New Zealand Law Foundation Research Reports
Last Updated: 2 April 2021
Regulation of new technology: Institutions and processes
Dr James Every-Palmer QC 28 March 2018
Table of Contents
Regulation of new technology: Institutions and processes
Dr James Every-Palmer QC,1 28 March 2018
1. Executive Summary
The rapid development and uptake of information technology has had a significant impact on how New Zealanders work and socialise, and on how our economy and occupations are organised.
Today we have almost continuous access to the Internet, increasingly through mobile devices and in particular smartphones. We use digital platforms to interact socially via Facebook and other social media sites, to access the news, to watch television and movies, and to shop. 70% of New Zealanders spend 2 or more hours online per day.2
Digital technologies have also had a disruptive effect on many industries. Uber competes against traditional taxi companies through a smartphone app which connects drivers and passengers; sets prices; determines travel routes; and allows for drivers and passengers to be rated. Google and Facebook dominate the digital advertising market by attracting users to their search and social media services, and offering advertisers the ability to target advertising based on the data they collect about their customers.
The uptake of these technologies has been extremely rapid. Although the Web has been in mainstream use since the mid-1990s, the technology which dominates our present online experiences is much more recent: Facebook was launched in 2004; the smartphone in 2007; and Uber in 2011. During its mere 14 years in existence, 2 billion users have joined Facebook.3
While consumers have quickly adopted these new technologies and platforms, Governments have generally taken a hands-off approach. Like many other countries, New Zealand has to date responded by taking relatively modest steps such as legislative adjustments to facilitate new technologies or address new harms,4 and by establishing reviews or industry groups to consider the implications of new technologies.5
3 Statista “Number of Facebook users worldwide 2008–2017” (accessed 26 March 2018).
The apparent imbalance between technological change on the one hand and regulatory change on the other, reflects the inherent adaptability of our existing regulatory systems. But, it is also likely a result of a lack of awareness and the difficulty in knowing what regulatory changes are appropriate. This means that, for example, despite the influence that platforms such as Facebook and Google wield in commerce and society, they are relatively unregulated compared with sectors that rely on more traditional technologies and business models such as schools, broadcasting, public utilities, payment systems etc.
Concerns are, however, increasingly being expressed in relation to the potential social harms from new technologies and the absence of regulations to address them.
Consider the following examples:
rise to concerns about the emergence of a “gig economy” where the traditional protections afforded to employees are lost.
As well as regulating to address new risks or to maintain an appropriate balance of rights and responsibilities, regulatory changes may be required to remove impediments to the introduction of new technologies so that we can unlock their value to society.
The electricity industry is a sector where the market rules may need to be updated in light of new technologies such as smart meters, electric vehicles, distributed generation and grid scale batteries. Traditionally consumers played a passive role of turning on the switch, using electricity which was remotely generated and then paying a monthly bill. Today it is easy to imagine transactions that were not contemplated when the current market was designed: a solar panel user may wish to gift their surplus
Amendment Bill 2017 now before a Parliamentary Select Committee), the Copyright Act review (assessing whether current law is fit for purpose in the context of new technology and changing patterns of creation, distribution and consumption of content, with MBIE due to release an issues paper in mid 2018), the Smart Grid Forum (a platform for discussion of emerging trends in the electricity and gas sectors, which has held lectures, workshops and panel discussions) and the AI Forum (a forum for discussion and investigation of artificial intelligence issues).
generation to a nearby school; or a neighbourhood may wish to use solar and wind generation with battery storage to become self-sufficient from the traditional electricity grid. These possibilities raise a number of issues in terms of ensuring that the market rules do not block desirable transactions or tilt the playfield in terms of particular outcomes.
The issues just discussed all exist today. The next wave of technologies – more advanced artificial intelligence; autonomous vehicles; the Internet of Things; and blockchain technologies – will no doubt be accompanied by another set of complex regulatory issues. Whether or not these particular technologies live up to their billings, there seems little reason to think that the following decades will be subject to less or slower change than we are currently witnessing.
The purpose of this paper is not to solve any of these particular issues.8 Rather, its purpose is to contribute to a conversation about the implications of technology change for good regulatory practice in terms of our institutions and processes. This paper asks whether some sorts of institutions and processes are likely to be better suited to accommodating and regulating technological change than others?
In order to address these questions, this paper is structured as follows:
technologies. Because new technologies are so diverse it is difficult to organise the main policy issues into simple categories, but the range of issue include updating existing regulatory frameworks; addressing new risks; regulating digital gatekeepers; data, privacy and algorithmic decision-making; competition issues and business disruption; and the broader impacts of technology on society.
institutions. In particular, this paper puts forward five proposals that are worthy of further consideration:
#1: Our existing regulatory systems should be kept as clear and up to date and as forward-looking, as possible.
#2: A central co-ordination function, such as a Technology Commission, should be established to co-ordinate regulatory efforts in relation to new technologies.
#3: We should articulate our expectations of the principles that online business should follow.
#4: We should explore the capacity for more rapid regulatory responses, for example, temporary approvals or bans.
#5: We should facilitate and harness consumer input.
The challenges of regulating new digital technologies
As noted by the Productivity Commission in its 2014 Report on Regulatory Institutions and Practices, regulation is a pervasive feature of modern life with coverage that “stretches from the workplace to the sports field, the home to the shopping mall, and from the city to the great outdoors. When it works well, it underpins our everyday transactions and interactions, allowing us to do such things as travel within and outside of New Zealand safely, buy and sell goods and services and invest with confidence, and start businesses with ease.”9
Regulations are typically drafted to deal with the problems of the day (and the technology of the day).10 Unsurprisingly, the bulk of the present regulatory system in New Zealand focuses on the physical/industrial economy and addresses 20th Century issues.
