Rethinking the Public Interest in Consumer Protection A Critical Comparative Analysis of Article 6 Rome I Regulation

Consumer protection directly influences the design of choice of law rules in the EU. Article 6 Rome I Regulation stipulates that the law of the consumer’s habitual place of residence applies, unless another law has been chosen. This choice may not deprive the consumer of certain rules of her “home law”, however. This likely requires a comparison of the involved laws, putting the foreseeability of the parties’ legal rights in jeopardy. Such comparison also raises issues for the public, as it decreases administrability (measured by the amount of work necessary to apply a conflict rule) and hence increases costs for the courts. Through comparative analysis between Article 6 Rome I and the latter’s different interpretations, this article investigates how consumer protection affects the administrability and foreseeability of choice of law rules. It is shown how simple changes to Article 6 Rome I could increase administrability and foreseeability for all involved stakeholders.

Introduction* The ever-growing legislative activity in the area of European Union (EU) consumer law, which started in the late 1960s, has not left Private International Law (pil) untouched. In 1973, only one legal system in Romance-language countries featured specific pil rules on consumer contracts: Quebec.1 European nations came much later to the party, with Austria introducing a special conflict rule for consumer contracts in 1978.2 On an EU level, the Rome Convention on the law applicable to contractual relationships first introduced special choice of law rules on consumer contracts (hereinafter: col) to many of the Member States (ms).3 It entered into force in 1991, almost 20 years after Quebec and 13 years after Austria, and has since been mostly replaced by the Rome I Regulation (Rome I).4 The comparatively late acknowledgment of consumer law in EU col is evidence for a debate that still exists today: Should substantive values play a role in col? Early evidence suggests that it should not, as it would "incur considerable costs without resulting in any benefits" for the efficiency of a conflict rule.5 But what about consumer law relationships? Article 6 Rome I is one example of a col rule where substantive values do play a role, as it aims to protect the consumer. It declares the law of the habitual place of residence of the consumer as applicable and restricts the freedom to choose the applicable law. Similarly in consumer law, the freedom of contract is restricted to ensure that the consumer is protected. From a public interest standpoint, it is undeniable that consumers need protection, and Article 6 Rome I Regulation grants them that protection. A limitation of party autonomy in col and the freedom of contract in EU consumer law is widely accepted. But the question that then arises is: At what cost? Have we managed to create a col rule that protects the consumer, and is foreseeable and administrable, or do we need to go back to the drawing board? 3 This article investigates the influence of European governance6 on col using the example of consumer contracts. It shows that consumer protection is a public interest and, as such, changes how col operates. By placing these changes in the right context, it is shown that low levels of administrability could lead to lower consumer protection due to high costs and a lack of foreseeability. Administrability is measured by the amount of work necessary to apply a conflict rule. The higher the workload, the less administrable is a rule. Foreseeability relates to the degree to which the parties can predict what legal rules cover their relationship (pre-contract conclusion as well as pre-litigation). Section 2 explores the development of consumer protection policy in the EU and briefly outlines the resulting debates within col. Section 3 investigates how the public interest of consumer protection has influenced the design of and scholarship about Article 6 (2) Rome I. I then analyse and try to rethink the current model in the light of the public interests in private relationships debate, focussing on issues of administrability and foreseeability (Section 4). Finally, I end this article with concluding remarks (Section 5).

Consumer Protection as a Public Interest
Public interests are meant to benefit the population as a whole instead of just individuals. Or as Ho puts it: "The public interest is the interest of 'the representative individual' -an imaginary person who forgot his identity and who imagined that he had equal chance of being anyone in society."7 While an individual consumer might have the private interest to be protected as best as they can be, this does not mean that the interest at stake is only of a private nature. I argue that we, as a society represented by our governments, have decided that consumers are worthy of protection.8 This Section explores the influence of the development of consumer policy in the EU (Subsection 2.1) on EU consumer col (Subsection 2.2).

Development in the EU
The initial protection of the consumer as a weaker party in the EU started as a rather incidental development.9 The original eec Treaty focussed on creating 4 new market freedoms for professional parties in the then Common Market,10 with which "consumer welfare (…) [appeared] quasi-automatically".11 While many European countries -and societies at large -had some form of consumer protection rules even before EU policy emerged,12 it was not before the late 1960s and early 1970s that consumer law became social policy in the national laws of many ms.13 The newly found freedoms of professional parties led consumers to be exposed to foreign traders, which called for regulation.14 With the 1972 Paris Summit, the eec recognised the need for a joint approach to consumer policy,15 resulting in the Council Resolution 14 April 1975 "calling upon the institutions of the Communities to strengthen and coordinate measures for consumer protection".16 According to the Resolution, consumer protection forms a part of the eec's goal for the "improvement of the quality of life".17 In 1987, consumer policy officially became a part of the internal market strategy through inclusion of what is now Article 114 tfeu.18 As Howells rightly notes, the harmonisation of consumer law benefits the development of the internal market for two reasons:19 First, differing consumer laws in the ms constitute a barrier of trade for a professional party, because that party would have to get acquainted with a large assortment of different laws, which is a costly exercise. Second, researching many laws reduces competition because not all market participants would have the means to do so. Howells makes clear, however, that the enhancement of an internal market would not justify "social justice".20 This 5 means, in essence, that a harmonisation of consumer law is warranted, but the content of that consumer law (which does not necessarily have to protect the consumer) is irrelevant for the internal market.
