Digital upgrade of German antitrust law - blueprint for regulating systemic platforms in Europe and beyond?
2019 was the year of large-scale scientific reports on competition policy in the digital era. In particular, the UK Furman Report,[1] the EU Special Advisors’ Report,[2] the Australian ACCR’s Digital Platforms Report,[3] the US Stigler Report[4] and the German Competition Law 4.0 Report[5] formed the academic basis for future reforms. In January 2020, the German government translated these preparatory works into action and published the first fully-fledged legislative proposal for reforming competition law (the 'Draft Bill'), tellingly dubbed the “Digitalization Act.”[6] Addressing issues such as data access and portability, cross-market leveraging, and intermediation power, the Draft Bill lifts the German competition law into the 21st century. France, Italy and Poland have already indicated[7] support of the core proposals of the Draft Bill. This makes it likely that these proposals will also be at the heart of the upcoming regulatory reforms at the European level. After all, with a new German President of the European Commission (Ursula von der Leyen), and Germany and France having the EU Council Presidency[8] in 2020 and 2022 respectively, the German proposals will have an excellent chance of serving as a blueprint for regulatory reforms in the EU. In fact, the EU Commission’s strategy of “shaping Europe’s digital future” published on February 19, 2020[9] already adopts some of the new approaches and terminology. The Draft Bill is likely to also influence some debates in other European and non-European countries on regulatory reforms in the tech area.
Overall approach: stronger focus on genuinely problematic mergers and abuse cases
Reforming the German Competition Act (GWB) was necessary to implement the European ECN+ Directive 2019/1[10] on the empowering of competition authorities of the EU Member States. In implementing this Directive, the Draft Bill contains new rules on: (i) the enforcement powers of the competition authorities, (ii) leniency programs and (iii) sanctions for competition law infringements as well as (iv) assistance for other competition authorities.
The Draft Bill’s proclaimed aim is to set up a regulatory framework that addresses the challenges posed by the digitalization of the economy. It suggests a “moderate” reform of the control of unilateral conduct, that is, the abuses of market power, with a view to more effectively cease the abuse of market power, particularly by digital platforms. To free up resources and allow the competition authority to focus on complex cases, the Draft Bill suggests lowering the threshold for merger notification. While, in the current framework, mergers are notified if one party has domestic revenues in excess of €25 million and the other party in excess of €5 million, in the future, the other party’s revenues need to exceed €10 million.[11] It is expected that this will lead to 270 fewer merger reviews per year. However, to ensure that “killer acquisitions” may still be prevented, the competition authority can order particular companies to notify every merger if there are indications that future concentrations may restrict domestic competition in a particular sector.[12]
Main proposals on control of market power
The Draft Bill centers on reform of the law on abuse of dominance. Currently, it is not just undertakings with a dominant position on a given market that are subject to the ban of abusive behavior (Sec. 19 GWB); undertakings with so-called “relative” market power vis-à-vis undertakings that depend on them are subject to the ban as well (Sec. 20 GWB). The Draft Bill would moderate amendments for both dominant companies and companies with relevant market power. In addition, the Draft Bill introduces a third category of market power that is subject to additional obligations - so-called “undertakings with paramount significance for competition across markets” (Sec. 19a GWB-Draft).
Reforms concerning companies with a dominant position
- “Intermediation power” as a new factor for determining dominance.
The 2017 German competition law reform already clarified the following: (i) a zero-price for any service does not preclude the finding of a relevant product market and (ii) for determining dominance on markets for multi-sided platforms, the following factors need to be considered: direct and indirect network effects, the level of single- and multi-homing, access to data relevant to competition, and competitive pressures from innovation (see Sec. 18 paras. 2a and 3a GWB).
The Draft Bill now further determines that, when assessing the market power of an undertaking, “acting as an intermediary on multi-sided markets, account should be taken in particular on the importance of the intermediary services it provides for access to supply and sales markets” (Sec. 18 para. 3b GWB-Draft). This proposal would clarify that dominance may not just exist on the supply side or the demand side; it may also exist at the level of intermediation between both sides. It is hoped that this clarification further addresses the significant power of intermediaries that results from their ability to determine, on the upstream intermediation market, the outcome of competition on downstream markets (that they intermediate) on which they may operate themselves.
