Volume 13 – Issue 4 – November 2017

IN Focus Regulatory Issues

Regulatory Stewardship: voice of the regulator Trends and Determinants of Top Pay in the New Stephanie Winson 3 Zealand Public and Private Sectors, 1995–2014 Are Regulated Parties Customers? Tim Hazledine, Michael Wang and Kar Yew Lee 53 Ben Wauchop and Keith Manch 10 Civics and Citizenship Education in New Zealand: a New Zealand’s Port and Harbour Marine case for change? Safety Code – a case study in co-regulation Todd Krieble and Danijela Tavich 60 Keith Manch 13 Insuring Property under Climate Change A Rail Tale Belinda Storey and Ilan Noy 68 Chris Ballantyne 21 Managing the Opportunities and Risks Associated Graduate Pathway Students: three contributions with Disruptive Technologies: space law in Lessons from an Internship at Waipa- District New Zealand Council: motivations and incentives Kirsty Hutchison, Katherine MacNeill, Peter Mumford Alice Denne 75 and Val Sim 28 Two Key Lessons from a Public Sector Internship: Rapid Developments in Artificial Intelligence: accountability in public sector contracting and how might the New Zealand government client-level data respond? Matthew Macfarlane 77 Matthew Boyd and Nick Wilson 36 New Zealand Productivity Commission Internship: Income Volatility in New Zealand social sector productivity Toby Moore 44 Danijela Tavich 80 Editorial Note This issue of Policy Quarterly commences with five The five regulatory articles traverse a number of articles on aspects of government regulation in New issues. Two are conceptual. The first, by Stephanie

Volume 13 – Issue 4 – November 2017 Zealand. A separate editorial note (see below) by Dr Winson, argues that regulators have a lot of value to Peter Mumford and Keith Manch summarizes the offer in the policy development and review process. Policy Quarterly (PQ) is targeted at readers purpose and content of these articles. My sincere You may or may not agree with Winson’s assertion that in the public sector, including politicians and their staff, public servants and a wide variety thanks to Peter and Keith, and also Dr Veronica policy analysts may ‘dismiss or may not appreciate’ of professions, together with others interested Jacobsen, for their efforts in overseeing the selection, the real world of regulatory compliance. However, in public issues. Its length and style are preparation and peer review of these articles. she makes a compelling argument that regulators intended to make the journal accessible to Collectively, these pieces provide a fascinating have unique insights that can be exploited through busy readers. insight into the evolving nature of public regulation in good leadership and collaboration across regulatory Submissions: The journal welcomes contributions of about 4,000 words, written this country, not least the important quest for better systems, underpinned by effective information on any topic relating to governance, public regulatory stewardship and improved regulatory management and competency. policy and management. Articles submitted practice. Ben Wauchop and Keith Manch tackle the will be peer reviewed. Please submit articles The remaining eight articles fall into two contentious subject of whether regulated entities to the Editor: [email protected]. Although issues will not usually have single categories. First, there are five articles on various (those who are required to comply with the law) are themes, special issues may be published from topical policy issues from a range of (mostly ‘customers’ of the agencies that are regulating them. time to time on specific or general themes, academic) contributors. These cover the following Does it matter what they are called? In his seminal perhaps to mark significant events. In such matters: the ethical, societal and policy implications work on the philosophy of language, Wittgenstein cases, and on other occasions, contributions of artificial intelligence; the extent of income volatility concluded that ‘Different words create different may be invited from particular people. Subscriptions: The journal is available in PDF in New Zealand; trends in top pay in New Zealand; worlds’. From this perspective it does matter as both format on the Institute for Governance and ways to enhance education on civics and citizenship; regulators and those they regulate may condition Policy Studies (IGPS) website: http://igps. and the implications of climate change for household their behaviours based on what they understand the victoria.ac.nz/publications/publications/list/10. insurance. term ‘customer’ to mean. Readers who wish to receive it by email should register as PQ subscribers igps@vuw. The final three articles are from students in A further two articles provide insights into the ac.nz. This service is free. the Graduate Pathway Programme in the School regulatory practice world. Both deal with critical For all subscription and membership of Government: Alice Denne, Matthew Macfarlane infrastructures – ports and rail, and both deal with enquiries please e-mail [email protected] or and Danijela Tavich. These articles are based on safety. The quest for better regulatory outcomes post to Institute for Government and Policy research undertaken by these students as part can involve experimentation and the ports and rail Studies, P.O. Box 600, Wellington. Electronic Access: The IGPS directs of their internships with various public and non- safety stories show how the regulatory regimes interested individuals to its website: governmental agencies. This is the first time Policy have evolved based on what both the regulators and www.igps.victoria.ac.nz where details of the Quarterly has published such material, but I trust it regulated entities have learned from experience. Institute’s publications and upcoming events will not be the last. My grateful thanks to Dr Barbara Keith Manch emphasises the features of co- can be found. Allen, who coordinates our School’s Graduate regulation that align the mandates, incentives Permission: In the interest of promoting debate and wider dissemination, the Pathway Programme, for suggesting the idea of and capabilities of the national regulator, local IGPS encourages use of all or part of the having these students contribute to the journal and government regulators and port companies. Chris papers appearing in PQ, where there is no for overseeing the preparation of their articles. Ballantyne shows how the regulator first had to element of commercial gain. Appropriate New Zealand has a new government. The February recognise the need and then change its regulatory acknowledgement of both author and source should be made in all cases. The IGPS issue of Policy Quarterly will, accordingly, include culture, capabilities and ‘rules of thumb’ to retains copyright. Please direct requests reflections on the political and policy implications implement effectively the safety case approach that for permission to reprint articles from this of the change of government. In the meantime, I was introduced through legislative change in 2005. publication to [email protected]. trust that readers will find much of interest in the Finally, New Zealand has implemented the most Editor: Jonathan Boston November issue. modern regime for regulating outer space activities Guest Editors: Peter Mumford and Keith Manch and high altitude vehicles originating from its territory. Editorial Board: Barbara Allen, Jonathan Boston Kirsty Hutchison, Katherine MacNeill, Peter Mumford Guy Beatson, Roger Blakeley, David Bromell, and Val Sim describe this regime and what it seeks to Simon Chapple, Jo Cribb, Girol Karacaoglu, achieve. Put into the context of regulating disruptive Gerald Minnee, Anneliese Parkin, Mike Reid and Andrea Schollmann Articles on regulatory issues technologies and markets so that the benefits are ISSN: 2324-1098 (Print) This issue of Policy Quarterly continues the regulatory secured while managing the risks, this article puts a ISSN: 2324-1101 (Online) theme that has emerged over the past couple of stake in the ground for a permissive approach which Volume 13, Issue 4 – November 2017 years. This recognises the critical role that regulation facilitates innovation and entrepreneurship, but with Copy Editor: Rachel Barrowman plays in maintaining a society where people are safe rigorous checks and balances to ensure that what is Design & Layout: Aleck Yee Cover Photography: Aleck Yee and secure, and an economy that supports growth permitted is safe and secure. Production: Alltex Design and prosperity. Recent regulatory failures have Proof Reader: Vic Lipski graphically demonstrated the costs when regulation Keith Manch and Peter Mumford on behalf of does not deliver on its promises, and parliament has the Government Regulatory Practice Initiative ‘upped the game’, by requiring under the State Sector (G-REG). With thanks to Dr Veronica Jacobsen for her Act that chief executives exercise ‘stewardship’ over contribution to the editorial review of the five articles. the legislation administered by their agency. Stephanie Winson

Regulatory Stewardship Voice of the effectively over time. In his 2014 article ‘Why departments need to be regulatory stewards’, Jonathan Ayto said the then Regulator initial government expectations for regulatory stewardship ‘can be viewed as introducing some very basic asset management concepts to the regulatory In 2014 the New Zealand Productivity Commission inquiry environment’ (Ayto, 2014, p.27). The government issued updated on regulatory institutions and practices concluded that ‘The expectations for regulatory stewardship performance of New Zealand’s regulatory system is in need when it recently published Government Expectations for Good Regulatory Practice, of improvement – in particular around developing and from which the above definition of regulatory stewardship is lifted. The maintaining the capability needed to effectively implement updated expectations describe the regulation and the need to oversee and manage the overall regulatory stewardship role as including responsibility for: system’ (Productivity Commission, 2014, p.2). Since then · monitoring, review and reporting on existing regulatory systems; there has been much talk of regulatory stewardship. This · robust analysis and implementation article considers what it is and the importance of the role of support for changes to regulatory systems; and the regulator in achieving it. · good regulatory practice (New Regulatory stewardship regulatory system(s) within which an Zealand Government, 2017a, p.3). Regulatory stewardship is the adoption of agency works (New Zealand Government, The updated expectations differ from a whole-of-system view, and a proactive, 2017a). It looks at systems as assets that the earlier iterations because they now collaborative approach to the care of the need to be well managed to deliver contain a part B which explicitly includes a section on good regulator practice. This Stephanie Winson is the General Manager, Legal and Policy at Maritime New Zealand. reflects a growing awareness that

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 3 Regulatory Stewardship: voice of the regulator

Figure 1: The regulatory cycle encompasses different sets of legislation that seek to achieve a commonly defined System strategic planning public interest objective. Defining the system in this way helps clarify that various agencies may have regulatory roles within Review and the same system, which, once identified, evaluation assists with establishing who contributes to system performance. onme nvir ntal However, it is also important to E annin sc g consider regulatory systems in a wider Agency strategic Policy planning and S e t development context, such as that identified by the f a c g e priortisation k n n e i e

d a r Functional h Productivity Commission, which

o b

o

t

m

a

i

r

regulatory l

d c

n o

k

f e

o described it as including ‘the institutions,

r system r

e s m

P principles and processes through which

l regulations are made, implemented, Operationa intelligence enforced and reviewed’ (Productivity Commission, 2014, p.28). These things influence how agencies participate and engage with each other in pursuit of the Operating model design Instrument design defined regulatory objectives. Both and implementation and preparation perspectives support a more holistic view of the ‘system’ which underpins regulatory Source: New Zealand Government, 2017b, p.5 stewardship. stewardship extends beyond design and · good regulatory practice and review of regulatory regimes. Regulators regulator competence; The missing link – the regulator will welcome this shift in focus because · the need for ongoing monitoring of A careful review of the updated the role played by them in the success or the regulatory system; and expectations and the government’s failure of regulatory systems is well · the need for transparent and effective regulatory management strategy reveals known. Julia Black’s analysis of past engagement with stakeholders and that they capture best practice concepts regulatory failures makes this point very affected parties. and tools to consistently manage, maintain clear: At its heart, regulatory stewardship and improve regulatory systems over time. means that all parties involved in the Many existing instruments and guidelines, a striking feature of all the regulatory regulatory system need to work together such as the impact analysis requirements disasters analysed here is the central to ensure that the system remains fit for (Cabinet Office Circular, 2017) or role played by failures of governance purpose. The key point is that regulatory the Legislation Design and Advisory and leadership within organisations, systems are created to serve a defined Committee guidelines (Legislation in both regulators and regulated firms public interest and policy agencies and Design and Advisory Committee, 2014), … Also striking are the consistent regulators must collaborate to deliver the reflect similar standards and expectations, failures of organisations, particularly defined public interest outcomes, with particularly in relation to the design and regulatory organisations, to due regard to the views of the public and review stages of regulation. In fact, most coordinate in the operation of the regulated parties. This is a continuous and agencies should be able to easily identify regulatory system, failures which are evolving situation which requires ongoing practices and processes they adopt to meet exacerbated the more complex the management and cooperation. The these expectations. system. (Black, 2014, pp.6-7) government’s regulatory management Despite this, there is general consensus strategy includes a diagrammatic that regulatory stewardship has not yet It is notable that part B of the updated representation of the regulatory cycle (see been collectively demonstrated or expectations captures all the key elements Figure 1) to illustrate how this should achieved. Why is this, and what is missing? that have featured in past regulatory occur in a functioning regulatory system To answer these questions it is useful to failures, including: that changes and evolves. consider the approach taken by · the need to adopt a whole-of-system departments in preparing their regulatory view; What is a regulatory system? stewardship strategies and reflect on how · active collaboration between different Departmental regulatory strategies they recognise and integrate the regulator government agencies; describe regulatory systems at regime level in meeting the described stewardship role · a focus on improving the quality of (such as the financial markets regulatory in the updated expectations. legislation; system or the workplace safety regulatory Current regulatory system oversight is system). This is important because it generally led by policy agencies that are

Page 4 – Policy Quarterly – Volume 13, Issue 4 – November 2017 structured as distinct departments. However, regulatory systems often exist across departments or agencies. A simple diagram (see Figure 2) illustrates how this arises. There are two issues that this presents. The first is that system oversight requires a clear cross-agency approach, and the second is that it also requires direct involvement of regulators with functions in that regulatory system, irrespective of which agencies they are in. This is important because it is well known that Steps to achieving regulatory stewardship needed clearer leadership and a more regulatory outcomes depend almost It could be argued that the fact that the active centre (Productivity Commission, entirely on regulators. The OECD has updated expectations include a section on 2014, p.402). If one accepts that various concluded: ‘Regulators are playing an good regulator practice addresses the need agencies may have a role to play in any increasingly important role in delivering to include regulators in the stewardship given regulatory system, it becomes even economic and societal objectives as well as effort. However, this alone will not be clearer that a focus on regulatory system being tasked with regulating more enough. Even where regulatory functions stewardship depends on system leadership complex situations … As such regulators exist within the same agency, the current to achieve this. This starts at the top and are key state actors with responsibilities approach to regulatory stewardship is requires senior state sector leaders to make and therefore are accountable for the often quite ineffective in ensuring that commitments to achieving regulatory delivery of policy outcomes’ (OECD, there is a true understanding and meeting stewardship. 2014, p.15). of minds in the regulatory management The State Sector Act supports this Where regulatory design, monitoring cycle. This issue has been recognised approach in two ways. The first is that and review is treated as a policy exercise by some agencies. The Department of under section 4A the state services that seeks ‘feedback’ from regulators and Internal Affairs has made a conscious commissioner has the responsibility to regulated entities, it does not recognise decision to adopt ‘regulatory stewardship’ ‘promote a culture of stewardship in the the pivotal role played by regulators. as a practical approach to regulation State services’. This provides an There is often a cavernous divide between (Department of Internal Affairs, 2017, opportunity to establish clear state sector these parties. Policy analysts generally rely p.12). This clearly seeks to leverage leadership for promoting a system view. on theoretically objective and logical the value of regulators for stewardship The second is the express stewardship constructs to develop the policy solutions purposes and illustrates that practical obligations on state sector chief executives that underpin regulatory frameworks. steps need to be taken to ensure that for, among other things, ‘the legislation The challenge is that these are intended to departments effectively engage the administered by the department or be deployed by practical technicians to regulator (or regulators) within the departmental agency’. Interestingly, the influence or change the often subjectively system. obligations of chief executives are not driven behaviours of people. The dynamic Specific steps that can be taken to collective and their stewardship focus is and complex problems that are often improve the involvement of regulators primarily on ‘legislation’, not regulatory faced on the ‘front line’ by regulators are should preferably be arranged around the systems. The obvious difficulty with this is not always susceptible to singularly clear four key themes that emerge from the that it is likely to be more narrowly and objective responses and may be updated expectations. They are: focused than system leadership if each significantly influenced by drivers and · modelling leadership that promotes a department meets this obligation by only beliefs that policy analysts may simply whole-of-system view; focusing on the legislative regimes that dismiss or not appreciate. · engaging in genuine collaboration they are expressly tasked to administer. In Searancke and others have noted that (between policy agencies, regulators fact, existing regulatory stewardship ‘Current thinking locates the regulator at and stakeholders); strategies seem to evidence exactly that, the pivot of a complex system that requires · adopting good information which may be because the current many actors in many different roles to all management practices to support departmental approach doesn’t necessarily play their part’ (Searancke et al., 2014, insight and improvement; and promote active involvement by the wider p.55). It is exactly this system complexity · promoting regulator competence. state sector in developing these strategies. which requires regulators to be intimately The 2017 regulatory strategy published involved, not least because they are best Leadership that promotes a whole-of-system by the Ministry of Business, Innovation placed to report on the system, and view and Employment (MBIE) notes that an identify what is needed when changes are The Productivity Commission concluded all-of-system view is one of the three required. that improving the regulatory system dimensions of regulatory stewardship and

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 5 Regulatory Stewardship: voice of the regulator

‘means understanding how all the parts of importance of relationships between regulatory outcomes is almost always a a regulatory system work together and are participants in a regulatory system co-operative effort: by the government, performing’ (Ministry of Business, (Ministry of Business, Innovation and amongst regulators, the regulated, and Innovation and Employment, 2017a, Employment, 2017a, p.20). These charters the broader community’ (OECD, 2014, p.10). Despite this intention, a more set out very clearly what the objectives of p.15). This applies to policy agencies and detailed review of MBIE’s strategy suggests the relevant regulatory system are and regulators alike. The important issue is that this is still a work in progress because how all agencies within that system will ensuring that engagement is done in the regulatory systems identified do not work together to monitor its performance. a genuinely co-operative and effective necessarily include all the regulatory It is crucial for effective regulatory manner. agencies that have roles in their identified stewardship that such frameworks are While there is much evidence of regulatory systems. For example, Figure 1 mandated and supported by senior leaders regular collaboration and engagement on page 10 of the strategy seeks to identify from all the agencies concerned, so that between parties, it is not uncommon to who carries out each regulatory function they can model the collaborative hear concerns about levels of consultation across MBIE’s regulatory systems. For the behaviours and culture needed to ensure or engagement in the regulatory financial markets and health and safety at that these charters are living documents. management process. Regulators often work systems, key regulatory agencies Actions to provide system leadership express concern that their views on policy such as the Reserve Bank of New Zealand, as described above are relatively new. As design and review are not being sought or heard, while many of the departmental Establishing a genuine collaborative stewardship strategies would suggest that policy departments have a different view. participation model with regulators (and There appears to be engagement asymmetry that results in these different other stakeholders) at the core of the perceptions. regulatory cycle is more likely to improve Work by Alberto Alemanno (professor of European Union law and regulation) the policy design process and ultimately helps to explain the cause of this phenomenon. He has concluded that regulatory management. ‘public engagement has not yet become part of the policy process’ (Alemanno, 2015, p.4). He identifies three types of Maritime New Zealand and the Civil they evolve, the key improvement will be engagement: Aviation Authority are not included. This to ensure that they truly encompass a (1) Public communication: policymakers omission will undoubtedly reduce the system perspective, because collective convey information to the public. reliability of the system performance view. governance and leadership that promotes Due to the one-way information flow, The result is that the strategies may not a system view is likely to be the single the public is not involved. Public identify cross-cutting regulatory interests most transformative step that can be input is neither foreseen nor expected. (or agencies) and therefore cannot fully taken to achieve effective regulatory (2) Public consultation: Policymakers identify ‘the system’ or its performance. stewardship. receive information from members of Collective leadership of all the Done well, it will require genuine the public within the framework of a stewardship obligations (at system level) collaboration between policy and government-initiated process. The is required to achieve a proper whole-of- regulatory agency leaders and their input gathered is perceived as system approach. Each agency can people, as discussed below. It will also representative of societal opinions on contribute to this by identifying the require individual agencies across the the subject. leadership commitment it will make to wider state sector to commit resources (3) Public participation: members of the regulatory stewardship. The Department more effectively to support stewardship public and policymakers exchange of Internal Affairs has implemented a endeavours. It should also result in the information. As a result, unlike with strategy that is ‘mandated and owned at a benefits identified by the Productivity the previous two forms of senior leadership level’ and supported by Commission, such as shared solutions engagement, public participation the establishment of a role responsible for and efficiencies (Productivityinvolves some form of dialogue the regulatory stewardship work Commission, 2014, p.29). between policymakers and programme (Department of Internal stakeholders. By involving some Affairs, 2017, p.12). This illustrates a clear Genuine collaboration deliberation, public participation may recognition by the department that strong Once the leadership framework is agreed lead to change in the opinions of both agency leadership is a crucial step. and established, there is a need to ensure parties. (ibid., p.10) MBIE has implemented regulatory that agencies collaborate effectively. The While Alemanno’s work focuses on charters, which it considers recognise the OECD has noted that ‘Achieving good engagement with the public, the concepts

