Forum Shopping and Patent Law—A Comment on TC Heartland

Case Comment - Volume 96 - Issue 1

The Supreme Court addressed rules affecting forum-shopping incentives in three cases during its 2016–2017 term.[1] This Essay focuses on one of those cases—TC Heartland LLC v. Kraft Foods Group Brands LLC.[2] In TC Heartland, the Court narrowly interpreted the patent venue statute, 28 U.S.C. § 1400(b), to restrict where patentees can file infringement suits. The case involved a technical issue of statutory interpretation, but one that implicated substantial questions of patent policy and promised serious real-world consequences affecting the future of patent litigation, the efficacy of patent law, and even the economic health of communities in East Texas, especially the town of Marshall, Texas. For these reasons, the case attracted widespread public attention. The Court’s unanimous opinion, however, ignores this broader context. It focuses narrowly on the statute and defends the holding with a largely textualist interpretation.

This is more than a little surprising. The contrast between the Court’s style of reasoning and the decision’s real-world consequences could hardly be more striking. It is not surprising that Justices firmly committed to textualism would insist on a textualist analysis even when statutory text offers very limited guidance. But where consequences are so significant, one might have expected Justices of a more pragmatic and functional bent, such as Justice Breyer, to have written a concurring opinion taking note of those consequences as part of a purposive interpretation of the statute. Yet, as I shall argue, constructing a convincing purposive interpretation is not easy to do. In the end, the Court’s decision to ignore the broader context might make more sense than it seems at first glance.

The aim of this Essay is to review the Court’s decision, assess its possible impact on patent litigation, and analyze its interpretive approach. Part I describes the factual background of the TC Heartland case, summarizes the Court’s holding, and explains the broader patent law context that made the case so important and newsworthy. Part II critically examines the Court’s reasoning. Part III offers some reasons why it might have made sense for the Court to ignore the broader patent law context despite its key importance to the case.

I. The TC Heartland Case and Its Broader Context

A. Background and Holding

TC Heartland is a lawsuit for patent infringement. The plaintiff-patentee, Kraft Food Group Brands LLC (Kraft), and the defendant, TC Heartland LLC (TC Heartland), are competitors in the market for flavored drink mixes.[3] Kraft sued TC Heartland and Heartland Packaging Corporation in the District of Delaware, alleging that the defendants infringed Kraft’s patents in liquid water enhancers.[4] The defendants filed a motion under Rule 12(b)(3) to dismiss the suit for improper venue, or in the alternative, to transfer the lawsuit to the Southern District of Indiana pursuant to 28 U.S.C. § 1406(a).[5]

Kraft is organized under the laws of Delaware with its principal place of business in Illinois, and TC Heartland is organized under the laws of Indiana with its headquarters in Indiana.[6] TC Heartland is not registered to do business in Delaware, nor does it have any supply contracts in Delaware, hire any local salespeople, or have any other significant “local presence” in Delaware.[7] However, TC Heartland ships allegedly infringing products into Delaware.[8]

The district court held that venue was proper based on the Federal Circuit’s interpretation of the patent venue statute.[9] TC Heartland filed a petition with the Federal Circuit for a writ of mandamus, and the Federal Circuit denied the petition.[10] The Supreme Court granted certiorari on the venue issue and reversed.[11]

The precise legal issue in the case is a technical one. Section 1400(b) creates two distinct grounds for venue: (1) in the judicial district “where the defendant resides,” or (2) in any district “where the defendant has committed acts of infringement and has a regular and established place of business.”[12] TC Heartland argued that the first ground could not support venue in the District of Delaware because it did not “reside” there, and Kraft disagreed.[13]

The venue issue thus turned on the proper definition of the word “resides” in § 1400(b). More specifically, the question before the Court had to do with the continuing vitality of a 1990 Federal Circuit decision, VE Holding Corp. v. Johnson Gas Appliance Co., [14] which held that § 1391(c) of the general venue statute supplies the definition of “resides” for § 1400(b).[15] Section 1391(c) states that a corporation “shall be deemed to reside, if a defendant, in any judicial district in which [it] is subject to the court’s personal jurisdiction.”[16] Since the District of Delaware had personal jurisdiction over TC Heartland, Kraft argued, the company resided in the District of Delaware for venue purposes under VE Holding.

TC Heartland, for its part, relied on a 1957 Supreme Court decision, Fourco Glass Co. v. Transmirra Products Corp.[17] Fourco Glass held that § 1391(c)’s definition of “resides” does not apply to § 1400(b)—which, in light of the precedent at the time, meant that a corporation “resides” only in the place of incorporation.[18] TC Heartland argued that since it was organized under the laws of Indiana and not under the laws of Delaware, it resided in Indiana under Fourco Glass, and not in Delaware.

The Supreme Court sided with TC Heartland. As explained in more detail in Part II below, the Court held that Fourco Glass still controls the definition of residence in § 1400(b) for domestic corporations, notwithstanding subsequent amendments to the venue statutes.[19] However, it left open the question of residence for unincorporated entities and foreign corporations.[20]

B. The Broader Context

One would hardly expect TC Heartland, with its rather dry technical issue, to attract much public attention. But it did. The Supreme Court received approximately thirty amicus briefs from a wide range of interested parties, including several major IP companies, organizations keenly interested in the future of IP law, IP scholars and economists, and a retired Chief Judge of the Federal Circuit (Judge Paul Michel).[21] The state of Texas even filed an amicus brief, joined by sixteen other states.[22] The national media also took an interest. The New York Times, for example, published at least one article about the case while it was pending in the Supreme Court.[23] This amount of attention is quite remarkable for a case involving such a narrow procedural issue.

The reason for the intense interest had to do with the real-world stakes of the Court’s decision. For many, the case implicated the proper functioning of the patent system, the success of the patent troll strategy, and even the future of Marshall, Texas.[24] To understand why, it is necessary to focus on how venue choices affect forum shopping by patent plaintiffs and forum selling by federal district courts.[25]

Before TC Heartland, a patent owner could file a patent infringement suit in virtually any federal district court in the country. This was the result of two Federal Circuit decisions, one having to do with personal jurisdiction and the other with venue.[26] In Beverly Hills Fan Co. v. Royal Sovereign Corp.,[27] the Federal Circuit upheld the exercise of personal jurisdiction over an out-of-state defendant that purposefully and regularly distributed allegedly infringing products in the forum state through an intermediary in an established distribution channel.[28] In VE Holding Corp. v. Johnson Gas Appliance Co.,[29] the Federal Circuit held that the definition of corporate residence in § 1391 applied to § 1400(b), thereby tying venue to personal jurisdiction.[30] Together these two decisions allowed patentees to sue almost anywhere that the defendant’s products were regularly sold.