In the 1990s, the activation of the confident consumer began. In the Maastricht Treaty, what is now Article 169 tfeu was added to ensure that "the Community shall contribute to the attainment of a high level of consumer protection." Shortly after, the ecj in 1994 officially recognised consumer protection as "one of the objectives of the Community".21 Howells calls this the "autonomous consumer policy":22 reaching the best possible protection for the consumer, regardless of internal market considerations. Through the Asturcom case, the cjeu has since made clear that consumer protection is not only "one of the objectives of the Community",23 but also a public policy objective.24 In this period, consumer protection policy focussed on encouraging consumers to acquire goods or services in other ms by creating minimum standards across the European Community.25 As a result, "consumers will be more willing to engage in cross-border transactions if they can be confident that the protective rules will be much the same wherever they choose to shop".26 It is assumed that the consumer is in a weaker position vis-a-vis the professional party and as a consequence, the equal footing between these parties needs to be restored.27 Legislative measures in the area of consumer law regularly use the legal base for the internal market (Article 114 tfeu),28 as explicitly mentioned in Article 169 (2) (a) tfeu.
At the beginning of the 21st century, the EU legislator grew increasingly unhappy with the current minimum harmonisation approach.29 After all, the use of Article 114 tfeu does not align well with that approach as it grants 21  ms the possibility to increase protection. This, in turn, creates barriers of trade yet again.30 The EU legislator saw the solution in making use of maximum harmonisation.31 The idea is that maximum harmonisation leads to an (almost)32 equal protection of consumers in the ms. Yet, one could argue that it is irreconcilable with autonomous consumer policy. Maximum harmonisation prevents ms from enacting more stringent consumer protection rules. In a sense, and contrary to what the term might suggest, maximum harmonisation instruments are more like minimum standards from which ms cannot deviate. Politically, in spite of the shift to qualified majority voting, it seems very unlikely that a majority of ms would agree to implement rules from the ms that currently offers the most protection. Compromises have to be made between governments from the most-and least-protective ms instead. This leads, in turn, to a reduction of consumer protection in the countries offering most protection.
Creating protective standards of any kind inevitably leads to a restriction of the freedom of contract. While all EU ms recognise that freedom in contract law,33 it was first mentioned in the EU in 2011.34 This late recognition of party autonomy in the EU could be one of the reasons why restricting party autonomy has not found much opposition, 35  readily limited the parties' freedom to contract. There are two views as to how one should deal with the principle of party autonomy on an EU level.36 The first view sees the freedom of contract as a fundamental right and requires a justification for any restriction thereof, while the second view does not recognise it as a freedom as such, but rather as a mean limited by public interests.37 Even though the principle is now recognised in the EU,38 it is unclear which view prevails. In any case, the interventionist character39 of consumer law is ample evidence for the effect of "drops of social oil"40 on contracts between professional parties and their non-professional customers. The above has shown that it becomes increasingly difficult to schematically separate the developments in EU consumer law. The incidental protection afforded to consumers through common market policies is the clear starting point. Yet, it seemed for some time that an autonomous consumer policy would replace that approach. The emergence of maximum harmonisation, however, could be evidence for a return by the EU legislator to policies based on internal market considerations.