Page 6 – Policy Quarterly – Volume 13, Issue 4 – November 2017 that underpin the three types of Keith Manch in this issue of Policy into work programmes to ensure that engagement are relevant to the Quarterly). This illustrates that genuine reviews actually take place, it is critical engagement between policymakers and engagement can work at different levels of that regulatory agencies form a cohesive regulators. All three types of engagement regulatory design and implementation view of the kind of information that is are important, but the partnership and have remarkable results. Focusing on needed to support this. concept in the public participation option this is likely to generate the most The regulatory stewardship strategies is likely to be the least common. Yet if significant opportunity for improvement being published by various departments deployed appropriately and genuinely it in this area. certainly demonstrate that progress is can be very effective. being made in this regard. However, many Traditional views of policy advisors Information management for insight and state that there is a lack of robust data and appear to favour input from regulators at improvement a limited evidence base, which challenges the point of evaluation and review but not For existing regulatory systems, their ability to draw reliable insights on necessarily as an integral part of the policy stewardship is aimed at ensuring that the system performance. Regulators can play design or regulatory strategy phase. The regulation remains fit for purpose. This a significant role in changing this because government regulatory management requires effective monitoring, evaluation they have the advantage of operational strategy expressly includes ‘stakeholder and review. The Productivity Commission knowledge. One way of doing this is to feedback’ as part of a functioning noted that ‘In-depth reviews of regulatory establish effective intelligence models regulatory system. If that feedback is obtained in the same way as in the public Effective regulator input into stewardship participation engagement model, it is more likely to achieve a fully rounded is not only dependent on the conduct of system view. Front-line experience gives regulators a considerable advantage in policy agencies; it is also conditional on understanding practical issues and likely sound regulator competence. behaviours or reactions to regulatory interventions. Regulators also know the limits of their own capacity and capability regimes have often followed a crisis, within regulatory agencies to ensure this to execute regulatory initiatives. Policy rather than being part of a systematic and occurs. While many have already started agencies must gain a deep understanding strategic approach to review’ (Productivity this process by setting up dedicated of these issues to make durable policy Commission, 2014, p.374). This is not intelligence units, it may be useful to decisions or inform redesign options. specific to New Zealand. In an expert consider this as part of a whole-of-system Establishing a genuine collaborative paper for the OECD, Cary Coglianese view to avoid duplication of effort and participation model with regulators (and pointed out that governments around inefficiency. This can be achieved through other stakeholders) at the core of the the world have ‘paid remarkably little information sharing and joint efforts on regulatory cycle is more likely to improve attention to analysing regulations after information collection instead of relying the policy design process and ultimately adoption or to evaluating the impacts of on individual agency efforts. The regulatory management. the procedures and practices that govern Government Regulatory Practice Initiative MBIE’s 2017 regulatory stewardship the regulatory process itself, so-called (G-Reg) has taken steps to address this strategy reveals that such mechanisms are regulatory policy’ (Coglianese, 2012). and recently hosted a workshop on being trialled (Ministry of Business, The obvious consequence of failing to effective collection, collation and Innovation and Employment, 2017a, do regular evaluation and review is that analysing of information. p.18), while Inland Revenue’s regulatory there is no ability to gain a reliable insight stewardship strategy indicates that aspects into the effectiveness of the regulatory Regulator competence of such an approach are already being system. Monitoring and evaluation Effective regulator input into stewardship applied through workshops and co-design depends on reliable information and is not only dependent on the conduct options (Inland Revenue, 2017, p.6). agreed frameworks for understanding and of policy agencies; it is also conditional Inland Revenue has described the benefits assessing system performance. A variety on sound regulator competence. The as follows: ‘Having policy and service of actors play a crucial role in ensuring updated expectations now include the design officials working throughout the that information is available to undertake expectation for regulators to ‘contribute Generic Tax Policy Process creates a better such a task. Policy agencies play a critical to wider regulator capability-building outcome’ (ibid., p.8). role in avoiding ‘imprecise regulatory initiatives within the state sector where Maritime New Zealand has gone a step objectives’ (Productivity Commission, there are common interests and benefits further and used this model very 2014, p.380), while regulators contribute from collective action and leadership’. successfully to review and redesign the to this by collecting data and information To address the issue of regulator New Zealand Port and Harbour Marine as they implement regulations. While an competence, G-Reg was set up. The Safety Code (see companion article by agreed cycle of review needs to be built initiative includes the introduction of

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 7 Regulatory Stewardship: Voice of the Regulator

Table 1: Summary of identified attributes of regulator culture national qualifications for regulatory Attributes of a functional regulator culture practitioners (from entry level to advanced practice), activities to share knowledge A culture embracing the organisation’s role as an educator and facilitator of compliance and good practice across agencies and A culture that places a high value on robust, evidence-based regulatory decisions sectors, and an arrangement with Victoria

A culture that values operational flexibility and adaption to changes in the regulatory University for a chair in regulatory environment practice. To ensure that regulator knowledge is A culture that values continuous learning effectively contributing to regulatory A culture where internal debate is normal stewardship, it is crucial that regulators have the right capability to understand A culture of transparency, openness and accountability policy objectives and evaluate actions and A culture that places great value on organisational independence and impartiality behaviours of others in pursuit of the public interest outcomes being sought. A culture recognising the significance of civic responsibility that accompanies coercive powers of the state This also requires competence to recognise patterns of behaviour that may point to Subcultures that align with the overarching objectives of the organisation more significant issues. John Braithwaite

Figure 3: Regulatory stewardship themes within the regulatory cycle

Design

Problem definition Regulatory impact assessment System view Improvement Regulatory options and leadership Regulatory role/objective setting

Report on vulnerabilities and opportunities for Policy agency improvement

Standard setting Guidance development Evaluation and Role clarification Evaluation assessment Regulatory Public Implementation agencies Regulator(s)

Evaluation frameworks Regulated Publishing procedures entities and requirements Education Review cycles included Genuine Enforcement in programmes Collaboration Audit

Insight Data collection information analysis Regulator capability Operational review and culture Test operational assumptions

Monitoring

Page 8 – Policy Quarterly – Volume 13, Issue 4 – November 2017 points to exactly such an example Another dimension of regulator meet these expectations, agencies focus on involving the American FBI not competence is the need to create the right embedding them in a way that promotes a recognising the pattern of small frauds culture. The Productivity Commission collaborative whole-of-system view. and loan arrangements which ultimately describes this as a ‘regulator culture’ To achieve this, it is suggested that led to the 2008 global financial crisis (Productivity Commission, 2014, p.86), efforts should be directed towards (Braithwaite, 2016, p.25). which is summarised in Table 1. These progressing the four themes mentioned Historically, many regulators have cultural elements match the updated above. The concrete steps that can be recruited their employees from the sector expectations for good regulator practice taken to do this are to establish a they are charged to regulate. While such very closely. The G-Reg initiative will framework for the regulatory cycle in sector technical knowledge is critical, in provide opportunities for improvement which the identified public interest the absence of regulatory skills they are across all of these. As it strengthens objectives of the system are placed at the compromised in their ability to engage regulator competence at individual and core, pulled together by genuine with people and exercise the significant organisational level, regulators will be collaboration between all parties, judgement required to make nuanced and much better positioned to participate in including competent regulators, who are complex regulatory choices. The director system leadership, genuine collaboration led from the top to promote a whole-of- of the Penn Program on Regulation, Cary and providing the greater insight that is system view in which good information Coglianese, notes: required for successful regulatory practices allow for reliable insight and stewardship. improvement over time. Figure 3 gives a Yet as vital as it is for a regulator to diagrammatic representation of this possess adequate technical skill and Conclusion framework. knowledge, such expertise is only one The updated government expectations In the meantime, a working group of necessary component of regulatory for good regulator practice are a marked the Government Regulatory Practice excellence. By itself it is not sufficient. improvement on earlier iterations Initiative is developing best practice To move from good regulation to because they recognise the role of the guidelines – from a regulatory practice excellent regulation, the regulator also regulator. They also clearly articulate perspective – in an effort to demonstrate needs to master the people side of sound ‘asset management concepts’ which how a strengthened contribution by regulation. Regulation, at its core is should improve regulatory stewardship. regulators to regulatory policy can relational. (Coglianese, 2017, p.12) However, it is important that in seeking to improve regulatory stewardship.

References Alemanno, A. (2015) ‘Stakeholder engagement in regulatory policy’, paper Government, http://www.mbie.govt.nz/about/our-work/roles-and- prepared for the consideration of the OECD, https://papers.ssrn.com/ responsibilities/regulatory-systems-programme/stewardship, accessed sol3/papers.cfm?abstract_id=2701675, accessed 26 August 2017 23 August Ayto, J. (2014) ‘Why departments need to be regulatory stewards’, Policy Ministry of Business, Innovation and Employment (2017b) ‘Regulatory Quarterly, 10 (4), pp 23-7 capability (G-Reg Initiative’, http://www.mbie.govt.nz/about/our-work/ Black, J. (2014) ‘Learning from regulatory disasters’, Sir Frank Holmes roles-and-responsibilities/regulatory-systems-programme/cross-cutting- Memorial Lecture, Policy Quarterly, 10 (3), pp.3-11 issues/regulatory-capability-g-reg-initiative?searchterm=Government+ Braithwaite, J. (2016) ‘Responsive excellence’, in C. Coglianese (ed.), regulatory+practice, accessed 31 August Achieving Regulatory Excellence, Washington, DC: Brookings Institution New Zealand Government (2017a) Government Expectations for Good Press Regulatory Practice, Wellington: New Zealand Government, http:// Cabinet Office Circular (2017) ‘CO 17(3): impact analysis requirements,’ www.treasury.govt.nz/regulation/informationreleases/pdfs/good-reg- Department of the Prime Minister and Cabinet, https://www.dpmc. practice.pdf, accessed 23 July govt.nz/publications/co-17-3-impact-analysis-requirements, accessed New Zealand Government (2017b) Building Effective Regulatory 23 August Institutions and Practices, Wellington: New Zealand Government, Coglianese, C. (2012) Measuring Regulatory Performance: evaluating the http://www.treasury.govt.nz/regulation/system/strategy, accessed 28 impact of regulation and regulatory policy, expert paper 1, Paris: OECD, August 2017 http://www.oecd.org/gov/regulatory-policy/framework-for-regulatory- OECD (2014) The Governance of Regulators, Paris: OECD Publishing, policy-evaluation.htm, accessed 23 July 2017 http://www.oecd.org/gov/regulatory-policy/the-governance-of-regulators- Department of Internal Affairs (2017) Regulatory Stewardship Strategy 9789264209015-en.htm, accessed 22 August 2017 2017/18, www.dia.govt.nz/Regulatory-stewardship-programme Productivity Commission (2014) Regulatory Institutions and Practices, final Inland Revenue (2017) Regulatory Stewardship Strategy, http://taxpolicy. report, Wellington: New Zealand Productivity Commission, http://www. ird.govt.nz/publications/2017-other-regulatory-stewardship/overview, productivity.govt.nz/sites/default/files/regulatory-institutions-and- accessed 26 August practices-final-report.pdf, accessed 22 August 2017 Legislation Design and Advisory Committee (2014) Legislation Advisory Searancke, G., P. Mumford, K. Simpson and M. Steel (2014) ‘Governing Committee Guidelines: guidelines on process and content of legislation, the regulators: applying experience’, Policy Quarterly, 10 (1), pp.54-8 http://www.ldac.org.nz/, accessed 23 August 2017 Ministry of Business, Innovation and Employment (2017a) MBIE’s Regulatory Stewardship Strategy 2017/2018, Wellington: New Zealand

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 9 Ben Wauchop and Keith Manch

Are Regulated Parties Customers? How does a regulator refer to the individuals or organisations for? Who benefits from its work? Whose interests should it please? it regulates? Are they customers, even though they are not It’s an ongoing question buying a product or service, and often have little choice in The debate around whether regulated the matter? Are they to be referred to as regulated entities, entities and people (regulated parties) should be referred to as and/or treated like obligatees, licensees, taxpayers, businesses, employers or one customers is a recurring one. The issue of a number of other terms of this kind, their identity defined originally arose during the late 1990s and early 2000s as governments worldwide by their specific rights and obligations under the law? But attempted to promote better delivery of government services through the guise of what does this mean for regulatory agencies implementing ‘customer-focused government’. A paper multiple regimes? from the British government is indicative of this time: Why does it matter? ers, and within organisations that have a This question matters because the current mix of functions which include, but are To deliver lasting results, emphasis on improving how government not limited to, regulatory activity. At organisations need to embed engages with citizens in the broad sense,1 worst, regulatory staff in organisations customer focus throughout the and on improving the quality of regulation with broader functions where a ‘customer system. Implementation must start by and regulatory practice specifically,2 is as focus’ mantra is embraced can feel understanding the needs, expectations strong as it has ever been. If these parallel disenfranchised and limited in their and behaviours of the public and then activities are to succeed, clear thinking is ability to do their jobs effectively. On by adjusting every aspect of the required. closer inspection, this tension should organisation to align with customer Although questions about how to refer perhaps not be surprising: different labels values. (Barker, 2001, emphasis to regulated parties may appear trivial, raise questions that go to the heart of added) they often provoke strong debate between some important questions of regulatory otherwise agreeable regulatory practition- practice.3 Who is the regulator working Although largely a focus of government strategies rather than academic literature, Ben Wauchop is a Principal Advisor in the Ministry of Business, Innovation and Employment. Keith Manch is the Chief Executive of Maritime New Zealand. the prominence of this approach provoked

Page 10 – Policy Quarterly – Volume 13, Issue 4 – November 2017 a number of responses from academics methodologies and techniques for a required between a regulator and with a regulatory practice interest. wide range of government services, regulated party. Malcolm Sparrow, writing in 2000, argued methodologies and techniques which that regulators needed to be wary about tend to refer to and utilise concepts of So what’s the answer? becoming customer focused, as the customer centricity;5 and One way to address this is described regulated entity is often ‘not paying for · the use of ‘customer frameworks’, by the United States Food and Drug the service, often does not want, and will ‘customer segmentation models’ or Administration (FDA) in a 2004 white not be pleased by it’. He was concerned ‘personas’ to inform the delivery of paper, ‘Defining the customer in a that a customer orientation may lead to regulatory services. regulatory agency’ (FDA Quality Resource ‘exclusion or neglect of enforcement These approaches present a number of and Development Team, 2004). In that capabilities’ (Sparrow, 2000). It is also opportunities for regulatory agencies. paper the FDA discusses relationships evident that referring to regulated parties Used appropriately, they can help with the industry it regulates as involving as customers (bearing in mind the regulators move beyond simple models a spectrum of customer interactions ‘normal’ meaning of the word and the focused on the individuals and and notes that its customer relationship often associated concepts such as ‘the organisations they regulate and identify with industry will differ depending on customer is always right’)4 will create the root cause of complex regulatory the transaction involved. Essentially (as certain expectations, specifically about problems. discussed in Appendix 2 of the paper), the whose interests are being served through regulatory activity. Alford, in 2002, detailed a number of Although contemporary service design academics who had criticised the customer-centric approach, particularly has expanded its concept of a service because it devalued the idea of ‘citizenship’. in a way that can … account for His contribution was to outline a number of reasons regulated entities were more of the work of a regulator, it is definitively not customers, including because they do not have choice and they still orientated around a ‘customer’ or do not consume the good or service the ‘user’ .... service is designed to deliver. Instead, he argued, regulated entities and government formed a more complex, three-way These methodologies and techniques proposition is that regulated parties are interaction, with the public as the third have been largely adopted from areas customers when receiving support and party (Alford, 2002). outside regulation, however, and bring guidance from the FDA, but not customers Despite the passage of time, the debate with them terminology and values that when subject to enforcement action. about whether to refer to regulated parties more easily relate to private sector service Readers of this article who are familiar as customers has not gone away. Within delivery, or public service delivery of non- with the Tip Top ice cream advertisements the Government Regulatory Practice regulatory services. Although from 2006 may be reminded of the Initiative – a network of New Zealand contemporary service design has ‘togs, togs, undies, undies’ television central and local government regulators – expanded its concept of a service in a way advertisement that asked how far you can the issue about whether or not to use the that can, with careful thought, account for go from the beach in your togs before you term is raised regularly. A quick survey more of the work of a regulator, it is still are regarded as wearing undies.6 shows that varying approaches are in orientated around a ‘customer’ or ‘user’ of Another approach would be to accept place. We can contrast two New Zealand that service. This does not necessarily that regulated parties can never be regulators’ approaches: Inland Revenue’s align with regulatory practice approaches considered customers in the ordinary Multinational Enterprises: compliance which focus on improving compliance sense of the word. Regulation is ultimately focus (Inland Revenue, 2016) mentions among regulated parties, where the value the exercise of the coercive power of the the word ‘customer’ six times; Maritime derived is in the form of regulatory state. It is intended to get people to do New Zealand’s Compliance Strategy outcomes that benefit the public. things they might not otherwise do, or (Maritime New Zealand, n.d.) mentions If care is not taken, agencies can fall stop people doing things they would the term not once. into a similar trap to the one Sparrow and otherwise do. The customer is not always More recent use of the term seems to Alford identified – applying broad right if the customer is being regulated. have slightly different origins to the late strategic frameworks that promote certain Regulators can, however, and should, 1990s/early 2000s high-level government regulatory practice techniques over draw on customer frameworks or service strategies outlined above. Drivers include: others, in a way that is not really design techniques to improve the way · the application and promotion of responsive, but predetermined by the they deliver the outcomes they are charged service delivery and service design ‘label’ chosen to define the relationship with delivering. There is little doubt that

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 11 Are Regulated Parties Customers? adopting innovative design thinking and How much more useful would it be for way. The debate will move on. The continuous improvement approaches regulators if the framing was as follows: purpose and nature of the relationship incorporating appropriate customer required between regulators and regulated service principles will make regulatory As a process, design thinking can help parties will first be defined through careful practice more successful, as long as it is fit regulators get closer to the parties they assessment of the relationships required for purpose for the particular regulate, uncover what they need to in what can be significantly different circumstances in which it is being used. support, encourage and require them to regulatory constructs,7 without the The widespread engagement in and comply, and develop innovative influence of labels that might be loaded support for the Government Regulatory approaches to achieve this. It is with meaning that is potentially unhelpful; Practice Initiative is evidence that particularly useful for addressing then practices can be put in place to regulators have a deep interest in ‘wicked’ problems, for being more ensure that those relationships deliver on improvement. person-centric and for encouraging the outcomes desired. A common lexicon This will require an acceptance that innovation. may not be entirely possible – or necessary regulation and regulatory practice has a – but a deeper understanding of the special character; it is not simply another Of course, it’s more complicated than underlying issues may be. line of government service delivery, for this. But the point is, if it is accepted that 1 Highlighted by the Design Thinking approach led by the the reasons set out above. In turn this regulation and regulatory practice is Department of the Prime Minister and Cabinet and the would require those leading and different from service delivery in the Continuous Improvement ‘better every day’ approach led by the State Services Commission. This process is described at influencing the way government agencies general sense, that the recipients of https://www.dpmc.govt.nz/our-programmes/policy-project/ policy-methods-toolbox/design-thinking/journey-mapping. undertake improvement activity to step regulatory services are not ‘customers’ in 2 Highlighted by the 2013 commissioning of the New Zealand back and consider how they can reframe the ordinary sense of the word, but the Productivity Commission to inquire into regulatory institutions and practices and the comprehensive government response their offerings to cater for regulators, same kinds of techniques and thinking to its subsequent report. alongside non-regulatory service delivery. that are applied to improving customer 3 In this article ‘regulatory practice’ relates to the range of activities that are undertaken to support, encourage Take, for example, this extract from services can be used, with appropriate and require compliance with regulatory obligations, from information and education to administrative sanction and the Department of the Prime Minister modification, to improve regulatory prosecutions. and Cabinet’s description of ‘design activity, then the suite of design thinking 4 The phrase ‘The customer is always right’ was originally coined in 1909 by Harry Gordon Selfridge, the founder of thinking’: and continuous improvement tools being Selfridge’s department store in London, and is typically used by businesses to convince customers that they used can be modified accordingly. This will get good service at this company and convince As a process, design thinking can help article might be thought of as an invitation employees to give customers good service (http://www. huffingtonpost.com/alexander-kjerulf/top-5-reasons-customer- public service providers get closer to to government’s leaders in customer service_b_5145636.html). customers, uncover their unmet needs, service improvement thinking to engage 5 For example, the Design Thinking approach led by the Department of the Prime Minister and Cabinet and the and develop innovative products and with leaders in regulatory practice to Continuous Improvement ‘better every day’ approach led by the State Services Commission. services to meet those needs. It is address the tensions and risks associated 6 Accessed 17 September 2017 https://www.youtube.com/ particularly useful for addressing with referring to regulated parties as watch?v=h-Lx2ihpGbc. 7 There is a wide range of different types of regulation, ‘wicked’ problems, for being more customers – by making those tensions and including prescriptive and performance-based; and positive and negative regulation, in the sense that in some areas a person-centric and for encouraging risks go away. licence is required to operate, while in others no licence is innovation. (Department of the With this approach we can move past required but you can be sanctioned if you do not do what is required. In some areas regulation is focused on individuals Prime Minister and Cabinet, 2017, labels, and innovation can be brought to and business; in others just on individuals. Each requires emphasis added) regulatory practice in a very considered different systems, processes, skills and engagement methods on the part of the regulator.