- Facilitating access to data held by dominant platforms.
Acknowledging the relevance of data for multi-sided digital markets, a central objective of the Draft Bill is to facilitate access to data held by dominant platforms in order to enable competitors to compete on up- or downstream markets. To this end, the Draft Bill would amend the German codification of the essential facilities doctrine (Sec. 19 para. 2 no. 4 GWB). While thus far, only the denial of access to essential networks and infrastructure qualified as an abuse, the refusal to supply “access to data” against an adequate remuneration might also constitute an abuse. According to the reasoning of the Draft Bill, this amendment would bring German law closer to the European case law on Article 102 TFEU.[13] Regarding the remuneration, in some cases, a price of zero could be considered “adequate,” in particular if the data was generated without significant investments.[14]
It is unlikely that the proposal will lead to a high number of (successful) data requests. For the substantive criteria of a duty to supply, the provision refers to the case law on Article 102 TFEU. However, if the restrictive conditions of Article 102 TFEU are fulfilled, an identical national rule provides no additional instrument to access data.
Reforms concerning companies with “Relative Market Power” vis-à-vis dependent companies
- Dependent company need not be small- or medium-size.
Currently, under German law, undertakings that have “relative market power” vis-á-vis small- and medium-sized companies in that the latter depend on the former in a lack of sufficient and reasonable possibilities to carry out business are prohibited from abusing this power. The Draft Bill would eliminate the criterion that the dependent company must be small- or medium-sized. Instead, the protection would be revoked only if dependency of the undertaking is offset by its corresponding countervailing power vis-à-vis the company with a strong market position.
- Relative market power to be assumed for intermediaries controlling access to markets.
In addition, it is proposed that relative market power shall also be assumed for “undertakings acting as intermediaries on multi-sided markets insofar as undertakings are dependent on their intermediary services with regard to access to supply and sales markets in such a way that sufficient and reasonable alternatives do not exist” (Sec. 20 para. 1 GWB-Draft). As a result, intermediaries that do not enjoy a dominant position will be subject to the ban of an abuse of dominance. While expanding the scope of undertakings that are now subject to the general provisions on abuse of dominance, the Draft Bill does not, however, contain any specific new prohibitions that address the challenges posed by intermediation power. In particular, the Draft Bill does not contain any general ban on intermediaries with market power to favor their own services.
- Relative market power to be assumed if undertakings depend on access to data to enter a market.
In addition to cases of a dependency on an intermediary to access customers, the Draft Bill would establish a new type of “relative market power” when undertakings are “dependent on access to data controlled by another undertaking for its own activities.” “Denying access to such data may constitute an unfair impediment even if no commerce has yet begun for such data” (Sec. 20 para. 1a Draft Bill). According to the Draft Bill, this may be the case within existing vertical relationships. If several companies contribute added value to a joint project (for example, with the Internet of Things), the company with the superior bargaining power could not withhold the data it collects in this process from its cooperation partner.[15] However, the provision will also be relevant outside existing co-operations, in particular for companies requiring single-source data from an undertaking with relevant market power in order to compete on any downstream market or aftermarket (for example, for the repair or maintenance of goods provided by the addressee of the norm).
- Impeding rivals’ attainment of positive network effects shall constitute an abuse if this is capable of triggering the tipping of a market.
The proposals mentioned above can be seen as moderate clarifications or evolutions of the existing concepts under German law on abuse of dominance. However, the Draft Bill also contains a new example of an abuse of market power that is especially designed for competition between multi-sided platforms. According to the proposed Sec. 20 para. 3a, an anti-competitive impediment of competitors “shall also be deemed to exist where an undertaking with superior market power on a [multi-sided] market […] impedes competitors’ independent attainment of positive network effects and thereby creates a serious risk of considerable restriction of competition on the merits.” Conceptually, this unprecedented provision aims at addressing one perceived weakness in the current concept of abuse of dominance by digital platforms.[16] Currently, an abuse can only be found once a company has already gained a dominant position. As a result, it is feared that authorities or courts may not be able to intervene early enough against unilateral conduct with the goal of an irreversible “tipping” of a market (which is subject to strong positive network effects) by means that do not reflect competition on the merits. By the time a company has achieved a dominant position on a relevant market (and thereby becomes subject to the ban of an abuse) as a result of an anti-competitive means to tip the market in its favor, any antitrust intervention may simply come too late to have any impact - the relevant competition “for the market” has already been decided.
