The Relevance of Ross to Geolocation and the Dormant Commerce Clause

Article - by and Online Edition - Volume 102

Introduction

In State Regulation of Online Behavior: The Dormant Commerce Clause and Geolocation, we argued that the validity of state laws regulating internet activity should turn in part on the feasibility of geolocation—the extent to which online services can reliably determine the state in which a user is located so that they can comply with the law of that state.[1] The article was published before the Supreme Court decided National Pork Producers Council v. Ross,[2] an important Dormant Commerce Clause (DCC) decision, in May 2023. This brief essay explains how Ross supports our central arguments.

I. Geolocation and the Dormant Commerce Clause

Our article argued that a state’s regulation of internet speech originating in another state does not necessarily violate the DCC, even if the regulation imposes significant costs in the originating state and elsewhere.[3]

We thus sought to refute an argument made in cases such as American Libraries Association v. Pataki,[4] the influential 1997 decision that used the DCC to invalidate a New York ban on intentionally using the internet “to initiate or engage” in certain pornographic communications deemed to be “harmful to minors.”[5] The court assumed that the internet was a borderless medium; that “[o]nce a provider posts content on the Internet, it is available to all other Internet users worldwide”;[6] and that since “no aspect of the Internet can feasibly be closed off to users from another state,” it was “impossible to restrict the effects of the New York Act to conduct occurring within New York.”[7] Based on these factual assumptions, the court ruled that the Act violated the extraterritoriality prong of the DCC since New York’s application of its law to internet communications “projected its law into other states whose citizens use the Net.”[8] And the court ruled that the act violated the Pike balancing prong of the DCC because the burdens were “extreme” since the law affected internet users everywhere.[9]

Our article made technological and legal arguments in response[10]—arguments that, as we showed, are supported by recent Supreme Court and lower court DCC cases.[11]

Most importantly, the Pataki court’s assumptions about the borderless internet, which might have had some validity in 1997, are not accurate in 2023. Geo-identification technologies—technologies widely deployed by firms to enhance their profitability—can now identify where an internet user is geographically located and can enable websites and other internet operations to treat users differently (and show different content) based on geography.[12] Contrary to what Pataki and similar cases assumed, internet firms regularly identify users based on geographical location and shape or limit the communications they receive based on that geography. They do so, we explained, for purposes of advertising, legal compliance, fraud detection, price discrimination, consumer profiling, and more.[13]

The pervasiveness of geolocation and filtering technology, we argued, affects DCC analysis, and gives states much more leeway to regulate internet communications than Pataki contemplated, because such technology permits internet speakers and content providers to shape their activities to comply with local law.[14] Geo-identification technology puts internet speakers and content providers in a similar position as real-space firms that must take steps and incur costs to ensure that their products comply with local laws everywhere they are available. Because geo-identification technology is so ubiquitous, we argued, “courts should not presume that internet operators have any greater difficulties than real-space operators in identifying internet users based on geography and tailoring their products to state law.”[15] We concluded that DCC analyses of state internet regulations must include a realistic assessment of the impact on interstate commerce based on the current state of geolocation and filtering technology.

On top of this argument from technology we made two arguments about relevant legal doctrine.

First, we argued that the so-called extraterritoriality prong of DCC cases should not be read to invalidate state laws merely because they impose legal and compliance costs on out of staters.[16] The extraterritoriality idea built on a dictum in Healy v. Beer Institute[17] that the DCC invalidates state laws that have the “practical effect” of controlling “conduct beyond the boundaries of the State.”[18] We maintained that this statement could not be taken seriously, since it “is widely accepted that, consistent with the Dormant Commerce Clause, a firm doing multistate business must bear the cost of discovering and complying with state laws—tort laws, tax laws, franchise laws, health laws, privacy laws, and much more—everywhere it does business.”[19] And we argued that Healy and related cases are best understood not as establishing a freestanding extraterritoriality doctrine, but rather as merely instantiating the core DCC tests that forbid (i) nondiscrimination against interstate commerce, and (ii) undue burdens on interstate commerce. We doubted that the extraterritoriality prong, standing alone, had any relevance to DCC analyses of state internet regulations.