References Alford, J. (2002) ‘Defining the client in the public sector: a social-exchange Inland Revenue (2016) Multinational Enterprises: compliance focus, perspective’, Public Administration Review, 62 (3), pp.337-46 Wellington: Inland Revenue, http://www.ird.govt.nz/ Barker, J. (2001) Customer-focused Government: from policy to delivery, resources/6/2/62414b82-6ab8-4017-b04d-cc5d950cab47/ London: Public Services Productivity Panel compliance-focus-2016.pdf Department of the Prime Minister and Cabinet (2017) ‘Design thinking’, Maritime New Zealand (n.d.) Maritime New Zealand Compliance Strategy, https://www.dpmc.govt.nz/our-programmes/policy-project/policy- https://www.maritimenz.govt.nz/about/what-we-do/compliance/ methods-toolbox/design-thinking documents/Compliance-Strategy.pdf FDA Quality Resource and Development Team (2004) ‘Defining the Sparrow, M. (2000) The Regulatory Craft: controlling risks, solving customer in a regulatory agency’, white paper, https://www.fda.gov/ problems, and managing compliance, Washington, DC: Brookings downloads/drugs/developmentapprovalprocess/manufacturing/ Institution Press odmanufacturingpracticescgmpfordrugs/ucm176384.pdf

Page 12 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Keith Manch

New Zealand’s Port and Harbour Marine Safety Code a case study in co-regulation

The inaugural annual national forum on the 2016 New Zealand’s ports and harbours (Ministry of Transport, 2007, p.7). While the Zealand Port and Harbour Marine Safety Code took place 2004 code was well intentioned and in July 2017 in Wellington. The 2016 code replaced a code created some improvement in the safety management culture in New Zealand’s originally put in place in 2004. Participants in the forum ports and harbours, over time concerns emerged about its relevance, effectiveness included the 2016 code partners: port companies, regional and efficiency as a way of managing councils/unitary authorities, Maritime New Zealand, as well successfully the risks it was intended to deal with. as maritime industry representatives, and other government These concerns were addressed by establishing the 2016 code as part of a co- agencies with an interest in maritime safety. The forum regulatory mechanism which is defined represented an important waypoint in the journey from the by a memorandum of understanding (MoU) (Maritime New Zealand, 2017a) development, implementation and review of the 2004 code, signed by representatives of port to the development and implementation of the 2016 code as companies, councils and Maritime New Zealand. While not legally binding, the a key part of the regulatory system that seeks to manage port MoU commits the parties to principles of tripartite joint ownership of the 2016 and harbour risks. code, voluntary adoption of mutually The 2004 code was put in place following Millennium. That ship grounded and agreed standards, collaboration, a series of shipping incidents in 2002– the subsequent investigation highlighted commitment of human resources to 03, including a serious incident in the the need for improved regulatory support delivery of an agreed work plan, Gisborne port involving the Jody F arrangements to manage risks in New information sharing, and joint funding of a secretariat. The 2016 code establishes Keith Manch is the Chief Executive and Director of Maritime New Zealand and has had a long involvement in regulatory practice capability development activity in New Zealand. clear responsibilities and accountabilities

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 13 New Zealand’s Port and Harbour Marine Safety Code – a case study in co-regulation

through a co-regulatory mechanism. That Box 1: History and role of pilots mechanism is established through legislation. The body responsible for its and pilotage implementation (the Gas Industry In its article ‘History of maritime pilots’, the French maritime pilots’ federa- Company Limited) is a formally tion notes that pilots have a long maritime history, starting around the time the constituted ‘industry body’ under the terms of the Gas Act 1992. However, the exchange of goods began in the Mediterranean basin, and reference to the term port and harbour safety co-regulatory pilot (as a guide of the ship) appeared in Ezekiel’s book in the 6th century BC2 mechanism is an entirely voluntary (Fédération Française des Pilotes Maritimes, 2017). The article notes that the arrangement which involves government ancestor of the maritime pilot can also be found in the travel literature, in a book (both central and local) and industry called Periplus of Erythraean Sea, written in the 1st century AD. Fast forward to establishing an MoU, developing a code, the 17th century and it is noted that ‘In Europe, since the beginning of the 17th and putting in place an implementation century, the pilotage is regulated in all the countries by national laws, called body and process. This mechanism “Pilotage Acts”, and by local regulations of ports, fixing the professional status of operates in addition to core legal the pilots in each country.’ requirements relating to safety and In New Zealand, in 1842 a Harbour Regulations Ordinance was issued, environmental protection in the port and divided into four parts: pilots and pilotage; quarantine; harbour regulations; and harbour environment. Effectively, the penalties. The modern version of this ordinance is the Maritime Rules, part 90: code adds value by providing guidance pilotage, the objective of which is to maintain the contribution of pilotage to the and support to parties with accountabilities and responsibilities under the law. safety of navigation, the protection of the marine environment and the efficiency of seaborne commerce, set minimum national standards while enabling port-spe- Why ports and harbours need to be cific risks to be addressed, provide a licensing regime for pilots and pilotage-ex- regulated empt masters within the maritime document provisions of the Maritime Transport In her keynote speech to the inaugural Act 1994, recognise and support industry best practice, and ensure the provision national forum, Belinda Vernon, deputy of pilotage services. chair of the Maritime New Zealand Authority,1 said: which operate in addition to the core Co-regulation New Zealand’s economy relies on the regulatory framework that underpins Referring to co-regulation as a mechanism safe operation of our ports and safety in the port and harbour is consistent with its position as part harbours. With more than 99% of environment. The MoU is clear that there of a taxonomy of the available forms exports and imports by volume is no constraint on the ability of councils of regulation as discussed by Freiberg moving through our ports, safety of or Maritime New Zealand carrying out (2010); and it reflects the generally shipping in New Zealand ports and their regulatory functions. understood description of co-regulation harbours is essential. It is important This article is a case study of the as involving industry and government not only economically but also in journey that led to the forum. It introduces (the regulator) collaborating to develop terms of potential risk to life, co-regulation as a regulatory mechanism, and administer arrangements that address property and environment from a canvasses why ports and harbours need to an area of risk, with those arrangements serious shipping accident. be regulated and identifies the key players accompanied by a ‘legislative backstop’, as Weather events, larger vessels, in the regulatory system. It describes the discussed by Compliance Experts (n.d.) more cruise ships, changing journey to the development of a co- and the Ministry of Consumer Affairs technology and conflicts between regulatory approach in which there is (1997). Co-regulation sits on a spectrum commercial and recreational use, all active acceptance of shared accountability of regulatory mechanisms which goes contribute to the changing marine for delivering agreed outcomes. It from full government regulation, through operating environment, creating concludes with reference to the key design co-regulation, to self-regulation, with challenges and opportunities. features of the co-regulatory mechanism each involving more or less government Managing the risks associated with and views of its effectiveness in the context intervention, as set out by Allen+Clarke these is paramount. (Vernon, 2017) of good regulatory practice. (2012). In essence co-regulation has In addition to being a case study of the features of both government regulation While this was a statement related to journey to co-regulation, it is a story of and self-regulation. the 2016 code, it describes well the nature central and local government agencies Within each mechanism there is also a of activity and risks associated with ports working together effectively to address spectrum of ‘levels’ of government and harbours. This does not in itself mean risks of national significance. intervention. For example, the gas that regulation is required to manage industry in New Zealand is regulated those risks.

Page 14 – Policy Quarterly – Volume 13, Issue 4 – November 2017 In plain terms, the rationale for regulation is that the risks won’t be Box 2: History and role of managed effectively if users are left to their own devices. The regulatory impact harbourmasters statement supporting changes to port and According to Captain Rinze K. Mast, writing in Port Technology International harbour safety regulation in 2013 (n.d.): (Ministry of Transport 2008) referred to The office of harbour master can be traced back to 317 BC at least. It the series of incidents that occurred in was in that year that a Cyprian businessman, with the name of Zeno of 2002-03. Those incidents revealed a Citium (c.336-c.264 BC), encountered a storm off the port of Athens, pattern of failures in port safety Greece. As at that time Zeno could not avail himself of the services of governance and operational procedures an Athenian Harbour Coordination Centre he was shipwrecked near (not only confined to ports where the port entrance and was rescued by a bookseller, who also held the incidents had occurred) that supported post of harbour master and bath-superintendent. … In more recent the need for regulation. years the harbour master’s office can be traced back to The Nether-

Key players in the port and harbour lands, Middelburg 1513 and Rotterdam 1554, the United Kingdom, 3 regulatory system Bristol 1669, and Belgian Ports about 1800. (Mast, n.d.) Port companies, councils and Maritime Closer to home, Maritime New Zealand’s history timeline on its website records New Zealand are all involved directly that in 1841 Auckland’s first harbourmaster was appointed. and/or indirectly in the regulatory system While the legislative underpinnings for harbourmaster functions referred that applies to ports and harbours. Two to above are not clear, today in New Zealand the role of the harbourmaster is of the primary means of addressing risks defined by reference to the functions and powers set out in the Maritime Trans- in the port and harbour environment port Act 1994, sections 33E and 33F (as amended in 2013). In summary, a are through the functions of pilots and harbourmaster’s purpose is to ensure maritime safety in relation to ports, har- harbourmasters, who are employed by bours or waters for which he or she has been appointed as a harbourmaster by port companies and councils respectively. a regional council. The harbourmaster has a variety of powers to achieve this, Maritime New Zealand has regulatory including giving directions regarding navigation practices and ships’ masters’ oversight of safety and environmental actions in respect to ship movements, and regulating and controlling traffic and matters, and pilotage and harbourmaster navigation in certain circumstances. functions, through the Maritime Transport Act 1994. Port companies came into being as a navigation safety issues. Their a significant risk to the safety of shipping result of the Port Companies Act 1988, the responsibilities in this area are typically and, potentially, to the marine purpose of which is to promote and met through employing harbourmasters environment. This is done through legally improve efficiency, economy and(or contracting the provision of enforceable interventions, should such performance in the management and harbourmaster services). While action prove necessary. Maritime New operation of the commercial aspects of harbourmasters provide services, they are Zealand also employs the harbourmaster ports. This gives a clue to their primary regulators. Box 2 sets out briefly the for the port at Taharoa, which is an focus being to operate commercially, not history and role of harbourmasters. offshore terminal used for loading to operate as regulators. Port companies In some cases harbourmaster services ironsands onto purpose-built vessels. are generally owned (in full or in part) by are provided by port company employees. The regulatory system in respect to councils, but operate at arm’s length. Port For example, in Nelson the harbourmaster ports and harbours is somewhat complex. companies employ, or contract, pilots to is also the marine operations manager for Given the integrated nature of the roles, assist in guiding ships into harbours. Port Nelson, and a pilot in that port and responsibilities and accountabilities, it is Pilots are service providers, not regulators. region. In Taranaki a similar situation apparent that clarity between parties, Box 1 sets out briefly the history and role exists, with the harbourmaster function supported by good communication and of pilots. being undertaken by the port marine cooperation, is necessary for the system to Councils’ responsibilities are set out in operations manager. As these combined work effectively. This provides fertile the Maritime Transport Act 1994 functions may be perceived as involving ground for a well-designed co-regulatory (following amendment in 2013) and conflicts of interest, arrangements are in mechanism. include harbour navigation and safety, oil place involving deputy harbourmasters to spills and other issues related to marine ensure that functionally independent The journey to a co-regulatory approach pollution. They are effectively the ‘local’ decision making occurs where necessary. The journey towards today’s co-regulatory regulators in the port and harbour Maritime New Zealand’s approach began with the Transport environment, with the ability to set by- responsibilities include taking action in Accident Investigation Commission and laws to manage a prescribed set of local situations where port management poses the Maritime Safety Authority (as of 2005

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 15 New Zealand’s Port and Harbour Marine Safety Code – a case study in co-regulation

Maritime New Zealand) investigations of The Code and Guidelines have been a It might be thought that this 2004 the Jody F Millennium and other incidents big step forward for improvements to approach was essentially a co-regulatory in 2002–03. Those investigations the safety of maritime operations for mechanism (i.e. involving industry and identified a pattern of problems, and ports and harbours. government (the regulator) collaborating noted that the current arrangements The new Code specifies the to develop and administer arrangements did not deliver sufficient rigour and responsibilities of all participants in that address an area of risk). However, the consistency in port and harbour safety the port and harbour system comment ‘Although the code is voluntary’ management. The statutory powers (the including regional councils, territorial followed by the list of steps required, and ‘legislative backstop’) available to deal authorities, port companies, and the the positioning of the Maritime Safety with these systemic problems were not MSA. Authority as signing off on risk considered to be of much use in doing so The Guidelines provide a clear assessments, periodically auditing (Ministry of Transport, 2007, p.7). steer on best practice for harbour safety management plans and The Maritime Safety Authority implementing measures for risk developing guidelines, challenges whether this was the case. In effect a regulatory mechanism had been put in place that was In retrospect it might be thought of as intended to operate through cooperation and be voluntary, but the authority’s imposed self-regulation, which is neither position as the ‘owner’ of the code government regulation, co-regulation nor (Ministry of Transport, 2007, p.8) and the responsibilities it held were essentially the self-regulation as those mechanisms are same as those that applied in its statutory regulatory function. This meant that its generally understood. true character was not one of co- regulation. In retrospect it might be thought of as imposed self-regulation, which is neither government regulation, recommended the development of a assessment, hydrographic surveys, co-regulation nor self-regulation as those national port safety code, modeled on the aids to navigation, and environmental mechanisms are generally understood. United Kingdom Port and Marine Safety factors affecting safe access and Another key aspect of the minister’s Code, or the making of maritime rules to operations in ports and harbours. speech was its announcing that: the same effect. However, the Maritime In terms of timings, Transport Act 1994 did not provide any implementation of the Code has A law review, programmed for the statutory authority for a mandatory code already begun. Although the code is coming year, will consider what new of the type proposed, so it would only be voluntary: or amended legislation is required to possible to establish such a code as an • Regional councils are being asked support the implementation and administrative measure, with voluntary to complete harbour risk ongoing development of the safety compliance (ibid.). Thus, in consultation assessments by June 2005 which system. A review of maritime rules with industry, local government and will include an assessment of dealing with pilotage will follow the relevant central government organisations, port-related marine operations in completion of regional risk the 2004 code was developed. It was their region. assessments and safety management accompanied by guidelines covering • Once MSA has signed off on these systems. matters such as port and harbour risk risk assessments, councils are will assessments and safety management move toward developing safety Ultimately, the review Duynhoven systems, hydrographic surveys, aids to management plans by June 2006 referred to didn’t commence until 2007. navigation, power line waterway crossings, which will be reviewed and However, this statement tended to support environmental factors and aquaculture approved by MSA. the possibility that the code could be areas and marine farms. There was no MSA will then periodically audit made mandatory at some point. body with any formal underpinnings harbour safety management systems Indeed, Maritime New Zealand (either legislative or voluntary) established to ensure compliance. expressed the view in response to the to oversee its implementation, and much Over the next two years, MSA has Ministry of Transport’s 2007 Port and of the responsibility sat with the Maritime programmed further work on Harbour and Navigation Safety Safety Authority. Code-related guidelines and Management discussion document Harry Duynhoven, minister of competency standards, covering vessel (Ministry of Transport, 2007) that a lack transport at the time, in a speech to the traffic services and harbourmasters. of legislative backing for the code left the National Ports Forum on 23 November (Duynhoven, 2004) system with an inherent weakness. 2004 made the following comments: However, changes to the Maritime

Page 16 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Transport Act 1994 that did occur finally Table 1: Key findings of the environmental scan in 2013 did not provide specifically for a Dimension Key findings in relation to the code mandatory code. Those changes did Relevance The code remained relevant but needed refreshing to renew sector improve the ‘legislative backstop’ by commitment to the process and position of the code, and to strengthen introducing what the minister for building nationally consistent safety management practices. and construction, Maurice Williamson, Use of the word ‘approval’ in respect to Maritime New Zealand’s role on behalf of the minister of transport in confirming that a safety management system met code requirements referred to in Parliament as: appeared to confuse accountability for safety outcomes. Effectiveness The code was an effective catalyst for establishing a culture of measures that specify the maritime safety management, and in some regions the level of enthusiasm for safety responsibilities of port collaboratively developing a safety culture was obvious. operators and the related powers of There were strong indicators, but insufficient evidence, to conclude that the Director of Maritime New the code had delivered a consistent standard of port and harbour safety management across New Zealand. Zealand to intervene in the interests Maritime New Zealand’s inability to resource its code implementation of maritime safety. These measures role had compromised effectiveness. will support and complement the Efficiency The leadership qualities and maritime knowledge of harbourmasters had voluntary New Zealand Port and a bearing on how efficiently code implementation occurred at the regional Harbour Marine Safety Code 2004. level. (Williamson, 2013) Maritime New Zealand’s inability to resource its code implementation role had compromised efficiency. Equity Councils vary in size and rating base and some found the financial Maritime New Zealand’s environmental and human resource requirements to meet their obligations were quite scan – the platform for a true co-regulatory onerous. approach Status The weight of opinion was strongest towards maintaining a voluntary As these legislative changes were emerging, code. Some smaller councils thought making it mandatory would ensure progress with the implementation of the that it was prioritised at the local level. 2004 code was increasingly a topic of discussion in the maritime community. have a funding stream to support it. Given Zealand/regional council steering group There were concerns about the level of that it was anticipated that a proper review that operates to address matters of attention being paid to it. In particular, and approval process took about 8–10 days common interest, commenced an discussion of concerns at a 2013 meeting of one person’s time, this was not an environmental scan6 to determine how of the Harbourmasters’ Special Interest insignificant matter (Ministry of Transport, relevant, effective, efficient and equitable 4 Group triggered Maritime New Zealand 2007, p.21). This situation was far from the code remained. The environmental to step back and consider the future of the satisfactory in terms of having appropriate scan also addressed whether the port and 2004 code. levels of documentation and assurance of harbour community considered the code An example of the basis for the concerns safety management systems, as was should remain voluntary.7 A representative was that, as of late 2013, despite the original intended by the 2004 code. of the Port Company Chief Executive intentions associated with the introduction However, the view was widely held Group8 was also invited to take part in of the 2004 code, while Maritime New that the advent of the 2004 code had overseeing this scan. Ultimately, the Zealand had approved safety management brought about a culture of systems-based collective leadership shown by this systems in eight ports and harbours, nine risk assessment and safety management in oversight group was an important ports and harbours were still without safety New Zealand’s ports and harbours, and contributor to the success of the review of management system approval. The there were good examples of this.5 For the 2004 code and development of the co- majority of the nine had submitted example, whether or not all safety regulatory mechanism and 2016 code, and documents to Maritime New Zealand, but management systems had been reviewed this continues to be the case. there was uncertainty about the status of and approved by Maritime New Zealand, The scan confirmed that, as discussed the safety management system documents as compared to the situation in 2004, above, while the development and in some cases (i.e. whether they were at a safety management systems did exist in all implementation of the 2004 code might standard that could be approved). The of New Zealand’s ports and harbours; and on the face of it appear to be an example reality of this was that Maritime New some regional councils and ports had of co-regulation, in comparison with the Zealand had not prioritised resources to strong and functional port and harbour 2016 approach that would emerge this fulfil the role it had under the code. safety meetings that drove collaborative was not really the case. Maritime New Zealand has a broad remit decision making around safety matters. and other priorities constantly emerged. Following internal consideration of Findings of the environmental scan With the 2004 code work not being part of the future of the code, Maritime New The findings of the scan were reported in its formal regulatory role, that work did not Zealand, through a joint Maritime New February 2014 (Maritime New Zealand,

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 17 New Zealand’s Port and Harbour Marine Safety Code – a case study in co-regulation

2014) according to the key dimensions key design features are: a tripartite a regulatory system or component that of: relevance; effectiveness; efficiency; memorandum of understanding; joint isn’t delivering net benefits. While no equity; and status. These findings ownership of the 2016 code, with clear quantitative cost-benefit assessment was identified key matters that would need objectives; voluntarily adopted and undertaken in respect to the 2004 code, to be addressed to make a refreshed mutually agreed standards; commitment concerns raised led to the environmental approach successful. They also provided of human resources to support delivery of scan. That scan highlighted problems and a platform for establishing a truly co- an agreed work plan; information sharing; weaknesses which meant that the express regulatory mechanism. The key findings and joint funding of a secretariat. intention, set out in the 2004 code, of are summarised in Table 1. ‘securing the future safety of marine Consideration of the findings led to a Effectiveness of this approach operations in New Zealand ports and set of recommendations that were The annual report of the New Zealand harbours’ was apparently not being met accepted by the joint oversight group and Port and Harbour Marine Safety Code (Maritime Safety Authority, 2004). informed the development of the co- (Maritime New Zealand, 2017b) included The expectations describe the features regulatory mechanism that is now in place the following statement from the code of a regulatory system that is more likely (including the 2016 code). The steering group: to deliver durable outcomes. Those recommendations, elaborated in Table 2, features include: clear objectives; were: The Code stakeholders can be assured flexibility of approach to accommodate 1. Address outstanding safety of the effectiveness of the voluntary the attitudes and needs of different management system (SMS) ‘approvals’, approach to the 2016 Code. In order regulated parties; proportionality, fairness to demonstrate that a high level of to test the level of satisfaction with and equity in the treatment of regulated safety management existed nationally. Code progress and performance, we parties; obligations set out clearly; and 2. Revive and strengthen joint consulted with the partner scope to evolve. These are all features of commitment of the port and harbour organisations during the year. They the design and implementation of the co- marine safety code, with the clear reported a high level of satisfaction regulatory mechanism that now exists; desire for it to be voluntary and with the rate of Code implementation and, importantly, they are features that provide a national safety standard. and Work Plan delivery. have arisen from a co-design process 3. Update the code and supporting We are confident that the 2016 involving government regulator (central guidelines, to reflect changes to the Code is moving us all in the right and local government working together) law, and provide for a style that was direction. The rate of progress this and regulated party engagement. less prescriptive than the original year is very encouraging. 1 The Maritime New Zealand Authority is the board of code and able to accommodate a Collectively we believe a voluntary Maritime New Zealand, the Crown entity tasked with broader concept of best practice risk Code applied through a systems undertaking safety, security, marine protection and other functions in a way which contributes to the aim of achieving assessment and safety management approach will enable us to an integrated, safe, responsive and sustainable transport system. practice. demonstrate that together in 2 This is a reference to Ezekiel, chapter 27 in the Bible. 4. Improve the code’s operational partnership we are stepping up to our 3 This article was kindly provided by Anne Carnegie, secretary, International Harbour Masters’ Association, PO Box 3111, efficiency, to minimise the compliance obligations and responsibilities as Lancing, BN15 5BQ, United Kingdom. burden while maintaining the guardians and managers of our 4 A forum that existed within the network of regional councils at the time to bring together harbourmasters and engage integrity of the ‘approval and audit’ marine environment. The 2016 Code with other key players such as marine managers, pilots and Maritime New Zealand; it has more recently been replaced system. is our pathway. by the Navigational Safety Special Interest Group, which has 5. Collect and communicate evidence of a similar purpose. 5 Indeed, it is this situation that had effectively provided some the code’s effectiveness, to provide for This is a significant shift from the comfort to Maritime New Zealand that it could focus on other priorities. statistical trend analysis of port and concerns being expressed in 2013. It 6 The term ‘environmental scan’ is used to refer to a quick harbour accidents and incidents, reflects a process of careful review of the assessment of what is working and not working for those engaged in a programme; environment in this context means provide data to champion the previous regime and consideration of the the broader setting. effective implementation of the code, design and implementation of the co- 7 This was probably a moot point as it was clear that the amendments to the Maritime Transport Act were not directly and support an annual meeting that regulatory mechanism that now exists. contemplated as making the code mandatory, although they did include a provision that enables the director of Maritime would provide for sharing of While this occurred prior to the release of New Zealand to impose conditions on port operators should information, learning lessons and Government Expectations of Good evidence support the need for that. 8 The Port Company Chief Executive Group coordinates continually improving. Regulatory Practice in April 2017, the common interests of the port companies in relation to approach taken reflects key expectations government. Key design features of the co-regulatory expressed in that document. mechanism The expectations indicate that Action on recommendation 1 set the regulatory systems should be assets, not scene for key design features for the co- liabilities: i.e. they must deliver benefits regulatory mechanism that now exists. that exceed their costs. They also As indicated in the introduction, the emphasise the need to remove or redesign

Page 18 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Table 2: Key design features flowing from consideration of recommendations Recommendation Design features

Address outstanding The first step taken was to address outstanding SMS ‘approvals’. An early decision was to acknowledge that SMS ‘approvals’ referring to ‘approvals’ did not fit comfortably with a voluntary, collaborative approach. This action was reframed as providing ‘confirmation’ to port company and council chief executives that the SMS met the requirements of the code (at that point, the 2004 version). In support of the intent of recommendation 1, as a trial a joint port, council and Maritime New Zealand working group was established. This working group established panels of experts to ‘confirm’ the SMSs that had not previously been ‘approved’. The experts were drawn from each of the partner groups (ports, councils and Maritime New Zealand).