The cited proposal is intended to overcome this issue by introducing three novel approaches. First, the proposal lowers the threshold for being subject to the prohibition to companies with “superior market power.” They do not need to be dominant yet. Thus, even companies that have no paramount market position in relation to its competitors, but are superior to small- or medium-sized rivals, are subject to the clause. Second, the proposal prohibits measures that have the goal of hindering competitors from attaining positive network effects. The provision is “consciously worded in an open manner and refrains from giving examples, so as to also cover any new measures to impede the attainment of network effects that are currently not known.”[17] The reasoning of the Draft Bill refers to a prohibition or a hindering of customers switching to a competing platform (that is, in an effort to multi-home) as an example. This may include measures to disable data portability or interoperability along with exclusivity clauses or tying practices. Such measures will only be prohibited, however, if they create a “serious risk of a considerable restriction of competition on the merits.” This wording is intended to clarify that the authority does not have to wait until it can demonstrate that competition has been foreclosed.[18] It is sufficient, but also necessary, to demonstrate that the conduct is likely to significantly impair competition.
New concept of undertakings with paramount significance for competition across markets
The prohibition to impede rivals’ attainment of positive network effects would apply to all companies with (at least) superior market power. Further limits are suggested for companies that are active to a significant extent on multi-sided markets and are of paramount significance for competition across markets. Proposed Section 19a GWB-Draft forms the heart of the legislative proposal. It contains five new types of an abuse of dominance. However, undertakings will not automatically (by statute) be subject to the ban of such new types of abuse. Rather, corresponding conduct will only be prohibited once the Bundeskartellamt (Federal Cartel Office) (i) has issued a decision declaring that an undertaking is of paramount significance for competition across markets, and then (ii) has issued a decision prohibiting that undertaking from any of the five new types of abuse.
In determining the paramount significance of an undertaking for competition across markets, the Bundeskartallamt shall take particular account of the following factors:
- its dominant position in one or more markets,
- its financial strength or access to other resources,
- its vertical integration and its activities on otherwise related markets,
- its access to data relevant for competition,
- the importance of its activities for third parties’ access to supply and sales markets, and its related influence on third parties’ business activities.
The Draft Bill expects that only “very few” companies will quality as having paramount cross-market relevance. In fact, within the first five years, the Bundeskartellamt is expected to declare just three companies as having such a status[19]
Once the Bundeskartellamt has identified an undertaking with paramount cross-market relevance, it may issue an order prohibiting this undertaking from any of the following (exhaustive) practices:
- Self-favoring: treating the offers of competitors differently from its own offers when providing access to supply and sales markets;
- Impeding competitors by leveraging market power: directly or indirectly impeding competitors on a market in which the respective undertaking can rapidly expand its position even without being dominant, provided that the impediment is likely to significantly obstruct the competitive process;
- Using third-party data to create barriers to entry: creating or raising barriers to market entry, or impeding other undertakings in another way by using data relevant for competition that has been collected from the other side on a dominated market, also in combination with other data relevant for competition from sources beyond the dominated market, or demanding terms and conditions that permit such use;
- Hindering interoperability and data portability: making the interoperability of products or services or data portability more difficult and thereby impeding competition; and
- Insufficient information about performance of customers: informing other undertakings insufficiently of the scope, the quality or the success of the performance they provide or commission, or making it difficult in other ways for them to assess the value of such performance.