With this many venue options available, patentee-plaintiffs had strong incentives to shop for a court that offered the most favorable procedures.[31] According to a number of commentators, these incentives generated a competition among federal districts eager to attract patent litigation, in which districts competed by offering pro-plaintiff procedures.[32] As a result, cases ended up concentrated in a few districts: those that valued patent law business enough and were able to adjust their procedures to offer the best deals to patentee–plaintiffs.[33]

The empirical evidence of case concentration is quite striking. According to one study, 48.9% of all patent suits filed from January 2014 through June 2016 were filed in only two federal districts: the Eastern District of Texas and the District of Delaware.[34] In fact, the Eastern District of Texas by itself captured 36% of the national filings over this period, and almost 44% in 2015 alone.[35] Indeed, most Eastern District cases were routed to a single federal district judge, Judge Rodney Gilstrap, located in the small town of Marshall, Texas. The empirical studies show that Judge Gilstrap handled almost 25% of all patent cases filed nationwide from January 2014 through June 2016.[36]

The Eastern District of Texas is hardly a hotbed of innovation or a central location for patent industries. Indeed, the small town of Marshall, Texas, where Judge Gilstrap sits, has a population of approximately 25,000.[37] According to commentators, the reason the Eastern District was so attractive has to do with its pro-patentee procedures, including a restrictive approach to granting summary judgment (making it harder for defendants to exit lawsuits) and a preference for broad and expedited discovery (increasing defendant’s costs relative to plaintiff’s).[38] These same commentators also point out that the Eastern District has a case assignment system that allows plaintiffs to select pro-patentee judges with high confidence, offers juries that tend to be generous with damage awards, and delays or denies transfer motions with regularity in order to lock in cases.[39]

Critics argue that these pro-plaintiff procedures impose considerable pressure on defendants to settle and that patent trolls—or, less pejoratively, “patent assertion entities” (PAE)—benefit greatly from this pressure. Patent trolls are companies that buy patents not to practice or commercialize them, but to assert them against others making productive use of the technology in an effort to leverage settlements.[40] According to the critics, many of these patents are of poor quality, the suits they support are weak, and the settlements they generate greatly exceed the patent’s contribution to the value of the infringing product.[41] PAEs are pervasive in the patent system; empirical evidence shows that they are responsible for more than half of all the patent infringement suits filed in the United States.[42]

Settlement is very important to the patent troll’s strategy. With weak patents, there is a slim chance of winning at trial, so success depends on pressuring defendants to settle by threatening high litigation costs. According to critics, the pro-plaintiff procedures of the Eastern District of Texas, and other patentee-friendly districts, play into this strategy and, as a result, patent trolls file in those districts.[43] Many of these critics believe that the problem is particularly serious because suits by patent trolls burden IP innovators and chill incentives to invest in research and development.[44] In sum, the concern is that the Federal Circuit’s liberal approach to venue and personal jurisdiction supports interdistrict competition, which leads to the concentration of patent cases in districts with patentee-favorable law, which in turn supports patent troll litigation that stifles innovation.

The patent troll problem has been a key issue for technology firms, patent lawyers, scholars, and politicians over the past decade.[45] In recent years, Congress has considered a number of legislative proposals designed to deal with the problem, including a cleverly named bill introduced in 2016, the Venue Equity and Non-Uniformity Elimination Act (VENUE Act), which would revise the patent venue statute to spread patent suits more evenly and reduce their concentration in patent-friendly districts.[46]

This is the reason TC Heartland was such an important case. Many believed that by adopting a narrow interpretation of the patent venue statute, the Supreme Court could do something about patent troll filings and case concentration without the need to wait for congressional action. Indeed, it is not much of an exaggeration to say that the patent community viewed TC Heartland as a patent reform case aimed at the patent troll problem.[47]

II. The Court’s Reasoning

Viewed in light of the high stakes for patent law, the Supreme Court’s opinion is surprisingly formalistic and remarkably thin. The Court treats the case as a straightforward exercise in statutory interpretation based on text, and ignores the broader litigation and patent law context. Moreover, the statutory interpretation analysis is unpersuasive even on its own terms.

The opinion for a unanimous court,[48] authored by Justice Thomas, relies on a simple line of argument. Stripped to its core, the argument is that the definition of “resides” in the 1957 Fourco Glass decision still controls because there is no clear indication that Congress intended to change it. The Court’s analysis, however, ignores rather strong evidence that Congress did intend to change it, evidence not only from legislative history but also from the text itself.

The following discussion first reviews the history of the venue provisions critical to the Court’s analysis and then explains how the Court uses and misuses that history to support its holding.

A. Venue History

The TC Heartland opinion begins by laying out the history of the patent venue statutes.[49] Congress adopted the first special venue statute for patent cases in 1897, and it amended that statute in 1948 to codify what is now § 1400(b).[50] Both the original and the amended versions recognized two grounds for patent venue: (1) the district where the defendant is an “inhabitant” (the 1897 version)—which was changed in 1948 to where it “resides”—and (2) any district where the defendant committed acts of infringement and has a regular and established place of business.[51] TC Heartland focuses on the meaning of “resides” in the first ground.[52]

Under the 1897 statute, a corporation was held to be an “inhabitant” of only the district where it was incorporated.[53] When “resides” replaced “inhabitant” in 1948, the question arose whether Congress intended “resides” to have a broader meaning. This question was complicated by the fact that the 1948 revision, in addition to amending § 1400(b), also altered § 1391, the general venue statute, by adding a new provision, § 1391(c). This new provision defined corporate residence to include districts where the corporation was licensed to do business or was doing business—in addition to districts where it was incorporated.[54]

Nine years after the 1948 revision, the Supreme Court addressed this interpretive question in Fourco Glass Co. v. Transmirra Products Corp. In that case, the Court held that the definition of corporate residence in § 1391(c) did not apply to the patent venue statute.[55] Examining the legislative history of the 1948 revision, the Fourco Glass Court concluded that Congress meant only to substitute “resides” for “inhabitant of” and not to make any substantive change.[56] Thus, the definition of “inhabitant” in § 1400(b), which had previously been limited to place of incorporation, carried forward to define “resides” as well.

If the venue statutes were the same today as in 1948, the TC Heartland Court would be justified in following Fourco Glass. But they are not. Congress revised § 1391 in 1988, and again in 2011.[57] In 1988, it amended § 1391(c) to change the definition of residence as follows:

(c) For purposes of venue under this chapter, a defendant that is a corporation shall be deemed to reside in any judicial district in which it is subject to personal jurisdiction at the time the action is commenced. In a State which has more than one judicial district and in which a defendant that is a corporation is subject to personal jurisdiction at the time an action is commenced, such corporation shall be deemed to reside in any district in that State within which its contacts would be sufficient to subject it to personal jurisdiction if that district were a separate State, and, if there is no such district, the corporation shall be deemed to reside in the district within which it has the most significant contacts.[58]

Two years later, the Federal Circuit, in VE Holding Corp. v. Johnson Gas Appliance Co., construed this amendment to overrule Fourco Glass. The Court focused on the preamble, “[f]or purposes of venue under this chapter,”[59] which it characterized as “exact and classic language of incorporation,” indicating a congressional intent to apply § 1391(c)’s definition to § 1400(b).[60] It also noted that the legislative history, while sparse, did not indicate a different intent and that the drafting history supported the view that § 1391(c) applied to patent venue.[61] Finally, the Court reasoned that the result of applying this “plain meaning” brought patent venue “more in line with venue law generally,” fit the legislative trend toward liberalizing venue outside of the patent context, and was consistent with the views of leading authorities.[62]

Congress amended § 1391 again in 2011.[63] In its TC Heartland decision prior to Supreme Court review, the Federal Circuit considered whether these 2011 amendments reflected a congressional decision to overrule its earlier VE Holding decision, and concluded that they did not.[64]

B. The Court’s Use and Misuse of Venue History

One might have expected the TC Heartland Court to support its holding with a careful analysis matching the careful analysis in VE Holding, especially as VE Holding had been the law for twenty-seven years and was decided by a court (the Federal Circuit) with broad power over the development of patent law. However, the Court’s analysis is extremely thin. The Court makes no effort to engage the legislative history or drafting background that influenced the Federal Circuit’s analysis in VE Holding. Instead, it focuses mainly on a single argument, namely, that Congress would have clearly indicated it was overruling Fourco Glass if that were what it intended to do. More precisely, the Court invokes a general proposition: when Congress intends to amend a provision indirectly by amending a different statutory provision, “it ordinarily provides a relatively clear indication of its intent in the text of the amended provision.”[65] There being no such clear indication in the 1988 amendments, the Federal Circuit’s interpretation in VE Holding must fail.