Keeping our regulatory systems up to date in response to 21st Century digital technologies appears to be challenging for four interrelated reasons.
First, is the increasing pace of change. This acceleration is illustrated by the following chart which shows how the decreasing number of years required for new technologies to be adopted by quarter of the American population (46 and 35 years for electricity and the telephone verses 13 and 7 years for the mobile phone and the web):11
10 Regulations are generally introduced in response to a perceived market failure. For example, to protect consumers from monopolies, to correct negative externalities (such as pollution), to provide public goods, or to address information asymmetries. Regulations are also introduced for more paternalistic reasons to assist consumers in making optimal decisions given difficulties with assessing risks and evaluating large information sets. See generally, Robert Baldwin, Martin Cave and Martin Lodge (eds) Oxford Handbook on Regulation (OUP, Oxford, 2010), Chapters 2 and 5; and Susy Frankel (ed) Learning from the Past, Adapting for the Future (LexisNexis, Wellington, 2011), Chapter 10.
11 The Economist “Happy birthday world wide web” (12 March 2014).
The rate of adoption of the smartphone illustrates a continuation of this trend. The first smartphone, the iPhone, was launched in the middle of 2007 and within eight years 70% of adult New Zealanders owned a smartphone.12 That is, in less than a decade our society had largely transitioned to a device that was novel in many ways: it provided geo-locational services and continuous connection to the Internet; it popularised the “app store” model; and it helped increase the usage of platforms such as Facebook.
Furthermore, these figures are likely to understate the true rate of change. That is, they focus on the physical uptake of smartphones whereas many of its features are defined in the software of its operating systems and apps, and updates and new apps are continuously available.
Secondly, there is the issue of complexity.
To continue with the smartphone example, this is a device that can substitute for a camera, a map, a music collection, a book, a phone-booth, videoconferencing equipment, a taxi company, a travel agent, and so on. While smartphones are easy and intuitive to use (just ask any parent of a toddler), the communications and geo- location technology, and the underlying business models, are much more complex and opaque than the technologies that are being displaced. So, for example, if you use a map app to search on your phone for a café in a new city, you may see the names of a few cafés close to the blue dot on your map. Nothing could be simpler to use but we may be oblivious to how the app knows where we are, why one café appeared but not another (is it curated as a form of targeted advertising?), or whether any further use is made of the information that we are in a particular location and looking for a café.
Thirdly, unlike medicines for example, digital technologies generally do not require any pre-approval process and nor are they subject to particular disclosure regimes. That is, we have treated the development and release of these technologies as matters
12 Apple introduced the first iPhone in 2007. In 2015 it was estimated that 70% of adult New Zealanders had a smartphone. See Holly Ryan “The Big Read: Showdown of the smartphones” (NZ Herald, 30 March 2017)
for private law and contract, and addressed regulatory issues that have arisen after the fact, rather than attempting to isolate any issues in advance. As a result many technologies are in widespread use before any thought is given as to the regulatory implications. Furthermore, there will typically be a significant gap between the average age of those adopters and of those in regulatory roles.
Finally, digital technologies tend to operate on a global basis. That is, New Zealand will often merely be one of the hundred or so countries that a digital platform operates in which may make it difficult to influence behaviour or enforce rules.13 Conversely, major regulatory initiatives from overseas, such as the European General Data Protection Regulation (GDPR) is likely to impact on data processesing and privacy practices around the world.
For these reasons – the increasing pace of change, complexity, the absence of pre- approval mechanisms and the global nature of digital technologies – 21st Century digital technologies have generally been adopted more quickly than regulators can form a view as to whether there may be any important regulatory consequences. As a result, it is a reasonable supposition that important gaps are starting to develop between our 20th Century regulatory frameworks and 21st Century technologies.
Literature on the regulation of new technologies
The literature on the regulation of new technologies tends to fall into one of two categories: studies set in a single technological context, or fairly philosophical and abstract analysis of how technology should interact with societal norms and understandings.
Research in the first category tends to looks at a single technology and on how it might be regulated.14 The technologies covered include, for example, healthcare robots,15 the sharing economy,16 3D printing17 and platforms such as Google and Facebook.18
13 See the discussion of issues of jurisdiction, applicable law and enforcement in Urs Gasser “Cloud innovation and the law: Issues, Approaches, and Interplay” (2014) Berkman Center Research Publication No 2014-7 at 15.
14 Lyria Bennett Moses “How to think about law, regulation and technology: problems with ‘technology’ as a regulatory target” (2013) 5 Law, Innovation and Technology 1 at 2.
15 Drew Simshaw and others “Regulating Healthcare Robots: Maximizing opportunities while minimizing risks” (2016) 22 Rich J L & Tech 3.
16 Sofia Ranchordas “Does Sharing Mean Caring? Regulating Innovation in the Sharing Economy” (2015) 16 Minn J L Sci & Tech 413. See also Vanessa Katz “Regulating the sharing economy” (2015) 30 Berkeley LJ 1067; Andrew Bond “An App for That: Local Governments and the Rise of the Sharing Economy” (2015) 90 Notre Dame L Rev 77; Hannah Posen “Ridesharing in the Sharing Economy: Should Regulators Impose Uber Regulations on Uber” (2015) 101 Iowa L Rev 405; Nayeem Syed “Regulating Uberification” (2016) 22 Computer & Telecommunications L Rev 14; Lucy Henderson “Innovators or Rule Breakers? Regulating Uber, Airbnb & The Sharing Economy” (LLB (Hons) Dissertation, University of Auckland, 2016).
17 James Beck and Matthew Jacobson “3D Printing: What could happen to products liability when users (and everyone else in between) become manufacturers” (2017) 18 Minn JL Sci and Tech 143.
