 510 U.S. at 359-60, 114 S.Ct. 855. But these cases dealt with fees that other users paid directly to the tolled facility (like rent), not funding that must be traced to the facility through general tax and licensing revenues.

Scheiner and Fulton make clear that the tolls must be evaluated on their own.³³

33 Even if the Court considered other contributions tolled and un-tolled users make toward the costs of the bridge, it would make no difference to the outcome. The State initially reconstructed the bridges with 80% federal funding and 20% state funding. June 10 Tr. 109:25-110:16. When toll revenues are collected, they reimburse these state and federal funds up to a certain percentage. Id. 110:18-111:3. The remaining costs for each bridge are repaid according to an 80/20 split, with the State paying 20% of those costs. By way of example, under the RhodeWorks program a $10 million bridge, with an 80% tractor trailer damage assessment, would be paid for using $8 million in tolls, $1.6 million in federal funds, and $400,000 in state funds once the tolls are fully collected. It is true that some un-tolled users of the bridge pay toward the costs of the project through, for example, gas taxes or registration fees. June 6 Tr. 195:10-197:2. But tolled users — both in-state and out-of-state — also contribute to state revenue used to fund the costs of the facilities through programs like the International Fuel Tax Agreement and the International Registration Plan. See May 26 Tr. 94:25-96:10, ECF No. 236. And there are some un-tolled users (e.g., those in lower vehicle classes registered out of state) who do not pay in any other way. Consequently, the other payments made by users not tolled under the RhodeWorks system do not render these exemptions reasonable.

b. What does it mean to use the bridges?

Having set the parameters of the inquiry, the next critical question is, who are the users of the facilities? Defendants contend the only (or the primary) users are Class 8+ trucks because these do most of the damage to bridges, in effect "consuming" the useful life of the bridge over time. If Defendants are correct, then allocating the entire tolling burden to these vehicles is fair. But, if not, the RhodeWorks tolling scheme fails. The Court finds that under every conceivable way of evaluating the question — the common sense understanding of what it means to use a bridge; Defendants' flawed "consumption method"; and Plaintiffs' comparison to HCAS studies — RhodeWorks fails the test.

Defendants have an unusual understanding of what it means to use a bridge. Obviously, in plain language, a "user" of a tolled bridge is a vehicle that crosses the bridge. See Trailer Bridge, 797 F.3d at 145 (defining "user fee" as a "charge assessed for the use of a government facility or service"). Dictionaries provide several definitions of the term "use." The more natural one here is "to carry out a purpose or action by means of [something else.]" Use, Merriam-Webster Dictionary, www.merriam-webster.com/dictionary/use. One uses a bridge to get to the other side of the span.

But another definition is "to expend or consume by putting to use." Id. Defendants' "consumption" approach relies on this latter notion of use, as if bridges are like bicycles or batteries. But bridges are not consumable goods; they are infrastructure that once constructed can remain operational for hundreds of years (the Brooklyn Bridge, for example³⁴) or quickly fall into dangerous despair and disuse amid disagreements about who's responsible for repair and maintenance (like the Fern Hollow Bridge in Pittsburgh³⁵). To use something does not, in this context, mean to exhaust it; Defendants' consumption approach defies common sense. In the Court's view, in the context of tolling bridges, use means to cross the bridge in a vehicle.

34 See Featured Video: Brooklyn Bridge, PBS, https://www.pbs.org/kenburns/brooklyn-bridge.

35 See John Shumway, Work well underway to replace the Fern Hollow Bridge, CBS Pittsburgh, Aug. 1, 2022.

It follows, then, that a tolling scheme can fairly approximate use by charging

each user a fee that corresponds with the number of times it crosses that bridge. In such a case, users in all vehicle classes would share in the toll burden when the user crosses the corresponding bridge or bridge group. Here, the record demonstrates that 97% of the vehicles making use of the tolled bridges — vehicles in Classes 1-7 — are not tolled. May 25 Tr. 114:12-13. Thus, RhodeWorks operates by design in a way that exempts not only most users, but nearly all users. Without looking much deeper, a system that places the entire toll burden on an extreme minority of users is inherently unfair and fails the test.

But a user fee may be permissible under the dormant Commerce Clause even if it exempts certain classes of users or draws distinctions between classes that are not "wholly unreasonable." See Evansville, 405 U.S. at 717-18, 92 S.Ct. 1349; Trailer Bridge, 797 F.3d at 145. Therefore, the fair approximation analysis contemplates the possibility of more than just a correlation based on frequency of an activity. And this makes sense: as the evidence shows, not every use is equal in impact. Indeed, the Supreme Court has sanctioned a variety of measures of use in the highway toll context, including some that vary fees based on weight or vehicle class. See Evansville, 405 U.S. at 715, 92 S.Ct. 1349 (collecting cases).³⁶

36 Defendants point to cases where states have permissibly varied highway tolls according to the weight or rated capacity of different vehicles and thereby, to some degree, used a load-based measure of use. See Evansville, 405 U.S. at 715, 92 S.Ct. 1349 (acknowledging "we have sustained numerous tolls based on a variety of measures of actual use, including: ... a manufacturer's rated capacity and weight of trailers"); see also Commonwealth v. B&W Transp. Inc., 388 Mass. 799, 448 N.E.2d 728, 734-35 (1983). Just because some tolls use classifications based on weight and are fair does not mean that every system that uses weight to apportion tolls does so in a way that satisfies the dormant Commerce Clause.

Defendants argue that the General Assembly reasonably measured use by vehicle weight class by focusing on which vehicles do the most damage to bridges and hence "consume them." They contend that so long as the State's methodology has "a rational basis in that it reasonably and fairly approximates large commercial vehicles' use," the Court must afford that methodology deference and not look to whether a different formula might more precisely reflect use of the facilities. Defs.' Br. 100. This methodology is reasonable, they say, because vehicle weight is an important consideration in regulating highway use and because vehicle weight measures have been used both as a way of allocating costs (in HCAS studies) and to estimate pavement damage.

Defendants rely on expert testimony from Dr. Andrzej S. Nowak, who testified that a "consumption methodology," such as that used in RhodeWorks, starts with the premise that "each passage of a vehicle takes a piece of a part of the lifetime which that bridge has." June 3 Tr. 4:7-17, 5:13-21, ECF No. 240. Heavy vehicles, according to this approach, take away more of the bridge's remaining life with every crossing than do lighter vehicles. Id. 5:21-22. And large commercial trucks "consume" 70-80% of bridge life. Id. 3:3-10. As for Class 6 and 7 trucks, Dr. Nowak testified that while their impact is not "insignificant," the volume of such traffic is small enough that it would not make a difference, id. 16:2-11, and the damage from light vehicles such as passenger cars is "practically negligible" — "so small that it doesn't make a difference in calculation of consumption,"³⁷ id. 14:7-13. The State's so-called consumption-based methodology is flawed from nearly every perspective:

37 It is worth noting that if use equals consumption, and cars do not consume the life of the bridges, two baffling conclusions follow: first, 90% of the vehicles on the bridges are not using them; and second, if only cars were allowed to cross the bridges, the bridges should last forever because they are not being consumed by trucks.