2.2
Influence on col Sauveplanne in 1985 recognised that consumer protection "constitutes [an] example of the modern evolution towards a 'social' col, responsive to the needs of society for specific rules which govern concrete situations".41 This indicates that col is not "social" by nature. Instead, it traditionally leaves "considerations of substantive legislative policy" out of the equation and thinks of itself as "neutral".42 The occurrence of social considerations in EU col can be dated back to roughly the same time that consumer policy started to develop within the EU and its ms.43 As noted in the Giuliano/Lagarde Report on the Rome Convention, consumer protection was "the present aim of several national legislators" in the mid-1970s.44 When the Council Resolution 14 April 1975 was adopted, no ms had special conflict rules for consumer contracts. By 1980, the then eec ms agreed on the Rome Convention on the law applicable to contractual relationships which finally entered into force in the first half of 1991. This Convention is not a Union instrument, but all then ms are a member of the Convention. Article 5 of the Rome Convention includes a specific conflict rule for consumer contracts which allows parties to choose the applicable law, but that choice is restricted. Article 6 Rome I has since "replaced"45 that article. This restriction is a political compromise "between freedom and social protection". 46 Internationally, "fundamental ideas of justice, morality or decency have served as a basis for introducing alternative reference rules"47 such as Article 6 Rome I. The reason, according to Roodt, "must be sought in the ideal of justice".48 At the very heart of this justice debate lies the question whether col is the right venue to protect the weaker party (in the case of this article, the consumer). For the longest time, col adhered to the "conflicts justice" view.49 Its followers are more concerned with finding the appropriate governing law rather than the proper -fairest -result.50 Appropriate, in the eyes of conflicts justice scholars, simply means that the legal relationship is somehow connected to a country, and this country's laws must be applied. This is also what the aforementioned "neutrality" refers to: a conflict rule should simply refer to a law, no matter its content. Klauer writes that "pil is concerned with the designation of the applicable law, not with the direct resolution of a legal dispute, in which a consumer is involved".51 Instead, says Zweigert, col "has only the value-free goal to determine the legal system, to which a question of law is 9 the most closely connected".52 The last century has seen the re-emergence53 of so-called "material justice".54 Within this school of thought, internal and external situations are seen as equal. Because a judge would try to reach the most just outcome for a party in a national dispute, the same should apply to international ones.55 The "appropriate" law is therefore the law that is most just for the parties -regardless of where it might come from. Within this view several different theories can be distinguished, of which Leflar's better law approach or Curries' governmental interest analysis are but two examples. 56 Recent literature suggests that this strict separation between "conflicts justice" and "material justice", some even speak of a "theory war",57 has no place in today's col debate. Banu concludes that "contrary to conventional accounts of the relationship between material justice and conflict justice",58 "the insights of the two theories (…) have a complementary role in setting out a unified account of justice in conflict of laws".59 Her study shows that conflicts justice is a lot more open to substantive considerations than commonly believed.60 Anecdotal evidence can be found in EU pil and col more specifically. Although Europe is traditionally seen as a conflicts justice fortress, rules protecting certain parties are regularly to be found in EU col, yet the vast majority of conflict rules retain their conflicts justice character. Weatherill asks himself why so many authors continue to argue that col is neutral when "ultimately, even the decision to disregard content is a value judgement".61 In the same vein, the simple reality is that EU col includes special rules for consumer contracts (Article 6 Rome I). To understand how these rules function, the following Section elaborates on how Article 6 Rome I has been designed to protect the consumer. Article 6 Rome I Rome I takes a paternalistic stance on the topic of consumer protection. Article 6 (1) Rome I designates the law of the country of the habitual residence of the consumer as the governing law, and Article 6 (2) restricts the choice of applicable law by making reference to that law. This choice for the "consumer's law" in and of itself is a steering instrument to restore equal footing amongst the parties. It therefore redistributes power back to the consumer and is not neutral. Rome I also recognises overriding mandatory rules in Article 9. Consumer law rules could form part of overriding mandatory rules, but it is generally accepted that Article 9 does not apply in addition to Article 6.62 In this Section, I offer a comparative analysis of the different interpretations of applying Article 6 Rome I. Article 6 of the Rome I Regulation represents a conflict rule that has been influenced by public interests. In fact, it would not exist without them. Whether the mere influence of public interests leads to less administrable and less foreseeable rules remains to be seen.63 In Subsection 3.1, the standard conflict rule for consumer contracts of Article 6 (1) Rome is analysed. Subsection 3.2 then dives into party autonomy in consumer contracts (Article 6 (2) Rome I). Lastly, Subsection 3.3 concludes this section.

3.1
Para. 1: General Conflict Rule When parties have not chosen a law to govern their contract, the general rule of Article 6 (1) Rome I applies. According to that rule, the contract "shall be governed by the law of the country where the consumer has his habitual place of residence". The place of habitual residence refers "to the place where the consumer ordinarily de facto resides".64 Till the cjeu order in MBank, it was generally assumed that the habitual place of residence of the consumer at the moment of contract conclusion informs the applicable law. In MBank, however, the cjeu ruled that the consumer's habitual place of residence at the moment of stepping to court is relevant, instead.65 If a Lithuanian business enters into a contract with a consumer living in France, but the consumer moves to Germany and subsequently establishes their centre of social integration there, Germany is seen as the habitual place of residence in the sense of Article 6 (1) Rome I. Additionally, frequent moving of a highly mobile consumer could make identifying the habitual place of residence difficult.66 This has foreseeability consequences for the business, which will be discussed further later on in this article.