18 Orla Lynskey “Regulating ‘Platform Power’” London School of Economics Law, Society and Economy Working Paper 1/2017. See also Daithi Mac Sithigh “App Law Within: Rights and Regulation in the Smartphone Age” Edinburgh School of Law Research Paper No 2012/22.
The second category contains work that focuses on issues such as the complex relationship between law, regulation, technology and society,19 the need for regulatory prudence, legitimacy and effectiveness in keeping regulations connected and engaged with new technologies,20 and the need for society and industry to develop a moral sense of the implications of new technology.21
This study is situated in between these two areas of focus.22 It attempts to provide a broad mapping of the issues faced by regulators today and a practical discussion of the sorts of institutions and processes that will operate to develop regulations for new technologies in the 21st Century.
3. What sort of regulatory issues arise from new digital technologies?
The goal of this study is to consider whether we have appropriate regulatory institutions and processes to best respond to new technologies. In order to address this question, it is helpful to first ask what substantive regulatory issues do new digital technologies typically give rise to?
As yet there is no generally accepted map of the policy issues which commonly accompany new technologies. This may be because of the novelty of these issues, but it also seems apparent that new technologies give rise to issues across the entire breadth of regulation and human endeavour, and so defy any attempt at simple classification.
Accordingly with due caution, in this section I set out six broad policy issues that often arise in relation to new digital technologies. While they do not purport to be exhaustive, they at least illustrate the variety and complexity of issues that arise and that need to be considered by our regulatory systems.
#1 Regulatory frameworks need to be updated as technologies and business models change
Regulations inevitably make assumptions about the world in which they are to be applied. If you were drafting a law in 1990 that required notice to be given in a way that was formal and could later be proven if necessary, it made sense to provide for the notice to be given “in writing”, that is with the information printed or written in a physical document. Today we can see that this is not a technologically neutral requirement, even though the draughtsperson could hardly be criticised for failing to
19 Lyria Bennett Moses “How to think about law, regulation and technology: problems with ‘technology’ as a regulatory target” (2013) 5 Law, Innovation and Technology 1.
20 Roger Brownsword “The shaping of our online worlds: getting the regulatory environment right” (2012) 20 Int J Info Tech 249.
21 Roger Brownsword “Lost in translation: Legality, regulatory margins, and technological management” (2011) 26 Berkeley Tech LJ 1321. See also Gary Marchant and others “Regulatory Frontiers: Integrating social and ethical concerns into regulatory decision-making for emerging technologies” (2010) 11 Minn J L Sci & Tech 345.
22 Other work in this vien includes Lyria Bennett Moses “Agents of change: how the law copes with technological change” (2011) 20 Griffith L Rev 763 and Julie Cohen “The Regulatory State in the Information Age” (2016) 17 Theoretical Inquiries in Law 369.
predict the development of email and other forms of electronic communications in the decade that followed.
Accordingly, it is quite common for new technologies to make new ways of doing things feasible that were not previously provided for or to otherwise clash with the assumptions and categories adopted in existing regulations. This may result in the new technology being hindered or blocked, or the new technology may be able to avoid the regulatory burdens faced by competitors. As new entrants facing obstacles, or incumbents being threatened, have a clear incentive to raise concerns with the existing regulatory framework, this is the most common way for new technology issues to come to our attention across all sectors.
The entry of Uber into the New Zealand passenger services market in 2014 gave rise to various “updating” issues. The regulatory framework is focussed on passenger and driver safety, and price comparability. At that time it was assumed that there would be two main types of service: taxi services and private hire services. Taxi services were subject to the greatest regulation: they could only be operated by approved taxi organisations, who needed to set meter tariffs, ensure safety standards were met, monitor complaints and provide a 24/7 service. Taxis could be hailed, selected at a rank or pre-booked. Private hire services, in contrast, could only be pre-booked and fare had to be agreed in advance. In both cases, drivers were required to hold a Passenger endorsement (P endorsement).
Uber (and other similar app-based platforms) provided a challenge to this model in two respects. First, incumbent taxi operators claimed that Uber was in reality offering a directly competitive hailed taxi service, but avoiding the burden of being an approved taxi organisation: the booking was made and the fare agreed in advance, but only minutes in advance, of the ride commencing.23 Secondly, Uber claimed that there should be no need for its drivers to hold a P endorsement since risks to passenger and driver safety were minimised by technology as: Uber pre-registered both passengers and drivers; drivers had no need to carry cash since payment occurred through the app; the recommended route was shown in real time; and the trip was recorded.
A review of the regime was commenced in 2015 and resulted in amendments to the Land Transport Act 1998 and related rules in 2017 to create a single category of “small passenger services”.24 The requirement for drivers to have a P endorsement was, however, retained.
Technological change is requiring regulatory frameworks to be updated in other sectors as well.
The electricity market has been regulated on the basis that there are clear delineations between different segments in the supply chain: electricity is generated at remote power stations, transmitted on Transpower’s national grid and via local distribution networks, and used by consumers who purchase electricity (and the associated supply services) from a retailer. In this basic model, the consumer is a very rudimentary participant and would not need to interact with anyone else in the market other than a single retailer: the consumer turns the lights on and off, and pays a monthly bill to the retailer.