There are several problems with this line of reasoning. First, TC Heartland is not a case where an amendment to one statutory provision is supposed to have amended an entirely separate statutory provision. Sections 1391(c) and 1400(b) are not entirely separate sections. The former defines a term, “resides,” that appears in the latter. Thus, the question is not whether an amendment to one provision—§ 1391(c)—implicitly amends a different provision—§ 1400(b). The question is whether a particular term (“resides”) is subject to a definition appearing elsewhere in the same statute.

Second, the Federal Circuit in VE Holding read Congress to give a relatively clear indication of its intent.[66] Indeed, it concluded that Congress adopted an explicit amendment, not an implicit one. That was, after all, the point of focusing on the preamble to § 1391(c), “[f]or purposes of venue under this chapter.”[67] Evidently, the TC Heartland Court believes this phrase is not clear enough, but it never explains why. Maybe the Court means that Congress should have been more explicit, but it nowhere justifies placing a clear-statement burden like this on Congress.

Third, the Court’s effort to draw on the current version of § 1391 (post-2011 amendments) to support its interpretation also fails. The Court argues that “[t]he current version of § 1391 does not contain any indication that Congress intended to alter the meaning of § 1400(b) as interpreted in Fourco.”[68] But it is not clear why this is relevant. When Congress amended § 1391 in 2011, VE Holding had been the law for more than two decades. Given this, Congress might reasonably have assumed that VE Holding defined the legal baseline and that the 1988 amendments had already overruled Fourco Glass. If so, there would have been no reason for Congress to say anything at all about Fourco Glass in 2011 or signal any intent to change the meaning of § 1400(b).

The Court attributes significance to the fact that the prefatory clause to the current § 1391(c) reads “for all venue purposes,” which is very similar to the phrase “for venue purposes” in place at the time of Fourco Glass.[69] Apparently, the Court believes that this similarity is evidence Congress did not mean to alter the Fourco Glass interpretation.[70] But the Court overlooks the statute’s history.[71] Recall that VE Holding relied on the longer phrase, “[f]or purposes of venue under this chapter,” inserted by the 1988 amendments.[72] When Congress shortened the phrase in 2011 to read “for all venue purposes,” it might have assumed that VE Holding was the law and not meant anything substantive by the change. In fact, Congress kept the longer phrase, “for purposes of venue under this chapter,” for § 1391(d), which is just the equivalent to § 1391(c) for multidistrict states.[73]

The Court also emphasizes the fact that the 2011 amendments inserted a proviso, except as “otherwise provided by law,” into § 1391(a).[74] Referring to this proviso as a “saving clause,” the Court argues that it saves the Fourco Glass interpretation of § 1400(b) because that interpretation counts as “otherwise provided by law.”[75] This argument, however, begs the question. There would be nothing to save if VE Holding was the law—not Fourco Glass. More precisely, the argument works only if Congress in 2011 assumed that Fourco Glass still defined corporate residence for purposes of § 1400(b). But it is at least equally plausible that Congress assumed VE Holding, not Fourco Glass, supplied the definition—especially as federal courts had assumed just that for more than two decades.

My point here is not to defend any particular interpretation of § 1400(b).[76] The Federal Circuit’s interpretation is at least as reasonable as TC Heartland’s on textualist grounds. My point is that the TC Heartland Court offers remarkably thin support for its conclusion.

III. The Problem with a Purposive Interpretation

Given the inadequacy of the Court’s reasoning, one might have expected that at least some of the Justices would have gone beyond text and relied on legislative history, statutory purpose, and the broader patent law context. Indeed, it is possible to construct a purposive interpretation that ties naturally into this broader context. Such an interpretation would start with the general purpose of the patent venue statute, which like all venue statutes, is to promote the “convenience of litigants and witnesses” with special concern for defendants who have “not chosen the forum.”[77] It would then draw on the broader patent law context to argue that a narrow interpretation of “resides,” which breaks up the concentration of cases in districts like the Eastern District of Texas, better serves the venue purpose because it assures greater fairness for defendants.

Admittedly, Justice Thomas, the author of the Court’s opinion, is uncomfortable with a purposive approach, but other Justices who are more comfortable could have written separately. Moreover, given the weakness of the Court’s reasoning, it is surprising that none of them chose to do so. Indeed, one—Justice Breyer—went so far as to question the relevance of the broader context at oral argument.[78]

Yet this choice might be less surprising than it seems. Anyone trying to construct a purposive analysis would have faced some serious problems. These include: (1) uncertainty about fitting statutory interpretation to the purpose of the patent venue statute, (2) uncertainty about the impact of narrower venue options on forum competition and case concentration, and (3) uncertainty about the judiciary’s ability to fashion an optimal solution to the problem. In the end, it is unclear how much TC Heartland will improve patent litigation or advance substantive patent policy. The patent troll problem calls for a more complex solution than the judiciary can provide through statutory interpretation.

A. The Purpose of Venue Rules

One problem with constructing a purposive interpretation has to do with bridging the gap between the general purpose of the venue statute and specific problems of case concentration in particular federal districts.[79] It is easy to state the purpose of venue limitations in general terms of fairness and convenience. It is much more difficult to apply these general norms to determine whether a particular forum qualifies as fair and convenient. General norms are not enough; one needs a more specific rendering of those norms.

For example, the mere fact that a forum burdens a defendant cannot be enough alone to condemn it. The litigation system gives plaintiffs considerable freedom to choose where to sue.[80] Obviously, a plaintiff has an incentive to choose a forum that burdens the defendant. Thus, one cannot condemn the plaintiff’s choice without also condemning the freedom to choose. This means that the unfairness or inconvenience of a specific forum depends not on the mere existence of a burden but on the magnitude of the burden, or more precisely, on the relative balance of burdens and benefits. It is not clear how to strike this balance. In short, we have no generally agreed-upon theory of forum selection that can guide the evaluation of particular forum choices.[81]

Many critics will insist, no doubt, that we do not need a fancy theory to determine that the Eastern District of Texas is a bad venue. But the question is what makes this district so obviously bad. It is not enough to cite case concentration or asymmetric procedural burdens. Case concentration is not always bad; indeed, it can be beneficial when it enables judges to develop expertise in patent law. Moreover, asymmetric procedural burdens are common in all types of litigation. Parties often use pleading, discovery, summary judgment, and other procedures strategically to impose burdens on their opponents, and those burdens are not always reciprocated in equal measure. To be sure, a procedural system that systematically imposes an asymmetric burden should be a matter of concern, but the appropriate level of concern depends on the magnitude of the burden.

If there is reason to worry about the Eastern District, it has to do with the consequences of case concentration and asymmetric burdens and, in particular, how those features encourage patent troll litigation that chills research and development. However, venue rules seem a poor way to solve this problem. As we will see in the following section, adjusting venue can backfire. For example, TC Heartland’s narrow interpretation of residence in § 1400(b) might just redirect many patent infringement suits to the District of Delaware, which according to some commentators, also has a history of competing for patent business with pro-patentee rules.[82] More generally, patent trolls flourish because of a number of perverse features of the patent system, which only a substantive patent law solution can adequately fix.[83]

Finally, amending general venue statutes is not an effective way to correct a forum-specific problem. Venue statutes like § 1391 and § 1400 operate at a high level of generality. They work by identifying forum-related parameters that correlate on average with fair and convenient forums. For example, if the defendant has a regular and established place of business in a district, it is less likely that litigation there will be seriously inconvenient or unfair to that defendant. Moreover, linking venue with personal jurisdiction assures that the forum is one with which the defendant has sufficient contacts to make it fair, just, and reasonable to defend there.[84]

Even so, a district that is fair and convenient for the average case might not be fair and convenient for a particular case. However, the venue statutes already contemplate this possibility. Section 1404(a) gives district judges discretionary authority to transfer a case from a district that has venue to another district where it “might have been brought” when doing so serves the “convenience of parties and witnesses” and the “interest of justice.”[85] To be sure, one complaint about the Eastern District of Texas is that judges often delay or deny transfer motions that they should grant.[86] But the solution to this problem is to change judicial behavior, not venue rules. Indeed, if judges are willing to ignore the law, the judicial system has much more to worry about than patent trolls.