Second, we argued that “in assessing the costs of compliance with state law for Dormant Commerce Clause purposes, a firm’s preferred national market structure is irrelevant.”[20] Firms, including internet firms, often find it advantageous to organize themselves to deliver products, including digital products, without regard to state borders. But the firm cannot invoke this preferred organizational structure, or the costs of adjusting it to conform to the state laws where it does business, as a reason to invalidate the state law under the DCC.

In support of this argument, we pointed to Exxon Corp. v. Governor of Maryland,[21] a Supreme Court decision that rejected a DCC challenge to a Maryland law that banned national oil producers from operating retail service stations in the state.[22] The Court there rejected Exxon’s argument that the law would “change the [national] market structure,” and might have “serious implications for their national marketing operations,” and held instead that the DCC does not “protect[] the particular structure or methods of operation in a retail market.”[23]

II. Ross and State Internet Regulation

Ross was not an internet case, but its analysis significantly bears on DCC arguments related to state internet regulations.

A. Background

Ross involved California Proposition 12, a law that banned the sale in California of whole pork meat from breeding pigs (or their offspring) that have been “confined” in a way that prevents them from “lying down, standing up, fully extending [their] limbs, or turning around freely.”[24] The proponents of the law claimed it would eliminate from California markets pork meat produced in inhumane conditions, and reduce health risks from pork products.[25]

A challenge to the law came before the Court, where plaintiffs made two arguments. First, they claimed that Proposition 12 violates the extraterritoriality component of the Dormant Commerce Clause because it imposed “substantial new costs on out-of-state pork producers who wish to sell their products in California.”[26] Second, they argued that Proposition 12 failed Pike balancing because “the benefits Proposition 12 secures for Californians do not outweigh the costs it imposes on out-of-state economic interests.”[27] The Court upheld the law, rejecting both arguments.[28]

B. Extraterritoriality

The Court, in an opinion by Justice Gorsuch, unanimously held that there is no extraterritoriality prong of the DCC beyond the DCC’s core anti-discrimination principle.[29] It did so largely for the reasons we gave in our article. The Court explained that the cases cited by petitioner—Healy and its predecessors[30]—did not in fact establish a doctrine forbidding state laws that have the practical effect of controlling out-of-state behavior, or that otherwise produce out-of-state effects.[31] The Court noted that this could not be the test for the DCC, since “[i]n our interconnected national marketplace, many (maybe most) state laws” have such extraterritorial consequences.[32] The Court instead read these cases to be applications of the core DCC “concern with preventing purposeful discrimination against out-of-state economic interests.”[33]

As our article explained, the primary doctrinal challenge to contemporary state regulation of internet communications in the lower courts has been the view that the DCC bans laws that regulate extraterritorial behavior. Ross announces that a standalone extraterritoriality test is not part of the DCC analysis.

C. Pike Balancing

1. The Court’s Analysis

The Court also rejected petitioners’ Pike balancing argument, though it splintered on the rationale. Justice Gorsuch garnered a majority for two modest propositions. The first was that the core aims of Pike balancing are (a) to flush out a discriminatory purpose in the state law and (b) to protect the “instrumentalities of interstate transportation.”[34] The second was a rejection of petitioners’ argument that the Pike cases “prevent a State from regulating the sale of an ordinary consumer good within its own borders on nondiscriminatory terms.”[35]

But on the bottom-line question of why the law was valid, the Court offered two non-majority opinions. Justices Gorsuch, Thomas, and Barrett concluded that the Court is institutionally incompetent to perform Pike balancing, at least where, as here, the benefits of the regulation (the advancement of moral and health interests) and the costs of the regulation (the costs of compliance) are incommensurable.[36] And in a separate opinion, Justices Gorsuch, Thomas, Sotomayor, and Kagan reasoned that petitioners’ alleged harm—that some pig producers’ preferred “methods of operation” would be more costly to maintain—does not state a cognizable injury to interstate commerce as required by Pike.[37]