While the focus of this analysis lies on the conflict rule contained in Article 6 Rome I and not the scope requirements that need to be fulfilled for Article 6 to apply in the first place, the infamous "pursuing or directing" test of Article 6 (1) (a) and (b) warrants at least a mention when talking about foreseeability and administrability. Under this test, the court needs to assess whether the professional party pursued or directed their activities to the country where the consumer has their habitual place of residence. Physical activities and branches in that country are straightforward examples of pursuing an activity there,67 but if that is not the case, assessing whether the professional party directed their activities towards that country requires more effort. The cjeu in Pammer & Hotel Alpenhof developed a "targeting" test.68 A professional party targets (and thus directs) their activities towards a ms if it is evident from their behaviour that they intended to do business in that country.69 For online sales especially, some elements to be taken into account are whether the professional party's website was available in other languages, allowed shipping to other ms, offered travelling instructions, featured payment in a currency other than the local currency, etc.70 Additionally, MBank has raised yet another question: What if the consumer moves to a country to which they take their habitual place of residence with them, but the professional party does not target that country? The cjeu dealt with this question in Commerzbank ag. It ruled that the targeting test must only be fulfilled at the moment of contract conclusion.71 If the consumer moves to another country afterwards, the targeting test need not to be re-applied. Interestingly, the court also accepted that the cross-border element needed for the applicability of col can appear at a later stage;72 but 66 D. Martiny the targeting test is still exercised at the moment of contract conclusion even though both parties are resident in the same country.
This can lead to two different places of habitual residence being relevant in a single proceeding. When assessing whether a professional targeted the country in which the consumer was habitually resident, the habitual place of residence at the moment of contract conclusion is decisive (Commerzbank ag). For deciding which law applies to the contract, however, the focus shifts to the habitual place of residence of the consumer at the moment of starting proceedings (MBank).

3.2
Para. 2: Choice of Law Rome I extends the freedom to choose the applicable law to consumer contracts. Article 6 (2) Rome I reads as follows: "Notwithstanding paragraph 1, the parties may choose the law applicable to a contract which fulfils the requirements of paragraph 1, in accordance with Article 3. Such a choice may not, however, have the result of depriving the consumer of the protection afforded to him by provisions that cannot be derogated from by agreement by virtue of the law which, in the absence of choice, would have been applicable on the basis of paragraph 1." Therefore, while a choice of law is possible, it may not deprive the consumer of the protection that they would have enjoyed under the provisions that cannot be derogated from by agreement in the country of the consumer's habitual place of residence. Looking at this provision, one question arises: What is meant by "depriving the consumer of protection"? The cjeu has yet to clarify this term. There are two main interpretations of this sentence, both of which influence Article 6's administrability to different degrees.
"Depriving the consumer of protection" is often understood as encompassing a legal comparison between the chosen law and the non-derogable provisions of the habitual place of residence of the consumer (Günstigkeitsvergleich  As the Advocate General observed in point 44 of his Opinion, the correct application of Article 8 of the Rome I Regulation therefore requires, in a first step, that the national court identify the law that would have applied in the absence of choice and determine, in accordance with that law, the rules that cannot be derogated from by agreement and, in a second step, that that court compare the level of protection afforded to the employee under those rules with that provided for by the law chosen by the parties. If the level of protection provided for by those rules is greater, those same rules must be applied."80 Another way to interpret the provision is by arguing that the non-derogable rules of the consumer's habitual place of residence remain applicable at all times. 81 The choice of law therefore only relates to the rules of the habitual place of residence that can be derogated from by agreement. This means, in essence, that the chosen law only complements the objectively applicable law. This interpretation is based on the "protection principle" (beschermingsbeginsel).82 Article 6 (2) Rome I would upset the balance between the consumer and the professional party if the for the consumer more favourable law would always be applicable,83 because then "the consumer would always gain protection".84 That the consumer gains protection would not be the goal of Article 6 however, but rather that the protection stays the same. This would guarantee a certain level of legal certainty and would avoid extensive legal comparisons. 85 In Amazon v. vki, A-G Saugmansgaard Øe seems to have followed this view arguing that "Article 6(2) […] enables the consumer to rely entirely on the mandatory provisions of the law of his State of residence, whether or not they are more favourable than the provisions of the chosen law from the point of view of their substance."86 It is unclear, however, whether by "enables" he means the mandatory application of that law or, alternatively, the right of the consumer to choose which of the two involved laws should be applied. Following the judgement in Amazon v. vki, in which the cjeu did not discuss how Article 6 (2) should be interpreted, the Austrian Supreme Court simply applied the law of the habitual place of residence of the consumer (Austria): "However, according to this provision [Article 6 (2)], the choice of law cannot have the effect of depriving consumers resident in Austria of the protection of mandatory provisions of Austrian law."87 "Since the plaintiff only alleges a violation of mandatory Austrian law, it is obvious to examine only this violation. If it is affirmed, it is irrelevant whether the clause would also be inadmissible under the chosen law or not."88 I find it unlikely that the cjeu would follow this line of argumentation.89 Instead, it would benefit the coherency of EU col if the cjeu would apply the same reasoning from Gruber also to consumer contracts.