To start, the methodology relies on equivalent single-axle loads ("ESALs") and an inapplicable GAO study from the 1970s. Testimony from both Plaintiffs' and Defendants' experts was that ESAL measurements are not an effective way to measure impact on bridges (as opposed to on pavement generally). June 2 Tr. 149:6-9 (Defendants' expert, Dr. Small, testifying that ESALs are not a measurement he would "use to look at bridges"); May 27 191:17-192:8, ECF No. 237 (Plaintiffs' expert, Dr. William Vavrik, testifying that "ESALs are not part of the bridge design process. ESAL was intended out of the Road Test to be a pavement design tool.... The bridge tool for load is the design truck."). With respect to the 1970s GAO study, Plaintiffs' expert Dr. Vavrik credibly testified that the purpose of it was to analyze overweight and oversized vehicles; further, the study's findings were based on an early road test for pavement design (not for bridges).³⁸ May 31 Tr. 15:8-16:22, ECF No. 244.

38 While courts undoubtedly owe some measure of deference to legislative findings, they may not shirk their "independent constitutional duty to review factual findings where constitutional rights are at stake." Gonzales v. Carhart, 550 U.S. 124, 165, 127 S.Ct. 1610, 167 L.Ed.2d 480 (2007). Indeed, courts have consistently rejected the proposition that legislative factfinding may wholly insulate a statute from constitutional review. See Sable Commc'ns of Cal. v. FCC, 492 U.S. 115, 129, 109 S.Ct. 2829, 106 L.Ed.2d 93 (1989); Lamprecht v. F.C.C., 958 F.2d 382, 392 n.2 (D.C. Cir. 1992). But see Clark v. Paul Gray, Inc., 306 U.S. 583, 594, 59 S.Ct. 744, 83 L.Ed. 1001 (1939) (requiring rational basis review of legislative findings).

Defendants' expert, Dr. Nowak, conducting his own analysis, nonetheless reached essentially the same damage estimate as the GAO study and RIDOT's ESAL-based study. See June 3 Tr. 6:16-7:7. Although his analysis did asses the impact on bridges, his methodology narrowly focused on one type of effect on a bridge (fatigue) without assessing other ways vehicles impact bridges.³⁹ Id. 35:5-24. Furthermore, bridges — and particularly those bridges on interstate highways — are designed to withstand the flow of heavy trucks. See May 31 Tr. 79:18-80:4 ("[I]n heavy bridges, high-traffic volume bridges, because we're designing for heavy trucks and we have all those factors of safety ... many of the elements within the bridge have an infinite fatigue life."); see also June 3 Tr. 50:24-51:5.

39 Plaintiffs attack Dr. Nowak's methodology in many other ways. To name a few, they argue that Dr. Nowak did not conduct a bridge-specific analysis; this type of methodology has never been used to allocate costs across vehicle classes; it excludes certain other traffic-induced loads; and it is based on flawed weight-in-motion ("WiM") data to measure traffic. These critiques all have merit, but the Court need not interrogate them given its holdings.

Even assuming load-based methodologies could appropriately capture how large commercial trucks "use" bridges, and further assuming Defendants are correct that large commercial trucks consume 70-80% of a bridge's life, the State's decision to place the entire toll burden on those users still does not pass the fair approximation test. While it may be that the government need not charge every "user" of a facility to pass the Evansville test, see Evansville, 405 U.S. at 718, 92 S.Ct. 1349 (holding that exempting certain users from paying fees is permissible so long as they are "not

wholly unreasonable" and "reflect rational distinctions"), this record evidence makes clear that the State failed to "reasonably draw[] a line between those it is charging and those it is not," Trailer Bridge, 797 F.3d at 145.

By its own calculations, Defendants' methodology exempts users who consume 30% of the bridge life from paying any charge at all for use of the tolled facilities. The most significant exclusion relates to Classes 6 and 7. Dr. Nowak explained that although a single crossing of a Class 6 or 7 vehicle may well cause a similar damage impact as a Class 8+ vehicle, the volume is too low to matter. See June 3 Tr. 99:12-100:3. The Court does not credit Dr. Nowak's "low volume" explanation for why Class 6 and 7 vehicles should be excluded. First, the State itself calculated that single-unit trucks account for 9-25% of the damage to the RhodeWorks bridges. See PX314-B (also listing the impact from passenger vehicles as between 1-6% of damage). Second, in his testimony, Dr. Nowak did not discuss these percentages specifically, so it is not clear what he was relying on for his statement about these classes having "low volume."⁴⁰ And other credible evidence contradicts it. Using Kapsch traffic data, Dr. Peters testified that in July 2020, Class 4-7 vehicles amounted to about 1.53% of traffic, compared to 3.05% for Class 8+ vehicles. See May 25 Tr. 114:3-22. The pre-enactment CDM Smith traffic data found that single-unit trucks (Classes 5-7) accounted for approximately 4.5% of traffic, compared to 3.7% of traffic for Class 8+ trucks. PX320 at 6. And when calculating the damage share attributable to single-unit trucks, RIDOT's Robert Rocchio used traffic data showing that at certain bridge locations single-unit trucks outnumbered tractor trailers (and did outnumber them overall). See PX328 at 7. Finally, weight-in-motion ("WiM") data show that a significant percentage of vehicles that exceeded the weight limits were in Classes 4-7, May 26 Tr. 16:17-25 (Dr. Peters testifying that "a large percentage of the vehicles that exceeded the Rhode Island weight limits were actually in Classes 4 to 7"); PX653 at 25 (reflecting that 86.4% of vehicles exceeding Rhode Island's weight-per-axle limit are in Classes 5-7), but Dr. Nowak excluded these overloaded vehicles from his analysis, June 3 Tr. 58:1-14.

40 Dr. Nowak's expert report includes a table showing all WiM data over a specific period, with counts specific to each class of vehicle. See Nowak Expert Rep. 16, ECF No. 198-4. These data show 778,638 Class 4 vehicles, 13,688,655 Class 5 vehicles, 1,631,978 Class 6 vehicles, and 136,235 Class 7 vehicles. Id. Class 8+ vehicles amount to 10,560,680 of the traffic count. Id. Dr. Nowak's report was not introduced as a full exhibit at trial, but it is in the case record as part of Plaintiffs' response to one of Defendants' Motions in Limine. The numbers in his chart do not support his statement and, rather, are consistent with Plaintiffs' expert Dr. Peters' findings.