Nevertheless, there might be a good reason to modify general venue statutes if defendants file transfer motions frequently enough. In that case, adjusting the general venue statute could reduce the frequency and cost of these motions and improve the efficiency of venue determinations.[87] However, this sort of reform requires a great deal of empirical information and fact-sensitive analysis, which makes it poorly suited to judicial implementation. Congress would seem the superior lawmaking institution for this purpose.

B. Effect on Case Concentration and Forum Competition

Suppose one concludes that districts where patent infringement suits concentrate, such as the Eastern District of Texas and District of Delaware, are unfair and inconvenient. To make a convincing purposive argument for TC Heartland’s narrow interpretation, one must still show that a narrow interpretation will substantially alleviate problematic case concentration. There are two reasons to question how much the TC Heartland decision will do in this regard. First, the Court’s decision might simply change the locus of case concentration rather than significantly reduce it. Second, the decision might encourage patent plaintiffs to switch to the second prong of § 1400(b). The following discussion addresses each of these possibilities in turn.

1. Shifting the Locus of Case Concentration.—TC Heartland’s interpretation of § 1400(b) might simply shift case concentration to other districts, especially the District of Delaware. Because many corporate defendants are incorporated in Delaware, equating “reside” with place of incorporation should increase case filings in Delaware, a district that some commentators have argued has a history of competing for patent infringement suits by offering pro-plaintiff procedures.[88] Apparently, some judges in the District of Delaware are moving in the direction of less biased procedures.[89] But if Delaware is inclined to compete for patent business, it is unclear how these judges will respond if TC Heartland strips the Eastern District of Texas of its market dominance.[90]

An empirical study by Professors Colleen V. Chien and Michael Risch supports this prediction. Their model forecasts that the Eastern District of Texas will likely sustain a substantial loss in its share of patent cases nationwide, from 44% to 14.7%, and the District of Delaware will enjoy an increase, from 9% to 23.8%.[91] Moreover, available data for the month of June 2017, which is after the TC Heartland decision, shows a significant reduction in patent filings in the Eastern District and a sharp increase in the District of Delaware.[92] The Chien–Risch study also predicts an overall distribution of patent cases that features the District of Delaware as the most popular venue (capturing 23.8% of all cases) followed by the Eastern District of Texas (14.7%), the Northern District of California (13.0%), and the Central District of California (6.1%).[93]

However, this does not necessarily mean that districts will compete for patent cases as vigorously as before TC Heartland.[94] VE Holding’s broad interpretation of § 1400(b) made lots of federal districts available to patent plaintiffs. As a result, districts could compete for the same cases, and each had a chance to capture the bulk of patent litigation. After TC Heartland, the residence provision of § 1400(b) limits venue to the defendant’s place of incorporation. Since most companies have only one place of incorporation and since that place is often Delaware, federal districts outside of the District of Delaware will have many fewer cases to capture and thus a presumably weaker incentive to compete.

This analysis focuses only on residence-based venue. Section 1400(b) also creates venue in any district where the defendant has a regular and established place of business and has committed acts of infringement.[95] For suits against large companies operating nationwide, this provision can open up a number of additional forum options. This is especially true for the digital technology cases that attract patent trolls. Many large computer, Internet, and software companies have regular and established places of business in several districts and are likely to have committed acts of infringement there.[96] This should increase the number of districts that can compete, which might lead to more vigorous competition and more pro-plaintiff procedures.

2. Switching to the Other Prong of Section 1400(b).—If patentees cannot use the residence prong of § 1400(b) to secure venue in the Eastern District of Texas, they will probably turn to the “regular and established place of business” prong. Moreover, Eastern District judges are likely to be receptive to this strategy and strongly disinclined to grant motions to dismiss or transfer. To be sure, the plaintiff must make a colorable argument that the defendant has a regular and established place of business in the district and committed acts of infringement there. However, these requirements are not all that difficult to satisfy.

At the time that TC Heartland was decided, there was considerable uncertainty about what qualifies as a “regular and established place of business.”[97] Under the then-existing precedent, it was possible to argue for venue based on rather slim connections between the defendant and the district. For example, a relatively small presence, such as a small store owned by the defendant or even a warehouse or supply center, might suffice.[98] Indeed, there was precedent at the time that even a salesperson working out of a home office in a district could create a regular and established place of business there.[99]

As for the second requirement—acts of infringement in the district—the Patent Act defines infringement broadly to include making, using, offering for sale, or selling the patented invention.[100] Even just one sale in the district can support venue, as long as the defendant is responsible for the sale.[101]

Thus, it would be reasonable for the TC Heartland Justices to have assumed that plaintiffs might be successful in keeping many suits in the Eastern District of Texas, regardless of what the Court decided. For example, Apple, Inc., a frequent defendant in patent troll suits, has stores in the Eastern District.[102] An Apple store surely qualifies as a “regular and established place of business,” and sales of allegedly infringing articles from the store would almost certainly constitute “acts of infringement in the district.” Indeed, in the wake of TC Heartland, a number of commentators predicted that plaintiffs bent on keeping cases in the Eastern District of Texas would switch to the second prong of § 1400(b).[103] And this appears to be exactly what is happening.[104]

In addition, the TC Heartland Justices could have assumed, quite reasonably, that Eastern District judges would interpret § 1400(b)’s second prong broadly. In fact, this is what happened about a month after the TC Heartland decision. Judge Gilstrap, the Eastern District judge with the most patent cases, upheld venue based on a sales representative operating from his home in the Eastern District.[105] In his opinion, Judge Gilstrap also took the opportunity to lay out a flexible balancing test for determining “regular and established place of business,” a test capable of supporting broad exercises of venue.

Since our purpose is to explain why the TC Heartland Court ignored the broader patent law context, the relevant timeframe is the period just before the TC Heartland decision. Nevertheless, it is worth noting that the Federal Circuit, in a September decision, vacated Judge Gilstrap’s broad venue ruling mentioned above, and in so doing limited the scope of § 1400(b)’s second prong.[106] While conceding the possibility that a salesperson operating from a personal residence might sometimes create a regular and established place of business, the Federal Circuit made clear that the place of business must be a physical location established or ratified by the defendant and a place where the defendant conducts business in a regular and stable way.[107] A careful analysis of the Federal Circuit’s decision is beyond the scope of this Essay, but it is reasonable to suppose that the decision will reduce the number of suits filed in the Eastern District. By how much remains to be seen.[108]

C. Limitations on Judicial Intervention

As a practical matter, the Supreme Court in TC Heartland was limited to two options: adopt VE Holding’s interpretation of the venue statute or Fourco Glass’s. Neither choice was optimal. The VE Holding interpretation would continue the existing interdistrict competition, and the Fourco Glass interpretation might just redirect that competition to different districts. Other, more promising approaches exist, but they are for Congress to implement, not the Court.[109] Moreover, Congress is better positioned than the Court to weigh the costs and benefits of alternative solutions. Given all of this, it would have been quite reasonable for the Justices to defer to Congress rather than try to address the problems through the TC Heartland case.[110]

It would also have been reasonable for the Justices to assume that Congress might trump any decision they reached, especially with the VENUE Act pending.[111] Given this possibility, getting the decision right might not have seemed quite so pressing. There is a rub, however. The Court’s interpretation can affect the likelihood of congressional action. In particular, by adopting a narrow interpretation, the TC Heartland Court has reduced congressional pressure to amend § 1400(b). Big technology companies that are the targets of PAE suits will have weaker incentives to push for congressional action if they believe the Court’s holding will substantially reduce case concentration and pro-plaintiff bias. Patent trolls might lobby for reinstatement of VE Holding, but they are unlikely to have much success given the current political climate and general hostility to patent trolls.