The four-judge opinion almost certainly states the Court’s controlling Pike rationale for the lower courts. Under Marks v. United States,[38] “the holding of the Court may be viewed as that position taken by those Members who concurred in the judgments on the narrowest grounds.”[39] Often the Marks rule is hard to apply.[40] But here it is not. The Gorsuch–Thomas–Barrett plurality calls into question a core element of Pike. The Gorsuch–Thomas–Sotomayor–Kagan plurality is a fact-bound application of Pike that keeps open the possibility that petitioners can replead.[41] The latter is clearly the narrower ground for decision, and thus the Court’s holding on Pike.[42] It is worth noting, for predicting future decision-making, that different majorities of the Court (including some members in dissent) affirmed important elements of Pike. As the Chief Justice correctly noted in his partial dissent, a majority of the Court affirmed that Pike is not limited to “laws either concerning discrimination or governing interstate transportation,” and that Pike balancing remains valid and within the Court’s competence, even for incommensurable costs and benefits.[43]

2. Exxon

The controlling opinion’s Pike balancing analysis thus extends the earlier Exxon holding: the costs to a firm’s preferred market structure of complying with a state regulation are irrelevant to DCC analysis.[44] And this has important implications for state internet regulations.

In Exxon, Exxon had alleged that a facially nondiscriminatory Maryland law banning oil producers from running retail gas stations in the State exclusively burdened companies operating in interstate commerce and threatened to force some to withdraw from the Maryland market or “incur new costs to serve that market.”[45] The Court in Exxon acknowledged that the Maryland law favored a certain type of business structure (independent gas station retailers), and would raise compliance costs for vertically integrated national production and for retail firms, thus leading some oil refiners to stop selling petroleum in the state.[46]

Yet the Exxon Court rejected the view that a “‘change [in] the market structure’” to vertically integrated oil companies would “impermissibly burde[n] interstate commerce.”[47] The Court rejected “appellants’ underlying notion that the Commerce Clause protects the particular structure or methods of operation in a retail market,” and added that the DCC does not protect “particular interstate firms . . . from prohibitive or burdensome regulations.”[48]

And in Ross the Court extended the Exxon legal principles to apply squarely to the flow of goods in interstate commerce. The Court’s analysis is worth quoting in full:

If Maryland’s law did not impose a sufficient burden on interstate commerce to warrant further scrutiny, the same must be said for Proposition 12. In Exxon, vertically integrated businesses faced a choice: They could divest their production capacities or withdraw from the local retail market. Here, farmers and vertically integrated processors have at least as much choice: They may provide all their pigs the space the law requires; they may segregate their operations to ensure pork products entering California meet its standards; or they may withdraw from that State’s market. . . .

In Exxon, as far as anyone could tell, the law threatened only to shift market share from one set of out-of-state firms to another. Here, the pleadings allow for the same possibility—that California market share previously enjoyed by one group of profit-seeking, out-of-state businesses (farmers who stringently confine pigs and processors who decline to segregate their products) will be replaced by another (those who raise and trace Proposition 12-compliant pork).

In both cases, some may question the “wisdom” of a law that threatens to disrupt the existing practices of some industry participants and may lead to higher consumer prices. But the dormant Commerce Clause does not protect a “particular structure or metho[d] of operation.” That goes for pigs no less than gas stations.[49]

And the same goes for internet content (and internet content providers) as well—at least insofar, as we argued, that the state law in question “provides a defense for reasonable efforts to keep forbidden internet content out of the state.”[50] One might equally say, with regard to (for instance) laws requiring internet platforms not to discriminate based on viewpoint as to posts seen by citizens of a state:

In Exxon, vertically integrated businesses faced a choice: They could divest their production capacities or withdraw from the local retail market. Here, [internet companies] have at least as much choice: They may provide all their [users freedom from viewpoint discrimination]; they may segregate their operations to ensure [that in-state users aren’t denied access to posts based on viewpoint]; or they may withdraw from that State’s market. . . .

In Exxon, as far as anyone could tell, the law threatened only to shift market share from one set of out-of-state firms to another. Here, the pleadings allow for the same possibility—that [the state’s] market share previously enjoyed by one group of profit-seeking, out-of-state businesses ([Internet platforms that engage in viewpoint discrimination and] who decline to segregate their [operations]) will be replaced by another (those who [do not discriminate based on viewpoint]).