23 Ministry of Transport “Future of small passenger services: consultation paper” (2015) at 6.
24 Land Transport Amendment Act 2017, subpart 5 of part 1.
However, the uptake of solar power, batteries and smart control systems by residential consumers opens up the potential for consumers to play roles, and engage in transactions, that were not envisaged when the market was created. For example: a consumer may wish to enter into a peer-to-peer sale of surplus solar generated electricity to a neighbour; a load aggregator may contract with multiple consumers to access their battery storage or control the timing of discretionary electricity consumption; or “micro grids” may emerge where a subdivision, street or suburb finds that it can more cheaply provide its own generation and distribution infrastructure than relying on incumbents. These new technologies and business models raise a number of issues in terms of access to metering data by multiple service providers and whether the present rules allow the transactions and create efficient incentives for investment decisions.25
Electoral advertising provides another example. Road-side billboards and TV advertisements are closely regulated by the Electoral Commission. The Commission also monitors social media, and referred a number of people to the police for posting content on election day in 2017.26
But deeper, and until recently more obscure, aspects of social media give rise to a new set of challenges in applying these rules. A simple trial carried out by Facebook in 2010 showed that by including an “I voted” stamp and including a banner showing friends who had voted could increase voter turn out by 0.4%.27 While this study was not intended to influence who voted or how votes were placed, it has been alleged that in the 2016 US Presidential election, Facebook was used to distribute Russian sponsored propaganda to significant numbers of the US electorate. It is believed that operatives posed as political activists and used the flash points of immigration, religion and race in an attempt to create division and dissent by posting on Facebook and purchasing targeted advertising.28 If Facebook was used to distribute targeted propaganda in New Zealand, to what extent does the Electoral Commission have the power, expertise and experience to investigate such actions and take appropriate action?
#2 Novel risks that may require a regulatory response
We have just seen a number of examples that illustrate the need for regulatory frameworks to be kept up to date as new technologies open up the possibilities for new transactions and new business models. A closely-related issue is that new technologies may give rise to novel risks that need to be assessed and potentially regulated through the creation of new regulatory frameworks.
Crypto currencies such as Bitcoin give rise to a number of potential regulatory issues. At the end of 2017, many people were asking whether the price of Bitcoin was being
25 Electricity Authority “Multiple Trading Relationships” (28 November 2017), requesting submissions on whether market rules inefficiently restrict flexibility for consumers who are able to use solar panels, smart controls and battery technology. See also Bridget Moon “Relative Progress of Smart Grid Development in New Zealand” (New Zealand Smart Grid Forum , 21 August 2016).
26 Electoral Commission “Referral to the police” (2 October 2017).
27 Robert Bond and others “A 61-million-person experiment in social influence and political mobilization” (September 2012) 489 Nature 295.
28 Emily Bell “Silicon Valley helped Russia sway the US election. So now what?” (The Guardian, 29 October 2017); Matt Apuzzo and Sharon LaFraniere “13 Russians Indicted as Mueller Reveals Effort to Aid Trump Campaign” (New York Times, 16 February 2018).
driven by a speculative bubble and whether consumers ought to have been protected in some way.
The price of Bitcoin rose from around USD $1,000 at the start of last year to a peak of just under USD $20,000 in December 2017 (the price as at the end of March 2018 was USD $8600). The price rise occurred despite good reasons to think that it was very unlikely that Bitcoin would ever become a mainstream crypto currency including: its limited transaction processing capacity; its high energy use per transaction; the absence of any remedies for lost or stolen passwords or transfers to incorrect addresses; the risk that the cryptography would become insecure in the face of future technological developments; the potential impact of future regulation or the issuance of State-backed crypto currencies; and the “51% risk” that a cartel of the largest miners (transaction verifiers) could exercise control over the ledger of transactions. Without minimising the difficulties involved, it seems reasonable to ask whether Governments and regulators globally could have acted earlier to spell out the risks for investors and to articulate how crypto currencies would likely be regulated in the future, for example, in terms of anti-money laundering requirements.29
The adoption of autonomous vehicles would create new risks for road users and would also change the roles of the manufacturer and driver in terms of road safety.30 Presently, vehicles must meet certain safety standards when they are imported, and after that the owner and driver of the vehicle is responsible for keeping the vehicle in a safe condition and driving it safely. Unsafe driving is discouraged in many ways, including by the potential imposition of criminal penalties. The question arises whether the manufacturer of an autonomous vehicle would face appropriate incentives to ensure road safety through the vehicle’s software, particularly with the potential absence of negligence liability due to the accident compensation scheme’s bar on civil claims for personal injury by accident.
More broadly, society faces a very different risk profile as a result of the widespread adoption of digital technologies in terms of the resilience of our social and commercial infrastructure.
Today the Internet and the various digital platforms built upon it underpin a huge variety of personal and commercial tasks. While email and online banking may be much more efficient than letters and cash/cheques, digital technologies are potentially less robust in that there are fewer work-arounds if something goes seriously wrong.
For example, in a cash-based society transactions can still be processed if the till stops working, but in a society dominated by online payments what would happen if the Internet went down for a week? In considering the resilience of our national infrastructure, it is important to consider how systems rely on digital technology and the Internet, how vulnerable these systems are to hacking or viruses, and how they would perform in a natural disaster.31
29 It will be interesting to review who the 2017 Bitcoin investors were or what their level of financial sophistication was and to ask what factors led to the price bubble: is there evidence of “pump and dump” manipulation; what role did advertorial style Bitcoin advertising play; and did the listing of Bitcoin related futures on formal exchanges give an appearance of tacit Governmental approval?
30 Michael Cameron “The ethical and legal implications of driverless cars are bigger than you think” (NZ Listener, 14 October 2017). See also Matthew Wansley “Regulation of emerging risks” (2016) 69 Vand L Rev 401 at 464–472
31 Note that the NZ Treasury’s “National State of Infrastructure Report” (2016), aimed at making our infrastructure resilient and able to contribute to growth and increased quality of life, focuses on real
#3 Regulating digital “gatekeepers”
Many digital markets tend to exhibit “winner takes all” tendencies where one company dominates. That is, despite the apparently low barriers to entry in relation to, for example, launching a social media network or an online search website or an online retailer, we see a market place where one or a small number of providers have a clear lead.32
In particular, a small number of companies – Google, Facebook, Apple and Amazon
– have grown to dominate their markets and to act as “gatekeepers” to commerce, news and social media for a large proportion of the Internet. For example, for many people Facebook is their main contact point with the Internet,33 and Google is the dominant search tool in relation to the web.34 Approximately 3 million New Zealanders have a Facebook account, with an average of 14 visits a day and with around 2.3 million people using Facebook every day. Globally, the average user spends 50 minutes a day on the site.35
A number of potential regulatory issues arise from the social, economic and political power of these platforms and their gatekeeper roles:
viewpoints and behaviour and is becoming a gateway to news for many users. However, it is not a straightforward source of news. “Fake news” – that is deliberate misinformation or hoaxes – can be harder to discern in an online environment and is often spread through social media. In addition, what
property, roading/transport and water access rather than the availability and security of the Internet and attendant digital technologies.