Revive and strengthen In August 2016, reflecting the success of the trial approach to addressing outstanding SMS ‘confirmations’, an joint commitment of MoU was signed by representatives of port company chief executives and regional council chief executives (that the Port and Harbour were listed individually in the MoU) and the chief executive/director of Maritime New Zealand (Maritime New Marine Safety Code Zealand, 2017a). While not legally binding, this commits the parties to principles of tripartite joint ownership of the code, voluntary adoption of mutually agreed standards, collaboration, commitment of human resources to support delivery of an agreed work plan, and joint funding of a secretariat. Critically, the MoU is clear that there is no constraint on the ability of councils or Maritime New Zealand to carry out their regulatory functions.

Update the code and Parallel with the trial, and the development and signing of the MoU, the 2004 code was refreshed and the 2016 supporting guidelines version released. The 2016 code provides clarity in respect to its objective (‘to ensure the safe management of ships navigating in New Zealand ports and harbours, including the prevention of: injury to people or loss of life, and damage to the environment, particularly to the marine environment, but also to property’). It also establishes clearly who the 2016 code applies to, what it covers, where it should be applied, the respective responsibilities and accountabilities of the parties to the code, and its relationship to legislative requirements. Consistent with the MoU, the 2016 code established a tripartite steering group, working group and a jointly funded secretariat. It established processes for joint peer review panels and confirmed that code activities did not prevent councils and Maritime New Zealand from independent exercise of their regulatory functions. The secretariat and working group have the task of developing a work plan for consideration by the steering group, including a schedule of SMS reviews, and review and development of supporting guidelines. Resources for these activities are provided by the partner groups.

Improve the code’s The key barrier to efficiency had been the difficulty Maritime New Zealand had in assigning resources to operational efficiency undertake 2004 code-related activities. It was also noted that the leadership qualities and maritime knowledge of harbourmasters had some bearing on how efficiently tasks could be carried out. The MoU and the 2016 code establish clear commitment to resources and sharing of the work load across all parties to ensure efficient execution of code tasks. Involvement of panels of experts supports the development of leadership and knowledge sharing among all code participants. The 2016 code also addresses potential conflicts of interest. Noting that in some cases harbourmasters are employees of port companies in addition to having harbourmaster responsibilities, the code provides guidance to ensure the independence of the exercise of harbourmaster functions. Conflicts of interest may also occur should a review panel member (for example, from Port ‘A’) have access to information that is commercially sensitive when reviewing the SMS of another port (Port ‘B’), particularly where ports A and B are in direct competition. This matter is addressed in the assignment of panel members, including a requirement to protect and respect commercial confidentiality.

Collect and The code establishes: communicate • a requirement for review panels to report their findings to the working group; evidence of the code’s • a requirement for annual self-assessments by ports and councils, which are to be provided to the working effectiveness group; • the need for the development of performance standards to measure the effectiveness of the code. The information provided from these processes is used to prepare an annual report on the effectiveness of the code. The MoU also provides for national meetings of 2016 code participants and stakeholders to assess national performance of the code. The inaugural national forum took place in July 2017 in Wellington. The first annual report was released in September 2017.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 19 New Zealand’s Port and Harbour Marine Safety Code – a case study in co-regulation

References Allen+Clarke (2012) A Quick Guide to Regulation, Wellington: Maritime Safety Authority (2004) Port and Harbour Marine Safety Code Allen+Clarke, www.allenandclarke.co.nz, accessed 24 August 2017 2004, Wellington: Maritime Safety Authority Compliance Experts (n.d.) ‘What do you mean by “self-regulation” and Mast, R.K. (n.d.) ‘The International Harbour Masters’ Association (IHMA)’, “co-regulation”?’, http://www.complianceexperts.com/faq-self- Port Technology International, 4, pp.169-73 regulation-vs-co-regulation, accessed 17 August 2017 Ministry of Consumer Affairs (1997) Market Self-Regulation and Codes of Duynhoven, H. (2004) ‘Address to the National Ports Forum, Wellington’, Practice, Wellington: Ministry of Consumers Affairs https://www.beehive.govt.nz/speech/address-national-ports-forum- Ministry of Transport (2007) Port and Harbour and Navigation Safety wellington, accessed 19 August 2017 Management: a discussion paper for public comment, November 2007, Fédération Française des Pilotes Maritimes (2017) ‘History of maritime Wellington: Ministry of Transport pilots’, http://public.pilotes-maritimes.com/en/history-of-maritime- Ministry of Transport (2008) ‘Regulatory impact statement: port and pilots/, accessed 6 August 2017 harbour and navigation safety’, Wellington: Ministry of Transport, Freiberg, A. (2010) ‘Re-stocking the regulatory tool-kit’, paper presented at http://www.transport.govt.nz/assets/Import/Documents/Regulatory- the Regulation in an Age of Crisis conference, Dublin, 17–19 June, Impact-Statement-Port-and-Harbour-and-Navigatin-Safety.pdf, http://www.regulation.upf.edu/dublin-10-papers/1I1.pdf accessed 8 September 2017 Maritime New Zealand (2014) Port and Harbour Marine Safety Code New Treasury (2017) Government Expectations for Good Regulatory Practice, Zealand, Wellington: Maritime New Zealand Wellington: New Zealand Government, http://www.treasury.govt.nz/ Maritime New Zealand (2016a) Maritime Rules, Part 90: pilotage, http:// regulation/informationreleases/pdfs/good-reg-practice.pdf, accessed 26 www.maritimenz.govt.nz/rules/part-90/Part90-maritime-rule.pdf August Maritime New Zealand (2016b) Port and Harbour Marine Safety Code New Vernon, B. (2017) ‘Address to New Zealand Port and Harbour Marine Zealand: 2016 edition, Wellington: Maritime New Zealand Safety Code inaugural forum’, Wellington: Maritime New Zealand Maritime New Zealand (2017a) Memorandum of Understanding NZ Port Williamson, M. (2013) Maritime Transport Amendment Bill, third readings, and Harbour Safety Code, Wellington: Maritime New Zealand Hansard, 13 October, vol. 694, p.13861 Maritime New Zealand (2017b) Port and Harbour Marine Safety Code Annual Report 2016–2017, Wellington: Maritime New Zealand

Victoria Professional and Executive Development High quality professional and executive development courses specifically designed for the public sector:

MACHINERY OF GOVERNMENT MANAGING STAKEHOLDERS: A SYSTEMS APPROACH -> Wed 13 March, 9am–4:30pm -> Mon 30 April, 9am–4:30pm

STRATEGIC THINKING FOR GOVERNMENT ADVANCED POLICY LEADERSHIP WORKSHOP -> Fri 13 April, 9am-4:30pm -> Mon 27 & Tue 28 November, 9am–4:30pm

PUBLIC SECTOR FINANCE FUNDAMENTALS STRATEGY BUILDER -> Mon 6 November , 9am-4:30pm -> Mon 28 May, 9am–4:30pm -> Mon 16 April , 9am-4:30pm USING DATA: DISCOVERY, ANALYSIS, VISUALISATION ECONOMIC PRINCIPLES AND APPLICATIONS IN PUBLIC AND DECISION-MAKING POLICY -> Mon 20 & Tue 21 November, 9am-5pm -> Wed 11 & Thu 12 April, 9am-5pm -> Mon 23 & Tue 24 April, 9am-5pm

ENGAGING EFFECTIVELY WITH YOUR STAKEHOLDERS SYSTEMS THINKING -> Fri 23 March, 9am-4:30pm -> Mon 14 & Tue 15 May, 9am-4:30pm

IMPROVING PUBLIC POLICY EFFECTIVENESS -> Mon 19 March, 9am-4:30pm

We can also deliver in-house courses, customise existing courses or design new programmes to suit your requirements. We now also run courses at our Auckland training rooms. For more course dates, further information and to enrol visit www.victoria.ac.nz/profdev or call us on 04-463 6556.

Page 20 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Chris Ballantyne

A Rail Tale In 2005, Parliament passed new legislation to regulate railway organisations and staff to make best use of their new regulatory tools and skillset. safety in New Zealand. Applying international best practice, As a result, the Transport Agency is changing the way it interacts with the the Railways Act took a goal-based approach that utilised organisations it regulates, and how it the Safety Case concept as the foundation for regulatory measures the success of regulatory interventions. Its experiences are not oversight. This article describes the Transport Agency’s unique among the regulatory community (New Zealand Productivity Commission, experience in implementing this regulatory approach, 2014). particularly the Safety Case concept. The change required the The Railway Story Transport Agency to first recognise that fully harnessing the Railways in New Zealand recently celebrated its 150th year – the first public legislation required a transformational response and then, railway joined Christchurch to Ferrymead along with the wider industry, address the challenges faced in 1863, just 23 years after the signing of the Treaty of Waitangi. The railway in developing and implementing an appropriate regulatory industry has been through peaks and troughs – both in its role as a driving force operating model behind the growth of our nation and in its success in keeping its workers and Railway safety is just one of many areas and continually growing instruction on passengers safe. in New Zealand where a goal-based safety requirements (Robens, 1972). approach has been adopted. Goal-based Regulators are fast building the 1863 – 1983: The emergence of a safety regulation offers the potential skillsets required in response to this nationalised rail network to provide organisations with more change in regulatory practice, supported 17 years after the Provincial Government flexibility in meeting their obligations. by cross-government capability building created the 7km Christchurch-Ferrymead It contrasts with prescriptive-based initiatives such as G-Reg. However, as the link, the network had grown to 1,900km regulation, where government seeks to Transport Agency has found, the change and New Zealand Railways, the national manage risks on industry’s behalf, not is not just about new skills. It is about rail provider, was created. By 1920 necessarily more reliably, by voluminous transformational change to enable railways ran the length of the country and carried 28 million passengers per year - in a country with a population of only a Chris Ballantyne is the Lead Advisor, Rail System Integrity, Transport Access Delivery at the New Zealand Transport Agency million.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 21 A Rail Tale

Table 1: Disasters in New Zealand’s early rail history This echoes the findings of a seminal Year Location Fatalities Event 1972 inquiry of worker health and safety 1923 Ongarue 17 Auckland to Wellington passenger service hit by Lord Robens in the United Kingdom. landslide and derailed Robens proposed focusing on the level of 1936 Ratana 7 Wellington to New Plymouth passenger service safety that must be achieved (the “goal”), overturned due to excessive speed on a corner rather than how it was achieved (Robens, 1943 Hyde 21 Cromwell to Dunedin passenger service 1972). He foresaw prescriptive regulation overturned due to excessive speed on a corner was fatally flawed in that, as complexities 1948 Seddon 6 Picton to Christchurch passenger service grew, it would unacceptably throttle safety overturned due to excessive speed on a corner improvements. Businesses wouldn’t be 1953 Tangiwai 151 Wellington to Auckland passenger service fell able to exploit safety advances or adopt into the Whangaehu River after a lahar sweeps customised approaches. away the bridge

Post 2005: The way forward? As the network continued to grow, the Former Treasury official John Wilson In 2005, Railways Act came into force, safety systems required to manage it grew (Wilson, 2010) wrote that, by 2001: regulating an industry that was far as well. Rail is an activity exposed to low different to that of the previous 140 frequency, high consequence accidents – Tranz Rail’s financial problems were years. Although the national rail provider movement of large objects at speed has now creating visible shortfalls in the returned to government ownership, always meant a complex system of capital asset replacement programme its staff numbers were down to 3,500 controls are required to prevent fatalities. and it operated a network of only 3,700 Vulnerabilities have been learnt the hard The prescriptive-based safety system km (from a peak, 50 years earlier, of way in some cases (see Table 1). strained with the pressures of rapid 5,689km). However, the wider industry The national rail provider took a downsizing, asset deterioration and had proliferated - today there are about strong, but prescriptive approach to safety increased production. Prescriptive-based 100 licensed rail participants using a huge – as all industries of the day did. It created regulatory approaches give a sense of variety of rail vehicles for commuting, a quality management system based on an security because they are comparatively freight, servicing, tourist and heritage ever expanding mountain of rules, policies simple to confirm compliance. However, purposes. and schematics to dictate safety. whether this compliance achieves safety is The commentary for the Railways Act By the 1950s, rail was starting to more difficult to confirm – changing did not identify it as a goal-based decline in popularity as a series of shocks circumstances, such as railways saw over approach, but the principles were there. hit the industry. The widespread uptake those two decades, can easily destroy this The Select Committee sought to enable a of motor cars cut passenger numbers, and link. co-regulatory framework and flexibility in deregulation in the 1970s and 1980s saw a After five Tranz Rail workers died in a how safety requirements were met, and considerable move of freight tonnage to 7 month period (a fatality rate eight times the Act imposed an overarching duty of road. New Zealand Rail had been a higher than the national worker average) care for all rail activities, primarily that government department for most of its the Wilson Ministerial Inquiry was the operator: existence but, struggling to cope with initiated (New Zealand Government, these shocks, was turned into a State 2000; Williams, 2000). The Wilson .. must ensure, so far as is reasonably Owned Enterprise in 1982 in an attempt Ministerial Inquiry identifiedpracticable, that none of the rail to turn the industry around. fundamental flaws in the railways activities for which it is responsible regulatory approach. causes, or is likely to cause, the death 1983 – 2005: A struggling system of, or serious injury to, individuals. Government efforts to cut costs and the number of [safety system] restructure to address rail’s profitability documents is in the range of 1,000 to The Act put in place a bespoke ultimately culminated in the sale of 2,000 and the number of drawings is regulator that oversees a licensing regime the national rail provider to a private in the region of 100,000. It would for those operating rail vehicles or consortium in 1993. Staff numbers, appear that the scale of controlling a network. To obtain a licence, peaking at 21,000 in 1982, had dropped to documentation is unmanageable operators must have a Safety Case below 5,000 by this stage. relative to the aim of securing a safe approved by the regulator. The regulator The following decade was one of the system … The sheer extent and has powers to conduct in-depth system hardest in New Zealand rail’s history as coverage of the [regulator] Approved assessments of the operator to verify ownership and investment drifted. The Safety System renders it unwieldly compliance with the Safety Case and other business was renamed Tranz Rail and and probably impenetrable to outside safety documentation. The Act allows the subject to aggressive profit strategies. observers. (New Zealand regulator to intervene more directly if Government, 2000, p.23)

Page 22 – Policy Quarterly – Volume 13, Issue 4 – November 2017 there are immediate safety risks or if non- compliances are detected. What is a Safety Case? However, the decade following the enactment of the Railways Act 2005 Safety Cases originated in the nuclear power industry in the late 1950s but their demonstrated the difficulty in applying a potential value for other high-risk, specialised industries was recognised after goal-based approach on a deeply accidents in the chemical industry (particularly Flixborough in 1974 and Seveso established industry without sufficient in 1976) and then more widely after the Cullen Inquiry into the 1988 Piper appreciation of the change it entailed. The Alpha explosion. potential of the approach was not fully recognised and the regulator was under- Principles of Safety Cases prepared to deliver the altered style and Safety Cases are based on three principles common to most goal-based ap- thinking required to fully exploit it. proaches (Leveson, 2011). Particularly, the oversight of Safety Cases, • Risks must be reduced below a specified threshold of acceptability a powerful tool for implementing a goal- • Those who create risks are responsible for controlling those risks based regime (see boxed text “What is a • Government set safety goals and operators decide the appropriate methods to Safety Case?”) was under-utilised. Instead, achieve those goals a process-based operating model was adopted that focused heavily on The Safety Case delivers on these principles in a way that allows the regulator mechanically auditing activities across all to focus on what is being achieved rather than how it is being achieved, while still operators. providing a degree of assurance as to the quality of the operator’s methods. A key element of a Safety Case is that it is an argument, not a summary. An Recognising the need for transformation argument convinces an audience of a premise – in this case, confidence that an Following the Wilson Inquiry and passing operator can achieve its safety commitments. As with an argument, the rationale in of the new Act, railway deaths and serious a Safety Case must be able to be tested. The United Kingdom’s Defence Standard injuries significantly reduced but, as 00-56 notes a Safety Case to be: has been found before, the absence of accidents is not proof of safety. There was a structured argument, supported by a body of evidence, that provides a uncertainty over the role of the regulator compelling, comprehensible and valid case that a system is safe for a given and the robustness of safety protections. application in a given environment. Change was needed, and the railway Because a Safety Case communicates this rationale for how the operator will industry was fortunate in that its impetus keep safe, it sits apart from the management systems to provide direction and for this change did not come from a integrate them. catastrophic accident, but as part of a When a regulator intervenes in a goal-based approach, it is on the basis general push towards improved regulatory that the operator has not demonstrated it has developed its approach in a robust services by the Transport Agency, the manner and so is unlikely to be able to meet its goals. The regulator is not critiquing industry and other government agencies, particularly in the wake of regulatory the safety approach taken but the method the operator used to determine it. failures such as the 2010 Pike River Mining Disaster. support arrangements within NZTA It was only once the Transport Agency, A number of internal and independent for such a safety critical independent while implementing these tactical reports helped the Transport Agency rail safety regulator function to be less changes, reflected more deeply on the core understand where the gaps lay and became than an acceptable standard. drivers for the problems these reviews the catalyst for a multi-year improvement (Australasian Transport Risk highlighted that it began to recognise that project that is still underway. In particular, Solutions, 2013, p.5) moving to a goal-based regime in rail a report by Australasian Transport Risk safety and delivering the required standard Solutions (2013) made 18 observations The various reviews recommended a of regulatory service was a more and commented: series of tactical changes, such as elevating fundamental shift. the regulator in the Transport Agency A dramatic shift in the regulatory ... there is considerable room for hierarchy, increasing regulatory staff framework, from a prescriptive to a goal improvement. This conclusion is numbers, broadening the skill-base and basis, can have a disorientating impact on based on the results of the performing greater analysis. However the regulators and the regulated if not international benchmarking analysis, none, including the Wilson Inquiry, recognised and can, ultimately, lead to feedback received from the explicitly highlighted the need to move to regulatory failure (Mumford, 2011; Black, stakeholders, and evidence found by a goal-based culture or the challenges this 2014). For instance, New Zealand’s leaky the review team which suggests that would entail. building crisis was partially attributed to the current administrative and such a change. Mumford (2011) noted that

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 23 A Rail Tale

implementing a pre-planned strategy. This The New Zealand Rail Industry reflected that the transformational nature of the change was only gradually recognised in 2017 after the simpler, task-driven change programme had been begun. In addition, 6,200 workers are employed by, or volunteer for, licensed rail participants in New with the rail industry going through parallel Zealand. Many of these workers are proud to be from a long lineage of rail work- changes, the approach needed flexibility to ers, have known rail all their lives and have seen a number of structural models navigate a shifting environment. come and go. The industry carries 34 million passengers and 18 million tonnes In addition, the change was relatively of freight every year, across 4,260km of rail network, on everything from brand- compartmentalised within the wider new commuter fleets to hand-restored museum pieces. Those licensed to operate Transport Agency. It sought direction railways include: from the senior leadership and values of • KiwiRail, a State Owned Enterprise that employs 3,700 staff. As the national the organisation, but it only had a rail provider, it controls and maintains the 3,744km national rail network and transformational impact on a small operates all national freight and passenger services. number of teams who had direct roles in • Transdev Auckland and Transdev Wellington operate the two metropolitan enabling a goal-based approach to rail commuter services, carrying 32 million passengers every year. safety. Achieving the change did not • Industrial businesses utilising rail to load and move freight on their sites or to require a whole-of-organisation shift. As a service their infrastructure. result, although the changes were deep, • Tourist and Heritage services operating rail services to showcase rail his- they were narrow and more manageable. tory and/or provide tourist experiences. Running excursions on the national Successful change required two network or their own railways, they operate a huge range of vehicles including significant perceptual changes to be made. full-sized heritage locomotives, trams, cable-cars, custom-built rail cars and modified golf-carts. Roles in the regulatory framework To understand its role in the regulatory In addition to this, more than 200 organisations operate rail vehicles under framework, the Transport Agency had to others’ licences, to provide vehicle and infrastructure maintenance services. understand what it meant for the operator to own the risks of its activities – a the haste to move away from an expensive approaches, it was exploring its role as concept that felt contrary to the regulatory (if reliable) standards-based approach to a manager of the national rail system under role. Regulation is enacted to provide more flexible regime resulted in discarding such a regulatory framework. A particular assurance that the risks that society finds previous safeguards without adequate challenge for it was reconciling how its unacceptable are addressed (New Zealand consideration of what was replacing them. safety outcomes related to those of the Productivity Commission, 2014). It is the Mumford commented: operators under its supervision on the responsibility of the regulator to provide network. this assurance, and it is therefore natural for New Zealand moved into an Smaller operators, on the other hand, a regulator to feel most comfortable when unknown future while burning the experienced difficulties in maintaining meeting this duty directly by controlling bridges to its past. the competencies and documentation to the risk itself (for instance, by approving comply with a regulatory framework safety measures) rather than entrusting To be successful, the Transport Agency more often applied to multi-national, management of the risks to the operator. needed to redefine its rail safety role, how high-risk industries than small hobbyist. The challenge became for the Transport it engaged with the industry, and how it They were still looking for clear, Agency to let go of this role and meet its exerted its influence over the industry. prescriptive expectations from the duty through more indirect methods. A cultural change such as this is regulator as to how to be compliant with Some in the industry were coming to referred to as transformational change. It the safety requirements rather than the same conclusion. The rail industry had fundamentally alters the strategy by which moving into a goal-based mind-set. been challenging themselves and the an organisation achieves its objective. Regulator and regulated alike were Transport Agency as to where Existing skills, behaviours and approaches struggling to understand their respective responsibilities and accountabilities lay for are no longer suitable and have to be responsibilities and approaches to safety, and what tools could support clearer adapted, enhanced or replaced meeting those responsibilities in a goal- accountabilities and allow the regulator to (Cummings and Worley, 2009; New based regulatory framework. have confidence the industry would deliver Zealand Productivity Commission, 2014). on their safety commitments. The national rail provider was also Making the changes To enable such an approach, some going through its own transformational The Transport Agency took a learning, regulatory tools, such as the Safety Case, change. Along with a heavy period of asset highly adaptive approach to its rail safety needed to be recognised as being of equal, renewal and developing new safety transformational change, rather than or greater, value to the industry than the

Page 24 – Policy Quarterly – Volume 13, Issue 4 – November 2017 regulator – a fact commented on by Lord Cullen in the Piper Alpha Inquiry. The RAF Nimrod mid-air explosion Haddon-Cave (2009) paraphrased his observations, noting that the purpose of The Haddon-Cave inquiry (Haddon-Cave, 2009) into the loss 14 aircrew after Safety Cases was: British Royal Air Force (RAF) Nimrod XV230 exploded in mid-air serves as one of the most well examined and illustrative studies of the potential weaknesses of to assure [the company] that its Safety Cases. operations were safe ... Whilst the The report made 12 observations, which included the following failures of the Safety Case had a further role in Safety Case process at the RAF demonstrating this to the regulatory 1. Safety Cases had become too long, bureaucratic and contained impenetrable body, this latter function was a matter detail, often simply to give a “thud factor”. of only secondary importance. 2. Safety Cases gave equal attention to minor hazards as they did catastrophic (p.166) hazards.