In both cases, some may question the “wisdom” of a law that threatens to disrupt the existing practices of some industry participants and may lead to [a different level of internet platform content moderation]. But the dormant Commerce Clause does not protect a “particular structure or metho[d] of operation.” That goes for [internet platforms] no less than [for pigs or] gas stations.[51]

After Ross, internet firms’ preferences for distributing content on a national scale are not privileged under the DCC, especially since they can exclude or shape content to conform to state regulations of that content, just as real-space firms do with regard to state regulations of non-digital goods.

3. Inconsistent Regulations and the Instrumentalities of Interstate Commerce

Ross implicates another issue that we touched on and that some courts have thought relevant to state internet regulations—the so-called “inconsistent regulations” cases. A 1987 decision, CTS Corp. v. Dynamics Corp. of America,[52] stated that the Court’s DCC cases have “invalidated statutes that may adversely affect interstate commerce by subjecting activities to inconsistent regulations,” and upheld an Indiana anti-takeover law in the face of this test.[53] But the DCC cases that the Court cited for this proposition were (i) the so-called extraterritoriality cases,[54] and (ii) the DCC cases that invalidate certain state regulations that unduly burden the “instrumentalities of interstate transportation.”[55]

Some lower courts, especially around the turn of the century, have treated the “inconsistent regulations” idea as an independent DCC test and on that basis struck down state internet regulations. We argued that this test, like the “extraterritoriality” prong of the DCC, “cannot be applied literally” since “[f]irms operating in different states typically must comply with scores of inconsistent regulations, even if doing so is more costly than complying with a uniform national rule would be.”[56] We argued that “the Supreme Court’s inconsistent-regulations cases require no more than an application of the broader [Pike] undue burden test,” and noted that many commentators and lower courts agreed.[57]

It is hard to see how the standalone “inconsistent regulations” test sometimes applied by the lower courts survives Ross. The majority opinion in Ross does not mention CTS Corp. (except in a reference to an article by Donald Regan)[58] or an inconsistent regulations test, even though the majority purports to describe the universe of DCC analyses. As noted above, the majority ruled that the so-called extraterritoriality cases are really instances of Pike balancing. The same opinion, and indeed every Justice on the Court, also acknowledges that the instrumentalities cases too are exemplars of Pike balancing.[59]

We did not in our original article separately discuss the instrumentalities cases, but it is clear that these cases are of little relevance to state regulation in the geolocation era. The modern instrumentalities precedents, the Court noted, involve state regulation of “trucks, trains, and the like.”[60] The core idea in these cases is that some regulations of the instrumentalities of the interstate transportation of goods—notably, the lengths of trailers and trains, and the size of mudguards for trucks and trailers—so burden the interstate transportation of goods, and contribute so little to a legitimate local concern, such as health or safety, as to fail Pike balancing.

The state internet regulations canvassed in our original paper regulate digital goods, especially digital content, and not the instrumentalities of internet communication, such as the physical infrastructure of the internet, or internet protocols. As the Ross Court noted, “at least some decisions in [the extraterritoriality] line might be viewed as condemning state laws that ‘although neutral on their face . . . were enacted at the instance of, and primarily benefit,’ in-state interests”;[61] that would not apply to, for instance, internet platform viewpoint-neutrality mandates.

And in any event, courts considering such regulations on viewpoint discrimination “do not face a law that impedes the flow of commerce.”[62] Just as “[p]igs are not trucks or trains,”[63] so social media posts are not trucks or trains, either. And to the extent that a platform viewpoint-neutrality mandate affects commerce, it does so in a constitutional way, for the reasons we described in our article: Because geolocation and filtering technology is pervasive, we argued, “courts should not presume that internet operators have any greater difficulties than real-space operators in identifying internet users based on geography and tailoring their products to state law.”[64] “Assessing the costs and benefits of complying with state regulations . . . must include realistically assessing compliance costs based on the current state of geolocation and filtering technology.”[65] Going forward, courts can only strike down such state regulations if, after assessing costs based on a contemporary and realistic assessment of geo-technology, the costs are “clearly excessive in relation to the putative local benefits.”[66]