32 There are a number of potential explanations. First, “network effects” exist in relation to many technologies where the value of the service to a consumer depends on how many others participate. For example, people will tend to join social media platforms that are already popular, buy smartphones that already have a wide range of apps available or use taxi-apps which have the most users. Secondly, there may be barriers from switching from one platform to another. For example, to transition from Facebook to a new social media may mean abandoning a history of messages and photographs. Similarly, switching between an iPhone and an Android-based phone may mean abandoning previous app purchases. Finally, it may be difficult for users to directly judge the quality of digital platforms (which search engine is the best or which social media site has the best data security?). In relation to such “credence” services, an established incumbent may have an advantage because new users assume its present popularity is an indicator of its quality.
33 Facebook accounts for about 25% of all internet traffic (Andre Staltz “The web began dying in 2014, here’s how” (30 October 2017)) and 19% of time spent on mobile devices (Maddy Osman “28 powerful Facebook stats your brand can’t ignore in 2018” (Sprout Social, 15 February 2018)) and is how many consumers obtain news, discover retail products, and find out about live events (Christina Newberry “A long list of Facebook statistics that matter to social marketers” (Hootsuite blog, 14 January 2018).
34 Google accounts for about 75% of all search engine activity on the internet: Carolanne Mangles “Search Engine Statistics 2018” (Smart Insights, 30 January 2018).
35 See Andy Fyers “Facebook is New Zealand’s second favourite leisure activity” (Stuff, 23 March 2017).
happens if an important story or a political viewpoint does not trend and so is largely invisible on social media?
These are all weighty issues and it is notable that they have received only limited attention until the last couple of years. Platforms such as Facebook and Google were not initially seen as giving rise to any particular regulatory concerns, and they grew up in the private law realm of contract, with only ad hoc brushes with the legal system.38 However, regulators are now starting to grapple with the steps that may be appropriate to respond to these issues for example by seeing them as analogous to providers of public utility services.
#4 Data collection, privacy issues and algorithmic decision-making
The businesses that we interact with online typically know vastly more about us than their real world counterparts. Over the last decades, personal data in digital form has become more valuable.
Many online business models rely on using information about customers in terms of targeted advertising or customising services based on past behaviour. Indeed, the key comparative advantage for platforms like Google and Facebook seems to be the ability to offer targeted advertising based on detailed data collected about their users.39
Other platforms use personal data to customise content, or make pricing decisions. Accordingly, there is strong commercial pressure for online businesses to harvest data derived from our clicks, searches and other online behaviour and to keep the most detailed records possible, tied to a user’s real identity.
36 European Commission “Factsheet” (27 June 2017).
37 European Commission “Fairness in platform-to-business relations” (25 November 2017).
38 In the private law realm, the State’s role is seen as facilitating private transactions subject to the generic rules about contracts, fair trading and privacy. In this realm, any regulatory issues are worked out after the fact, see “The future of drones depends on regulation, not just technology” (The Economist, 10 June 2017).
39 Facebook’s Business page states that advertisements can be targeted based on demographics (age,
gender, relationship status, education, workplace, job titles, and more), location, interests and behaviours. For example, Facebook can targets ads for jewellery or flowers to people with a wedding anniversary coming up in the next month.
At the same time, it has become cheaper to collect, store and process data.40 Data can be collected from browsing patterns, searches, and registration requirements.
Consumers also generate information by uploading information to blogs and social media or by responding to social media.41
The collection of mass data has been accompanied by the growth of algorithmic decision-making in Government and business. That is, recommendations or decisions affecting individuals may be made by a computer algorithm. This may be on the basis of explicit decision-making rules or through techniques such as machine learning where an algorithm is trained on past cases to produce an implicit set of criteria.
There are a number of potential implications from the increased collection and automated use of personal data:
is based on the idea that information should be collected and used only in ways that are authorised by the data subject. However, the “consent” based model seems difficult to apply when general descriptions provide limited transparency over what data is harvested, who it is shared with, how it is used and what the ultimate consequences may be.44
bookseller or airline may be able to use our past purchase behaviour to predict whether we have a high or low willingness to pay for the next transaction. In theory such a firm could maximise its profitability by price discriminating, that is charging customers individualised prices depending on its estimate of their willingness to pay. While we see price discrimination in many industries and it can mean lower prices being offered to people who would not otherwise use
40 Andrew W Bagley & Justin S Brown Limited Consumer Privacy Protections Against the Layers of Big Data (2015) 31 Santa Clara High Tech L J 483; Rebecca Balebako and others Is Notice Enough: Mitigating the Risks of Smartphone Data Sharing (2015) 11(2) IS J Law & Policy Info Soc 279.
41 Information can also be gathered by data scraping Jeffrey Kenneth Hirschey “Symbiotic relationships: pragmatic acceptance of data scraping” (2014) 29 Berkeley Tech LJ 897.
42 There is evidence that people state they have a strong interest in data protection, but will provide information about personal information for a fairly modest benefit. In one experiment, most participants would disclose their date of birth and monthly income in return for a €1 discount. See Federico Morando, Raimondo Iemma and Emilio Raiteri “Privacy evaluation: what empirical research on users’ valuation of personal data tells us” (2014) 3(2) Internet Policy Review.