3.3
Analysis Under Article 6 (2) Rome I, one single legal obligation (e.g. delivery) could be partly governed by law A and partly by law B. This applicability of (at least) two laws also leads to foreseeability issues: how can either party really know what rules the judge will apply in the end? The extent of the reduction in foreseeability and administrability depends on the interpretation followed. The application of the favour principle and its (likely extensive) legal comparison would require the parties to exercise that comparison before entering into the consumer contract if a good level of foreseeability should be maintained. Rühl calls this the "pre-litigation predictability".90 Doing so is impractical. Not only are most consumer contracts performed without a dispute arising out of them, but a comparison would also require extensive work for each individual contract. This goes hand in hand with high costs. For this reason, the more extensive the legal comparison is, the less administrable and the less foreseeable the question of applicable law becomes.
The other interpretation based on the protection principle does not require comparisons to be made. The non-derogable rules of the consumer's habitual residence apply and are complimented by rules of the chosen law that do not have mandatory equivalents in the consumer's country of habitual residence. This reduces the impact on administrability significantly, but does not remove it completely.91 As a consequence, this interpretation also leads to a higher degree of foreseeability because both parties know for certain that the non-derogable rules of the consumer's habitual place of residence will apply.
Another interesting aspect is the tendency of judges to apply the lex fori.92 Within the EU it seems likely that the consumer would sue in their country of habitual residence as per Arts. 17-19 Brussels Ibis Regulation. The professional party is even required to sue in the country of the consumer's habitual place of residence. The place of jurisdiction thus often coincides with the otherwise applicable law, which is the habitual place of residence of the consumer. A recent study of Dutch case law involving Article 6 (2) Rome I suggests that judges tend to apply the law of the habitual place of residence of the consumer which happens to be the lex fori.93 It seems that they negate the possibility of providing the consumer with better protection through the chosen law.94 If this tendency really exists, why follow the favour principle?
Lastly, regard should be paid to the harmony of decisions created under the application of Article 6 (2). The existence of (at least) two different interpretations means that not every court will apply the conflict rule in the same manner. Consequently, one court might decide that the non-derogable rules should be applied in any case, while another chooses to let the more favourable provisions of the chosen law apply. Dutch case law shows that even courts within one country apply the conflict rule differently.95 The recent judgement in Gruber is a great indicator that it is most likely that the cjeu would apply the favour principle to consumer contracts. For the remainder for this article, it is therefore assumed that Article 6 (2) Rome I does indeed require a comparison between the chosen law and the non-derogable rules of the consumer's habitual place of residence.

Rethinking the Current Model
With the transformation of the Rome Convention into a EU Regulation, the general conflict rule in now Article 6 (1) Rome I has been mainly left untouched, but party autonomy has widely been debated. At the time, the abolishment of party autonomy for consumer contracts was seen as an option, but it ultimately never arrived,96 because some involved legislatures believed that it would restrict trade too much.97 The UK Consultation Paper feared that the lack of a choice of law would require parties to thoroughly know the law of every ms.98 Martiny argues that restricting party autonomy (freedom of contract) in EU substantive consumer law but not in col would be incoherent with EU law, as professional parties -often through contracts of adhesion -could circumvent that substantive EU consumer law.99 Lando and Nielsen write that if a professional party is unhappy with the double protection regime of Article 6 (2), they could simply choose not to include a choice of law in their terms and conditions.100 They "see no advantage for the business enterprise in eliminating their option to select another law than that of the consumer", because "many of them may prefer [choosing their own law], even if the operation of their own law is modified by the mandatory provisions of the consumer's law".101 I will first discuss the arguments presented above before expanding my analysis to the issues of foreseeability and administrability. First, the argument that cross-border trade would be negatively impacted due to an increase in workload holds only partly true. In my opinion, the mere applicability of a single law over a multitude of laws actually decreases workload. After all, the professional party would have to do research only into that single (unknown) law, instead of needing to search in that law for non-derogable provisions that protect the consumer. Once such provisions would be found, the professional party would need to compare them with the chosen law (most likely its place of incorporation or a law most beneficial to the professional party) to establish which law protects the consumer better. In essence, this exercise requires the professional party to do exactly what the UK Consultation Paper expected from an abolishment of party autonomy: the need to research an entire foreign law. And even worse, the professional party needs to exercise a complicated legal comparison. The outcome is the legal equivalent of Frankenstein's monster: a bit of this law, a bit of that law, and voilà. Furthermore, professional parties have every right to limit their operations to certain ms, using the cjeu judgements in Pammer and Hotel Alpenhof,102 Mühlleitner,103 and Emrek104 as guidance.105 Recent cjeu case law such as MBank and Commerzbank ag makes not targeting a certain ms at all impossible, however, because the consumer could always move to a country that has not been target by the company. Limiting operations would indeed lead to a negative impact on cross-border trade if businesses would actually stop serving certain ms simply because of their laws. I find this unlikely, because under the current rules the consumer would most likely already be protected under these laws in any case, as important consumer protection provisions can regularly not be derogated from.106 Second, I agree with the argument that full party autonomy would lead to incoherency between EU substantive law and EU col. EU substantive law respects the freedom of contract after all,107 and parties are only limitedthrough instruments such as grey and black lists108 -in the selection of contractual clauses they can use. In my opinion, abolishing party autonomy altogether is even more coherent with EU substantive law. This would "force" the application of EU substantive law which, in turn, restricts the parties' freedom of contract. I doubt that completely abolishing party autonomy would also be seen as an inconsistency.