This Essay began with a question. Given the generally accepted and widely publicized link between the venue issue in TC Heartland and the patent troll problem, why did the Court’s opinion completely ignore the broader patent law context? This omission is especially puzzling because the Court’s reasoning is so thin and unpersuasive. Indeed, at first glance, a purposive interpretation drawing on the broader patent law context would seem to provide stronger support.

On closer examination, however, it is not readily apparent how to construct a convincing purposive interpretation. One problem lies in building a connection between case concentration and purpose. Another problem stems from uncertainty about the effect of the Court’s interpretation on case concentration. And a third problem has to do with the Court’s limited ability to assess the relevant empirics, which makes it sensible to leave the issue to Congress.

Thus, the Court’s unanimous support for a formalistic opinion with a thin textualist justification might be less surprising than it seems. Those Justices with a pragmatic bent were probably quite pleased with a result that made it harder for patent trolls to sue in their favorite districts, and they had no obvious way to strengthen the reasoning. Perhaps endorsing VE Holding’s broad interpretation of residence would have made Congressional action more likely, but that is highly speculative. In the end, TC Heartland might be good enough after all, despite its thin rationale.

  1. .The three cases are: BNSF Ry. Co. v. Tyrrell, 137 S. Ct. 1549 (2017) (general jurisdiction over the person); Bristol-Myers Squibb Co. v. Super. Ct. of Cal., S.F. Cty., 137 S. Ct. 1773 (2017) (specific jurisdiction over the person), and TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514 (2017) (patent venue).
  2. .137 S. Ct. 1514 (2017).
  3. .Id. at 1517.
  4. .In re TC Heartland LLC, 821 F.3d 1338, 1340 (Fed. Cir. 2016).
  5. .Id. TC Heartland also moved to dismiss under Rule 12(b)(2) for lack of jurisdiction over its person. Id. The district court rejected this ground for dismissal, and the Federal Circuit affirmed. Id. at 1341. The Supreme Court did not address the personal jurisdiction issue. See generally TC Heartland, 137 S. Ct. 1514.
  6. .TC Heartland, 137 S. Ct. at 1517. While TC Heartland is a limited liability company, the Supreme Court decided the venue issue as if it were a corporation because that is how the parties presented the case. See id. at 1517 n.1 (“Because this case comes to us at the pleading stage and has been litigated on the understanding that petitioner is a corporation, we confine our analysis to the proper venue for corporations. We leave further consideration of the issue of petitioner’s legal status to the courts below on remand.”).
  7. .Id. at 1517; In re TC Heartland, 821 F.3d at 1340.
  8. .In re TC Heartland, 821 F.3d at 1340 (noting that this amounted to about 2% of TC Heartland’s total 2013 sales).
  9. .See id. at 1340–41 (explaining that congressional amendments to the patent venue statute did not undo prior Federal Circuit precedent).
  10. .Id. at 1341.
  11. .TC Heartland, 137 S. Ct. at 1518.
  12. .28 U.S.C. § 1400(b) (2012).
  13. .See In re TC Heartland, 821 F.3d at 1340–41 (summarizing the case’s procedural history).
  14. .917 F.2d 1574 (Fed. Cir. 1990).
  15. .TC Heartland, 137 S. Ct. at 1517–20.
  16. .28 U.S.C. § 1391(c)(2) (2012). Section 1391(d) applies to cases in multidistrict states. 28 U.S.C. § 1391(d).
  17. .353 U.S. 222 (1957).
  18. .Id. at 223–24, 228–29.
  19. .TC Heartland, 137 S. Ct. at 1520–21.
  20. .Id. at 1517 n.1, 1520 n.2; see Maxchief Inv. Ltd. v. Plastic Dev. Grp., LLC, No. 3:16-cv-63, 2017 WL 3479504, at *2 (E.D. Tenn. Aug. 14, 2017) (applying TC Heartland’s definition of residence to unincorporated associations).
  21. .See TC Heartland LLC v. Kraft Foods Group Brands LLC, SCOTUSblog, [] (listing the amicus briefs that were filed).
  22. .See Brief for the State of Texas et al. as Amici Curiae Supporting Petitioner at 2, TC Heartland LLC v. Kraft Food Brands Grp. LLC, 137 S. Ct. 1514 (2017) (No. 16-341) (arguing that the Federal Circuit erred in departing from the Supreme Court’s holding that 28 U.S.C. § 1400(b) is the exclusive statute governing venue over corporations in patent cases). Arizona, Colorado, Connecticut, Hawai’i, Illinois, Iowa, Maine, Maryland, Michigan, Nebraska, North Carolina, Ohio, South Carolina, Vermont, Virginia, and Wisconsin joined the brief. Id. at 1.
  23. .Adam Liptak, Supreme Court Considers Why Patent Trolls Love Texas, N.Y. Times (Mar. 27, 2017), []. Moreover, although it predates the Federal Circuit decision in TC Heartland, it is worth mentioning that comedian John Oliver did a segment on patent trolls as part of his HBO show in April 2015, which included a discussion of the concentration of lawsuits in Marshall, Texas. Patents: Last Week Tonight with John Oliver, YouTube (Apr. 19, 2015), [https://perma
  24. .For a description of patent trolls and the patent troll strategy, see infra notes 40–46 and accompanying text.
  25. .See Daniel Klerman & Greg Reilly, Forum Selling, 89 S. Cal. L. Rev. 241, 242 (2016) (“When plaintiffs have a wide choice of forum, . . . judges have incentives to make the law more pro-plaintiff because plaintiffs choose the court with the most pro-plaintiff law and procedures.”). As all litigators know, choice of forum can have a major effect on outcome, which is why parties invest a lot in battling over where a suit is litigated. Kevin M. Clermont & Theodore Eisenberg, Litigation Realities, 88 Cornell L. Rev. 119, 121 (2002) (noting that “[t]he name of the game is forum-shopping” and that “[f]orum is worth fighting over because outcome often turns on forum”).
  26. .There are three requirements that must be satisfied for a federal district court to be a proper forum: subject matter jurisdiction, jurisdiction over the person, and venue. Sections 1331 and 1338 each confer federal subject matter jurisdiction over a patent infringement suit. 28 U.S.C. §§ 1331, 1338 (2012). This leaves personal jurisdiction and venue.
  27. .21 F.3d 1558 (Fed. Cir. 1994).
  28. .Id. at 1564, 1572; see 8 Donald S. Chisum, Chisum on Patents § 21.02[3][a][i] (2017). Rule 4(k)(1)(A) of the Federal Rules of Civil Procedure authorizes personal jurisdiction over a defendant “who is subject to the jurisdiction of a court of general jurisdiction in the state where the district court is located.” Fed. R. Civ. P. 4(k)(1)(A). This means that the exercise of jurisdiction must comply with the state long-arm statute and the Fourteenth Amendment’s Due Process Clause. In Beverly Hills Fan, the Federal Circuit applied federal law to give the state long-arm statute a relatively broad reach. Beverly Hills Fan, 21 F.3d at 1571. As for the due process analysis, the Court held that Federal Circuit law applies rather than the law of the circuit in which the district court sits, and that Federal Circuit law endorses a broad stream-of-commerce theory. Id. at 1564–65 (“The creation and application of a uniform body of Federal Circuit law in this area would clearly promote judicial efficiency, would be consistent with our mandate, and would not create undue conflict and confusion at the district court level.”).
  29. .917 F.2d 1574 (Fed. Cir. 1990).
  30. .At least for single-district states. Id. at 1580.
  31. .Substantive patent law offers little reason to forum shop because it is controlled mostly by the Patent Act and Federal Circuit and Supreme Court decisions and thus is relatively uniform nationwide. See J. Jonas Anderson, Court Competition for Patent Cases, 163 U. Pa. L. Rev. 631, 634, 684 (2015) (noting that the “uniformity of patent law throughout the country forces forum-shopping plaintiffs to seek out advantageous case-management norms and procedural differences”).