43 Alex Preston “The death of privacy” (The Guardian, 3 August 2014).
44 Alessandro Acquisti, Curtis Taylor and Liad Wagman “The Economics of Privacy” (Sloan Foundation Economics Research Paper No 2580411, 8 March 2016) at 43; Orla Lynskey “Regulating ‘Platform Power’” London School of Economics Law, Society and Economy Working Paper 1/2017 at 24.
the service, some people would feel that their past transactional information should belong to them in a sense, and not be used against them in the form of higher individualised prices.45
#5 Competition issues and digital disruption
Competition (or antitrust) law promotes market-based competition by prohibiting collusive conduct such as price fixing between competitors and by controlling mergers which may lessen competition.
New technologies have made applying the traditional tools of competition law more challenging. For example, novel business models (such as offering search and social media services to one set of customers for free in order to provide access to that group to advertisers) and the rapid evolution of technology can make it more difficult to define the affected markets and predict future market dynamics.46
In addition, there are questions over whether the competition toolkit will need to be expanded in the future or supplemented by targeted regulation. For example:
may also call for regulatory interventions to reduce switching costs and encourage contestability. For example, a common feature of many successful digital incumbents is that they hold large amounts of user data. Giving the user more control over this data may reduce a significant barrier to switching.
45 A related concern is the risk of data leaks, such as the leak of 50 million Facebook profiles to Cambridge Analytica that was revealed in March 2018: Olivia Solon “’A grand illusion’: seven days that shattered Facebook’s façade” (24 March 2018).
46 Particular concerns have arisen around powerful incumbents acquiring smaller businesses that do not directly compete, but which would otherwise have the potential to grow into competitors over time. For example, competition authorities have been criticised for allowing Facebook to acquire businesses such as Instagram and WhatsApp which could have evolved into competitors: see Barry Lynn and Matt Stoller “How to stop Google and Facebook from becoming even more powerful” (The Guardian, 2 November 2017).
Competition from new business models using digital technologies has given rise to “digital disruption” in many industries.47 We have seen the digital camera displace film-based businesses such as Kodak. And Uber shows how the combination of a number of technological developments (ubiquitous smartphones, online maps, credit card payments and driver/rider rating systems) allowed an app to substitute for a traditional taxi company in setting fare rates and dispatching cars to passengers.
The disruption to traditional industries brought about by new technologies is generally just a result of the usual process of competition.48 That is, we would normally welcome the introduction of a technology that can provide a more attractive offering to consumers. However, a number of regulatory concerns may still arise. For example:
infrastructure that has been regulated as a natural monopoly may in fact become subject to competition. For example, assets belonging to electricity distribution businesses may face decreased demand due to the increasing prevalence of solar power and batteries in households. As a result, it may become increasingly important to ensure that such frameworks are fair to infrastructure investors but do not result in new competitive technologies being supressed.
#6 Broader impacts on society
The adoption of new digital technologies over the last two decades has been rapid and based largely on market forces.
As a result, only limited attention has been paid to the broader impact of these technologies on society. There has been little opportunity to reflect on whether
47 Clayton Christensen in The Innovators Dilemma (Harvard Business Review Press, 1997) draws a distinction between sustaining technology and disruptive technology. Changes to transaction costs have resulted in new models such as Trademe and Uber. Tom Goodwin comments “Uber, the world’s largest taxi company, owns no vehicles. Facebook, the world’s most popular media owner, creates no content. Alibaba, the most valuable retailer, has no inventory. And Airbnb, the world’s largest accommodation provider, owns no real estate. Something interesting is happening.” “The Battle is for the Customer Interface” (Techcrunch, 4 March 2015).
48 In entering the market, Uber has decreased the value of taxi medallions New York City: Winnie Hu “Taxi Medallions, Once a Safe Investment, Now Drag Owners Into Debt” (New York Times, 10 September 2017). Judge Posner comments that sort of effect is just “the normal consequence of replacing a cartelized with a competitive market” and “‘Property’ does not include a right to be free from competition”; see Joe Sanfelipo Cabs v City of Milwaukee (7th Cir 2016) and Illinois Transportation Trade Association v Chicago (7th Cir, 2016).
49 Whether or not this classification as independent contractors rather than employees is legally correct is presently being litigated in a number of jurisdictions. See Omri Ben-Shahar “Are Uber Drivers Employees? The Answer Will Shape The Sharing Economy” (Forbes, 15 November 2017).
particular technologies enhance or diminish shared social values such as connectedness, inclusion, the development of social and intellectual skills, respect for others, rights to participate, rights to transparent decision-making, protection of our environment and the promotion of sustainability, physical security and a sphere of privacy.
To date most technological development has followed the dictates of consumer sovereignty, with the result that we have communications systems and social media platforms that fight to gain and retain our attention. One policy question is whether there are detrimental effects from a world of constant digital distractions where we check our phones a hundred times a day,50 work invades our personal space, and it becomes harder to be “in the room” during our personal time because we are simultaneously connected to the rest of the outside digital world and its likes, tweets and other alarms. If this is indeed the case, then how can we tilt the field in favour of technologies that are more likely to contribute to human flourishing?51
4. Implications for regulatory institutions and processes
Section 2 of this study noted the regulatory challenges posed by the pace of change, the complexity of new technologies, the absence of pre-approval mechanisms and the global nature of digital technologies.
Section 3 then considered the sort of substantive policy issues that are emerging in relation to new digital technologies. It is apparent that new technologies may not fit well with existing frameworks, that a single technology may give rise to a wide range of policy issues, that the issues are complex and that it may take some time for appropriate regulatory responses to emerge.