To the extent that consumption or damage-based calculations factor into a tolling program in some way, users having more than a "negligible" impact on the tolled facilities must be assessed a fee for their use, so that all users pay some portion of their fair share. And if the State's own expert, Dr. Nowak, is correct that Class 6 and 7 vehicles cause damage akin to Class 8+ trucks per crossing, their "consumption" of the bridges is far from negligible. It was unreasonable for the State to exempt entire classes of similarly impactful vehicles. See Bridgeport & Port Jefferson Steamboat Co., 567 F.3d at 87-88 (finding that imposing the total cost of a service on one group of users (ferry passengers) but no charge on another group of users (pleasure boats) failed the fair approximation

test).⁴¹

41 If more is needed — and the Court doubts more is — the toll caps (which are discussed in more detail below) illustrate another flaw in the consumption methodology because, when triggered, they break the relationship between the tolls charged and the actual use of the bridge. A large commercial truck passing over the same bridge five times in one day "consumes" the bridge each time it crosses, but it only pays compensation for the first "consumption" in each direction of that bridge. Although fair approximation does not require precision fairness, the RhodeWorks caps run directly counter to the stated premise of the State's methodology.

The analysis could end with Defendants' approach, but it is worth considering Plaintiffs' evidence, too. Plaintiffs' HCASs likewise show that RhodeWorks fails to fairly approximate the costs associated with each user's use of the facilities, they just do so more dramatically.

As told by Dr. Vavrik, HCASs consider the different ways users impact highways and bridges, allocating common costs across all users and incrementally adding responsibility to users who require additional infrastructure or design features. See May 27 Tr. 186:14-188:11, 196:8-198:9, ECF No. 237. Said another way, HCASs are a means to allocate costs to users respective of their use of roads and bridges. Id. 188:1-11. Indeed, the FHWA HCAS states that its study was intended to analyze "the costs occasioned in design, construction, rehabilitation, and maintenance of Federal-aid highways by the use of vehicles of different dimensions, weights, and other specifications, and by the frequency of such vehicles in the traffic stream." PX12 at V-1 (internal quotation marks omitted).

The federal HCAS estimates that tractor trailers account for 29.4% of total highway costs, while single-unit trucks account for 10.7%, and passenger vehicles account for 59.9%. May 31 Tr. 28:17-29:12; PX12 at VI-2. It further estimates that tractor trailers are responsible for 21.4% of all bridge costs. May 31 Tr. 29:7-12; PX12 at V-16. Bridge costs are broken down into minor rehabilitation, major rehabilitation, reconstruction, and new bridges. May 31 Tr. 29:13-20; PX12 at V-16. Under this sub-analysis, tractor trailers are responsible for 32% percent of bridge replacement costs. May 31 Tr. 29:13-20, 49:17-50:14.

Dr. Vavrik also examined nine state HCASs, finding that these states apportioned responsibility to tractor trailers anywhere between just over 20% to just over 40% of highway costs. See May 31 Tr. 27:23-28:3; PX640 at 51. After reviewing these studies, Dr. Vavrik "benchmarked" these data to estimate Rhode Island's situation.⁴² May 31 Tr. 39:20-40:14. He testified that, based on an adjustment for the low volume of Class 8+ trucks, Rhode Island's cost allocation for Class 8+ vehicles would be, at the high end, 11% of costs. Id. 55:11-18; PX640 at 65.

42 Rhode Island has not commissioned a HCAS. May 31 Tr. 165:2-4, ECF No. 244.

Defendants took great pains to discredit the individual HCASs as dissimilar to Rhode Island, but their attacks fall flat. These studies are not meant to be precise analogues to Rhode Island; rather, they are useful for understanding how different government authorities have compiled data and assigned responsibility among users for whom highways and bridges were designed and built. Taking even the highest of these HCAS figures (attributing 40% of bridge costs to large commercial trucks), a tolling system that places 100% of the tolls and 70% percent of the repair costs on those trucks fails to fairly apportion the costs of its bridges among all the users. The HCASs bolster Plaintiffs' position that large commercial trucks should not bear

the entire tolling burden associated with defraying the costs of the tolled bridges and that this failure of apportionment does not pass the fair approximation test.

For these reasons, the Court finds that by any relevant measurement the RhodeWorks toll charges do not fairly approximate the use of the tolled facilities and therefore fail this part of the Evansville test.

3. Does RhodeWorks discriminate against interstate commerce?

The First Circuit has framed the relevant contours as follows: "Discrimination under the Commerce Clause means differential treatment of in-state and out-of-state economic interests that benefits the former and burdens the latter[.]" Fam. Winemakers of Ca. v. Jenkins, 592 F.3d 1, 8 (1st Cir. 2010) (internal quotation marks and citation omitted). "A finding that state legislation constitutes economic protectionism may be made on the basis of either discriminatory purpose or discriminatory effect."⁴³ Gwadosky, 430 F.3d at 35 (quoting Bacchus Imps., 468 U.S. at 270, 104 S.Ct. 3049) (internal quotation marks omitted).

43 RhodeWorks does not discriminate on its face and is silent on its application to in-state versus out-of-state vehicles. See § 42-13.1-1 et seq.

When discriminatory, a statute is "virtually per se" invalid. Or. Waste Sys., Inc. v. Dep't Env. Quality of State of Or., 511 U.S. 93, 99, 114 S.Ct. 1345, 128 L.Ed.2d 13 (1994). In that event, the State must "show that [the statute] advances a legitimate local purpose that cannot be adequately served by reasonable nondiscriminatory alternatives." Id. at 100-101, 114 S.Ct. 1345 (quoting New Energy, 486 U.S. at 278, 108 S.Ct. 1803). This is effectively a strict scrutiny analysis. In contrast, "[a] state statute that regulates evenhandedly and has only incidental effects on interstate commerce engenders a lower level of scrutiny." Gwadosky, 430 F.3d at 35 (applying something resembling rational basis review as applied in Pike v. Bruce Church, Inc., 397 U.S. 137, 90 S.Ct. 844, 25 L.Ed.2d 174 (1970)) (internal quotation marks omitted); see also Wayfair, 138 S. Ct. at 2091.

a. Was RhodeWorks enacted with a discriminatory purpose?

Plaintiffs assert both that RhodeWorks was enacted with a discriminatory purpose and that discriminatory intent alone can render a statute unconstitutional. See Bacchus Imps., 468 U.S. at 270, 104 S.Ct. 3049 ("Examination of the State's purpose in this case is sufficient to demonstrate the State's lack of entitlement to [the Pike balancing test].").

"[W]hen evaluating whether a state statute was motivated by an intent to discriminate against interstate commerce," the Court looks "first to statutory text, context, and legislative history, as well as to whether the statute was closely tailored to achieve the non-discriminatory legislative purpose asserted by the State."⁴⁴ Am. Trucking Ass'n, Inc. v. Alviti ("ATA II"),

44 The Court once again rejects Defendants' argument that discriminatory purpose is irrelevant to the dormant Commerce Clause analysis. Although the First Circuit at an earlier phase of this litigation doubted both the probative value of the evidence of intent and whether a statute could be struck down based solely on discriminatory purpose, it did not dispense with this inquiry. See Am. Trucking Ass'n, Inc. v. Alviti, 14 F.4th 76, 89-90 (1st Cir. 2021). Because the Court finds that Plaintiffs have demonstrated that the RhodeWorks toll system does not represent a fair approximation of the use of the tolled facilities and discriminates in effect, it need not determine whether discriminatory purpose alone is sufficient to find a statute unconstitutional under the dormant Commerce Clause. Id. at 90.