  32. .See Anderson, supra note 31, at 649–61 (describing the competition for patent cases in a number of federal districts); Klerman & Reilly, supra note 25, at 243 (discussing how judges in the Eastern District of Texas “have distorted the rules and practices relating to case assignment, joinder, discovery, transfer, and summary judgment in a pro-patentee (plaintiff) direction”). There are a number of reasons why federal judges might want to attract patent litigation. See Klerman & Reilly, supra note 25, at 270–77 (discussing some reasons, including the challenge offered by patent suits, the reputational opportunities from specializing in patent litigation, the economic benefits for the local community, and the professional benefits for the local bar).
  33. .Klerman & Reilly, supra note 25, at 248–49. Defendants tried to escape these pro-plaintiff districts by filing motions to transfer, but Eastern District judges tended to delay or deny these motions. See Klerman & Reilly, supra note 25, at 260–63; Brian J. Love & James Yoon, Predictably Expensive: A Critical Look at Patent Litigation in the Eastern District of Texas, 20 Stan. Tech. L. Rev. 1, 22–23 (2017) (noting the differences across districts for rulings on motions to transfer); see also 15 Charles Alan Wright et al., Federal Practice and Procedure § 3848 (4th ed. 2013) (discussing the presumption in favor of plaintiff’s choice of forum).
  34. .Love & Yoon, supra note 33, at 8; see also Mathew Sag, IP Litigation in U.S. District Courts: 1994–2014, 101 Iowa L. Rev. 1065, 1096–99 (2016) (documenting the “remarkable ascendancy” of patent litigation in the Eastern District of Texas between 1994 and 2014 and noting that “but for the Eastern District of Texas and the District of Delaware, the geographic distribution of patent litigation over the past two decades would look remarkably stable”).
  35. .Love & Yoon, supra note 33, at 8.
  36. .Id. at 6 (“[O]ne judge—Judge Rodney Gilstrap of Marshall, Texas—saw almost one quarter of all patent case filings nationwide . . . , more than all the federal judges in California, New York, and Florida combined.”). In 2015 alone, Judge Gilstrap was assigned 1,686 patent cases. Jonas Anderson, Judge Shopping in the Eastern District of Texas, 48 Loy. U. Chi. L.J. 539, 539 (2016).
  37. .See Marshall, Texas, City-Data, [] (noting a population of 24,701 in 2014).
  38. .For a detailed description of these and other pro-plaintiff procedures, see Klerman & Reilly, supra note 25, at 251–70, and Love & Yoon, supra note 33, at 15–25.
  39. .Klerman & Reilly, supra note 25, at 254, 260–61.
  40. .See generally Mark A. Lemley & A. Douglas Melamed, Missing the Forest for the Trolls, 113 Colum. L. Rev. 2117, 2118–46 (2013) (discussing patent troll business models, including settlement techniques employed by patent trolls).
  41. .Id. at 2120, 2124, 2126; see generally Robert P. Merges, The Trouble With Trolls: Innovation, Rent-Seeking, and Patent Law Reform, 24 Berkeley Tech. L.J. 1583, 1587–88, 1591, 1599–1600 (2009) (discussing problems created by patent trolls).
  42. .Lemley & Melamed, supra note 40, at 2123.
  43. .For the period covering January 2014 through June 2016, about 93.9% of the patent infringement cases filed in the Eastern District of Texas were filed by patent assertion entities. Love & Yoon, supra note 33, at 9.
  44. .See Lemley & Melamed, supra note 40, at 2124–25 (listing costs imposed by patent trolls that discourage innovation). It is worth mentioning that not all commentators are hostile to PAEs or patent trolls. See, e.g., David L. Schwartz & Jay P. Kesan, Analyzing the Role of Non Practicing Entities in the Patent System, 99 Cornell L. Rev. 425, 427 (2014); James F. McDonough III, Comment, The Myth of the Patent Troll: An Alternative View of the Function of Patent Dealers in an Idea Economy, 56 Emory L.J. 189, 223 (2006).
  45. .See Lemley & Melamed, supra note 40, at 2118–19 (cataloging public and private entities that have publicized the patent troll problem or taken action against patent trolls).
  46. .Venue Equity and Non-Uniformity Elimination Act of 2016, S.2733, 114th Cong. (2016) (pending). For a description of the bill, see Colleen V. Chien & Michael Risch, Recalibrating Patent Venue, 77 Md. L. Rev. (forthcoming 2018) (manuscript at 23–24).
  47. .Many in the media characterized the Supreme Court’s decision as a blow to patent trolls. See, e.g., Brian Fung, The Supreme Court’s Big Ruling on ‘Patent Trolls’ will Rock Businesses Everywhere, Wash. Post (May 23, 2017),
    wp/2017/05/23/the-supreme-court-just-undercut-patent-trolls-in-a-big-way []; Adam Liptak, Supreme Court Ruling Could Hinder ‘Patent Trolls’, N.Y. Times (May 22, 2017), [].
  48. .Justice Gorsuch did not participate in the decision. TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514, 1521 (2017).
  49. .Id. at 1518 (noting that the statutory history is “important context for the issue in this case”).
  50. .Id.
  51. .Id. at 1518–19.
  52. .Id. at 1517.
  53. .Id. at 1518.
  54. .28 U.S.C. § 1391(c) (1948) (“A corporation may be sued in any judicial district in which it is incorporated or licensed to do business or is doing business, and such judicial district shall be regarded as the residence of such corporation for venue purposes.”) (amended 1988).
  55. .Fourco Glass Co. v. Transmirra Prods. Corp., 353 U.S. 222, 229 (1957). The Court relied on its earlier decision in Stonite Products Co. v. Melvin Lloyd Co., 315 U.S. 561 (1942), which treated the patent venue statute as completely independent of the general venue statute. Fourco Glass, 353 U.S. at 225.
  56. .Fourco Glass, 353 U.S. at 226–28.
  57. .Federal Courts Jurisdiction and Venue Clarification Act of 2011, Pub. L. No. 112-63, § 202, 125 Stat. 758, 763; Judicial Improvements and Access to Justice Act, Pub. L. No. 100-702, § 1013, 102 Stat. 4642, 4669 (1988).
  58. .Judicial Improvements and Access to Justice Act § 1013(a).
  59. .VE Holding Corp. v. Johnson Gas Appliance Co., 917 F.2d 1574, 1580 (Fed. Cir. 1990). The words “this chapter” referred to Chapter 87 of Title 28, which included § 1400(b), the patent venue statute. Id.
  60. .Id. at 1579; see also id. at 1580 (“In the case before us, the language of the statute is clear and its meaning is unambiguous.”).
  61. .Id. at 1581–82. The 1988 amendments were based in large part on recommendations by the Judicial Conference Committee on Court Administration. Notably, Professor Edward Cooper, the Reporter of the subcommittee responsible for the proposal that became § 1391(c), strongly suggested in a December 4, 1986 memorandum to the subcommittee that the new definition of corporate residence applied to all the venue provisions in Chapter 87. Id. at 1582; see Paul R. Gugliuzza & Megan M. La Belle, The Patently Unexceptional Venue Statute, 66 Am. U. L. Rev. 1027, 1047–49 (2017) (also arguing that changes in the prefatory clause and other revisions to the statute support VE Holding).
  62. .VE Holding, 917 F.2d at 1583–84.
  63. .Federal Courts Jurisdiction and Venue Clarification Act of 2011, Pub. L. No. 112-63, § 202, 125 Stat. 758, 763–64 (codified as amended at 28 U.S.C. § 1391 (2012)).
  64. .In re TC Heartland LLC, 821 F.3d 1338, 1341 (Fed. Cir. 2016) (characterizing the 2011 amendments as “minor”).
  65. .TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514, 1520 (2017).
  66. .See supra notes 59–62 and accompanying text (discussing how the Federal Circuit interpreted congressional intent in VE Holding).