The question then is whether there are any implications for how we manage our regulatory systems. That is, what steps should we consider taking now to be in the best position to adjust our regulating settings to accommodate and/or regulate new technologies in the future?
In this section I put forward five proposals for further consideration.
#1 Ensure that existing regulatory systems are well-tuned and as forward-looking as possible
Any list of principles for good regulation will include the importance of having clearly articulated policy goals and drafting rules which are as clear, simple and forward-
50 Alex Mayyasi “Which generation is most distracted by their phones?” (Priceonomics, 26 February 2016).
51 See for example Jessica Brown “Is social media bad for you? The evidence and the unknowns” (BBC, 5 January 2018); Jesse Mulligan “Why the pursuit of pleasure is making us sad” (Radio New Zealand, 17 October 2017); and Jean M Twenge “Have Smartphones Destroyed a Generation?” (The Atlantic, September 2017); James Stewart “Facebook Has 50 Minutes of Your Time Each Day. It Wants More” (New York Times, 5 May 2016); and Paul Lewis The tech insiders who fear a smartphone dystopia (The Guardian, 6 October 2017).
looking as practicable.52 However, these ideas seem to be of particular salience at a time of rapid technological change: the tidier our existing regulatory frameworks, the easier it ought to be to accommodate new technologies.
One aspect of this is ensuring clarity of our present frameworks: Are they clearly written or have they been subject to multiple amendments so that they have become hard to follow? Do they use language and concepts that are technologically neutral? Is there an appropriate balance between general principles (which may be more adaptable for the future) and prescriptive rules (which may be easier to apply in particular cases)? And is there as much consistency as possible within and between regulatory systems?
While “regulatory stewardship” is part of the Government’s strategy for regulatory management,53 the question is whether we should make a step change in resourcing the rationalisation of our regulatory frameworks which often evolve organically from periodic amendments?
Secondly, it seems desirable to minimise fragmentation where responsibility for the framework for a particular sector is split between multiple bodies. Many sets of market rules (such as those in the electricity and telecommunications markets) have both a component in legislation and a component determined by one or more responsible regulator. While such a division may make sense in a static environment where it is possible to demarcate roles appropriately, it may cause difficulties in times of change as it may not be clear who has the mandate and responsibility for considering technological change, and any required amendments may involve both the regulator(s) and Parliament.
Finally, our regulatory frameworks and our regulators should be as forward-looking as possible. Examples of proactive consideration of new technologies include the Electricity Authority’s work on the implications for the market of increased uptake of technologies such as solar power, batteries and electric vehicles,54 and the Financial Markets Authority’s work on “robo-advice” and the status of Initial Coin Offerings.55 In this regard, it is also appropriate for the Commerce Commission to be empowered to undertaken “market studies” to have the opportunity to better understand the dynamics of new technology markets rather than waiting for complaints to emerge.56
52 See for example OECD “Guiding Principles for Regulatory Quality and Performance” (October 2011; NZ Treasury “Expectations for Regulatory Stewardship” (2013 and 2017 update); Productivity Commission Regulatory Institutions and practices (June 2014); and Ministry of Business, Innovation and Employment Building a Digital Nation (March 2017).
53 NZ Treasury “Expectations for Regulatory Stewardship” (2013 and 2017 update).
54 Electricity Authority “Multiple Trading Relationships” (28 November 2017).
55 Financial Markets Authority “Consultation: Exemption to enable personalised digital advice” (November 2017). This consultation led to an exemption being granted to enable the provision of personalised robo-advice services. Financial Markets “Initial coin offers” and “Crytocurrency services” (October 2017) provides commentary on how these new developments are regulated.
56 Legislation to amend the Commerce Act 1986 and to allow for the Commission to conduct competition studies was introduced to the House on 28 March 2018. The Australian equivalent to the Commerce Commission, the Competition and Consumer Commission, has a power to conduct market studies (s 28(1)(ca)) under the Competition and Consumer Act 2010 (Cth).
#2 A central co-ordination function should be established in relation to the adoption and regulation of new technology
In my view, there would be significant benefit from establishing a central institution to monitor and co-ordinate the way that we respond to digital technologies.57
While existing bodies will still need to address technological issues within their remits, a “Technology Commission”58 could play a number of important roles:
one of many countries facing the same issues, a global perspective is appropriate. This could encompass: analysing the regulatory issues and proposing appropriate solutions; seeking to influence the global companies; seeking to avoid a “race to the bottom” between countries in terms of content controls and tax policy; and seeking to ensure effective remedies for consumers regardless of where service providers are based. It may also be appropriate for New Zealand to attempt to borrow regulatory approaches from other countries more commonly than it does today. A Technology Commission
57 In a similar vein (but on a smaller scale), the Government has recently established a new Chief Technology Officer position to assist in the development of a digital strategy and to help “ensure that New Zealand has the right foundations in place to seize the benefits, and mitigate the challenges, of increasing technological change.” See Ministry of Business, Innovation and Employment “Chief Technology Officer” (9 February 2018).
58 Referring to the body as a “Commission” is not intended to make assumptions about the type of entity that could be formed
59 Sofia Ranchordas “Does Sharing Mean Caring? Regulating Innovation in the Sharing Economy” (2015) 16 Minn J L Sci & Tech 413 at 451, observing that experimenting on a small scale is a useful way to ensure rules keep up with changes in technology and society and to obtain information.
60 For example, the Ministry of Transport’s review of Uber and other small passenger services focussed on licensing issues, but not issues such as competition (Ministry of Transport “Future of small passenger services: consultation paper” (2015)). By contrast, the England and Wales Law Commission considered competition concerns in recommending Uber be treated as a separate market to traditional hail-able taxis: Law Commission for England and Wales “Taxi and Private Hire Services” (Cm 8824, May 2014).
would provide a natural point of contact for such inter-Governmental initiatives.