14 F.4th 76, 90 (1st Cir. 2021) (internal quotation marks omitted and cleaned up). "The plain meaning of the statute's words, enlightened by their context and the contemporaneous legislative history, can control the determination of legislative purpose." Edwards v. Aguillard, 482 U.S. 578, 594, 107 S.Ct. 2573, 96 L.Ed.2d 510 (1987). Although the Court may weigh "the specific sequence of events leading to passage of the statute," id. at 595, 107 S.Ct. 2573, it exercises caution in using "evidence of individual lawmakers' motives to establish that the legislature as a whole enacted RhodeWorks with any particular purpose," ATA II, 14 F.4th at 90.

As Plaintiffs see it, their trial evidence proved that RIDOT conceived of, and the General Assembly passed, a revenue-generating system designed to raise funds primarily from out-of-state tractor trailer trucks while protecting local businesses as much as possible. Specifically, they point to the selection of a tolling system (as opposed to other revenue generating options like a taxation regime), pre-enactment and pre-implementation studies focused on the in-state/out-of-state vehicle comparison, comments made by Director Alviti and Senate President Dominick J. Ruggerio to the General Assembly, the decision to exclude Class 7 and smaller trucks, the lack of effort to justify tolling only large trucks, and more.

To start, that the General Assembly chose to enact a statute implementing a toll system to generate revenue to repair its bridge infrastructure, as opposed to a different revenue-generating option (like a gas tax), does not on its own show that lawmakers intended to burden out-of-state interests. Further, the Court agrees with Defendants that two state-commissioned studies — the Blue Ribbon study and the Regional Economic Models, Inc. ("REMI") study — have only marginal relevance. The 2008 Blue Ribbon study examined Rhode Island's funding system across the entire transportation infrastructure and found it was insufficient to support future needs. PX30 at 5. This study proposed several options to generate revenue, one of which was tolling I-95 at the borders. Id. at 24. Although the study may be helpful to understand the need for the RhodeWorks legislation, and was referenced in the legislative findings, it cannot be said that the Blue Ribbon Panel's observation that tolling the I-95 corridor would have little impact on Rhode Islanders reflects on the General Assembly's intent when enacting the RhodeWorks statute nearly seven years later. Id. The 2015 REMI study examined, at RIDOT's behest, the economic impact of the proposed RhodeWorks legislation and noted that "the tolling financing regime shifts a segment of the cost of the RhodeWorks project onto semi-tractor trailer trucks that pass through the state without stopping." PX218 at 4, 6. This study is closer in time to the passage of RhodeWorks and is consistent with the findings of protectionism discussed below. But it is still a step removed from the General Assembly's enactment of the statute.

Next, although Plaintiffs suggest that RIDOT's selection of which bridges to toll evinces a discriminatory intent because RIDOT downgraded certain bridges on the interstate highways in order to toll them, Plaintiffs have not demonstrated that the selected bridges did not meet the 29 U.S.C. § 129 requirements. See PX317 at 5 (indicating all bridges meet the Title 129 definition); June 1 Tr. 136:17-137:5, 140:9-15. As Director Alviti testified, RIDOT also looked to invest in bridges that would become structurally deficient. June 1 Tr. 133:21-135:2. Nothing in the law requires the State to select bridges in the worst condition. And since ISTEA specifically allows for excess toll dollars to be redirected

toward other bridges, it makes economic sense to toll high-volume bridges even if they are not the worst of the worst. This evidence adds little.

And it is true that RIDOT, with the help of its consultants, carefully evaluated the difference in impact on in-state and out-of-state trucks prior to RhodeWorks' enactment. PX190. Clearly it was important to state officials and the General Assembly to understand the extent to which Rhode Islanders would be impacted by the tolling program. Id. (in an email, explaining that "RIDOT leadership will not take 'no it can't be done' as an answer and told Rocchio that if [CDM Smith] couldn't come up with a plan, he needs to find someone who can"). And at the same time, RIDOT's "process" for identifying a justification for placing the burden on large commercial trucks was less than robust. See Garino Dep. 15:12-16:8 (recalling that the concept of a trucks-only tolling provision was divined at the family dinner table); May 24 Tr. 105:13-108:12 (recounting that justifying sources were found through an Internet search). This evidence weighs in favor of an intent to shift the burden to out-of-state trucks, but only marginally so.

Beyond these minor points, however, Plaintiffs presented compelling evidence that the General Assembly sought to protect local businesses with its decision to toll only Class 8+ trucks. In 2015, lawmakers made public the first version of RhodeWorks, which would have tolled all Class 6+ trucks. PX83. The inclusion of Class 6 and 7 trucks was based on a recommendation made by RIDOT consultant CDM Smith. PX80 at 1-2. However, only a couple weeks later, a second draft was proposed with two important changes: (1) Class 6 and 7 vehicles were exempted from tolling and (2) a limitation was added to address repeat per-day visits to a single gantry. PX101. Lawmakers and state officials alike specified at the time that these changes were made to address concerns raised by local businesses and that both changes would reduce the impact of the tolls on local industries.⁴⁵ PX146I (Senate President Ruggerio testifying that "the revised bill takes several important steps to improve upon the original bill to address the concerns of the local trucking industry. It applies to trucks Class 8 and above, these are the larger commercial trucks, not your local landscaper truck or not your local, the usual pick-up trucks."); PX146H (Director Alviti testifying that "[i]t's important for you to know that these changes came as a result of us listening to the various stakeholders and transportation industries in Rhode Island. There were — there was some criticism as to the impact that this legislation, our original legislation, might impose on various industries, we listened to those industries, we've met with them and we've changed our legislation to improve it. But not only improve it, to actually provide economic incentives to those industries to do business in Rhode Island."). Although the Court is mindful that comments made by individual legislators are not always the most probative evidence bearing on legislative purpose, see ATA II, 14 F.4th at 90, these statements are not, as Defendants claim, wholly irrelevant because they were made by Senate President Ruggerio and Director Alviti — two key officials behind the RhodeWorks

45 Defendants suggest that these statements are irrelevant in part because vehicles in Classes 4-7 contribute to bridge costs in other ways, including by paying certain fees and gas taxes in Rhode Island. But as addressed above in the context of fair approximation, these vehicles' contributions by way of general fees and taxes have little relevance. The question here concerns whether these provisions were enacted with the intent of local protectionism.

legislation.⁴⁶

46 In Alliance of Auto Manufacturers v. Gwadosky, the First Circuit rejected an attempt to show that the legislative process demonstrated intent. 430 F.3d 30, 36 (1st Cir. 2005). There, the legislature deleted the at-issue language, but ultimately restored it after an "intense lobbying campaign." Id. The court held that the legislative process evidence was "indeterminate." Id. at 38-39. It continued, "[a]s a general rule, statutory interpretation cannot safely be made to rest upon inferences drawn from intermediate legislative maneuvers. In this instance, such reliance would be especially problematic because there are countless reasons why the state legislature may have altered its position." Id. at 39 (internal citation omitted). The Court is cautious that it must not "lose[] sight of the forest while searching for trees." Id. at 37. But, as the First Circuit stated, "context is a critically important interpretive tool," and this legislative process helps put RhodeWorks in context when discerning its purpose from the statute "as a whole." Id.