  1. . Jack Goldsmith & Eugene Volokh, State Regulation of Online Behavior: The Dormant Commerce Clause and Geolocation, 101 Texas L. Rev. ‌1083, 1088, 1107 (2023). ↑
  2. . 598 U.S. 356 (2023). ↑
  3. . See Goldsmith & Volokh, supra note 1, at 1099 (“But when a company distributes material into a state, including online, applying state tort law to that material likely doesn’t violate the [Dormant Commerce] Clause.”). ↑
  4. . 969 F. Supp. 160 (S.D.N.Y. 1997). ↑
  5. . Id. at 163. ↑
  6. . Id. at 167 (quoting Am. Civ. Liberties Union v. Reno, 929 F. Supp. 824, 844 (E.D. Pa. 1996), aff’d, 521 U.S. 844 (1997)). ↑
  7. . Id. at 171, 177. ↑
  8. . Id. at 177. ↑
  9. . Id. at 179. Pike balancing is the test from Pike v. Bruce Church, Inc., 397 U.S. 137 (1970). The Pike Court held that “where the statute regulates even-handedly to effectuate a legitimate local public interest, and its effects on interstate commerce are only incidental, it will be upheld unless the burden imposed on such commerce is clearly excessive in relation to the putative local benefits.” Id. at 142. ↑
  10. . Goldsmith & Volokh, supra note 1, at 1110. ↑
  11. . See, e.g., S.D. v. Wayfair, Inc., 138 S. Ct. 2080, 2098, 2100 (2018) (finding that the DCC did not invalidate a state tax law requiring out-of-state merchants to collect and remit taxes on in-state sales because of the availability of software to help merchants manage the burden); see also Greater L.A. Agency on Deafness v. Cable News Network, 742 F.3d 414, 433 (9th Cir. 2014) (finding that the DCC did not invalidate a state law requiring closed captioning on programs accessed on the internet in California because the network could avoid extraterritoriality issues by targeting which states’ users had the option to see closed captions). ↑
  12. . Goldsmith & Volokh, supra note 1, at 1104, 1107. ↑
  13. . Id. at 1105–06, 1108. ↑
  14. . Id. at 1110, 1111. ↑
  15. Id. at 1111. ↑
  16. . See id. at 1125 (“[C]ourts should . . . play only a limited role in striking down, on ‘extraterritoriality’ grounds, state laws that apply to internet transactions.”). ↑
  17. . 491 U.S. 324 (1989). ↑
  18. . Id. at 336. ↑
  19. . Goldsmith & Volokh, supra note 1, at 1090. ↑
  20. . Id. at 1111. ↑
  21. . 437 U.S. 117 (1978). ↑
  22. . See id. at 128. ↑
  23. . See id. at 127–28. ↑
  24. . Cal. Health & Safety Code §§ 25990, 25991(e) (West 2018). ↑
  25. . Graham Vyse, Over PETA’s Objections, California Voters Pass Strict Animal Protections, Governing (Oct. 17, 2018), https://www.governing.com/archive/gov-california-meat-sales-ballot-peta-animal-rights.html. ↑
  26. . Nat’l Pork Producers Council v. Ross, 598 U.S. 356, 371 (2023). ↑
  27. . Id. at 377. ↑
  28. . Id. at 375–76, 380. ↑
  29. . Id. ↑
  30. . Id. at 371 (first citing Baldwin v. G.A.F. Seelig, Inc., 294 U.S. 511 (1935); then citing Brown-Forman Distillers Corp. v. N.Y. State Liquor Authority, 476 U.S. 573 (1986); and then citing Healy v. Beer Inst., 491 U.S. 324 (1989)). ↑
  31. . And the Court, as we did, invoked Pharmaceutical Research and Manufacturers of America v. Walsh, 538 U.S. 644, 669 (2003), as support for its narrow reading of the “extraterritoriality” cases. Nat’l Pork Producers Council v. Ross, 598 U.S. 356, 373–74 (2023). ↑
  32. . Id. at 374. ↑
  33. . Id. at 371. ↑
  34. . Id. at 380; see also id. at 389 n.4 (“When it comes to Pike, a majority agrees that heartland Pike cases seek to smoke out purposeful discrimination in state laws (as illuminated by those laws’ practical effects) or seek to protect the instrumentalities of interstate transportation.”). ↑
  35. . Id. at 390–91; see also id. at 389 n.4. ↑