None of these practices would be prohibited if the conduct is objectively justified. However, to facilitate the work of the competition authority, and to reflect the typical information asymmetry vis-à-vis the undertakings in question, unlike in other cases of an abuse of dominance, the burden of proof for the objective justification shall lie with the undertaking.[20]
Any conduct fulfilling these criteria will only become unlawful ex nunc once the Bundeskartellamt has issued an order prohibiting the conduct and such order becomes legally binding. This means that there can be no fine for any corresponding conduct prior to such order. It is also not the case that affected parties can turn to civil courts to seek redress prior to a decision of the Bundeskartellamt. This has been criticized as being inefficient and counter-productive[21] because the Bundeskartellamt may not have the resources to intervene in time to address the various types of conduct of the various undertakings. Moreover, given the lack of any risk of a fine or damages claims, the legal provisions may have insufficient deterrent effect. In its defense, the current concept is intended to mitigate the consequences of any legal uncertainty created by the novel provisions for the undertakings with cross-market relevance.[22]
Flanking procedural proposals to speed up investigations
To facilitate more rapid intervention by the Bundeskartellamt, the Draft Bill would lower the threshold for interim measures. Instead of a prima-facie case and the requirement of a risk of a serious irreparable damage to competition, it would suffice if an infringement “appears predominantly probable” and interim measures are “necessary for the protection of competition or because of an imminent threat of serious harm to another undertaking” (Sec. 32a GWB-Draft).
Further, to increase legal certainty for the roll-out of new digital business practices or the sharing of relevant data or inputs, undertakings would be entitled to ask the Bundeskartellamt to confirm that, subject to new findings, it currently saw no ground to intervene.
Blueprint for reforms in other jurisdictions?
Some have suggested that Germany’s pioneering role may increase the risk of legal fragmentation within Europe, and that, in the interest of a single digital market, Germany should wait for antitrust reforms at the EU level before it acts. Yet, Germany is the largest Member State of the EU. It cannot (always) wait for European consensus to move ahead. In addition, history has shown that European legislation benefits from the experience of national forerunners.
In fact, the most significant proposal of the German reform, that is, to subject undertakings with paramount cross-market relevance to additional obligations, has already gained support from the largest EU Member States. On February 4, 2020, Germany, France, Italy, and Poland invited Margrethe Vestager, Vice-President of the European Commission, in a joint letter to “identify systemic actors against objective criteria taking into account specificities of digital markets, such as emergence of digital platforms with paramount importance for competition, that should be subject to specific scrutiny, and, if relevant, to a specific regulatory framework, in accordance with the particular responsibilities of these players.”[23] In Germany, the term “systemic actors” had always been considered as a synonym for “undertakings with paramount cross-market relevance”.[24]
Germany, France and Poland had already agreed on similar wording in a joint memorandum on “Modernizing EU Competition Policy” in July 2019.[25] The wording reflects similar terminology used to address the same concerns in the UK Furman Report[26] and the EU Special Advisors’ Report.[27] With this academic backing, and with Germany and France having the EU Council presidency in 2020 and 2022 and further support from Italy and Poland, it would be surprising if this joint approach would not be adopted by the Commission or by other European institutions.
In fact, the European Commission’s agenda on “Shaping Europe’s Digital Future,” published on February 19, 2020, suggests that the proposals are already being considered for European legislation. Describing plans for a “fair and competitive economy,” the agenda points out that “some platforms have acquired significant scale, which effectively allows them to act as private gatekeepers to markets, customer and information. We must ensure that the systemic role of certain online platforms and the market power they acquire will not put in danger the fairness and openness of our markets.”[28]
This is backed up by the European Parliament’s Annual Report on Competition Policy, published on February 25, 2020, calling on the Commission “to identify the key digital players and establish a set of indicators to define their systemic nature,” with a view to introducing “targeted regulation when practices become systemic,” including a ban on self-preferencing.[29]
Of course, amending the law on abuse of dominance at the EU level, that is, Article 102 TFEU, is not particularly likely because it constitutes primary EU law that can only be amended through a new European treaty (between all Member States). However, the European Commission can initiate secondary legislation (directives or regulations) to give effect to the principles set out in Article 102 TFEU[30] or to harmonize the law governing the digital economy.
Disrupt antitrust laws – or regulate systemic platforms?
While legislative amendments at the EU level (and beyond) to address the challenges posed by Big Tech are just a matter of time, the actual question now is whether it is sufficient and adequate to amend general antitrust laws, or whether it would not make more sense to set up a specific regulatory framework for digital gatekeepers.