But the most important evidence of the intent of the General Assembly to protect local businesses and residents at the expense of interstate commerce is found in both the structure of law itself and in the data. The trial evidence showed that lower-classed trucks are more likely to be Rhode Island-plated than Class 8+ trucks. See May 26 Tr. 13:8-25 (comparing Rhode Island share of Class 4-7 vehicles at "42 percent, 38 percent" with Rhode Island share of Class 8+ vehicles at "generally an 18 to 20 percent range"). This means that approximately 80% of the tolled vehicles are from out of state. And the less than 20% of tolled Class 8+ trucks from Rhode Island disproportionally receive the benefit of the toll caps, which the Court finds below discriminate in effect. These discriminatory effects "strengthen[] the inference that the statute was discriminatory by design." Fam. Winemakers, 592 F.3d at 14. Furthermore, both the exclusion of lower-classed trucks and the ceiling imposed by the toll caps make the statute less tailored and effective in fulfilling its express purpose of funding the repair of its failing bridges.

On the whole, the record evidence demonstrates a common theme of placing the lion's share of the tolling burden on out-of-state large commercial trucks and protecting Rhode Island commercial interests as much as possible through the exemption of Class 4-7 vehicles and the utilization of toll caps. This intent to mitigate the effect of tolling on local businesses was clearly protectionist.⁴⁷

47 Plaintiffs also rely on PX128 — and particularly a comment made on the draft Rhode Island Office of Management and Budget document — to show discriminatory purpose. Pls.' Post-Trial Br. ("Pls.' Br.") 52, ECF No. 228. The Court admitted this document over Defendants' objection for the limited purpose of demonstrating a level of understanding regarding the shifting burden of tolls between in-state and out-of-state users. May 26 Tr. 40:20-43:25. As for the comments, the Court indicated they were "beyond that limited purpose" and had "limited relevance and limited usefulness because it's not even known who is making" them. Id. 43:18-25. Given this ruling, the Court does not rely on the comment now (nor is it necessary to the Court's conclusion).

b. Does RhodeWorks discriminate against interstate commerce in effect?

Typically, "[a] state law is discriminatory in effect when, in practice, it affects similarly situated entities in a market by imposing disproportionate burdens on out-of-state interests and conferring advantages upon in-state interests." Fam. Winemakers, 592 F.3d at 10. In other words, a statute discriminates within the context of the dormant Commerce Clause when it disfavors competitors based on their location outside a state's borders. See id. "[T]he mere fact that a statutory regime has a discriminatory potential is not

enough to trigger strict scrutiny under the dormant [C]ommerce [C]lause." Cherry Hill Vineyard, 505 F.3d at 37 (cleaned up). Plaintiffs must prove an actual adverse impact.⁴⁸ Id. at 36.

48 In Cherry Hill Vineyard, LLC v. Baldacci, the First Circuit considered what kind of evidentiary showing was needed for discriminatory effect. 505 F.3d 28, 36 (1st Cir. 2007). It concluded that when a statute is facially neutral and wholesome in purpose, the showing of discriminatory effect "must be substantial." Id. at 36. Later, the First Circuit noted but did not decide "whether a lesser showing might suffice when a law is allegedly discriminatory in both effect and purpose." Fam. Winemakers, 592 F.3d at 11 n.11. Either way, Plaintiffs meet their burden.

Plaintiffs argue three strands of evidence demonstrate the discriminatory effect of the RhodeWorks tolling program: (1) the differential impact of the toll caps; (2) the locations of the tolls; and (3) the classes of vehicles tolled versus those not tolled.

i. The Toll Caps

Plaintiffs' contention is that the toll caps discriminate against interstate commerce because local interests disproportionately reap the benefit of the caps and so the structure of the RhodeWorks tolls inevitably burdens interstate travel more heavily than intrastate travel. Although Plaintiffs acknowledge that the RhodeWorks tolls do not, on their face, look exactly like the unconstitutional flat fees at issue in Scheiner or the transitory trailer fee struck down in Trailer Marine, the tolls nonetheless "share the essential characteristics that made those [] flat fees unconstitutional." Pls.' Post-Trial Br. ("Pls.' Br.") 29, ECF No. 228.

Defendants, for their part, dispute the applicability of Scheiner and Trailer Marine Transport Corp. v. Rivera Vazquez, 977 F.2d 1 (1st Cir. 1992), and argue the toll caps are instead analogous to the frequency-based discounts held constitutional in Doran. And, alternatively, even if the Court finds the caps to be like flat fees, they are more like those found constitutional in American Trucking Ass'ns v. Michigan Public Service Commission, 545 U.S. 429, 125 S.Ct. 2419, 162 L.Ed.2d 407 (2005).

As the parties acknowledge, none of these cases are directly on point. And the fact-dependent nature of dormant Commerce Clause challenges has long been understood as part of what makes them difficult. See Scheiner, 483 U.S. at 280, 107 S.Ct. 2829 (noting series of judicial responses to specific state tax measures resembled a "quagmire"). That said, the Court agrees that these four cases — Scheiner, Trailer Marine, Doran, and Michigan — are the essential guideposts for the inquiry. And, as explained below, the Court agrees with Plaintiffs that RhodeWorks' toll caps discriminate against out-of-state interests by failing Scheiner's "internal consistency test" and by creating real, documented financial incentives to benefit local trucking interests.

For this reason, and as the First Circuit has already indicated, Scheiner is the most important analogue. See ATA II, 14 F.4th at 90 (calling Scheiner the "the most factually analogous precedent cited" by Plaintiffs). In Scheiner, the Supreme Court assessed the constitutionality of two Pennsylvania statutes that amounted to "flat taxes" on out-of-state users of the Commonwealth's highway system.⁴⁹ 483

49 The first was a $25 "marker fee," which exempted Pennsylvania-registered vehicles because the fee was deemed part of the cost of registration. Am. Trucking Ass'ns, Inc. v. Scheiner, 483 U.S. 266, 274, 107 S.Ct. 2829, 97 L.Ed.2d 226 (1987). The second was an axle tax with an accompanying reduction for Pennsylvania registration fees in the same amount. Id. at 274-75, 107 S.Ct. 2829. Both fees were assessed upon all trucks operating in Pennsylvania, regardless of whether they were registered in Pennsylvania or elsewhere, but the net effect was that out-of-state trucks paid the fees, while trucks registered in Pennsylvania received a rebate equivalent to the fees on their registration, and so paid nothing. Id. at 283, 107 S.Ct. 2829.