Becoming a risk-based regulator 3. Safety Cases were routinely outsourced and reduced to mere back-office With the focus on high-level outcomes, paperwork with little appreciation or personal interest in the subject of the moving to a goal-based approach facilitates Safety Case. the Transport Agency in becoming a 4. Safety Cases were a compliance document that looked for evidence to justify a risk-based regulator, as expected of all predetermined answer that the system is safe, rather than look for evidence as agencies (New Zealand Government, to why it might not be. 2017). Previously the Transport Agency 5. Safety Cases were “shelfware”, rather than living documents to keep abreast had taken an across-the-board approach to oversight, working across all areas of of hazards and cultural changes. harm and pursuing serious issues as they Haddon-Cave commented: arose. This was an ineffective approach The Nimrod Safety Case was a lamentable job from start to finish. It was (Sparrow, 2008; Baldwin and Black, 2007) riddled with errors… Its production is a story of incompetence, complacency as it spread Transport Agency resources too thinly and intervention was too and cynicism… The best opportunity to prevent the accident to XV230 was, transient to have lasting effect. tragically, lost. Again, this challenge can generate uneasiness. Becoming a risk-based regulator means consciously de- the Transport Agency and the industry. Safety Cases, in particular, require prioritising other potential sources of The Transport Agency has gained several discipline by the Transport Agency to harm. These sources of harm may still be insights regarding its regulatory role and maintain its regulatory role by following important, but just not as important as being a risk-based regulator through goal-based principles and resist using its others. In addition, there is uncertainty in focusing on this tool. rail technical expertise to double-check decisions as to where the greatest risk is. These insights are similar to many the operator has chosen the appropriate The result is it places a great deal of other industries’ experiences with Safety safety approach. Instead, to ensure that an responsibility on the regulator - what if Cases, such as the Nimrod mid-air operator owns the risks of its activities, the Transport Agency makes the wrong explosion (see boxed text), the Gulf Oil approval needs to be based on the strength choice and it is a different risk that leads disaster (United States, 2011) and as of the operator’s argument. to the next catastrophic accident? Being surveyed by the United Kingdom Health A consequence of this is that the clear about the evidence base (and and Safety Executive (Vectra Group regulator requires broad, rather than limitations of it) and each party’s Limited, 2003). deep, expertise to assess a Safety Case. accountabilities in respect of the targeted Rather than relying solely on rail and risks and general risks helps move beyond safety expertise, to critique the argument the inertia this can cause. Avoid owning the risk the Transport Agency staff also have to A Safety Case doesn’t work well when have expertise in such aspects as change Reflecting these changes through Safety used in a prescriptive way. The natural management, assurance, safety systems Cases tendency is for an operator to describe and governance. The Transport Agency’s re-positioning how it is managing its safety. However, of the Safety Case’s role in the regulatory when a Safety Case descends into nothing Underpin trust with evidence framework has been a core outcome of more than a summary of how the safety In the traditional use of Safety Cases for the change process. Better utilisation of measures are being carried out, all it offers heavy equipment installations, the Safety this tool is both reflecting and driving is paper safety. Case provides confidence by providing many of the changed perspectives within direct evidence of safety. However, for an

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 25 A Rail Tale

Figure 1: The Safety Case providing a trust framework face complex risks, the flexibility and safety-focus of a goal-based approach remains more effective. In essence, a good • Have they made measurable safety commitments? Commitment Safety Case is as simple as proving an • Have they set clear accountabilities for these commitments? organisation can make good decisions. Focusing its efforts according to risk, • Have they demonstrated that they understand the context they the Transport Agency is looking towards Context will operate in – particularly the risks and barriers to achieving pragmatic solutions that simplify their commitments? requirements without losing the core principles of Safety Cases. This is not • Do they make a robust argument that their management about turning the Safety Case back into a systems can achieve safety commitments? safety system summary or providing • Have they demonstrated that they have the determination, Confidence detailed templates that result in a paint- capability and capacity to carry out their safety activities? by-numbers exercise. The Safety Case still • Are there the hallmarks of a strong safety culture? has to demonstrate the operator’s own safety argument for it to have any value. • Can they justify their decisions? Conclusions Confirmation • Is safety awareness widely embedded? The move away from prescriptive • Are there any warning bells? regulation to the effective use of goal- based tools such as Safety Cases has both empowered and challenged the Transport organisational Safety Case like in railways, such as integration, communication and Agency to better carry out its role in such definitive evidence is not available - capability. achieving the Government’s objectives for particularly when first licensed. A framework-based approach also rail safety. To be successful it had to But rather than requiring blind trust, helps the industry understand they have recognise that the transition is not just a the Safety Case provides the Transport to earn that trust – it’s not just a case of matter of better regulatory practice, but Agency with a framework to structure its “hands-off regulation”. organisational transformation – for both confidence around – a robust and the regulator and the regulated. evidenced argument, embedded in the Simplify but don’t compromise safety cases The Transport Agency is still organisation, that the operator will The Railways Act has imposed a undergoing this journey, but is already achieve its safety commitments (see sophisticated regulatory approach over an recognising important lessons: Figure 1). industry that includes operators with only 1. The move to goal-based legislative This allows the Transport Agency to basic competencies and comparatively approaches is positive for safety, but better understand the value-add of its low risks. The original users of Safety the cultural shift required should not role. Its engagement with operators is Case approaches were high risk, heavily be underestimated – transformational about building confidence in the Safety resourced and complex operators – such change approaches should be adopted Case argument, rather than approving the as nuclear, chemical and extractives and a learning environment safety system approach. industry. Parts of the rail industry are, encouraged. As an example, approval of Safety in contrast, hobbyists. This challenge 2. The industry needs to be involved – it Cases involves site visits. Previously, the has been observed in other sectors where is not something the regulator does operator may have thought (rightly or Safety Cases or similar tools have been behind closed doors. wrongly) that the Transport Agency was applied homogeneously over a diverse 3. As a cultural shift, the change has to approving its specific safety activities. industry (Deighton-Smith, 2008). be deep but it can be achieved However, the Transport Agency visits not Some rail operators may have the right through a focus on the critical as auditors but to gain evidence for a approaches and culture in place, but regulatory staff, rather than more subjective judgement as to whether struggle to join them up as a system and organisation wide. the operator’s safety argument stands up demonstrate it on paper. An operator may 4. Being a goal-based regulator is an and it will achieve its safety commitments. simply not have the competency to uneasy experience – assurance of Regulatory staff may still observe the develop bespoke safety solutions and so safety is based on trust (underpinned adequacy of a rail vehicle or the network rely on standard industry practices. It can with evidence) in the capability and control approach, but as part of building be argued that a more prescriptive willingness of the regulated party, a picture of the organisational approach should be applied to such rather than having the ability to competence, along with their operators (Deighton-Smith, 2008). dictate the “right” solution. observations of more general aspects However, the Transport Agency is 5. Goal-based regulatory approaches confident that, given these operators still support the transition towards

Page 26 – Policy Quarterly – Volume 13, Issue 4 – November 2017 risk-based regulation by stressing the surveys and the Transport Agency’s own doesn’t create it. As Lord Cullen operator’s overall accountability for experiences have highlighted that a commented (Jeffrey 2013): their risks and allowing the regulator cultural transformation among the to focus on the high-level outcomes. regulator and the regulated community is A Safety Case should reflect the Safety Cases and goal-based legislation what enhances safety. A Safety Case helps organisation’s safety culture. If that are not an automatic panacea to demonstrate this has been achieved – it culture is sound and healthy – it organisational safety. Various inquiries, should show.

References Australasian Transport Risk Solutions Pty Ltd. (2013) Independent Review New Zealand Government (2000) Ministerial Inquiry into Tranz Rail Report into Rail Systems Team, downloaded from www.nzta.govt.nz/ Occupational Safety and Health, Wellington: New Zealand Government assets/site-resources/content/commercial/docs/atrs-rail-systems-team- New Zealand Government (2017) Government Expectations for Good review.pdf Regulatory Practice, downloaded from www.treasury.govt.nz/regulation/ Baldwin, R. & Black, J. (2008). Really Responsive Regulation, Modern expectations Law Review, 71, pp. 59 - 74 New Zealand Productivity Commission (2014) Regulatory Institutions and Black, J. (2014) Learning from Regulatory Disasters, Policy Quarterly, Practices, downloaded from www.productivity.govt.nz/sites/default/files/ 10(3), pp. 3 – 11 regulatory-institutions-and-practices-final-report.pdf Cummings, T. G., & Worley, C. G. (2009). Transformational Change, in Robens, A. (1972) Safety and Health at Work: Report of the Committee Organization development and change, pp 505 – 534, Mason, Ohio, 1970–72, London: Her Majesty’s Stationery Office Thomson/South-Western. Sparrow, M. (2008) The Character of Harms: Operational Challenges in Deighton-Smith, R. (2008) Process and Performance-Based Regulation: Control, Cambridge: Cambridge University Press Challenges for Regulatory Governance and Regulatory Reform, in United States. (2011) Deep Water: The Gulf Oil Disaster and the Future of Carroll, P. (Ed) Minding the Gap: Appraising the Promise and Offshore Drilling : Report to the President. Washington, D.C.: National Performance of Regulatory Reform in Australia, pp. 89 – 104, ANU Commission on the BP Deepwater Horizon Oil Spill and Offshore EPress: Canberra Drilling. Haddon-Cave, C. (2009) The Nimrod Review, London: The Stationery Vectra Group Limited. (2003) Literature Review on the Perceived Benefits Office and Disadvantages of UK Safety Case Regimes, downloaded from Jeffrey, K. (2013) Review: Lord Cullen - What Have we Learned from Piper www.hse.gov.uk/research/misc/sc402083.pdf Alpha?, downloaded from www.findingpetroleum.com/n/Review_Lord_ Williams, M. (2000) Industry of Myth and Might (Accidents, Ownership Cullen_what_have_we_learned_from_Piper_Alpha/044b5113.aspx and Change on the Railways), The Dominion, 12/05/2000 (2), pp. 13 Leveson, N. (2011) The Use of Safety Cases in Certification and Wilson, J. (2010) A Short History of Privatisation in New Zealand, Regulation, Journal of System Safety, 47 (6) pp. 1 - 5 downloaded from www.treasury.govt.nz/downloads/commercial/ Mumford, P. J. (2011) Enhancing Performance-Based Regulation: Lessons mixed-ownership-model/mom-shppnz-wilson-dec10.pdf from New Zealand’s Building Control System, Wellington: Institute of Policy Studies School of Government Brown Bag seminars – open to all Join lively, topical presentations and Zealand Primary Health Care Strategy discussions in an informal setting at the • Strategic public procurement: a School of Government. These Brown Bag research agenda sessions are held the first Monday of most • What role(s) for Local Government: months, over lunchtime. Past topics have ‘roads, rates and rubbish’ or ‘partner included: in governance’? • Intergenerational wellbeing and public • Human capital theory: the end of a policy research programme? • A visual exploration of video • How do we do things? surveillance camera policy We would welcome your attendance and practice and/or guest presentation, if you are • The role of financial risk in the New interested. Contact us to go on the mailing list for upcoming sessions at [email protected]

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 27 Kirsty Hutchison, Katherine MacNeill, Peter Mumford and Val Sim

Managing the Opportunities and Risks Associated with Disruptive Technologies space law in standardisation and mass production of small satellites have also reduced barriers to entry and driven innovation in space- New Zealand related services and applications. Growth in the small satellite industry has in turn created demand for small A key challenge for regulators is how to enable satellite launch vehicles. Developments in space technologies and space business entrepreneurship and innovation while managing the risks models means that space is now open to a associated with rapidly evolving technologies and associated new generation of entrepreneurs and enthusiasts, and countries around the market change. A number of technologies could be used world are keen to share in the full range of economic development and social benefits to illustrate this problem, such as the internet, gene editing that space offers. and driverless vehicles. However, the Ministry of Business, In 2015 the New Zealand government Innovation and Employment (MBIE) recently had to face decided to enable space launches from New Zealand this specific issue in relation to space activities. Rocket Lab, a United States corporation Space was once the domain of a small investment associated with building, with a subsidiary in New Zealand, is the number of nation states and large launching and operating satellites. main commercial player in New Zealand’s corporations. This reflected that only a Today, many more players are able to emerging space industry. Rocket Lab few players had access to the advanced access space. Advances in technology has developed the Electron space launch technologies required to launch objects have enabled the production of smaller, vehicle to provide a dedicated launch into space, or could afford the significant cheaper and more powerful satellites. The service for small satellites. The company’s mission is to remove commercial Kirsty Hutchison is a Principal Policy Advisor, Labour, Science and Enterprise Group in the Ministry of barriers to space. Rocket Lab recognised Business, Innovation and Employment. Katherine MacNeill is Manager, International Strategy Labour, that New Zealand offered an attractive Science and Enterprise Group in the Ministry of Business, Innovation and Employment. Peter Mumford is the Policy Director, International Environment, Labour, Science and Enterprise location for space launch activities, due Group in the Ministry of Business, Innovation and Employment. Val Sim is the Director, Legal to our innovation-friendly business Services Branch in the Ministry of Business, Innovation and Employment. environment, strong science and

Page 28 – Policy Quarterly – Volume 13, Issue 4 – November 2017 research and development system, skilled obligations and manage risks associated The Outer Space Treaty workforce and suitably remote geography with space activities (including risks to The Treaty on Principles Governing the (including low volumes of air and sea safety, the environment and national Activities of States in the Exploration traffic). security). and Use of Outer Space, including the Space is a multi-billion dollar global Moon and Other Celestial Bodies (known industry and it has huge strategic International space treaties as the Outer Space Treaty) provides the importance. Satellites enable the provision In the 1960s and 70s, New Zealand ratified principles that govern the exploration of critical everyday services and three international space treaties: and use of outer space. It requires that infrastructure, including banking, · Treaty on Principles Governing the the exploration and use of outer space, transportation, electricity, telecommun- Activities of States in the Exploration including the moon and other celestial ications, navigation, remote sensing and and Use of Outer Space, including the bodies, is carried out for the benefit and in national security. The industry is an Moon and Other Celestial Bodies – the interests of all countries, irrespective important source of innovation and high- ratified 1968; of their degree of economic or scientific tech and high-value jobs. The development · Convention on International Liability development, and is to be the province of of a New Zealand-based space industry for Damage Caused by Space Objects all mankind. Outer space, including the would enable New Zealand to participate – ratified 1974; moon and other celestial bodies, is to be directly in this new economy and ensure · Agreement on the Rescue of free for exploration and use by all states that all New Zealanders could benefit Astronauts, the Return of Astronauts without discrimination of any kind, on a from the opportunities that the use of space and our participation in the global space economy have to offer. However, If space launches were to take place New Zealand had no specific space from New Zealand, a new law would be regulation, in contrast to many other countries. necessary to implement the international In a little under two years New Zealand has gone from having no national space space treaties. law to having a new act to regulate New Zealand space activities. The Outer Space and High-altitude Activities Act was passed in July 2017. The act governs the and the Return of Objects Launched basis of equality and in accordance with launch of space objects such as rockets into Outer Space (the Rescue international law, and there is to be free and satellites into outer space from New Agreement) – ratified 1969. access to all areas of celestial bodies. Zealand (and by New Zealanders overseas) At the time that these treaties were There are also restrictions on certain and it regulates launch facilities. The act ratified, New Zealand’s policymakers and activities in space, such as placing nuclear also introduces a regime to manage legislators clearly deemed legislation weapons or weapons of mass destruction certain high-altitude activities that take unnecessary to implement them. As noted in orbit or installing military bases on place from New Zealand, such as high- above, at the time space activities were celestial bodies (some of which found altitude balloons. (High-altitude vehicles carried out by states rather than their way into the purpose statement in operate above controlled airspace but do commercial operators, and New Zealand the legislation). To this end, the Outer not go into outer space.) policymakers could be confident that no Space Treaty imposes a number of This article sets out the main steps in national space activities were taking place. obligations on states. Most importantly, the process of developing the new Fifty years later, Rocket Lab’s activities article VI imposes international regulatory regime and describes the key provide a graphic illustration of how responsibility on states for their national considerations that influenced the content developments in technology have changed activities. It is clear from the Outer Space of the regime. the space industry and made it accessible Treaty itself, and to a lesser extent from to a wider group of participants. This the terms of the later Liability Convention Legislation was necessary to meet necessitated a change of view on the need (discussed below), that national activities New Zealand’s international obligations for legislation to implement the rights include the launch of space objects from associated with space activities and obligations of the space treaties. It New Zealand and the launch of space One of the first steps in the process was also led us to consider how to ensure that objects by its nationals. to assess whether New Zealand’s existing legislation would provide a balance If space launches were to take place domestic law was adequate to manage between risk management and not from New Zealand, a new law would be space activities or whether a new law would inhibiting economic development and necessary to implement the international be required. It quickly became clear that innovation. This is discussed further space treaties. The legislation would also a new law was needed to ensure that we below in the context of the two need to have some extraterritorial effect. could comply with certain international determinative international space treaties. This is because, under the Outer Space

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 29 Managing the Opportunities and Risks Associated with Disruptive Technologies: space law in New Zealand

Treaty, each party bears international overall design of our domestic legislative Act and the Exclusive Economic Zone responsibility for its national activities in regime. and Continental Shelf (Environmental outer space, whether those activities are Effects) Act) would be adequate to conducted by public or private entities The Technology Safeguards Agreement (TSA) manage safety and environmental risks in and whether they are conducted in New In order for Rocket Lab to commence New Zealand. However, the novel aspects Zealand or in another jurisdiction. space launch activities from New Zealand, of space activities in New Zealand meant it had to seek approval from the United that there are other interests that needed The Liability Convention States government to transfer sensitive to be considered and built into the design Under the Convention on International technology to New Zealand. The US would of the New Zealand regime. Liability for Damage Caused by only allow the transfer of this technology For example, the long-term Space Objects (known as the Liability if New Zealand concluded a treaty-level sustainability of the space environment Convention), the launching state is Technology Safeguards Agreement with and the management of orbital debris absolutely liable to pay compensation the US government. have become increasingly important for any damage its space objects cause The TSA imposes certain obligations internationally, given the increase in space to other parties, or to third states, on the on New Zealand in relation to the safe and activities and the significance of space surface of the Earth or to aircraft in flight. secure transfer, use and management of applications to modern societies. This Liability for damage caused elsewhere US space launch technologies. The needed to be factored into New Zealand’s domestic regime. In addition, issues around national One of the biggest challenges with new security are highly relevant. Space technologies is that innovative activities applications have the potential to benefit, and also pose risks to, national security. or products ... rarely become visible to Many space technologies are dual purpose, which means that they can be used for governments in time for governments peaceful as well as military objectives. A to legislate before they are ready to new domestic law was needed to allow the benefits but also put in place appropriate emerge. safeguards to ensure that space activities were aligned with New Zealand’s national interests, including its national security.