  67. .See supra notes 59–60 and accompanying text. It is also worth mentioning that Congress converted § 1391(c) from a substantive venue provision with a definition into a purely definitional section. See Gugliuzza & La Belle, supra note 61, at 1048–49. This change further supports the conclusion that Congress meant the § 1391(c) definition to apply. See id. After all, it would be perfectly sensible for someone seeking a definition of “reside” in § 1400(b) to look to a purely definitional section, and therefore reasonable as well to suppose that Congress might have contemplated that result. I am indebted to Professor Gugliuzza for alerting me to this point.
  68. .TC Heartland, 137 S. Ct. at 1520.
  69. .Id. at 1520–21.
  70. .Id.
  71. .It also gives insufficient weight to the word “all” in the current statute. “All” suggests a comprehensive application. The Court simply asserts that “for venue purposes” is as comprehensive as “[f]or all venue purposes.” Id. This is an embarrassing move for a textualist.
  72. .See supra notes 59–60 and accompanying text.
  73. .In 2011, Congress divided § 1391(c) into two parts—§ 1391(c) still defines corporate residence for single-district states, and the new § 1391(d) defines corporate residence for multidistrict states. Federal Courts Jurisdiction and Venue Clarification Act of 2011, Pub. L. No. 112-63, § 202, 125 Stat. 758, 763–64 (codified as amended at 28 U.S.C. §§ 1391(c)–1391(d) (2012)). There is no reason why Congress would have wanted a narrower definition of corporate residence for single-district states than for multidistrict states.
  74. .TC Heartland, 137 S. Ct. at 1521.
  75. .Id.
  76. .For a strong argument that VE Holding’s interpretation is the correct one, see Gugliuzza & La Belle, supra note 61, at 1046–52.
  77. .Wright et al., supra note 33, § 3801; accord Leroy v. Great W. United Corp., 443 U.S. 173, 183–84 (1979) (explaining that venue rules “protect the defendant against the risk that a plaintiff will select an unfair or inconvenient place of trial”).
  78. .See Transcript of Oral Argument at 14–15, TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514 (2017) (No. 16-341),
    argument_transcripts/2016/16-341_8njq.pdf [] (asking “what’s this got to do with this?” where the first “this” refers to case concentration, questioning the “relevance” of case concentration, and later cutting off counsel’s effort to discuss case concentration by saying “might be other people are interested in that”). Other Justices did ask questions about the Eastern District of Texas later in oral argument, indicating that they were at least aware of the broader policy concerns. Id. at 43–48.
  79. .Another problem has to do with how to characterize statutory purpose. For example, some commentators argue that Congress’s purpose in adopting the first patent venue statute in 1897 was to favor patent plaintiffs by giving them broader venue options than plaintiffs bringing other federal-question cases had at the time. See Gugliuzza & La Belle, supra note 61, at 1035–36 (arguing that because the 1897 venue statute allowed a plaintiff to sue a defendant in any district in which the defendant committed acts of infringement and had a regular place of business, it may be appropriate to interpret it as affording plaintiffs broad forum options). One could argue that Congress has not changed its original purpose and that the current patent venue statute should therefore be construed broadly to further that purpose, which means applying § 1391(c)’s definition of resides. See id. at 1052–53 (presenting a similar argument). My point is not to endorse this argument. My point is that any purposive argument must begin with a characterization of congressional purpose. I am grateful to Patrick Woolley for alerting me to this point.
  80. .See Wright et al., supra note 33, § 3848 (reviewing the various judicial formulations of the degree of deference given to the plaintiff’s forum choice).
  81. .A theory of this sort should be able to explain what constitutes an optimal forum, how much choice plaintiffs should have in forum selection, when defendants should be able to trump plaintiff choice, and how much deference contractual forum selection should receive. Some scholars have made efforts along these lines. See, e.g., Daniel Klerman, Rethinking Personal Jurisdiction, 6 J. Legal Analysis 245, 245–47, 249 (2014) (arguing that personal jurisdiction rules, which limit forum selection, should aim to minimize the sum of litigation costs and error costs).
  82. .It seems that the District of Delaware’s venue competition is not limited to patent cases. Apparently, it has also been an aggressive competitor for large corporate bankruptcy cases. Klerman & Reilly, supra note 25, at 291–96.
  83. .These features include a multiplicity of broad patents on small improvements, fragmented patent ownership, excessively generous patent damages rules, and high costs of patent litigation. Lemley & Melamed, supra note 40, at 2172–76.
  84. .See Int’l Shoe Co. v. Washington, 326 U.S. 310, 316–17 (1945) (referring to “traditional notions of fair play and substantial justice” and to what is “reasonable, in the context of our federal system of government”). This is one reason why it is difficult to condemn VE Holding’s broad interpretation of § 1400(b) on venue-policy grounds, since it links patent venue with personal jurisdiction. If this allows plaintiffs to make unfair forum choices, perhaps the problem lies with broad stream-of-commerce jurisdiction rather than with venue.
  85. .28 U.S.C. § 1404(a) (2012).
  86. .Chien & Risch, supra note 46, at 18–19; Klerman & Reilly, supra note 25, at 260–63.
  87. .The two-part structure of venue rules—a general venue rule coupled with a case-sensitive standard permitting exceptions—is efficient as long as most cases are handled by the general rule. However, if too many cases require exceptions, then a more efficient cost–benefit balance might be achieved by adjusting the general rule.
  88. .See Klerman & Reilly, supra note 25, at 281–82 (arguing that “personal gain [by judges entering private practice] may be a motive for attracting patent litigation” to the District of Delaware). In recent years, the District of Delaware has ranked second only to the Eastern District of Texas as a venue for patent suits in general and suits by nonpracticing entities in particular. Chien & Risch, supra note 46, at 13, 26 n.118. In fact, Kraft sued in the District of Delaware. TC Heartland LLC v. Kraft Foods Grp. LLC, 137 S. Ct. 1514, 1515 (2017).
  89. .Klerman & Reilly, supra note 25, at 282–83.
  90. .The incentives to adopt pro-plaintiff procedures depend on the expected benefits and costs of doing so. With the Eastern District of Texas dominating the market and presumably willing to counter any serious competition threat, the District of Delaware would not have expected to benefit as much from pro-plaintiff procedures as it did before the Eastern District’s market dominance. This might be a reason for its reversal of course. If so, one might expect a shift back toward pro-plaintiff procedures and more vigorous competition with the Eastern District’s grip on the market weakened by TC Heartland.
  91. .Chien & Risch, supra note 46, at 37. Moreover, the authors predict that the proportion of cases filed by PAEs will decline from 64.1% to 19.0% in the Eastern District and rise from 7.3% to 25.8% in the District of Delaware. Id. at 35.