U.S. at 270-71, 107 S.Ct. 2829. The Supreme Court struck down those fees as unconstitutional after applying the so-called "internal consistency" test. Id. at 284, 107 S.Ct. 2829. To be constitutional under this test, the Court explained, "a state tax must be of a kind that, 'if applied in every jurisdiction, there would be no impermissible interference with free trade.'" Id. (quoting Armco Inc. v. Hardesty, 467 U.S. 638, 644, 104 S.Ct. 2620, 81 L.Ed.2d 540 (1984)). The Pennsylvania fees had an "inevitable effect ... to threaten the free movement of commerce by placing a financial barrier around" Pennsylvania. Id. Because "[i]n practical effect, ... they impose a cost per mile on [out-of-state] trucks that is approximately five times as heavy as the cost per mile borne by local trucks, the taxes are plainly discriminatory." Id. at 286, 107 S.Ct. 2829.

RhodeWorks differs from the law struck down in Scheiner in that it does not impose a simple flat fee for the privilege of using all Rhode Island bridges or roads. Instead, the State assesses a series of single user fees in varying amounts for traversing specific bridges, which accrue unless or until the user crosses that bridge more than once in a single day (in each direction) or reaches $40 in total tolls across the RhodeWorks system in a single day.⁵⁰ Perhaps recognizing this distinction, Plaintiffs do not argue that the entire program compares to Scheiner. Rather, they argue that when the limitations on tolling are triggered, the toll caps effectively operate as a series of daily flat fees. Plaintiffs describe a hypothetical example to make the point: say two trucks travel the same number of miles and cross the same bridges per day, one doing so exclusively within Rhode Island and the other passing through the state. The "local truck will pay for the privilege of using Rhode Island's roads at a lower effective rate per mile and per bridge-crossing than will the interstate truck." Pls.' Br. 27. This, they argue, is because the local truck will have made greater use "of the State's bridges as compared to the truck that is passing through as part of an interstate trip." Id. It is of no moment, they say, that some in-state trucks will pay more in the aggregate because, on average, the caps will inevitably cause the tolls to fall more heavily on out-of-state trucks.

50 There is also a third statutory toll cap, but the Court concludes it is irrelevant. Section 42-13.1-4(c) prohibits the State from charging a commercial truck using an RFID more than $20 on a border-to-border trip on I-95. RIDOT's pricing decisions set the total tolls for this journey at $17.75, see DX250, which prevents this statutory mandate from ever having an impact. The Court will not hypothesize (in either direction) about how RIDOT could have implemented the tolls on this corridor or the potential effect of those decisions.

Under RhodeWorks, a fee is assessed when a large commercial truck makes use of a bridge or bridge group specifically associated with the RhodeWorks program. Thus, the toll cap discounts are not based on whether a truck uses Rhode Island's bridges, as in Scheiner, but with how they do so. Although perhaps unlikely, an out-of-state large commercial truck could enter Rhode Island and travel freely around it without ever accruing any toll at all. And such a truck could likewise enter Rhode Island and travel across various tolled bridges and receive exactly the same benefit through the caps as a Rhode Island-based

truck. Thus, unlike Scheiner, the tolls do not appear at first blush to create a financial barrier around the state. 483 U.S. at 284, 107 S.Ct. 2829.

But Scheiner's internal consistency test gets to the rub. It asks whether, if the same system was applied in every jurisdiction, it would cause an "impermissible interference with free trade." Id. at 283, 107 S.Ct. 2829. Imagine an interstate highway system with tolls and caps like RhodeWorks across every state. And consider a trucking company based in Pennsylvania. A tractor trailer might be able to travel 600 miles in a day (ten hours per day at sixty miles per hour, to keep it simple), in which case it could make the trip between Harrisburg and Pittsburgh three times (about 200 miles each way). But such a driver might instead bypass Pittsburgh and go on to Indianapolis, Indiana (about 540 miles). Assuming a toll system like RhodeWorks with a toll cap that limited the tolls to one toll per gantry crossing per day in each direction, the Harrisburg to Pittsburgh route driver would pay only the tolls from the first 200 miles to Pittsburgh and the 200 miles back to Harrisburg (400 miles total). The interstate driver would also pay the Pennsylvania tolls but would lose the benefit of the caps when she crossed into Ohio, and then again when she crossed into Indiana. And, assuming each state also has a maximum daily toll like RhodeWorks, this too could come into play. So long as a driver confines her trips to one state, the total fare will not be proportional to the miles driven or bridges crossed. This system does not permit a vehicle to "pass among the States as freely as it may roam the State in which it is based," and neither does it "maintain state boundaries as a neutral factor in economic decisionmaking." Id. Rather, such a system "exerts an inexorable hydraulic pressure on interstate businesses to ply their trade within the State that enacted the measure rather than 'among the several States.'" Id. at 286-87, 107 S.Ct. 2829 (quoting U.S. Const., Art. I, § 8, cl. 3).

The discriminatory effects shown by Plaintiffs are not merely hypothetical. See ATA II, 14 F.4th at 89 (quoting Associated Indus. of Mo. v. Lohman, 511 U.S. 641, 654, 114 S.Ct. 1815, 128 L.Ed.2d 639 (1994)) ("The Supreme Court has 'repeatedly... focused [its] Commerce Clause analysis on whether a challenged scheme is discriminatory in effect,' and 'emphasized that equality for the purposes of ... the flow of commerce is measured in dollars and cents, not legal abstractions.'"). Like the fees in Trailer Marine, the structure of the RhodeWorks tolls with their attendant caps ensures that the financial impact falls more heavily on large commercial trucks traveling primarily in interstate commerce than on similar local trucks. 977 F.2d at 10-11.

Dr. Peters testified in practical terms that Rhode Island-plated trucks, in the aggregate, received a 23% markdown (on average) from full fare tolls, while out-of-state-plated trucks received only an 8% markdown (on average). May 26 Tr. 55:6-16; PX653 at 19, 34 (based on Emovis data). And Rhode Island vehicles received a disproportionate share of the benefits: in-state vehicles accrued 39.9% in discounts, despite only amounting to 18.6% of the transactions. May 26 Tr. 54:5-25, 56:15-24 (indicating the Emovis data for months besides July 2020 reflected a similar pattern); PX653 at 16, 18 (based on Emovis data).

Looking at the caps individually, Dr. Peters estimated that Rhode Island-plated vehicles account for approximately 48% of the once per-day/per-gantry cap. May 26 Tr. 59:5-23; PX639 at 21. He cited an extreme example of a Rhode Island-plated truck that crossed a gantry thirteen times

in each direction, accruing more than $150 in tolls, and paying only $13. May 26 Tr. 60:17-61:15; PX639 at 22. The user most likely to benefit from this type of cap, he testified, is a local one who uses the facilities frequently throughout the day. May 26 Tr. 62:7-19. As for the $40 daily cap, Dr. Peters testified that Rhode Island-plated vehicles received 36% of the discounts (on 18.6% of transactions), while out-of-state vehicles received 64% of the discounts (on 81.4% of transactions). Id. 65:13-66:16; PX639 at 24 (reflecting percentages of 36.08 and 63.92, respectively); PX639 at 16 (giving percentages of transactions). This means, according to Dr. Peters, that Rhode Island vehicles are "more frequently getting the discount," May 26 Tr. 66:17-19, and that once Rhode Island vehicles hit the $40 maximum, they then "accrue more benefit," id. 69:3-6.