– for example, after a space object is majority of the obligations on the New The risks associated with space activities launched into orbit – is fault-based. The Zealand government are to ensure needed to be managed while a new regime term ‘launching state’ means ‘(i) A State compliance by Rocket Lab and third was being developed which launches or procures the launching parties, such as Rocket Lab’s contractors, One of the biggest challenges with new of a space object’ and ‘(ii) A State from with the provisions of the TSA. technologies is that innovative activities whose territory or facility a space object The TSA also protects New Zealand’s or products, while they might take time is launched’. laws and sovereignty over space launch in development, rarely become visible to Having discussed the interpretation of activities from New Zealand. New Zealand governments in time for governments to the Liability Convention with our is able to veto launches from New Zealand legislate before they are ready to emerge. counterparts in a number of overseas that are contrary to our domestic laws, This was the case with Rocket Lab. There jurisdictions, it became apparent that regulations and policies. While many of was no real opportunity to legislate for states are struggling to interpret the the obligations under the TSA could be its space activities before its planned first international liability rules in a world managed through contractual launch. where, increasingly, space activities are arrangements with Rocket Lab and the The solution was to regulate Rocket carried out by commercial rather than existing criminal law, legislation was Lab’s activities through a contract until state actors. Perhaps even more necessary to fully implement the TSA. legislation could be enacted. This allowed importantly (since the Liability the government to pass on the Convention has been invoked on only one Legislation was necessary to manage risks requirements under the TSA through its occasion),1 the way in which states have associated with space activities contractual arrangements and to manage dealt with the international liability in As noted earlier, space activities the risks associated with Rocket Lab’s their domestic regulation has had an effect create risks to public safety and the activities while the legislation was on the competitiveness of space regimes environment. For the most part, we developed. In this respect, the and their effectiveness in establishing a anticipated that New Zealand’s existing government reserved the ability to veto domestic space industry. As discussed laws (including the Health and Safety at any launch of a space object that would below, these were important factors in the Work Act, the Resource Management not be in the national interest during the

Page 30 – Policy Quarterly – Volume 13, Issue 4 – November 2017 contract term, the intention being that controls over participation in space Although the law had to manage risk, it was licensing requirements would be in place activities, including powers to prevent or important that it did not inhibit economic by the time the contract expired (with stop space launches in appropriate development careful arrangements to manage the circumstances. A licensing regime also The content of the new regulatory regime transition). enables financial risk to be transferred to is conditioned by what is necessary to A contract cannot be a substitute for where it can best be managed, a key comply with New Zealand’s international legislation. A contract is only binding on consideration given our obligations under obligations and to manage risk. However, the parties to it. While it applies directly the international space treaties. However, we had choices around how prescriptive to Rocket Lab, Rocket Lab can only the approach taken to the new regime was or permissive the regime was. regulate its customers indirectly through not solely concerned with managing risk, its ability to veto their payload. In but rather about managing opportunity A prescriptive approach addition, criminal offences can never be and risk. If the primary objective was to manage created by contract. However, it has risk, the government might have adopted proved to be a very valuable tool for Space activities create opportunities for a prescriptive approach, with detailed managing Rockets Lab’s activities economic development and innovation provisions in the primary legislation pending the development of legislation In order to quantify the opportunities that specified requirements pertaining and its coming into force. for New Zealand arising from Rocket to implementation of our international There was another advantage of Lab’s activities, MBIE commissioned obligations (including liability), safety regulating Rocket Lab through the contract. We did not necessarily expect this at the time, but it seems obvious with The establishment of a rocket industry the benefit of hindsight. It gave us an in New Zealand presented a strategic opportunity to engage with our international counterparts without opportunity to build New Zealand’s hindering Rocket Lab’s activities. That meant that we could benefit from the capacity and expertise in space knowledge and experience of our activities. international counterparts and bring it to bear when developing our domestic legislation. Sapere Research Group to undertake an and the environment. The advantage of It was also important to take account of assessment of the economic benefits of the this approach is that it provides space developments in international space law and development of a space launch industry industry participants with certainty in the practice in New Zealand. Sapere’s report (June short term about what requirements they A key consideration in the development of 2016) estimated that Rocket Lab could have to meet. However, there are certain New Zealand’s space law was to ensure that contribute between $440 million and disadvantages with a highly prescriptive it was informed by international space law $1,550 million to New Zealand over 20 approach. Having detailed requirements and practice. This was important to ensure years. The Sapere report focused primarily in the primary legislation would also that we developed an internationally on launch activities. The benefits from a limit the ability of the law to adapt to credible regime that positioned New satellite industry could be significantly rapidly evolving space technologies and Zealand as a responsible player in the higher. new markets. This reduces flexibility, international space community. Aligning The establishment of a rocket industry which imposes unnecessary costs for the content of New Zealand’s space law in New Zealand presented a strategic industry and increases the risk of the with international best practice would opportunity to build New Zealand’s law becoming obsolete. A prescriptive also ensure that we could conform to our capacity and expertise in space activities. approach also requires the regulator to international obligations. Over time, the benefits could include the have detailed knowledge of the activities In addition to reviewing the potential for different launch providers to being regulated. This was not practicable approaches taken in other jurisdictions, operate out of New Zealand, as well as for given that space activities were new to we also engaged Professor Steven Freeland, New Zealand to design, manufacture and New Zealand. an international space law expert at the launch its own satellites. The decision to An example of a prescriptive approach University of Western Sydney, to give us have the minister for economic to space law, particularly in relation to the benefit of his knowledge and development as the responsible minister questions of liability and insurance, is experience. under the act (albeit to be reviewed in provided by Australia’s Space Activities Act International practice is for countries three years) indicates the importance that 1998 (part 4). However, it was relevant to to put in place licensing or permitting the government places on the economic our consideration of this approach that the systems that provide the necessary and innovation benefits of space activities. Australian government had announced a

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 31 Managing the Opportunities and Risks Associated with Disruptive Technologies: space law in New Zealand review of their act.2 The rationale for the account whether an applicant has an providing for economic development was review included the fact that space overseas license when considering a key consideration in the development technologies had advanced significantly whether to grant a New Zealand license of the act. For this reason, the objective since the act was introduced and there was (this is discussed in more detail below). In to facilitate the development of a space a need to ensure that Australia’s space the initial two–three years of the regime industry was built into the purpose regulatory regime was facilitative of we expect to rely on foreign licensing statement of the act. Section 3 of the act innovation and investment in this growing organisations in considering the safety states that the purpose of the act is to: industry whilst effectively meeting and technical competence of space launch (a) facilitate the development of a space Australia’s international obligations in vehicles and payloads. This will provide industry and provide for its safe and managing the space environment. time for the New Zealand regulator to secure operation: build up the necessary technical expertise (b) implement certain international A performance-based approach and regulatory capability. We will not rely obligations of New Zealand relating The alternative to a prescriptive approach on foreign licensing organisations to to space activities and space is a performance-based approach. undertake national interest or national technology: Performance-based regulatory regimes security analysis, however, as this will be a (c) without limiting paragraph (b), focus on setting intended outcomes that New Zealand-centric question. implement the obligations in the Outer Space Treaty not to – (i) place in orbit around the Earth Performance-based regimes are ... more any objects carrying nuclear likely to facilitate innovation and to avoid weapons or weapons of mass destruction, install such weapons unnecessary regulatory compliance on celestial bodies, or station such weapons in outer space in costs. any other manner: (ii) establish military bases, installations, or fortifications on celestial bodies: must be achieved, rather than prescribing The act establishes a flexible regime with (iii) test any type of weapons or detailed processes and procedures to be a presumption in favour of economic conduct manoeuvres on celestial followed. Performance-based regimes development bodies: are therefore more likely to facilitate The government’s preferred approach (d) manage any potential or actual innovation and to avoid unnecessary was for a flexible, outcomes-based regime liability that may arise from the space regulatory compliance costs. which would manage opportunities as industry: A performance-based approach to the well as risks. Consistent with this, the act (e) establish a system for the regulation primary legislation allows the regulator’s provides for: of space activities and certain decisions to be tailored to particular cases. · a decision-making framework of high-altitude activities: It is therefore more amendable to changes which one of the purposes is to (f) preserve New Zealand’s national in technology, risks and market conditions. facilitate the development of a space security and national interests. The more prescriptive elements that are industry; The act also gives power to the minister required are provided for in regulations, · an ability to treat authorisations for economic development to issue which are more easily changed than granted in a country other than New licences for space activities: launch primary legislation, and in guidance. Zealand as meeting the requirements licences (i.e. licences for launch companies, One important consideration in relation of the act; such as Rocket Lab, that launch space to designing a performance-based regime is · risk-based and proportionate decision objects into space) (ss7, 23), payload that while performance-based regimes making that tailors requirements to permits (i.e. permits for satellites that will apparently avoid imposing unnecessary allow a graduated approach to risk be launched by companies such as Rocket compliance costs on industry, if a regulator management rather than a one-size- Lab) (ss15, 31), launch facility licences is truly going to be able to tailor decisions to fits-all approach – e.g. flexibility to set (on the assumption that in the future individual cases, a significant investment is conditions; and there could be space ports with a different required in regulatory capability. Without · future-proofing to accommodate person operating a launch facility than the that, it is not possible for the regulator to do changes in technology and markets. person who carries out the launch) (s38) their job adequately. and high-altitude vehicle licences (vehicles Recognising that MBIE currently has A decision-making framework with economic that operate at very high altitudes are limited technical competence in relation development as one of the objectives increasingly innovative, with some to space activities, the new act (s51) allows Having regard to the economic benefits carrying out much the same activities as the responsible minister to take into of space activities outlined earlier, satellites) (s45).

Page 32 – Policy Quarterly – Volume 13, Issue 4 – November 2017 The minister must decline licences if The detailed requirements imposed in regulatory regime we became aware of a he or she is not satisfied that certain risks relation to these tests are prescribed in range of new technologies being developed are not being managed, and may decline a the regulations that support the act (the to operate at very high altitudes (above licence if he or she considers that the Outer Space and High-altitude Activities the upper limit of controlled airspace) launch is not in the national interest. The (Licences and Permits) Regulations 2017). and performing similar functions to minister has been given an explicit power As outlined above, the minister can also satellites, including Earth observation and to weigh up the economic benefits decline a licence if he or she considers that internet connectivity. In order to future- associated with the activity against other it is not in the national interests. proof the space activities regime for these national interests when determining However, the primary legislation also new technologies, and to ensure that whether a licence should be declined on enables the decision maker to apply a similar applications and services provided national interest grounds (clauses 2 and 3 graduated approach to risk management. by different technologies are treated of ss9, 17, 25, 33, 40, 47). For example, flexibility is provided consistently, the act brings high-altitude through the ability to set licence vehicles within scope of the regulatory An ability to recognise overseas licences conditions that can be varied according to regime. New Zealand is the first country One of the ways the act keeps compliance the circumstances. in the world to establish a high-altitude costs down is by enabling the decision Discretion is also explicitly provided activities regime to govern these activities. maker to take into account foreign licences for in the language of the act. For example, The act’s regulation-making powers for the activity when deciding whether to with respect to managing New Zealand’s also provide the necessary scope to deal grant a licence for launches, payloads or launch facilities. This approach has also been taken to enable New Zealand’s space One of the ways the act keeps regime to operate effectively from day one compliance costs down is by enabling – recognising that the initial applications for a New Zealand licence will come from the decision maker to take into account Rocket Lab and its customers, who are subject to the US licensing regime. foreign licences for the activity when The ability for foreign licences to be deciding whether to grant a licence for taken into account has influenced the nature of the regulations, particularly the launches, payloads or launch facilities. nature of the information that needs to be provided with licence applications. For example, there is no need for an applicant to provide information to satisfy the potential liability, the act provides that the with changes to technologies and markets. minister that the applicant has the minister may require a licensee, as a For example, the act establishes technical capability to carry out an condition of the licence, to indemnify the regulation-making powers to prescribe activity, that the activity is safe, or that Crown in whole or in part against any that a thing is or is not, for the purposes of orbital debris mitigation requirements claim brought against the Crown under the the act, a launch vehicle, a space object or will be met if a competent body in another Liability Convention or the Outer Space a high-altitude vehicle (s88(1)(11)). jurisdiction has already undertaken this Treaty, or any other claim brought against While powers of this kind that allow assessment. The aim of this is to reduce the Crown under international law (ss10, regulations to override a statute (often unnecessary duplication and cost. 18, 26, 34). The experience in other described as Henry VIII clauses) are International cooperation arrangements jurisdictions has been that, unless the generally frowned upon, both the with certain foreign regulators will liability/indemnity regime is properly Legislation Advisory and Design facilitate this aspect of the regime. managed, innovation is easily stifled. The Committee and subsequently the Foreign minister’s ability to set tailored indemnity/ Affairs and Defence Select Committee Risk-based and proportionate decision insurance requirements having regard to accepted that in areas such as this, where making the nature of the risks associated with the legislation struggles to keep up with rapid The act provides that the decision maker particular launch is intended to ensure that innovation, broad regulation-making may grant a licence if he/she is satisfied innovation is not stifled by onerous powers are justified (albeit with that certain threshold tests have been insurance requirements that are not safeguards, such as consultation met.3 These tests address key risk areas justified by the risks associated with the requirements and a requirement for associated with space activities, such as activity. reasons) (s88(1) and (2)), which were risks caused by an applicant’s lack of incorporated into the legislation at the technical capability, risks to public safety, Future-proofing the act to deal with emerging suggestion of the Legislation Advisory risks to the space environment and risks technologies and activities and Design Committee. of breaching international obligations. During the development of the space

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 33 Managing the Opportunities and Risks Associated with Disruptive Technologies: space law in New Zealand

Carrying through the philosophy of the act reason, some important aspects of the · minimisation of the potential for into implementation law, such as the requirements for orbital on-orbit break-ups; The act comes into force on 21 December debris management and safety cases for · minimising the risk of collision; 2017. Regulations to support the act come space activities, were left to secondary · minimising the risk on Earth and in into force on the same day. Although the legislation. The approach taken in the space through post-mission disposal. act contains broad regulation-making regulations – for example, to prescribing Specifying the outcomes in regulations powers, not all of them must be used requirements for orbital debris mitigation provides certainty for applicants about initially or are relevant for every activity; plans and safety cases – is consistent with the outcomes that the debris mitigation they have been built into the act to future- the outcomes-based approach in the plan will need to achieve while also proof it and to enable a flexible approach primary legislation. providing flexibility about how the to managing risk. However, certain The regulations were developed in requirements are met. There are a range regulations are necessary to implement close consultation with other New of international norms and standards for the act when it comes into force. These Zealand government agencies which will orbital debris mitigation that will meet include: be involved in undertaking the these outcomes. These include the United · requirements for licences and assessments required by the act, and with Nations Debris Mitigation Guidelines, the permits, particularly the information prospective applicants to ensure that the Inter-Agency Space Debris Coordination that applicants provide; regulations provided the necessary degree Committee (IADC) Space Debris Mitigation Guidelines and the International Standards Organisation While the potential for conflict between (ISO) 24113. However, as the standards Rocket Lab’s activities and the interests are evolving and there is not yet a single internationally recognised standard, it is of airlines and shippers is minimal, not practicable to enshrine a particular standard in regulations. Instead, the if space activity increases this may regulations will require an orbital debris become an issue with which we will mitigation plan to state what standards have been applied and whether the plan have to grapple. has been independently verified.

Safety cases The requirement for a safety case · requirements for orbital debris of certainty and transparency whilst not (schedule 3(10)) puts in place the mitigation plans; imposing unnecessary compliance costs. regulatory mechanisms to enable the · requirements for safety cases for regulator to be satisfied of the safety of launch licences, launch facility Orbital debris mitigation plans the proposed activity. At the highest level, licences and (non-aircraft) high- Orbital debris poses a significant threat the safety case provision will require the altitude vehicles; to sustainable access to space and New applicant to identify their approach to · the circumstances in which certain Zealand wants to position itself as a managing safety, including ensuring that vehicles that go into high altitude are responsible regulator by ensuring that known significant risks are identified and not high-altitude vehicles and hence space activities from New Zealand do not appropriate controls in place. won’t require a licence. unacceptably contribute to the problem. To inform what should be required for The decision about what went into the However, it is also important that New a safety case we looked at the regulations act and what went into the regulations Zealand does not prescribe detailed developed to manage safety in other was informed by the Legislation and orbital debris mitigation requirements regimes dealing with high-risk activities, Design Committee guidelines on the which could have the effect of deterring such as the Health and Safety at Work allocation of power between Parliament space activities from New Zealand and (Major Facilities Regulations) 2016 and and the executive. In times of rapid potentially inhibit innovative approaches parts of the Railways Act 2005. From these innovation and technology change, the to managing orbital debris. we distilled the key elements that ability to change the law quickly is one of The regulations prescribe the high- applicants will be required to include in a the most important considerations in this level outcomes that a debris mitigation safety case. However, as space activities allocation of power. In the case of space plan would need to achieve, consistent are new to New Zealand, the safety case activities, we were confronted not only with international best practice requirements have been pitched at a with rapidly evolving technology but also (regulation 13). The outcomes proposed relatively high level. On balance, we with evolving standards internationally, are: judged that this was preferable to setting which made the ability to be able to · limitation of debris released during highly detailed requirements upfront change the law quickly essential. For that normal operations; which might stifle innovation, and also

Page 34 – Policy Quarterly – Volume 13, Issue 4 – November 2017 carried high risks of us getting it wrong. Zealand’s regulatory requirements. A key The act requires that there be a review This is consistent with the approach taken focus for MBIE as the regulator is to as soon as practicable after the expiry of more generally in relation to the ensure that we can meet industry’s three years from the commencement of regulations. expectations of a responsive regulatory the act (s86). This will provide an regime while we build our knowledge and opportunity to adjust the primary Learning as we go and adapting the regime capability. This will also ensure that our legislation to take account of new as required recommendations to ministers about any developments in technology and/or It is not often that you get to develop conditions on licences are commensurate changes to international space law and a new regulatory regime from scratch, with the risks posed by the activities. practice, as well as (we hope) a wealth of let alone a regime that has to balance MBIE will also be interested in public/ experience gained from regulating a New obligations arising from international community expectations of the way that Zealand-based space industry. treaties entered into 50 years ago with the the space regime works. Associated with 1 The Liability Convention was invoked by Canada through present-day challenges of regulating an this is how trade-offs are made between diplomatic channels after the re-entry and subsequent crash industry undergoing rapid technological objectives related to the space regime and of the RORSAT Kosmos 954 on 24 January 1978 in north- west Canada, and led to a settlement for the costs of the and market changes. other regimes, such as the Civil Aviation clean-up and damages. 2 The review of the Space Activities Act 1998 was announced We will continue to work with other Act 1990 and the Maritime Safety Act by the minister for industry, innovation and science on 24 New Zealand regulators who can bring 1994. October 2015. 3 Even if satisfied of these matters, the minister may not knowledge and expertise about different An issue we identified early on as issue a licence if he/she is not satisfied that the launch is in aspects of the regime, including the something to monitor is the potential for the national interest, or if national security risks have been identified, which would result in the issuing of a certificate Ministry of Transport, the Civil Aviation competing interests in the use of airspace with the effect of vetoing the proposed activity. Authority and WorkSafe New Zealand. and the sea between space activities and We will also build our relationships with aviation and maritime activities. While Acknowledgement overseas space agencies and regulators. the potential for conflict between Rocket The authors wish to thank Professor In order to implement a responsive Lab’s activities and the interests of airlines Steven Freeland for his helpful comments regulatory regime, we will also continue and shippers is minimal, if space activity on an earlier version of this article. to work with industry players and space increases this may become an issue with participants to build relationships and which we will have to grapple. ensure that people understand New Fast-track your career in the public sector with a top-level professional qualification with a: Master of Public Policy or a Master of Public Management • Try the graduate pathway professional programme for a Master’s in Public Policy or Public Management with the School of Government. • This is a one-year course after the completion of a Bachelor qualification. School of • Public policy and public management are examined at Government the postgraduate level.

for more details visit victoria.ac.nz/sog/study

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 35 Matthew Boyd and Nick Wilson

Rapid Developments in Artificial Intelligence how might the New Zealand government respond?