Google’s favoring of its service Google Shopping alone has been investigated for many years by the FTC, the European Commission, and the Brazilian and Turkish competition authorities – all with different outcomes. The different outcomes, however, were not the consequence of significantly diverging legal standards – they were based on different findings of fact, including the intensity of gathering relevant evidence and assessing technical details. The various functionalities, algorithms, designs, and interdependencies of multi-sided intermediaries are so complex, and the information asymmetry between competition authorities on the one hand and globally-active, super-dominant tech giants with unlimited resources and opaque business practices on the other are so high, that even the most motivated and highly skilled competition authority quickly reaches its limits. It is becoming even more complicated when it comes to remedying any identified self-favoring or manipulation in the crawling, indexing, ranking, or display of commercial results. As a consequence of the “in-built asymmetry here between the one doing the illegal behavior and the law-enforcer,” in a speech of March 2, 2020, Competition Commissioner Vestager also signaled a need for ordering remedies before a market has tipped rather than after a lengthy investigation, when the remedy may come too late.[31]
Outlook: ex-ante regulation of systemic platforms at the EU level
Against this background, it is understandable, if not inevitable, that in adopting the initiatives in Germany (supported by France, Italy and Poland), both the European Commission[32] and the European Parliament[33] suggest complementing antitrust law with sector-specific ex-ante regulation. The Commission acknowledges that “[c]ompetition policy alone cannot address all the systemic problems that may arise in the platform economy. Based on the single market logic, additional rules may be needed to ensure contestability, fairness and innovation and the possibility of market entry, as well as public interests that go beyond competition or economic considerations.”[34] Consequently, in Q4 2020, the Commission will further explore “ex ante rules to ensure that markets characterized by large platforms with significant network effects acting as gate-keepers, remain fair and contestable for innovators, business, and new market entrants.”[35] This finds its backing in the European Parliament calling on the Commission “to assess the possibility of imposing ex ante regulatory obligations where competition law is not enough to ensure contestability in these markets, therefore avoiding competitors’ foreclosure and ensuring that emerging bottlenecks are not perpetuated by the monopolization of future innovation.”[36]
It would, of course, not be the first time that antitrust laws and specific regulation go hand in hand in reaching the same economic objectives. The same combination was globally successful between 1990 and 2010 in establishing competition in previously monopolized network industries. In this historical example, while the “essential facilities doctrine” developed under antitrust laws initially provided the general foundation for access to bottleneck infrastructures, sector-specific ex-ante access regulation took over later on to deal with the technical details and price settings that antitrust authorities and (even more so) courts were deemed to be ill-suited to supervise on a day-to-day basis.[37] The same development can be foreseen for ensuring the fair, non-discriminatory provision of systemic intermediation services by platforms with significant market power.
Footnotes
[1] Furman et al., Unlocking digital competition, Report of the Digital Competition Expert Panel, May 2019, https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/785547/unlocking_digital_competition_furman_review_web.pdf.
[2] Crémer/Montjoye/Schweitzer, Competition policy for the digital era, 2019, https://ec.europa.eu/competition/publications/reports/kd0419345enn.pdf.
[3] Australian Competition and Consumer Commission, Digital Platforms Inquiry – Final Report, June 2019, https://www.accc.gov.au/system/files/Digital%20platforms%20inquiry%20-%20final%20report.pdf.
[4] Stigler Committee on Digital Platforms, Final Report, September 2019, https://research.chicagobooth.edu/-/media/research/stigler/pdfs/digital-platforms---committee-report---stigler-center.pdf.
[5] Report by the Commission ”Competition Law 4.0.” A New Competition Framework for the Digital Economy, September 2019, https://www.bmwi.de/Redaktion/EN/Downloads/a/a-new-competition-framework.pdf.
[6] https://www.bmwi.de/Redaktion/DE/Downloads/G/gwb-digitalisierungsgesetz-referentenentwurf.pdf. An (unofficial) English translation of key provisions can be found here: https://www.d-kart.de/wp-content/uploads/2020/02/GWB10-Engl-Translation-2020-02-21.pdf.
[7] Joint Letter of P.Altmaier, S.Patuaneli, B. Le Maire and J.Emilewicz, February 4, 2020, https://g8fip1kplyr33r3krz5b97d1-wpengine.netdna-ssl.com/wp-content/uploads/2020/02/Letter-to-Vestager.pdf.
[8] Legislative proposals of the Commission must be adopted by the EU Parliament and the Council (which represents the Member States). The EU Council Presidency allows for the coordination of the Council.