Third, while a professional party can indeed choose not to include a choice of law in their terms and conditions, this argument completely ignores the litigative realities of the scenario when the professional party does include it.109 102 Pammer and Hotel Alpenhof (n 68 This is certainly the case when EU law has been implemented. Minimum harmonisation Directives allow ms to enact laws offering higher protection which would then be mandatory, and maximum harmonisation Directives still have some room in regards to the remedies. 107 Be it through the values of its ms or autonomously. 108 See for example: Caruso (n 40). 109 To this effect: Schmitz (n 93). A conscious decision to include a choice of law clause will -in practicemean that the business knows of the consequences and can bear the extra work involved. But what about the consumer and the court? That professional parties have the power to decide whether or not to include a choice of law clause can be seen as a choice of law in and of itself. The consumer, as has always been the case with contracts of adhesion, has little choice. She will be confronted with two legal systems that apply to the contract. How is she, as a weaker party, meant to bear the extra workload that the professional party consciously decided to bear, presumably after extensive weighing of their interests? I believe that she cannot. As a result, the consumer is exposed to the arbitrary choice, made by the professional party, to include or not to include a choice of law clause. Additionally, it is not only the parties themselves that bear the consequences of the applicability of Article 6 (2) Rome I. After all, it is the courts that have to apply it in case of litigation. While this plays a smaller role in countries such as Ireland where the judge is not investigating the law herself,110 judges in jurisdictions such as Germany and The Netherlands need to exercise the legal comparison that flows forth out of Article 6 (2) Rome I. This costs time and money, as I will elaborate now.

4.1
The Three Options Rethinking Article 6 Rome I should focus on para. 2 thereof, as the analysis in Section 3.1. shows that the conflict rule of para. 1 is relatively simple. Issues of establishing a consumer's habitual place of residence can exist, however, and will be discussed throughout this and the next Section.
In regards to Article 6 (2), three options arise:111 First, party autonomy is removed completely. Second, the restricted party autonomy approach stays intact. Third, parties are granted full party autonomy. I will compare these three options on their foreseeability, administrability, costs, impact on courts, and level of consumer protection.112 110 Though, it should be argued that the judge must: On the ex officio application of EU col, see: F. Ferrari, Concise Commentary on the Rome I Regulation (Cambridge: Cambridge University Press, 2020) 176; Rühl (n 73) 255. 111 Campo Comba also discusses a fourth option, that being to allow a choice of law but only for specific jurisdictions (M. Campo Comba, The Law Applicable to Cross-border Contracts involving Weaker Parties in EU Private International Law (Cham: Springer, 2021) 83f), as also put forward by the Green Paper on the conversion of the Rome Convention of 1980 on the law applicable to contractual obligations into a Community instrument and its modernisation (n 97) 30f. 112 Aspects of procedural consumer protection are excluded from the score of consumer protection, as they are represented by the other aspects (foreseeability, administrability, and costs). It is important to mention that if the substantive consumer laws of the ms are the same or similar enough, a choice of law would be quite irrelevant. In intra-EU cases, the consumer would enjoy the same protection regardless of which law applies. We have not reached that point yet, though, as the many minimum harmonisation directives allow ms to adopt stricter consumer protection rules, and national laws continue to exert their influence through open norms and nationally-enacted remedies.113 For the sake of this analysis, I will therefore assume that the laws of the ms provide different levels of consumer protection. This is certainly true for laws of third countries.