‘Faced with the possibility of an intelligence explosion, how can we maximize the chances of a desirable outcome?’ — Chalmers, 2010

Introduction Intelligent systems are here and are likely to bring about a ‘fourth industrial Advances in artificial intelligence (AI) have opened revolution’ (Government Office for opportunities in a range of human endeavours (NSTC Science, 2015; Helbing et al., 2017; NSTC Committee on Technology, 2016). Committee on Technology, 2016). In response to the speed Systems with general intelligence, more of these developments there has been a burst of analysis capable than humans at most tasks, are more probable than not within 20–30 and dialogue in New Zealand. The New Zealand Institute years (Muller and Bostrom, 2016). According to Nicholas Davies, head of of Directors commissioned a white paper (Chapman Tripp, the World Economic Forum society and 2016); the Ministry of Business, Innovation and Employment innovation department, such systems will ‘fundamentally change the way we published Building a Digital Nation and the Strategic create value and do business, and value Science Investment Fund 2017–24 business plan (Ministry ourselves as human beings’ (New Zealand Herald, 2017). of Business, Innovation and Employment, 2017, 2016), AI is a global issue and presents great opportunities for benefit, but also great and supports the new Artificial Intelligence Forum of New risk. Risks range from economic to social Zealand (www.aiforum.org.nz). and psychological, to existential. We argue that these risks are insufficiently Matthew Boyd is the Director of Adapt Research Ltd, Wellington. Nick Wilson is a Professor of Public articulated in New Zealand government Health in the Department of Public Health, University of Otago, Wellington. reports to date, and there is an obligation

Page 36 – Policy Quarterly – Volume 13, Issue 4 – November 2017 for New Zealand government agencies to capable of things we cannot yet conceive. The risks of bias and injustice consider deep questions about the kind of If we soon design AI systems that are very Algorithmic bias leads AI systems to society we wish to live in and our role in much better than humans at designing AI demonstrate unexpected behaviour on the emerging global transition to an AI systems then an explosion of intelligence occasion. Microsoft’s Twitter chatbot world. We explain AI briefly, identify four may occur spectacularly thereafter. Rapid learned to be racist from human data key risks that the policy response needs to evolution of intelligence may occur as well sets (Devlin, 2017; Gibbs, 2016). AI is not reflect, and conclude that New Zealand through Darwinian selective processes immune to human prejudice (IBM, 2016). needs to support serious research into the (Association for Computing Machinery, The public may be much less forgiving of risks of AI. We need informed debate, 2017). But subsequent generations of AI a biased machine than a biased individual, formal public engagement and emergent may contain ‘bugs’ that lead to unforeseen especially in critical domains. policy. consequences, which humans are unable The New Zealand government makes to remedy. use of an integrated data infrastructure What is AI? The applications of AI are catalogued (IDI) to help target social investment by AI systems are digital systems that elsewhere (Chapman Tripp, 2016; using past data to predict likely actions or automate or replicate intelligent Government Office for Science, 2015; qualities of different groups. Certain behaviour. AI systems use data to infer IBM, 2016; NSTC Committee on objective facts about individuals make the patterns and learn solutions to problems. Technology, 2016). It is very clear that AI probability of them needing social Programmers provide AI systems will influence policy, but we need policy assistance very much higher (McLeod et with goals rather than strict methods. Present intelligent systems may augment the intelligence of a user, or provide A UK government report on AI states domain-specific expertise (narrow AI). that ‘at present there is generally AI development also has the potential to produce artificial general-purpose agreement that there should be a intelligent systems (AGI) – even sentient systems according to some speculators. human in the loop … the nature of their One reason AI has such enormous role is likely to evolve as the technology potential is that the world has been stockpiling information. This information develops’... takes the form of our internet search histories, purchase histories, social media posts, blogs, media reports, GPS data, government databases, marketing about AI. Policy around AI is going to al., 2015). Statistical profiling is also used databases, sensor databases; any form of have to be flexible to accommodate rapid in insurance, law enforcement and many data you can imagine that is stored change and will need to be based upon other domains. The IDI is a governmental digitally. The quantity of information principles of upholding core societal database, and subject to government available is doubling year on year (Helbing values. ethics, but databases outside government et al., 2017; IBM, 2016). Such vast data In this article we intentionally avoid control do not necessarily benefit from sets arise in part because of the the kind of ‘theological speculation’ found such protection. Even within government, ‘surveillance capitalism’ that pervades the in many radical assessments of AI ethical checks may not always occur in globe (Zuboff, 2015). We have created a (Chalmers, 2010; Muller, 2014), and focus intelligence or law enforcement activities. rich representation of social reality, filled on the social and personal risks of rapid, This is especially a risk if humans don’t with sequences of cause and effect, unfettered development and fully understand the systems. Even ethical associations, beliefs, emotions, goals, implementation of AI. safeguards may not be able to overcome hopes, dreams, memories and behaviour. the bias of machines that learn from These data are what AI systems learn from What are the plausible risks of AI? human data sets. and the methods and rules that AI Broad and non-specific risk statements Sophisticated AI systems, analysing employs in learning and making inference are easy to ignore. Details matter and the stockpile of data representing the are often opaque to human observers. are necessary. Inspired by the New structure of human social reality, may AI brings great economic and social Zealand Institute of Directors report, we infer great insight about the probabilities possibility, with potential for deep insight now elaborate on the potential for AI to of risks to individuals, augmenting human and productivity gains, novel methods, radically transform our world. We cannot social work, police work or health care delegation of decision making to catalogue all the risks of AI here, so we and improving human decisions through automatic systems, task competence have chosen four particularly challenging prescience. Social media can already be exceeding expert humans, and the risks, not yet detailed in key New Zealand used to predict some future events possibility of artificial superintelligence policy documents. (Phillips et al., 2017).

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 37 Rapid Developments in Artificial Intelligence: how might the New Zealand government respond?

A UK government report on AI states Indeed, our psychology is already potential to undermine not only truth, but that ‘at present there is generally exploited by advertising, propaganda and free will, autonomy and democracy agreement that there should be a human rhetoric. Connectivity and digital (Helbing et al., 2017). If the AI systems of in the loop … the nature of their role is platforms make it much easier to share interest groups become proficient at likely to evolve as the technology develops’ and spread information. It can be exploiting patterns of cause and effect we (Government Office for Science, 2015, relatively easy to manipulate the public’s aren’t even aware of, reality may recede in a p.10). What happens when AI is perception of reality, and technological storm of artificial content while we remain demonstrably more reliable than human manipulation of public opinion is a daily oblivious to our own manipulation decision makers? Do we remove the occurrence. The ‘fake news’ phenomenon (Woolley and Howard, 2017). human from the loop? No system of illustrates this (Gu, Kropotov and It may be that attempts to control prediction is perfect and the inference Yarochkin, 2017; Woolley and Howard, opinions are doomed to fail; however, rules of AI systems may be inaccessible to 2017). In 2017, 45% of Twitter activity in what results from such attempts is unpre- human understanding. Ought such Russia was estimated to be automated dictable. No one knows where ‘persuasive systems be used in insurance, law (Woolley and Howard, 2017). computing’ and ‘computational propa- enforcement, social investment, or in the Furthermore, content is also becoming ganda’ may lead us. interests of profit? What happens when more individualised. human decisions and AI conflict? New Private and public entities already use The risks of economic chaos and the transformation of work AI has the potential to massively disrupt Even if humans remain in control of our core economic systems. Many the intelligent systems we design, ... reports detail mass unemployment due to automation. New jobs created may AI technologies threaten to make us not be jobs that New Zealand’s labour market is equipped to capitalise upon. A vulnerable, alienated and, paradoxically, 2017 OECD report cited New Zealand’s ‘automated masters’ of our creations. low productivity and weakness in mathematics as barriers (OECD, 2017). The twin forces of job loss and profit gain create widening inequality. This is a major Zealand’s chief science advisor Professor ‘big nudging’ to provide information that policy concern given the relationship Sir Peter Gluckman notes that: exploits the relationship between between socio-economic conditions and psychological biases and behaviour health (Marmot and Allen, 2014). While prediction based on risk factors (Benartzi et al., 2017). Such ‘mind hacking’ Whether society values something, or is a key objective … such predictive works probabilistically on a population someone, is contingent on the norms, approaches will identify risk and level. We also know that ‘fake news’ can beliefs and needs of the time (Sandel, resilience factors based on group drive real behaviour. 2010). There is a risk that present systems characteristics and there are Psychological exploitation is possible will become unfair with the arrival of AI. significant limits and dangers in on an unprecedented scale with the help A new and just distribution of resources is extrapolating this to a specific of intelligent machines exploiting the needed as many workers begin to suffer individual. (Morton, 2017) structure and function of social media through no fault of their own. We risk and vast data sets. For example, a Trend having large numbers of economically These situations (which could involve Micro report claims that it costs $200,000 valueless citizens and a minority of issues of health care or prejudice) raise to generate fake social media and technologically literate people acquiring significant questions about liability, authentic-seeming news that results in a unprecedented wealth and influence. control, fairness, privacy and society. Will real-life demonstration about an issue No one really knows where this will we, and ought we to, accept more and that doesn’t exist (Gu, Kropotov and end up. But the endgame of an intelligence more delegation of authority to machines? Yarochkin, 2017). AI systems could enact explosion might be a post-scarcity such hacks much more efficiently with economy, where economic growth rates The risks of AI dominance of media discourse fabricated truth ‘that panders to its increase dramatically (Bostrom, Dafoe Human rationality is bounded. We are audience’s ideologies … enough to and Flynn, 2016), supply outstrips subject to biases and heuristics of thinking compel people to join an imagined cause’ demand and the value of money collapses that control the information we believe (ibid., p.60). (Starkey, 2017). One possible solution to and may confound our best intentions Combine big nudging, fake content, a this issue is a universal basic income (Gigerenzer, 2008; Kahneman, 2011; greater understanding of human psychology funded by taxing robots that supplant Richerson and Boyd, 2004). This means and its vulnerabilities, and the ability of AI human workers (James, 2017; Nauman, that our psychology is exploitable. to individualise content: this has the 2017). This is the preferred position of

Page 38 – Policy Quarterly – Volume 13, Issue 4 – November 2017 global tech leaders such as Bill Gates and that could be controlled and interfered If [humans] question the advice they Elon Musk. with through an internet connection. receive, however, they may be thought Even if humans remain in control of the These critical systems include power reckless, more so if events show their intelligent systems we design (as opposed to grids, food supply chains and quarantine decision to be poor … departments being influenced and swayed by them), AI systems. They could potentially include will need to be transparent about the technologies threaten to make us vulnerable, future geoengineering systems that could, role played by artificial intelligence in alienated and, paradoxically, ‘automated if interfered with, cause ecological havoc. their decisions. (Government Office masters’ of our creations. We risk falling into Pervading the four risks outlined here for Science, 2015, p.10) a state where we lack know-how, and are (and other risks identified elsewhere) are dependent on algorithmic processes that a set of moral and ethical themes, which Some of the solutions proposed in the control our lives and undertake the beg for debate and policy. These themes UK report are local (such as certification meaningful work we once did. This ‘tragedy centre on the locus of power and control for AI engineers) and so may not address of the master’ (Coeckelbergh, 2015) has (at levels of society and human–AI global risks in a connected digital world. profound implications for power, knowledge interaction). There are themes of privacy, The UK government also has a Data and experience. Increasing dependence on freedom and autonomy, liability, Science Ethics Framework (UK Cabinet AI could ultimately lead to loss of meaning as human work is replaced by robots (Nauman, 2017) and we voluntarily submit, AI is a problem space where letting algorithms rule our lives (White, 2015). ideologically diverse parties must come together over ethical issues, and New Security and existential risks The US Intelligence Community outlines Zealand has a history as a flexible the risks of AI in a 2017 report (Coats, 2017). These include the vulnerability of legislator and innovator in the space of AI systems to cyber attack, and advances in social protection foreign weapon and intelligence systems (in particular, autonomous weapon systems). Autonomous weapon systems could be made extremely difficult to ‘turn regulation and safety, and curtailing Office, 2016) which goes some way to off’ to evade enemy interference, but this malicious intent. guiding public sector data use, but we could make them inherently dangerous. need regulation around the private sector Some of the risks of AI seem to be International response to the risks of too. A Royal Society report identifies genuinely existential (Bostrom, 2014; artificial intelligence ‘social issues’, ‘implications for data use’ Chalmers, 2010; Danaher, 2015). These are The US government calls for a whole-of- and ‘security and control’ as issues (Royal particularly concerning given that AI government response to AI, and outlines Society, 2017); but it contains little actual research and development might be faster 23 recommendations to ensure that the detail of what these risks entail. than expected and catch policymakers off long-term consequences of AI are beneficial The Canadian government has initiated guard as we face systems we do not (NSTC Committee on Technology, 2016). a $125 million Pan Canadian AI Strategy understand. Existential risk from AI could Identified risks, alongside regulatory ones, aimed at making Canada a world leader in be possible in one of several scenarios: first, include inequality, employment disruption, AI and attracting top talent (Canadian if AI is programmed to do something challenges in trying to understand and Institute for Advanced Research, 2017). devastating; second, if AI chooses a predict the behaviour of AI systems, and Google, Amazon, Facebook, IBM and destructive or perverse method to pursue the safety of AI ‘when exposed to the full Microsoft have created the Partnership on benevolent goals (Bostrom, 2014). In either complexity of the human environment’ Artificial Intelligence to Benefit People and case, very competent AI would pursue goals (ibid., p.2). Society to conduct research, including on that are misaligned to those of humans. The US recommends mandatory ethi- ethics and fairness (Hern, 2016). IBM Alternatively, AI could pose an cal training of AI practitioners, policy identifies issues of safety, control and trust, existential threat by doing something consistent with international humanitari- and that a fact-based dialogue is needed to accidental or unexpected (think firing an law, monitoring of milestones in AI inform progressive social and economic nuclear weapons without a human-like development, bilateral talks with foreign policy (IBM, 2016). Elon Musk is funding grasp of the consequences, or devising a governments, and ongoing public engage- open AI to advance digital intelligence in a potent biological pathogen without ment. way that is most likely to benefit humanity knowing it will infect humans). AI systems A UK government report identifies as a whole. The Future of Life Institute has don’t have to be robotic to pose a physical similar risks, and in our opinion a critical published an open letter regarding the safe existential threat to humans; there is a lot risk pertaining to advice from AI: development of AI and a list of research

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 39 Rapid Developments in Artificial Intelligence: how might the New Zealand government respond? priorities to ensure that AI is beneficial · allocation (all people are exposed to report calls for a high-level working group (Russell, Dewey and Tegmark, 2016). the risks of AI so there must be to research these issues, and for a whole- A collective of European academics resource shuffling to recognise risk of-government and whole-of-nation has recently published an opinion piece externalities and the need for justice, approach. We agree with these calls to in Scientific American offering warnings given that we are behind a veil of action. These are necessary – but not about some of the most insidious risks of ignorance regarding a post-AI world); sufficient – responses to the risks posed by AI (Helbing et al., 2017). Their views are · population (we must consider how to AI. The New Zealand government needs critical warnings about change in the treat AI systems and what kinds of new to see AI as a wider issue than merely an nature of society and human reality. The entities to bring into existence); and instrumental tool for increasing GDP, and European Union (EU) has taken official · context transformation needs to be transparent in communicating steps towards implementing civil law (responsibility and wisdom are the changes AI poses for society. rules on robotics and requirements to needed in a radically unfamiliar The Ministry of Business, Innovation register advanced robots (European environment). and Employment writes that we ought to Parliament, 2017). Also, the EU’s new By ‘governance’ the authors refer not ‘accelerate the safe adoption of AI General Data Protection Regulation only to the actions of states but also to technologies’ (Ministry of Business, effectively creates a ‘right to explanation’, transnational governance. Innovation and Employment, 2017, p.7). whereby a user can ask for an explanation It favours collaboration between the government, Callaghan Innovation and industry, and supports the AI Forum – yet The New Zealand Law Foundation ... nascent – to undertake research into AI to identify opportunities and mitigate risks. is funding a $400,000 University of The AI Forum has an agenda for open Otago project which aims to explore the discussion around policy and awareness of AI, the economy, and capability and possible implications of AI innovations skills needs. It also aims to balance the conversation by providing evidenced for law and public policy in New arguments against AI doomsayers. The Zealand. forum’s first research project is a stocktake of issues around the economy, society, education and government. This includes New Zealand’s readiness for AI, the direct of an algorithmic decision that was made What has been the response to AI risks in and indirect impacts, skills needed and about them. This will drive global New Zealand? government opportunities. The AI Forum standards for anyone who wants to We argue that in comparison with the looks set to provide important deploy their AI products in the EU response of some nations, New Zealand information for the government to (Goodman and Flaxman, 2016). lacks a governmental response. We also consider in policymaking. More of this It is somewhat surprising that policy argue that the global response yet lacks sort of activity focused on the New documents produced by governments pay the coordination required to deal with Zealand context is needed – the sooner little attention to the outputs of truly global risks. the better. organisations such as the Centre for The New Zealand Institute of Directors, These sentiments were reinforced in Public Impact, the AI Initiative of the noting the lack of dialogue about AI in the 2017 Royal Society of New Zealand’s Future Society at Harvard Kennedy New Zealand policy, published a horizon regional lecture series, where Professor School, the Future of Humanity Institute scan of AI in New Zealand (Chapman Alistair Knott spoke of employment at Oxford University, and others. Many of Tripp, 2016). This report surveys the issues, machine bias, transparency, these academics concur that we need opportunities and risks and identifies accountability and ethics. He argues that some form of global governance board potential inequality, unemployment, and interdisciplinary structures are required, (Bostrom, Dafoe and Flynn, 2016). Given legal and regulatory needs. However, there which would include AI researchers, AI the risks, the transition to machine is little discussion of the threat to freedoms companies, economists, lawyers, social superintelligence requires a set of ‘policy and autonomy, to social power, of the risk scientists and ethicists (AI Forum New desiderata’, these authors argue. of autonomous weapons, or the many Zealand, 2017b). The New Zealand Law Policymakers must pay attention to: other potential risks of AI. Foundation, an independent charitable · efficiency (providing technological The Institute of Directors’ report trust, is funding a $400,000 University of opportunity, mitigating AI risk and poses two critical questions: what ethical Otago project which aims to explore the ensuring global stabilisation, e.g. challenges does widespread use of AI possible implications of AI innovations through the use of a single AI raise?, and what controls and limitations for law and public policy in New Zealand. governance body); should be placed on AI technology? The The study is a collaboration between the

Page 40 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Faculty of Law and the departments of What New Zealand might do Research the risks and impact of AI: Philosophy and Computer Science. The arrival of AI is a collective choice Government should fund research However, the New Zealand problem at a national and global level. It and reports on AI that include the government remains strikingly upbeat is not simply a matter of ensuring that ethical/philosophical/social and about AI and articulates few risks. New Zealand stays ‘ahead’, as the Institute psychological issues; fund Coupled with Ministry of Business, of Directors white paper argues. The issue engagement with the public and a Innovation and Employment’s strategy of AI bears much in common with climate societal discourse; and ensure that we ought to promote New Zealand as change. The New Zealand government mechanisms so that the findings of a test bed for emerging technologies, this and private and public organisations need studies such as Otago University’s can is concerning. We don’t yet know whether to focus on the relationship between AI inform policy. ‘Funding could come a fully informed New Zealand public and core values. More ventures like the AI from the government’s $410m would concur with this position. Forum are needed, along with extensive Strategic Science Investment Fund’ public engagement. (Ministry of of Business, Innovation More is needed and more global coordination If we agree that ‘[i]t is totally and Employment, 2016). We need to create national and global unacceptable … to use these technologies Inform and engage the public: The norms surrounding AI. We need to ensure to incapacitate the citizen’ (Helbing et government has a responsibility to that current regulation around data access, al., 2017, p.15), then we need to negotiate digest the outputs of research use, privacy and consent are robust at a new social contract, with a policy initiatives such as Google Brain, international level. These are truly global issues, and regulating within borders will not prevent abuse across borders. AI is an ... for what world of affordances do we opportunity for us to revisit inequality and justice on a global scale. want to be held accountable? Key existing policy recommendations include, but are not limited to, the following: · monitor AI development milestones framework which sees citizens as OpenAI, the Machine Intelligence to aid prediction; partners and protects the right of people Research Institute and the Future of · devise mechanisms to ensure fairness to clean information, to allow them to Humanity Institute, and a range of of benefit distribution; lead the truthfully informed, self- academic publications, and translate · support informational self- determined lives critical to a functioning these so that the New Zealand public determination and popular democracy. remains informed. Embedding a participation; Key questions that require a local programme of digital ethics within · improve transparency and remove answer include: the New Zealand educational ‘information pollution’; · Do our present legal tools provide curriculum is another option. · improve collaboration at national and suitable options for dealing with the Produce clear recommendations: global levels; issues posed by AI? Existing policy needs to be analysed, · promote responsible behaviour · Ought the world to permit international policy co-opted as through digital literacy and digital autonomous weapon systems? appropriate, and new policy around ethics. (Bostrom et al., 2016; Helbing · Ought we to permit individually the legitimate and low-risk use of AI et al., 2017; NSTC Committee on targeted persuasion systems that developed. Policy needs to cover risks Technology, 2016) threaten to undermine a truthfully of: bias and injustice, dominance of Given the risks, policy development is informed public? media discourse, autonomy, economic critical so that we don’t throw away · What are the limits of nudging in chaos and security. advances in democracy and human rights the public interest, and the Take a global lead: The government by succumbing to insidiously anti- permissibility of nudging for private support the formation of a single democratic risks like persuasive computing. interests? global body on AI similar to the Workers and other ordinary citizens · Is a universal basic income one of Intergovernmental Panel of Climate must be engaged or decisions will be made the solutions to the potential for Change, or an ‘AI Club’ as suggested for them by people who care more about dramatic inequality? by Nordhaus for addressing climate personal interests. We may need to move · Are there emerging tools which change (Nordhaus, 2015). It should toward more collective notions of might offer solutions to some of the advocate for social justice and equity responsibility, and this needs to have the threats posed by AI? on the global stage; and advocate for means and scope to include non-human Given the above, we suggest that New ethics around AI and protection of actions (White, 2015). A global response Zealand policymakers ought to pursue the rights, privacy and safety, so we are is needed (Lee, 2017). following five actions: not forced to follow the lead of others.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 41 Rapid Developments in Artificial Intelligence: how might the New Zealand government respond?