The sum of this evidence is that both in-state and out-of-state vehicles can benefit from the caps, but Rhode Island-plated vehicles benefit to a significantly greater degree based on their traffic share than out-of-state vehicles. Although no record evidence specifically translated this difference into a per-mile or per-bridge crossing calculation like the court relied on in Trailer Marine, the burden on out-of-state Class 8+ trucks is clear and significant. 977 F.2d at 10 (summarizing the aggregate evidence as showing that "the transient trailer pays almost half the ordinary flat fee, effectively paying five or six times as much per accident"). Plaintiffs have proven that the caps violate the internal consistency test and inevitably shift the toll burden to out-of-state interests. The Court concludes the caps discriminate in effect.

Doran does not counsel a different result. See 348 F.3d at 320. In Doran, the First Circuit analyzed a toll program giving discounts to drivers who purchased a Massachusetts E-Z Pass transponder. Id. at 317. The toll scheme was constitutional because the discount was available "on identical terms" to all drivers, and the drivers' tolls remained directly tied to their use of the highways, notwithstanding their residences. Id. at 319. And although non-participating drivers without an E-Z Pass may pay slightly more per-gantry crossing, such a distinction was not based on residence, but on participation in the program. Id. But here the toll caps sever this pivotal connection to actual use of the bridges. After the caps kick in, a driver pays nothing no matter how many times they drive through the gantries, unlike the discount in Doran where the tolls remained tied to gantry use. And, as in Scheiner, the break between how a user uses the roads/bridges and what they pay, practically functions to give a substantial discount to in-state interests. As the Court has already noted, this is just as the caps were intended to function.

Similarly, Defendants' analogy to Michigan falls short. They contend that the toll caps apply in the same way as the $100 Michigan assessment because both impose(d) fees based purely on intrastate activity. See 545 U.S. at 434, 125 S.Ct. 2419. The Michigan fee applied to trucks engaging in "point-to-point" commercial transactions, which the Court deemed "purely local activity." Id. at 437, 125 S.Ct. 2419. Here, unlike in Michigan, it cannot be said that the fees apply to "purely local activity." See id. at 438, 125 S.Ct. 2419 ("An interstate firm with local outlets normally expects to pay local fees that are uniformly assessed upon all those who engage in local business, interstate and domestic firms alike."). Importantly, the burden of a widely applied RhodeWorks toll program would fall on interstate commerce, whereas the burden in Michigan falls on increased intrastate commerce. Finally, Defendants' argument that Plaintiffs failed to identify specific market competitors is also inapt. There is no doubt that, in some dormant Commerce Clause cases, the market matters when asking whose economic interests are treated discriminatorily. See, e.g., E. Ky. Res. v. Fiscal Ct. of Magoffin Cnty., Ky., 127 F.3d 532, 543-44 (6th Cir. 1997); Gen. Motors Corp. v. Tracy, 519 U.S. 278, 298, 117 S.Ct. 811, 136 L.Ed.2d 761 (1997) ("Conceptually, of course, any notion of discrimination assumes a comparison of substantially similar entities."); Cherry Hill Vineyard, 505 F.3d at 36, 38 (comparing effects on in-state wineries to those on out-of-state wineries). Here, Plaintiffs rely on the undifferentiated market of businesses that transport goods on interstate highways using trucks of varying types. Their argument, as it relates to the tolled vehicles, is that the caps inevitably shift the economic burden to out-of-state carriers while shifting the benefit to instate carriers. And this is enough. Just as the First Circuit noted in Trailer Marine, such an "imbalance in favor of local interests ... is a proper concern of the Commerce Clause whether or not the market participants are direct business rivals." 977 F.2d at 11. Thus, the toll caps discriminate against interstate commerce in effect.

ii. Location

Next, Plaintiffs contend that in placing the toll locations on the interstate freight corridors, the State "gerrymandered the tolls" so that the gantries are more likely to capture interstate travel despite that the bridges on state roads are more in need of repair. Pls.' Reply 15. Defendants say no, the placement of the gantries is not evidence of discrimination because Congress specifically authorized the tolling of existing bridges on the interstate and other federally-funded roads through ISTEA. The Commerce Clause does not require states to set up tolls in a manner that achieves "road-by-road apportionment." Defs.' Br. 50. This time, Plaintiffs' argument fails.

It is true that Plaintiffs' data show that over 90% of toll revenue is collected from the I-95 and I-295 corridors, as of July 2020. May 25 Tr. 131:15-24; PX652 at 12. Dr. Peters testified that these corridors are heavily used in interstate commerce, including by out-of-state users and so-called through trippers. May 25 Tr. 131:15-132:15. Indeed, as noted above, he calculated that more than 80% of the RhodeWorks tolls were billed to out-of-state large commercial trucks. PX653 at 31; see also June 3 Tr. 143:2-18 (Dr. Saraf confirming data showing more than 80% out-of-state truck transactions); DX327 at 15 (same).

But this does not show a discriminatory effect any more than it shows a discriminatory purpose, an argument the Court has already addressed. Plaintiffs may be correct that many of the RhodeWorks bridges were in better condition than others in the state, but it is not for this Court to ensure perfect equity in bridge selection or to weigh whether RIDOT should have chosen other bridges because they may be in worse structural condition. As Plaintiffs concede, ISTEA permitted the State to choose bridges connecting parts of interstate highways as its tolling locations. See 23 U.S.C. 129(a) (permitting reconstruction or replacement of a toll-free bridge and conversion into a tolled facility on federal highways). Plaintiffs have not shown that the locations of the tolls discriminate in effect.

iii. Class 4-7 Exemption⁵¹

51 Plaintiffs style this argument as related to Classes 4-7, but the supporting evidence relates primarily to Classes 6 and 7.

Finally, Plaintiffs say that by exempting Classes 4-7 — which are more likely to be

Rhode Island-plated than Class 8+ vehicles, May 26 Tr. 14:9-15 — from tolling, RhodeWorks shrank the pool of locally-owned trucks eligible for tolling by 70% even though these vehicles bear significant weight-per-axle and are less likely to take interstate trips. Plaintiffs point again to Dr. Peters, who relied on data collected by the American Transportation Research Institute ("ATRI"), CDM Smith, and Emovis transaction data. In addition to causing the tolling burden to fall disproportionately on out-of-state vehicles, Plaintiffs also say exempting Class 6 and 7 trucks amounts to giving a competitive advantage to Rhode Island business and truck owners. Defendants retort that the only relevant comparison here is between large commercial vehicles.

On the first point, Plaintiffs again are correct. The record reveals that there are more than double the number of Rhode Island-registered trucks in Classes 5-7 than there are combination trucks (Classes 8+). See PX752; June 10 Tr. 175:1-24 (discussing with Director Alviti FHWA statistics stating that Rhode Island's fleet includes 10,518 straight trucks and 4,223 combination trucks). Further, combination trucks are used more often on "long-haul" trips, whereas lower-classed trucks are used more (but not exclusively) for local, intrastate trips, e.g., delivery, garbage service, or cement mixing. See May 25 Tr. 160:11-161:13. Shifting the tolling burden away from these classes of similarly situated vehicles (in terms of axle-weight) that are owned in higher percentages by local interests and which travel more frequently intrastate onto another group that is far more likely to be owned by out-of-state interests and to travel interstate both reduces the overall burden on local businesses and increases the tolls paid by tolled (mostly non-local) businesses. The Court finds the statute discriminates against interstate commerce in this way.