Abolishing Article 6 (2) Rome I altogether would mean that Article 6 (1) would apply to every consumer contract. This brings a high level of foreseeability. Both the consumer and the professional party will know which rules apply at the moment of contract conclusion. Only the business needs to be cautious. It cannot prevent the application of a law that it did not expect, as the consumer is free to establish herself elsewhere. The law of that place would then become applicable to the contract (Commerzbank ag). There is no need for a legal comparison, because only the law of the habitual place of the consumer applies. This is highly administrable. Yet, a court might be confronted with foreign law that applies in its entirety. Consequently, the impact on the courts is somewhere between low and high (= medium). The accessibility of resources in the consumer's native language also means that the costs for the consumer remain low.114 This assumes a correlation between habitual place of residence and language, which is not always true today. Yet, I argue that consumers are a lot more likely to speak the language of the country that they are habitually resident in than that of an arbitrary third country. They can also more easily find lawyers in the country of their habitual place of residence. Additionally, these lawyers do not have to be specialised in foreign law, which could lead to lower fees. For businesses, this would mean a one-time analysis of the legal system of 113 This has also been recognised in the Commission's Green Paper on the conversion of the Rome Regulation into Rome I: "But it must be borne in mind that these directives cover only certain aspects of legal rules which protect the consumer. In addition, there can still be differences from one country to another, in particular where a Member State fails to transpose a directive (…)": Green Paper on the conversion of the each country they want to conduct business in.115 As these are one-off costs, I argue that this approach would leave businesses somewhere between low and high costs (= medium costs). In regard to the level of consumer protection, the consumer enjoys the exact same protection for purchases made in other countries as they do in their habitual place of residence. In my view, this fulfils the goal of cross-border consumer law: to protect the consumer from adverse effects based on foreign law. Leaving Article 6 (2) Rome I untouched means that the uncertainties as discussed in Section 3 remain unsolved. In that analysis, I established that the level of foreseeability of the restricted party autonomy approach is low, because it would require both the consumer and the professional party to conduct research into the chosen law and the law of the consumer's habitual place of residence before contract conclusion. Without such exercise, the parties could not reasonably foresee what laws apply to their contract. This is unrealistic, of course. Additionally, studies suggest that judges have a tendency to apply the lex fori, reducing foreseeability even further: Will the judge apply the law that is actually more protective, or their own law as it is likely the easiest to apply? To make matters even worse, the consumer could habitually establish herself elsewhere after contract conclusion and the law of that new residence then becomes applicable according to Commerzbank ag. This comparison exercise leads to a low administrability of the current approach, as it takes a lot of time and efforts. Furthermore, the used resources are wasted if a dispute never arises. The comparison also has a high impact on the courts and on the costs of the parties because their legal counsels116 (in case of the court, possibly expert opinions or external reports) do, in general, exercise that comparison.117 These high costs could prevent consumers from starting proceedings in the first place, as they would have to carry these costs in case they lost.118 The current approach provides for a high level of consumer protection because either the mandatory provisions of the consumer's habitual place of residence apply or the more protective provisions of the chosen law. Next to abolishing party autonomy in its entirety, another option is to offer completely unrestricted party autonomy. Parties would be allowed to choose any law that they wish, including laws of non-EU ms. This influences the viability of this option significantly. To illustrate the difficulties, I will first discuss the scenario in which the parties choose a ms law, and then the scenario in which the choose a non-ms law.
Choosing a law of an EU ms leads to a high level of foreseeability for the parties, because it is clear from the beginning which law applies. Having only one law apply to a contract makes full party autonomy a highly administrable option, too, because no comparisons need to be made. Yet, it seems likely that proceedings would be brought before the courts of the consumer's habitual place of residence. This means that courts could frequently be confronted with a foreign law into which they must conduct research. As a consequence, the impact on courts is medium. On the one hand, the costs for the consumers are high, as they cannot (readily) rely on information about the chosen legal system in their language. They need to find a lawyer that speaks both the language of the consumer and the language of the chosen law, which takes effort and possibly costs more. Next to this, the consumer would regularly be confronted with contracts by different parties including choice of law clauses for different laws, leading to recurring costs as every professional party uses another law. On the other hand, the professional party can enjoy low costs as it chooses either the law of its place of incorporation, or alternatively, a law that is most beneficial for it. In any case, professional parties can focus on a single law for their entire customer base. The level of consumer protection when allowing full party autonomy varies, as businesses could choose laws for many reasons. Some of them might feature very favourable laws while others will not. Therefore, the level of consumer protection is generally low.
Things get more complicated when the law of a non-EU ms is chosen. Parties could escape mandatory EU law if that choice was fully unrestricted. This could be solved in two ways: Either the chosen law must be that of a ms, or non-derogable EU law remains applicable. In any case, this would lead to a restriction of the freedom to choose the applicable law once again and is thus not really an option. In Ingmar, the cjeu ruled that even though a non-EU ms law was chosen, a court of an EU ms must nonetheless apply mandatory EU law to the B2B relationship at stake.119 This approach would likely also apply to consumer contracts if the provision featured an unrestricted choice of law. Consequently, foreseeability of both parties would decrease as they cannot be 119 Case sure how EU law will be applied to their conflict. After all, it is the ms courts that will be tasked to apply EU law. The question then arises is: Is the implementation of EU law in the country of the forum decisive? Minimum harmonisation would allow ms to go beyond EU law after all. If one would follow the reasoning in Unamar, the forum's implementation would be decisive.120 If EU law is to be applied strictly according to its texts, issues can still arise as to their different interpretations but also legal consequences. This is an administrability nightmare for all parties.