  92. .Bloomberg Law reports a 47% reduction for the Eastern District compared to May 2017 and a 62% reduction compared to June 2016. Malathi Nayak & Peter Leung, Ruling Could Halt Drop in Texas Court Patent Complaint Filings, Bloomberg BNA (July 10, 2017), []. It also reports an increase in filings for the District of Delaware from 36 filings in May 2017 to 66 filings in June (although the authors point out that this is less than the same figure one year earlier, in May 2016, which is consistent with a general reduction in patent suit filings nationwide). Id.; see also Malathi Nayak, Swelling Docket Pushing Delaware Judges to Transfer Patent Cases, Bloomberg BNA (Sept. 20, 2017), [] (reporting that 79 cases were filed in the month of August and that the resulting case congestion is prompting judges in the “shorthanded” District of Delaware, which has two vacancies and only two “active judges,” to transfer patent cases elsewhere).
  93. .Chien & Risch, supra note 46, at 37.
  94. .See supra notes 31–39 and accompanying text. Some commentators raise a different concern. They worry that more cases will be litigated in districts favorable to defendants, thereby creating the opposite unfairness concern. Adam Mossoff, ‘Examining the Supreme Court’s TC Heartland Decision’: Testimony Before the House Judiciary Committee, Subcommittee on Courts, IP, and the Internet 7–8 (George Mason Law & Economics Research Paper No. 17-29, 2017), []. This is probably not a big problem for TC Heartland, however. Local judges and juries might be favorably predisposed to defendants that are major employers and operate economically substantial facilities in the district, but I doubt they care much about incorporation.
  95. .28 U.S.C. § 1400(b) (2012 ) .
  96. .See Chien & Risch, supra note 46, at 40 (describing the large number of companies headquartered in Northern and Central California). Apple, Google, and Yahoo, for example, are all located in the Northern District of California, and Apple owns stores across the country. See Kathy C. Leong, Apple’s Eye-Catching New Home Disrupts Silicon Valley, Boston Globe (July 5, 2017), [] (identifying Apple’s headquarters in Cupertino, California); see also Mike McPhate, California Today: Google’s Idea for a New Silicon Valley, N.Y. Times (June 20, 2017), [] (identifying Google’s headquarters in Mountain View, California).
  97. .See In re Cray, Inc., 871 F.3d 1355, 1359 (Fed. Cir. 2017) (observing that trial courts after TC Heartland “have noted the uncertainty surrounding and the need for greater uniformity on this issue [i.e., the issue of what constitutes a regular and established place of business]”). One reason for this uncertainty has to do with the paucity of case law construing the requirement. After VE Holding expanded the residence prong, few patent plaintiffs relied on the “regular and established place of business” prong, so courts had little need to address its meaning in the new digital and Internet age. See id. (noting the change in business practices from the “brick-and-mortar model”).
  98. .See Chisum, supra note 28, § 21.02[2][d] (2017) (“Generally, any physical location at which business is conducted will suffice, no matter what the amount or character of the activity.”); In re Cordis Corp., 769 F.2d 733, 737 (Fed. Cir. 1985) (noting, while denying mandamus petition, that “the appropriate inquiry is whether the corporate defendant does its business in that district through a permanent and continuous presence”). Moreover, although the matter is contested, there is precedent for the rule that the place of business need not have any relationship with the infringement alleged. Chisum, supra note 28, § 21.02[2][d].
  99. .See, e.g., In re Cordis Corp., 769 F.2d at 735–37 (holding in connection with denial of mandamus petition that employees working from home qualified as a regular and established place of business and that the appropriate inquiry is not whether there is a fixed formal office in the district); Shelter-Lite, Inc. v. Reeves Bros., Inc., 356 F. Supp. 189, 195 (N.D. Ohio 1973) (“[A]n unyielding rule that a regular and established place of business cannot arise by virtue of a salesman operating out of his residence is at odds with the practicalities and necessities of the business community.”).
  100. .35 U.S.C. § 271(a) (2012).
  101. .See Chisum, supra note 28, § 21.02[2][e][i] (“Any sale of an accused product within the district will meet the act-of-infringement requirement of Section 1400(b).”). However, a defendant does not commit an act of infringement in the district simply by selling to an intermediary outside the district, who then resells the allegedly infringing product in the district. See id. (discussing the consummated sale doctrine, which states that there is no act of infringement unless the defendant completes a sale within the district).
  102. .Apple has stores in Plano and Frisco, both of which fall within the Eastern District of Texas. Jan Wolfe, Patent Plaintiffs See Way Around U.S. Supreme Court Ruling,
    Reuters (May 23, 2017), [].
  103. .Gene Quinn, Industry Reaction to SCOTUS Patent Venue Decision in TC Heartland v. Kraft Food Group, IP Watchdog (May 22, 2017),
    industry-reaction-scotus-patent-venue-decision-tc-heartland-v-kraft-food-group/id=83518/ [https:
    //]; see Wolfe, supra note 102. Also, some have predicted an increase in suits against independent retailers or even the addition of independent retailers to suits against manufacturers or distributors in an effort to keep the latter in the district. Id. (reporting expert opinions that manufacturers or distributors may have to indemnify retailers who sell infringing material).
  104. .See In re Cray, Inc., 871 F.3d 1355, 1359 (Fed. Cir. 2017) (“Following the Supreme Court’s recent decision in TC Heartland, litigants and courts are raising with increased frequency the question of where a defendant has a ‘regular and established place of business.’”).
  105. .Raytheon Co. v. Cray, Inc., No. 2:15-CV-01554-JRG, 2017 WL 2813896, at *7–8 (E.D. Tex. June 29, 2017), vacated by In re Cray, Inc., 871 F.3d 1355 (Fed. Cir. 2017). Judge Gilstrap held that the “regular and established place of business” requirement was met in the case because Cray employed a sales representative in the Eastern District of Texas, id. at *7–8, and that the acts-of-infringement requirement was met because Cray induced infringement in the Eastern District by selling a supercomputer to the University of Texas at Austin, which was used by researchers at university branches located in the Eastern District, id. at *5.
  106. .In re Cray, Inc., 871 F.3d 1355, 1366 (Fed. Cir. 2017).
  107. .Id. at 1362-63.
  108. .As mentioned above, several of the typical computer-company defendants in patent troll suits have stores or other places of business in the Eastern District. Also, plaintiffs can still argue for personal jurisdiction as the basis for venue for unincorporated entities, such as LLCs, and foreign corporations. The TC Heartland Court was clear that it only addressed venue for domestic corporations. TC Heartland LLC v. Kraft Foods Grp. Brands LLC, 137 S. Ct. 1514, 1517 n.1, 1520 n.2 (2017). But see Maxchief Inv. Ltd. v. Plastic Dev. Grp., LLC, No. 3:16-cv-63, 2017 WL 3479504, at *2 (E.D. Tenn. Aug. 14, 2017) (applying TC Heartland’s definition of residence to unincorporated associations).
  109. .See, e.g., Klerman & Reilly, supra note 25, at 303–05 (discussing various solutions and defending their own).
  110. .See Gugliuzza & La Belle, supra note 61, at 1056 (“Simply put, although the VENUE Act may not be the perfect solution to forum shopping in patent cases, putting this problem in Congress’s hands makes more sense than resorting to a questionable interpretation of the venue statute that could have unintended consequences beyond patent litigation.”).
  111. .See supra note 46 and accompanying text.