Plaintiffs' second argument is that by exempting Classes 6 and 7 from tolling and increasing the tolls paid by out-of-state Class 8+ trucks, Rhode Island business owners gain a competitive advantage through lower operating costs and because those local business owners could switch their vehicles to avoid the tolls. See May 26 Tr. 20:9-21:2, 28:18-29:1. Plaintiffs rely on Dr. Peters' testimony to argue it is enough that some Class 6 and 7 trucks compete with Class 8+ trucks; it, Plaintiffs say, is a matter of common sense that certain freights can be carried by smaller and larger trucks alike. Id.

Dr. Peters' testimony is mostly speculative. Plaintiffs did not put forth any concrete evidence demonstrating an increase in Rhode Island-based companies' use of un-tolled trucks, changes in vehicle fleets, diversion,⁵² or any other data demonstrating that lower-classed trucks compete in the same market as large commercial trucks. See May 27 Tr. 23:3-25:6 (Dr. Peters testifying that he has neither quantified

52 Defendants point out that Plaintiffs have not presented concrete evidence of diversion around the tolled bridges. The Court agrees. Although there is some evidence in the record that prior to implementation CDM Smith estimated around 25% of trucks would divert around the tolled facilities, see May 26 Tr. 29:20-30:8, Plaintiffs have not shown how diversion has affected tolling, see May 23 Tr. 167:2-25, ECF No. 236 (President of Rhode Island Trucking Association stating that while diversion has come up informally in conversation, he has no knowledge that any members are actually diverting); June 13 Tr. 22:1-3, ECF No. 243 (testifying Cumberland Farms has not diverted to avoid tolls), 35:15-18 (testifying M&M Transport is not diverting around the tolls), 52:1-13 (having no knowledge of whether ATA members have diverted to avoid tolls).

how likely it is that Class 8+ owners would switch to Classes 6 and 7 or what percentage would if they did, nor conducted any studies or economic analyses on that inquiry, nor observed any change in the traffic data since RhodeWorks' tolls went live in 2018, nor heard any anecdotal evidence suggesting such a switch). Neither did Plaintiffs introduce any evidence that in-state competitors have gained a competitive advantage at the expense of out-of-state competitors that use Class 8+ trucks.⁵³

53 Defendants additionally argue that the decision to toll only large commercial vehicles does not evidence discrimination because (1) non-tolled vehicles pay in other ways; (2) there is no evidence that these users only travel intrastate; and (3) the decision to toll certain classes is for the General Assembly. These arguments were addressed above in the fair approximation section.

But in the end, none of this matters: As noted above, under Trailer Marine, overtly protectionist legislation such as RhodeWorks offends the Commerce Clause even in the absence of a specific market impact. 977 F.2d at 11. There is no question that the RhodeWorks legislation excluded lower-classed trucks to reduce the financial burden on in-state businesses and this, in turn, increases the tolls on the mostly interstate Class 8+ vehicles. In sum, the exclusion of similarly-situated vehicles like those in Classes 4-7 not only violates the fair approximation test and supports the Court's finding on discriminatory purpose, but, as with the toll caps, discriminates against interstate commerce in effect.⁵⁴

54 Because the Court holds that RhodeWorks' tolling scheme discriminates against interstate commerce, it need not decide whether it also fails under Pike v. Bruce Church, Inc., 397 U.S. 137, 90 S.Ct. 844, 25 L.Ed.2d 174 (1970). See Fam. Winemakers, 592 F.3d at 10.

c. Strict Scrutiny

Having found RhodeWorks discriminates in purpose and effect, the Court applies strict scrutiny, asking if the State can "show that it advances a legitimate local purpose that cannot be adequately served by reasonable nondiscriminatory alternatives." Or. Waste, 511 U.S. at 100-01, 114 S.Ct. 1345 (quoting New Energy, 486 U.S. at 278, 108 S.Ct. 1803) (cleaned up). This analysis is straightforward. Rhode Island has a legitimate — even compelling — interest in the maintenance of its ailing bridges. But there is no reason that interest cannot be served by a tolling system that does not offend the Commerce Clause. Indeed, many states have implemented tolling systems that fairly apportion their costs across various users and do not discriminate against interstate commerce. Applying strict scrutiny, RhodeWorks' tolling program fails the test.

III. Conclusion⁵⁵

55 For convenience, the Court summarizes the various rulings made throughout this decision: Plaintiffs' Motion to Reopen, ECF No. 221, is GRANTED; Defendants' Motion for Judgment on Partial Findings, ECF No. 222, is DENIED; Defendants' objection related to Dr. Saraf's testimony, see June 6 Tr. 68:17-72:19, 73:15-75:4, is SUSTAINED and that testimony is stricken; Defendants' objection to the admission of Exhibit DX236 is SUSTAINED; Defendants' Motion in Limine No. 1, ECF No. 166, is DENIED; Defendants' Motion in Limine No. 2, ECF No. 168, is now DENIED in full; and Plaintiffs' Motion to Exclude Evidence Related to the Rhode Island Department of Transportation's Road System Budget, ECF No. 181, is DENIED.

Because RhodeWorks fails to fairly apportion its tolls among bridge users based on a fair approximation of their use of the bridges, was enacted with a discriminatory purpose, and is discriminatory in effect, the statute's tolling regime is unconstitutional under the dormant Commerce

Clause of the United States Constitution. Plaintiffs' request for a permanent injunction is GRANTED, and RhodeWorks' tolling system is enjoined. The Court specifically orders the following:

1. Under 28 U.S.C § 2201, judgment shall enter declaring R.I. Gen. Laws § 42-13.1-4(a), which precludes tolling any vehicle except large commercial trucks, along with its implementing regulations, unconstitutional.

2. Defendants are permanently enjoined from charging or collecting tolls pursuant to R.I. General Laws § 42-13.1-4, or from enforcing nonpayment of such tolls through penalty for nonpayment or avoidance under R.I. Gen. Laws §§ 42-13.1-11, 12. This injunction shall take effect 48 hours following entry of final judgment.

3. The clerk is directed to enter final judgment in favor of Plaintiffs.

4. The Court defers ruling on Plaintiffs' request for attorneys' fees and costs (including expert-witness fees) pursuant to 42 U.S.C. § 1988. The motion for fees and costs shall not be filed until after any appeal is decided and a mandate is issued or, if no appeal is taken, until the period for entering a Notice of Appeal expires. See Local Rule 54.1 (setting deadline for motion for attorneys' fees "[u]nless otherwise ordered by the Court"). From that time the prevailing party will have fourteen days to file a motion for fees and costs, with an accompanying bill of costs, as set forth in Local Rules 54.1 and 54.

IT IS SO ORDERED.