Be it as it may, in my final analysis I will take as the starting point for option 3 a completely unrestricted choice of law. EU consumer law could then indeed be circumvented, and this is reflected in the low consumer protection score of option 3.

4.2
A Way Forward When comparing the three options (Table 1), it becomes evident that the second option (current restricted party autonomy approach) fares the worst in all aspects but level of consumer protection. It offers the consumer the same high level of protection as the first option (no party autonomy). My decision not to grade the level of consumer protection higher than that of the first option is a conscious one for two reasons: First, while it may be true that the second option could grant the consumer a higher level of protection, I question whether the consumer would often benefit from this situation. On the 120 Case off-chance that the businesses' law offers the consumer more protection, the latter would indeed lose protection if the current approach was abolished. Second, while the level of substantive consumer protection may be high, the level of actual protection will be a lot lower, because the consumer has to invest money and other resources into resolving a dispute, which could stop her from pursuing her protection in the first place. Both the first (no party autonomy) and third option (free party autonomy) offer high levels of foreseeability and administrability. Yet, a completely unrestricted right to choose the applicable law most likely leads to high costs for the consumer and a low level of protection. The possibility of the professional party to completely avoid EU law lowers the consumer's protection immensely. Even in cases when a favourable law is chosen, the consumer would first need to be able to know what that law contains. The first option (no party autonomy) provides lower costs to consumers and a high level of consumer protection, instead. The only two downsides of this option are a slightly increased cost profile for businesses compared to the third option, as they would have to research the laws of each country in which they intend to conduct business.121 Additionally, the business cannot reasonably foresee whether the consumer will, at some point in the future, change her habitual place of residence. Smits convincingly argues however that "it is more likely that differences in tax law or procedural law would form a much greater barrier to engage in cross-border trade than consumer contract law".122 In any case, worries by lobbying groups should be eased as the costs of applying option 1 are a lot lower than with option 2 (restricted party autonomy), the currently applicable system. No party autonomy and full party autonomy help decrease the workload of the courts fundamentally, as the judge would simply have to apply a single law (instead of a combination of many). While it is likely that consumers will step to court in the country they habitually reside in and that that court will know their own law best, continuing the current mixed approach would still require it to compare their own laws with that of the chosen other law. I argue that it is simpler to apply a single law, even be it foreign, than comparing two laws with another. From a pure consumer protection standpoint, the first option is clearly superior to the other two. While the interests of professional parties would fit most to option 3, my analysis shows that even they benefit from a more simplistic option such as the first. Having no party autonomy decreases their costs after all. Lando and Nielsen are right in stating that businesses can choose not to include a choice of law clause. But, as mentioned previously, the mere 121 Depending on which view one follows in regard to Article 6 (2) Rome I, businesses would have to do that already as the more protective provisions might apply. 122 Smits (n 31) 8. power to decide on this question empowers their legal position compared to that of the consumer. After all, they themselves decide when to expose the consumer to complicated legal comparisons and their consequences (unforeseeability, high costs, etc.) I believe that completely removing party autonomy for consumer contracts would benefit all involved stakeholders (consumers, professional parties, and courts). There are two remaining problems that should be addressed before making Article 6 (1) Rome the sole conflict rule for consumer contracts, however. First, the recent case law of the cjeu in Mbank and Commerzbank has a negative influence on foreseeability. A revised version of Article 6 (1) Rome I should clearly mention that the moment of contract conclusion is relevant for the determination of the consumer's habitual place of residence. This might not be the most consumer-friendly solution, but it could help balance interests of the professional party and the consumer. Second, there are situations in which Article 6 Rome I is applicable but the consumer's habitual place of residence is in a third country. A possible solution could be to determine as the applicable law the law of the country in which the cross-border connection to the EU exists (for example, place of establishment of the business or country in which the consumer concluded the contract).
Option 1 thus satisfies the public interest of consumer protection, but also the private interests of the parties involved. To use the words of recital 4 Consumer Rights Directive: Option 1 strikes "the right balance between a high level of consumer protection and the competitiveness of enterprises".

Conclusion
This article investigated the relationship between the public interest of consumer protection and the degree of administrability and foreseeability in the col rule for consumer contracts: Article 6 Rome I. EU consumer protection policy clearly influenced the design of Article 6 Rome I and its conflict rule. However, my analysis shows that the current restricted party autonomy approach misses the mark, and that there are better alternatives available. These alternatives could not be more different: either introduce completely unrestricted party autonomy or abolish party autonomy altogether. To paraphrase the cjeu in Heininger, low legal certainty is not reason enough to reduce consumer protection.123 But why not increase legal