[9] European Commission, Shaping Europe’s digital future, COM(2020) 67 final, February 19, 2020, https://ec.europa.eu/info/sites/info/files/communication-shaping-europes-digital-future-feb2020_en_4.pdf.
[10] Directive (EU) 2019/1 of 11 December 2018 to empower the competition authorities of the Member States to be more effective enforcers and to ensure the proper functioning of the internal market, OJ L 11, January 14, 2019, p. 3-33, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32019L0001.
[11] See Section 35 para. 1 No 2 Draft Bill.
[12] Section 39a Draft Bill.
[13] Reasoning of the Draft Bill, p. 74.
[14] Reasoning of the Draft Bill, p. 72.
[15] Reasoning of the Draft Bill, p. 80.
[16] Cf. UK Furman Report (fn. 1), para. 4; EU Special Advisors‘ Report (fn. 2), p. 4, 51; BMWi, Modernisierung der Missbrauchsaufsicht für marktmächtige Unternehmen, 2018, p. 40.
[17] Reasoning of the Draft Bill, p. 83.
[18] Ibid.
[19] Reasoning of the Draft Bill, p. 64.
[20] Section 19a para. 2 sentence 2 Draft Bill.
[21] See summary of the first hearing of the Draft Bill amongst industry associations at: https://www.d-kart.de/blog/2020/02/19/conference-debriefing-18-gwb10-verbaendeanhoerung/; Kühnen, WuW 3/2020, Editorial; Höppner, WuW 2/2020, 71-79 (in German) https://research.owlit.de/document/55f9c868-9344-3ed3-ab8e-603d78aa630e?searchId=123924937
[22] Reasoning of the Draft Bill, p. 62.
[23] Joint Letter of P.Altmaier, S.Patuaneli, B. Le Maire and J.Emilewicz, February 4, 2020, https://g8fip1kplyr33r3krz5b97d1-wpengine.netdna-ssl.com/wp-content/uploads/2020/02/Letter-to-Vestager.pdf
[24] This was explained by A. Mundt, president of the Bundeskartellamt, on January 16, 2020 in its speech at the 8th L&W-Wettbewerbstag 2020 in Hamburg.
[25] https://www.bmwi.de/Redaktion/DE/Downloads/M-O/modernising-eu-competition-policy.pdf.
[26] UK Furman Report (fn. 1), p. 5, 9 et sub., identifying particular concerns through undertakings with “strategic market status.”
[27] EU Special Advisors’ Report (fn. 2), p. 60 et sub., pointing out the particular concerns posed by “platforms as regulators.”
[28] Shaping Europe’s digital future (fn. 9), p. 8.
[29] European Parliament, Report on competition policy – annual report 2019 (2019/2131(INI)), Committee on Economic and Monetary Affairs, A9-0022/2020, https://www.europarl.europa.eu/doceo/document/A-9-2020-0022_EN.pdf at paras 30, 31, 41 and 43.
[30] See Article 103 TFEU.
[31] Speech, March 2, 2020, Keeping the EU competitive in a green and digital world, https://ec.europa.eu/commission/commissioners/2019-2024/vestager/announcements/keeping-eu-competitive-green-and-digital-world_en with additional answers reported by MLex, March 2, 2020, Dominant tech companies may face new EU curbs to stop ´tipping` into their laps.
[32] Shaping Europe’s digital future (fn. 9).
[33] Report on competition policy (fn. 29).
[34] Shaping Europe’s digital future (fn. 9), p. 9.
[35] Ibid., p. 10.
[36] Report on competition policy (fn. 29), para. 30.
[37] See Verizon Communication, Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 415 (2004) rejecting the essential facilities doctrine because “[n]o court should impose a duty to deal that it cannot explain or adequately and reasonably supervise. The problem should be deemed irremediable”. This quote was taken from P. Areeda’s famous article on the doctrine, where the sentence continues with the following crucial clarification: “[...] irremediable by antitrust law when compulsory access requires the court to assume the day-to-day controls characteristic of a regulatory agency” (Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, Antitrust Law Journal 1989, 841, 853); similarly Byars v. Bluff City News Company, Inc. 609 F.2d 843, 864 (6th Cir. 1979) “difficulty of setting a price [...] might well justify withholding [antitrust] relief altogether”.
*Prof. Dr. Thomas Höppner is a Partner in the Berlin office.