  36. . Id. at 380–82; see also id. at 393–94 (Barrett, J., concurring in part). ↑
  37. . Id. at 386–87 (plurality opinion). ↑
  38. . 430 U.S. 188 (1977). ↑
  39. . Id. at 193 (quoting Gregg v. Georgia, 428 U.S. 153, 169 n.15 (1976)). ↑
  40. . Richard M. Re, Beyond the Marks Rule, 132 Harv. L. Rev. 1942, 1944–45 (2019). ↑
  41. . Ross, 598 U.S. at 386 (“Further experience may yield further facts.”). ↑
  42. . Justice Kavanaugh in dissent agreed with this proposition. Id. at 410 (Kavanaugh, J., concurring in part and dissenting in part). And Justice Gorsuch in his opinion did not disagree. See id. at 389 n.4 (plurality opinion). ↑
  43. . Id. at 396–97 (Roberts, C.J., concurring in part and dissenting in part). The primary partial dissent also noted that “A majority of the Court agrees that—were it possible to balance benefits and burdens in this context—petitioners have plausibly stated a substantial burden against interstate commerce.” See id. at 1172. This is because Justice Barrett agreed with the primary dissent that petitioners stated a Pike claim but disagreed that the Court was competent to do Pike balancing in this context. See id. at 394 (Barrett, J., concurring in part). ↑
  44. . We take no position here on whether the Court’s application of Exxon to the different economic situation posed in Ross was true to Exxon or to prior Dormant Commerce Clause jurisprudence. The important point for present purposes is that the Court’s controlling opinion extended and applied Exxon. ↑
  45. . Id. at 383 (plurality opinion). ↑
  46. . Exxon Corp. v. Governor of Md., 437 U.S. 117, 122–24, 127. ↑
  47. . Id. at 127. ↑
  48. . Id. at 127–28. ↑
  49. . Ross, 598 U.S. at 384–85 (internal citations omitted) (paragraph breaks added). ↑
  50. . Goldsmith & Volokh, supra note 1, at 1112. ↑
  51. . Ross, 598 U.S. at 384–85 (internal citations omitted) (paragraph breaks added). ↑
  52. . 481 U.S. 69, 94 (1987). ↑
  53. . Id. at 88. ↑
  54. . Namely Brown-Forman Distillers Corp. v. N.Y. State Liquor Authority, 476 U.S. 573, 583–84 (1986). The Court also cited the plurality opinion of Justice White in Edgar v. MITE Corp., 457 U.S. 624, 642 (1982). ↑
  55. . Kassel v. Consolidated Freightways Corp., 450 U.S. 662, 671 (1981) (plurality opinion) (Powell, J.); Southern Pacific Co. v. Arizona, 325 U.S. 761, (1945) (noting the “confusion and difficulty” that would attend the “unsatisfied need for uniformity” in setting maximum limits on train lengths); Cooley v. Board of Wardens, 53 U.S. 299, 319 (1851). ↑
  56. . Goldsmith & Volokh, supra note 1, at 1093. ↑
  57. . Id. ↑
  58. . Nat’l Pork Producers v. Ross, 598 U.S. 356, 376 n.1 (2023). ↑
  59. . Id. at 373; see also id. at 389 n.4 (“[A] majority agrees that heartland Pike cases . . . seek to protect the instrumentalities of interstate transportation.”); id. at 392 (“Pike claims that do not allege discrimination or a burden on an artery of commerce are further from Pike’s core”) (Sotomayor, J., concurring in part); id. at 395–96 (Roberts, C.J., concurring in part and dissenting in part). ↑
  60. . Kassel v. Consol. Freightways Corp., 450 U.S. 662, 671 (1981) (plurality opinion) (length of trailers); Raymond Motor Transp., Inc. v. Rice, 434 U.S. 429, 447 (1978) (length of trailers); Bibb v. Navajo Freight Lines, 359 U.S. 520, 530 (1959) (type of mudguards for trucks and trailers); S. Pac. Co. v. Arizona ex rel. Sullivan, 325 U.S. 761, 783–84 (1945) (length of railroad trains). ↑
  61. . Ross, 598 U.S. at 379 n.2. ↑
  62. . Id. ↑
  63. . Id. ↑
  64. . Goldsmith & Volokh, supra note 1, at 1111. ↑
  65. . Id. ↑
  66. . Pike, 397 U.S. at 142 (emphasis added). ↑

Download PDF