In my view, there is the potential for a significant benefit from having a central body developing regulatory expertise on technology matters and it is worthy of further consideration in terms of the nature and size of the body, and the sort of powers that would be appropriate.
#3 We should articulate our expectations of the principles that online businesses should follow
Given that regulation has not addressed many of the substantive policy issues discussed in section 3 because of their novelty and complexity, many online businesses are trying to fill the gaps. This may be as part of being a good corporate citizen, but also too avoid a potential consumer or regulatory backlash in the future. Examples include Facebook banning crypto currency advertisements61 and taking steps to address the issue of fake news on its network,62 and the Partnership on Artificial Intelligence which has been formed by industry participants to study and formulate best practices on AI technologies including fostering the use of AI for the benefit of society.63
If we could articulate the principles that we expect online businesses to follow, this might an effective way of prodding technology in the right direction in relation to areas which are difficult to regulate. This is most likely to be effective if such a charter is created in collaboration with other countries.
A charter of technology norms could seek to address the sort of substantive policy topics discussed in section 3. For example:
about the use of automated decision-making?65 If and when should there be a right to a human decision-maker?
61 Alex Hern “Facebook bans cryptocurrency adverts because so many are scams” (The Guardian, 31 January 2018). Some credit card some issuing banks have also started preventing their cardholders from buying crypto currencies on their credit cards. See Lawrence White and Emma Rumney “Banks ban use of credit cards to buy bitcoin as value tumbles below US$7000” (NZ Herald, 6 February 2018).
62 John McDuling “Inside Facebook’s crusade against fake news” (Stuff, 24 December 2017).
63 Partnership on AI.
64 Jack Balkin and Jonathan Zittrain “A Grand Bargain to Make Tech Companies Trustworthy” (The Atlantic, 3 October 2016).
65 General Data Protection Regulation 2016/679 (EU), article 22. For discussion see Gianclaudio Malgieri and Giovanni Comande “Why a Right to Legibility of Automated Decision-Making Exists in the General Data Protection Regulation” (2017) 7(4) International Data Privacy Law 243.
66 General Data Protection Regulation 2016/679 (EU), article 20 contains a “right to data portability”. The Australian Productivity Commission in recommended a similar right (Australian
One vehicle for articulating these principles would be in the form of a digital bill of rights as recently proposed as part of the “D7”, a group of seven countries that collaborate on digital issues.67 This seems like an ideal forum for articulating guiding principles for technology companies by inputting the values and experiences of a number of technologically advanced nations. A bill or charter of rights and expectations would not need to have legal teeth to influence online business. Indeed, it is probably better that such a document does not have legal imprimatur at least initially so as to allow more flexibility and room for it to evolve.
#4 We should explore the capacity for rapid regulatory responses
Our present regulatory reform processes evolved when technological change was much more gradual than it is today. While exceptional matters can be dealt with urgently, a reform process involving consultation, a report and implementing legislation would typically take two to four years or longer.
Given the speed at which new technology can be adopted, particularly if the technology is an app or other form of software, the risk of regulation lagging behind seems likely to increase substantially in the coming years.
A “regulatory sandbox” is one way of allowing technologies to be deployed on a small scale basis for testing and experimentation, with exceptions to usual regulations. Such a framework would be implemented in legislation and tied to a particular area of regulation, such as financial services.
A more radical idea is to devolve legislative type powers to a regulator. In particular, a regulator (for example, the Technology Commission) could be given powers to temporarily approve or temporarily prohibit the use of new technologies in a way that overrides existing legislation. Regulators could thereby obtain a lot of information that would otherwise be difficult to obtain from the uptake and response to such temporary measures.68 Such powers would be very controversial, and would offend the “Henry VIII principle” that delegated legislation should not have the power to override Parliamentary legislation.69 Nevertheless, it seems desirable to consider whether the importance of agility and urgency in responding to new technologies warrants delegation of such powers and whether safeguards, such as judicial oversight, could ameliorate some of the concerns.
Productivity Commission Data Availability and Use (31 March 2017) at 35. This is in progress towards a legislated Consumer Data Right: Shashank Venkat “Australia mulls over new Consumer Data Right legislation” (Cerillion, 29 November 2017).
67 Minister for Government Digital Services “Media statement: Digital nations put digital rights at the heart of agendas” (23 February 2018); and Tom Pullar-Strecker and Katie Kenny “Kiwis could get right akin to ‘judicial review’ of decisions made by computers” (Stuff, 22 February 2018).
68 See Sofia Ranchordas “Does Sharing Mean Caring? Regulating Innovation in the Sharing Economy” (2015) 16 Minn J L Sci & Tech 413 at 451.
69 Mary Harris and David Wilson (eds) Parliamentary Practice in New Zealand (4th ed, Oratia Books, Auckland, 2017) at 465–466.
#5 We should facilitate and harness consumer engagement
There is a real risk that the regulatory dialogue in relation to digital technology issues is dominated by technology companies. It can be difficult to get meaningful consumer engagement, especially as regulations for new technology are likely to be complex and multi-faceted. However, it is important to attempt to create an ongoing deliberative process of open and informed dialogue that has the ability to influence policy and includes diverse viewpoints and values.70
Consumer involvement is also important in that the solution to many of the issues relating to new technologies is likely to be found in engaged consumers who can, for example, recognise fake news and make informed choices about privacy settings.
There are not easy solutions to getting consumer more engaged, but some options include:
One of the goals of consumer engagement should be to widen the window of regulatory options that are open for discussion. That will feed into a wider reaching and more flexible process for adopting regulations fit for the 21st Century.
70 L Carson & J Hartz-Karp “Adapting and combining deliberative designs: Juries, Polls and Forums” in J Gastil & P Levine. (eds) The Deliberative Democracy Handbook, (Jossey-Bass, San Francisco, 2005) at 122.