Maintain a vision for New Zealand: Conclusion (relating to autonomy, vulnerability, We must maintain a vision of New The risks of AI range from comparatively disconnection from decision processes Zealand as a society of equality, minor issues of privacy and liability, and the ethics of manipulation), let alone empowerment and autonomy, with through major societal and economic existential issues. rights to truthful information, where issues, to issues of existential risk. In New Zealand punches above its weight we are protected from weapons of general, the lack of detail on risk in on global issues and has been a world mass destruction. These are non- government reports proffers a false leader on women’s suffrage, nuclear policy partisan issues. sense of security and of the absence of and addressing colonial injustice. New fundamental risks to society. This appears Zealand can encourage and work with We must prepare for a qualitatively to be especially the case in the limited New other countries to move in the right different kind of society and move on Zealand policy material on AI produced direction, but we need to decide from present thinking. The critical so far. One important reason that this is collectively what that direction looks like. question is, ‘for what world of affordances concerning is the fact that governments This article is not the place for reaching do we want to be held accountable?’ are not immune from causing accidental, normative conclusions, but these (White, 2015). Issues of foreseeability and or indeed intended, harm. Many of questions need New Zealand answers. negligence must be central to these the examples we have presented focus With policy decisions being heavily discussions. AI is a problem space where on threats from the private sector, but dependent on values and high uncertainty, ideologically diverse parties must come governments can be just as capable of New Zealand can act as a global honest together over ethical issues, and New AI-driven ‘Big Brother’ social control as broker for forging international policy Zealand has a history as a flexible legislator private entities. solutions. and innovator in the space of social None of the responses to the risks of protection (for example, the Accident AI we have seen fully addresses the Compensation Corporation). problem of profound social change

References Association for Computing Machinery (2017) ‘Researchers are using Coeckelbergh, M. (2015) ‘The Tragedy of the Master: automation, Darwin’s theories to evolve AI, so only the strongest algorithms vulnerability, and distance’, Ethics and Information Technology, 17, survive’, press release, https://cacm.acm.org/news/214830- pp.219-29 researchers-are-using-darwins-theories-to-evolve-ai-so-only-the- Danaher, J. (2015) ‘Why AI doomsayers are like sceptical theists and why strongest-algorithms-survive/fulltext it matters’, Minds and Machines, 25 (3), pp.231-46, doi: 10.1007/ AI Forum New Zealand (2017) ‘How should society prepare for advances s11023-015-9365-y in artificial intelligence?’, https://aiforum.org.nz/news/2017/8/16/ Devlin, H. (2017) ‘AI programs exhibit racial and gender biases, research how-should-society-prepare-for-advances-in-artificial-intelligence-ai, reveals’, Guardian, 13 April, https://http://www.theguardian.com/ retrieved 21 August technology/2017/apr/13/ai-programs-exhibit-racist-and-sexist-biases- Benartzi, S., J. Beshears, K. Milkman, C. Sunstein, R. Thaler, M. Shankar, research-reveals W. Tucker-Ray, W. Congdon and S. Galing (2017) ‘Should governments European Parliament (2017) ‘Robots and artificial intelligence: MEPs call invest more in nudging?’, Psychological Science, 28 (8), pp.1041-55, for EU-wide liability rules’, press release, http://www.europarl.europa. epub 5 June, doi: 10.1177/0956797617702501 eu/news/en/press-room/20170210IPR61808/robots-and-artificial- Bostrom, N. (2014) Superintelligence: paths, dangers, strategies, Oxford: intelligence-meps-call-for-eu-wide-liability-rules Oxford University Press Gibbs, S. (2016) ‘Microsoft’s racist chatbot returns with drug-smoking Bostrom, N., A. Dafoe and C. Flynn (2016) Policy Desiderata in the Twitter meltdown’, Guardian, 30 March, https://http://www. Development of Machine Superintelligence, Oxford: Future of Humanity theguardian.com/technology/2016/mar/30/microsoft-racist-sexist- Institute, Oxford Martin School, Oxford University chatbot-twitter-drugs Canadian Institute for Advanced Research (2017) ‘Canada funds $125 Gigerenzer, G. (2008) Rationality for Mortals: how people cope with million Pan-Canadian artificial intelligence strategy’, press release, uncertainty, Oxford: Oxford University Press retrieved from http://www.newswire.ca/news-releases/canada-funds- Goodman, B. and S. Flaxman (2016) ‘European Union regulations on 125-million-pan-canadian-artificial-intelligence-strategy-616876434. algorithmic decision-making and a “right to explanation”’, retrieved html from arXiv:1606.08813 Chalmers, D. (2010) ‘The singularity: a philosophical analysis’, Journal of Government Office for Science (2015) Artificial Intelligence: opportunities Consciousness Studies, 17 (9–10), pp.7-65 and implications for the future of decision making, London: Chapman Tripp (2016) Determining our Future: artificial intelligence Government Office for Science opportunities and challenges for New Zealand: a call to action, Gu, L., V. Kropotov and F. Yarochkin (2017) ‘The fake news machine: how Auckland: Institute of Directors propagandists abuse the Internet and manipulate the public’, Trend Coats, D. (2017) Worldwide Threat Assessment of the US Intelligence Micro, 13 June Community, Washington: Office of the Director of National Intelligence

Page 42 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Helbing, D., B. Frey, G. Gigerenzer, E. Hafen, M. Hagner, Y. Hofstetter, J. Nauman, Z. (2017) ‘AI will make life meaningless, Elon Musk warns’, New van den Hoven, R. Zicari and A. Zwitter (2017) ‘Will democracy York Post, 17 February, http://nypost.com/2017/02/17/elon-musk- survive big data and artificial intelligence?’, Scientific American, 25 thinks-artificial-intelligence-will-destroy-the-meaning-of-life/ February New Zealand Herald (2017) ‘How to survive the jobs apocalypse’, New Hern, A. (2016) ‘“Partnership on AI” formed by Google, Facebook, Zealand Herald, 26 May, http://www.nzherald.co.nz/business/news/ Amazon, IBM and Microsoft’, Guardian, 28 September, https://http:// article.cfm?c_id=3andobjectid=11863989 www.theguardian.com/technology/2016/sep/28/google-facebook- Nordhaus, W. (2015) ‘Climate clubs: overcoming free-riding in amazon-ibm-microsoft-partnership-on-ai-tech-firms international climate policy’, American Economic Review, 105 (4), IBM (2016) Preparing for the future of artificial intelligence: IBM response pp.1339-70 to the White House Office of Science and Technology Policy’s request NSTC Committee on Technology (2016) Preparing for the Future of for information, IBM Artificial Intelligence, Washington, DC: US National Science and James, M. (2017) ‘Here’s how Bill Gates’ plan to tax robots could actually Technology Council happen’, Business Insider, 20 March OECD (2017) 2017 OECD Economic Survey of New Zealand, Paris: OECD Kahneman, D. (2011) Thinking, Fast and Slow, New York: Farrar, Straus Phillips, L., C. Dowling, K. Shaffer, N. Hodas and S. Volkova (2017) ‘Using and Giroux social media to predict the future: a systematic literature review’, Lee, K. (2017) ‘The real threat of artificial intelligence’, New York Times, retrieved from arXiv:1706.06134 24 June, https://http://www.nytimes.com/2017/06/24/opinion/sunday/ Richerson, P. and R. Boyd (2004) Not by Genes Alone: how culture artificial-intelligence-economic-inequality.html transformed human evolution, Chicago: University of Chicago Press Marmot, M. and J. Allen (2014) ‘Social determinants of health equity’, Royal Society (2017) Machine Learning: the power and promise of American Journal of Public Health, 104 (supplement 4), pp.S517-9 computers that learn by example, London: Royal Society Ministry of Business, Innovation and Employment (2016) Strategic Russell, S., D. Dewey and M. Tegmark (2016) ‘Research priorities for Science Investment Fund: investment plan 2017–2024, Wellington: robust and beneficial artificial intelligence’, retrieved from Ministry of Business Innovation and Employment arXiv:1602.03506 Ministry of Business, Innovation and Employment (2017) Building a Digital Sandel, M. (2010) Justice: what’s the right thing to do?, London: Penguin Nation, Wellington: Ministry of Business Innovation and Employment Starkey, D. (2017) ‘Elon Musk: automation will force universal basic McLeod, K., R. Templeton, C. Ball, S. Tumen, S. Crichton and S. Dixon income’, Geek.com, 29 May, https://http://www.geek.com/tech- (2015) Using Integrated Administrative Data to Identify Youth Who Are science-3/elon-musk-automation-will-force-universal-basic- at Risk of Poor Outcomes as Adults, Wellington: Treasury income-1701217/ Morton, J. (2017) ‘Top scientist discusses big data, social policy’, New UK Cabinet Office (2016) Data Science Ethical Framework, London: Zealand Herald, 19 June, http://www.nzherald.co.nz/nz/news/article. Cabinet Office cfm?c_id=1andobjectid=11878987 White, J. (ed.) (2015) Rethinking Machine Ethics in the Age of Ubiquitous Muller, V. (2014) ‘Risks of general artificial intelligence’, Journal of Technology, Hershey, PA: IGI Global Experimental and Theoretical Artificial Intelligence, 26 (3), pp.297- Woolley, S. and P. Howard (2017) Computational Propaganda Worldwide: 301 executive summary, Oxford: Oxford Internet Institute, University of Muller, V. and N. Bostrom (2016) ‘Future progress in artificial intelligence: Oxford a survey of expert opinion’, in V. Muller and N. Bostrom (eds), Zuboff, S. (2015) ‘Big other: surveillance capitalism and the prospects of Fundamental Issues of Artificial Intelligence, Berlin: Springer an information civilization’, 30, pp.75-89

Forthcoming Events Title Speaker Date The Ground Between: Sefton Darby Friday 17th November Navigating the oil In association with 12:30 – 1:30pm and mining debate Bridget Williams Books Old Government Building, lecture theatre 3, in New Zealand 55 Lambton Quay RSVP: [email protected] Putting leadership in Inaugural lecture by Professor of Tuesday 21st November its place Public and Community Leadership, Lecture at 6pm. Brad Jackson Rutherford House lecture theatre 2, 23 Lambton Quay, Wellington RSVP: Please phone 04 463 7458 For further information on SOG Events visit our website http://www.victoria.ac.nz/sog

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 43 Toby Moore

Income Volatility in New Zealand ‘Economic risk is a lot like a hurricane. Hurricanes strike shaping individual incentives and levels of human capital, their exposure to shocks of powerfully and suddenly. They rip apart what they touch; various sorts and their ability to absorb or property, landscape and lives … And although they can be adapt to such shocks. The question of economic security for New Zealanders is prepared for, they cannot be prevented.’ These sentiments, therefore not simply one of social assistance, but of whether the system from Yale political scientist Jacob Hacker, explain why overall results in buffers of a public and economic risk is a concern for households, and why the private nature that are adequate for the challenges of the current environment. extent of that concern depends a great deal on how well This article examines this question through an analysis of the volatility of households are protected against risk. The potential for individual incomes in New Zealand. It individual bad luck to lead to hardship has meant that society begins by explaining the concept of income volatility and its importance as an has, in many instances, determined that individual risk indicator of economic security. The data, should be borne collectively through systems of social welfare methodology and findings of this analysis will then be explained, followed by a or social insurance (Hacker, 2008, p.5). consideration of the limitations of this approach in terms of both measurement It is useful to remind ourselves of this Though the redistributive role of and the conclusions that can be drawn. fact, because much public discussion of government is important, there is a risk The discussion section will address how economic security has a distinctly static that, in Nicholas Barr’s analogy, the this analysis relates to existing work on quality to it. Inequality, poverty and ‘Robin Hood’ aspect of the welfare state is income mobility, examine other evidence hardship are often debated as if the emphasised at the expense of its role as that might contextualise the findings and involved parties were frozen in position, the collective ‘Piggy Bank’ (Barr, 2001). suggest areas for future research. like figures on a complex and occasionally The state’s role in contributing to the cruel carousel. This tendency towards smooth life trajectories of its citizens is Measuring volatility in incomes treating social classes as relatively fixed not limited to welfare payments. The state Income volatility is the variance in means that social welfare can come to be also shapes the institutional setting in personal incomes over time. The seen as based more in charity than in which individuals and households international literature on income reciprocal obligations and risk sharing. operate. It therefore has some hand in volatility is substantial and growing, and mostly focused on year-to-year volatility, Toby Moore is a teaching fellow at the School of Government, Victoria University of Wellington. though there is a smaller body of work

Page 44 – Policy Quarterly – Volume 13, Issue 4 – November 2017 considering volatility over a shorter time results in low or inadequate incomes, then by cutting back on something more frame. As part of the 2017 ‘state of the the social, psychological and health- or fundamental than fine foods or holidays. state’ report, Fit for the Future: boosting education-related consequences Finally, income volatility measures resilience in the face of uncertainty, Deloitte associated with poverty will be of provide us with an indirect look at the and Victoria University of Wellington relevance (Boston and Chapple, 2014). extent to which the welfare system is constructed an index of income volatility However, downward shifts in income are redistributing across the life cycle. As for New Zealanders between 2001 and important for additional reasons, beyond noted by Hills (2015), the variance of 2014. This is the first study to construct the possibility of falling into poverty. living standards at different stages of a year-by-year index for New Zealand, Greater variability in incomes makes it people’s lives was a core rationale for the though other studies have addressed the much harder for households to plan for creation of the welfare state. In 1899 question indirectly (e.g., Carter, Mok and the future; to be able to make major Seebohm Rowntree identified the ‘five Le, 2014). investment, educational or lifestyle alternating periods of want and Measuring income volatility can decisions with reasonable confidence that comparative plenty’, based around the provide some insight into both the they will still be in a position to manage presence of children in the household and frequency and the impact of the shocks these in the years to come. An income loss the individual commencing, and faced by individuals, as well as the may be unanticipated, and one that retirement from, their working life. In the effectiveness of government transfers in households are ill-prepared for. This may model proposed by William Beveridge in offsetting economic loss. Disruption to force them to cut back on consumption, 1942, the effective transfer of resources income can be of a direct nature: for instance, a worker being made redundant; or more indirect, such as a serious health ... income volatility measures provide problem that prevents someone from working. Taking a broad view of the us with an indirect look at the extent likelihood of individual income loss does to which the welfare system is not tell us what the cause of that loss is. Nonetheless, it is useful to look at the level redistributing across the life cycle. and trend of income volatility, if only because macroeconomic trends may not tell us the full story of what is happening at the micro level, and averages can often disguise changes in distribution. deplete what savings they have or resort to over time could help to smooth out the When assessing the social and borrowing. periods of want by providing child economic context in which households Furthermore, when households are allowances to those with additional operate, there is a tendency to resort to faced with falls in their income, changes mouths to feed and pensions to those in broad narrative accounts such as within the basket of goods and services retirement. The same logic applied to ‘globalisation’, ‘the great moderation’ or consumed by households become very misfortune which might intermittently ‘the great recession’. (More recent important. Though we are living through threaten the livelihood of individuals and examples would include ‘secular a period of low inflation, there has been a their families, principally periods of stagnation’, ‘the rise of the robots’ and ‘the noticeable divergence in the relative costs unemployment and ill health that impair new normal’.) However, it is not of necessities and luxuries. Education, one’s ability to work. immediately clear that such terms health, rental accommodation and In practice, there are areas where accurately reflect the experiences at the homeownership costs1 have all increased private insurance deals with risk relatively household level (Dynan, Elmendorf and at a faster rate than the Consumers Price effectively, particularly regarding the Sichel, 2012). ‘The great moderation’ was Index (CPI) overall since 2006 (Deloitte protection of assets (e.g. home, car and the term used to describe the reduced and Victoria University of Wellington, contents). On the other hand, dealing volatility of key macroeconomic 2017). At the same time, the CPI has been with risk associated with future earnings indicators in the United States in the held down by the falling price of imported seems to pose additional problems for decades prior to the global financial crisis manufactured goods, such as computing markets. These problems include moral (Stock and Watson, 2002); yet this was and cell phone technology. When incomes hazard, imperfect information and also a period of increasing volatility of are rising faster than prices, this may not adverse selection. The fact that a private household incomes, according a number be a great concern. However, it does affect consumer is in fact generally not well of US studies (Hacker and Jacobs, 2008; the balance of discretionary and non- informed about how long they will live, or Gottschalk and Moffitt, 2009). discretionary spending in the household their future chances of unemployment, Much has been written about why low budget, and therefore whether households means that risk in these areas has tended incomes are a problem for the families need to respond to a shortfall in income to be managed through direct state that must live off them. If volatility also involvement, or at the very least a

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 45 Income Volatility in New Zealand

Table 1: Absolute and proportional income impact of a two-income decile fall (2014 data) by people whose incomes fall due to Decile midpoint two-decile loss percentage loss retirement at age 65 or over, when they are 1 [under $6,940] entitled to New Zealand Superannuation. 2 $10,210 Individuals who were classified as ‘absent’ in the data set (i.e., not recording any 3 $14,805 income at all) for either the current year, 4 $18,245 $8,035 44% or previous year, for any of the year-to- 5 $23,395 $8,590 37% year movements are not counted in this 6 $31,655 $13,410 42% analysis. This therefore excludes people 7 $42,050 $18,655 44% who have died, or left the country between 8 $53,595 $21,940 41% one year and the next. 9 $70,160 $28,110 40% LEED includes data on New 10 [over $80,350] Zealanders’ income from wages, self- employment and most government considerable state role in regulating for those towards the lower end of the transfers, including income-tested private markets (Barr, 2001). income distribution such a loss would benefits, student allowances, paid parental Measuring income volatility can give have much greater repercussions in terms leave, New Zealand Superannuation and us a better idea of how the risk-sharing of meeting basic human needs. It should ACC. It does not include income from function of the welfare system is operating, also be emphasised that this level of investments, government transfers that in terms of providing assistance in times income loss is the minimum of what we are not taxable (such as childcare of need. But perhaps more importantly, it are measuring. Some individuals will have payments, the accommodation provides a measure of the extent to which fallen further than two deciles, and thus supplement and disability allowances) or economic loss is an inevitability for some experienced an even more severe loss of tax transfer payments by Inland Revenue part of the population at any given time. income. or Work and Income New Zealand. This Social assistance, on this account, should The reason for confining the analysis latter category includes the Working for be understood not so much as to drops of at least two deciles is to remove Families tax credits, which are payments redistribution between static groups in some of the ‘noise’ that we might see by structured around the entitlement of society, but rather as redistribution across including those who have dropped one families rather than individuals. As the people’s own lives, from good times to bad decile. This group might include people data is based on tax information, we (Hills, 2015, p.52). who were just above the bottom limit for cannot account for undeclared income. an income threshold in one year, and have The coverage of LEED beyond just Data and methodology fallen into a lower bracket in the following market income means we are capturing To get an idea of the degree of volatility of year due to a small drop in income, or by not only the initial hit to an individual’s New Zealanders’ incomes, we looked at the being overtaken by increases in the bottom income, but also the effectiveness of the share of the working-age population who limit for their previous bracket. This has social welfare system in offsetting that fell two or more income deciles from one the effect of dampening the degree of loss. It is, however, notable that as LEED year to the next between 2000 and 2014, measured volatility, but we would includes ACC payments, and the based on Statistics New Zealand’s linked otherwise have no easy way of determining percentages of income that equate to a employer–employee data (LEED). By way whether what we are measuring actually two-decile drop in Table 1 are all well over of example, this would be someone in the corresponds with a meaningful shock to 20%, ACC’s model of compensating for top 10% of income earners one year whose individual incomes. up to 80% of pre-injury income (up to a income fell to somewhere in the bottom This approach also means that we do capped sum) should exclude all but the 80% in the second year, or someone in not capture downward shifts in income top decile from the share of New the fifth decile (between the 40% and for people in deciles 1 and 2, who do not Zealanders suffering a two-decile fall. 50% points of the income distribution) have two deciles below them. For the sake This picture may change slightly if we who fell into the bottom three deciles of clarity, the exclusion of these deciles were able to account for all social transfers, (the lowest 30%). This analysis is at the relates to the first year of the income including those detailed above not individual, rather than household, level. transition measurement, meaning those included in LEED. However, it is Table 1 shows how big an income loss falling from a decile to a lower one. The important to understand that some a two-decile drop equates to. Taking the analysis does include individuals who fell transfers are not designed as automatic midpoints of each income decile,2 a two- into deciles 1 and 2 in the second year of stabilisers. Their effectiveness as a buffer decile drop represents a loss of around the measurement (assuming that fall was against income loss is therefore 40% of an individual’s income. Though of two deciles or more). questionable. For instance, a number of this proportion varies somewhat, such a Confining our analysis to New payments under the Working for Families loss represents a substantial hit to Zealanders between the ages of 20 and 64 tax credits have a paid work requirement household incomes. For obvious reasons, means that the picture is not complicated of 20 hours per week for a single parent,

Page 46 – Policy Quarterly – Volume 13, Issue 4 – November 2017 or a combined 30 hours per week for Figure 1: Proportion suffering a fall of two income deciles or more, individuals couples. This means that falling below the aged 20–64 years in deciles 3–10 required working hours would also effect 15% a household’s entitlement to this assistance. In some circumstances the loss 14% of entitlement to a work-related tax credit would be offset (partially or wholly) by an 13% increased entitlement to the means-tested family tax credit. However, there may be 12% situations in which a small decrease in the hours worked within a family leads to a 11% change in Working for Families entitlement that actually amplifies the loss 10% of household income, due to the steep Chance of two-decile or more drop 9% ‘cliff’ associated with the strict work requirements of the in-work tax credit 8% and minimum family tax credit. 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Deciles 3-10 Findings Figure 1 shows that close to one in Figure 2: Comparison of levels of income volatility – lower income vs higher income nine working-age New Zealanders will individuals, aged 20–64 years suffer a significant fall in income in any 17% given year. The volatility of incomes follows the business cycle, peaking in 2009 at approximately 12.5% and 15% declining steadily thereafter. Despite this broad replication of overall economic conditions, it is notable that volatility was 13% rising from 2005 onwards, prior to the onset of the global financial crisis. There has not been a trend of increased 11% volatility over the period measured. The

volatility of the most recent years Chance of two-decile or more drop measured is in fact lower than that of any 9% year during the decade beginning in 2001. In terms of total numbers of individuals, 7% the average number of New Zealanders 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 seeing this degree of income loss between Lower Income (Deciles 4-6) Higher Income (Deciles 7-10) 2001 and 2014 is over 214,000 each year. We are also able to examine where this in risk between the higher and lower and Sichel, 2012). The New Zealand aggregate level of volatility is concentrated income groups set out in Figure 2, and data, seen in Figure 4, shows the same in the income distribution. As Figure 2 displays a clear trend towards an increasing pattern. The volatility of female incomes shows, deciles 4–6, or those receiving gap between the two. The additional risk is typically around one percentage point between $15,720 and $35,550 in 2014, experienced by lower income groups has higher. Male incomes also seem to be have a notably higher level of volatility increased from somewhere below 5% more sensitive to the business cycle than than the higher income deciles.3 For this from 2003 to 2009, to above 7% in the female incomes. The volatility of male group, the year-on-year risk of a more recent years measured. This is due incomes rose more quickly with the onset substantial income drop is generally primarily to a falling level of risk among of the global financial crisis (and is higher between 15% and 17%, or somewhere higher income groups, rather than risk at than for female incomes in 2010), and between a one-in-six and one-in-seven the lower end of the income distribution also subsided more quickly as economic chance. increasing in absolute terms. conditions improved. The volatility of incomes in this lower This picture changes considerably income group has also been slower to Income volatility by gender and age when we look at the volatility of males and subside in the wake of the global financial The international literature tends to find females in different income groups (Figure crisis than that of the measured population that females have higher volatility of 5). Dividing our eight-decile population as a whole. Figure 3 shows the difference incomes than males (Dynan, Elmendorf into a lower income group (deciles 3–6)

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 47 Income Volatility in New Zealand

Figure 3: Difference in levels of volatility faced by higher and lower income groups on households, and those that likely in Figure 2 understate it. 8% Factors overstating volatility (or its impact) 7% Dealing first of all with the former, there are some key qualifiers to treating this 6% measure of income volatility as a direct proxy for economic risk or insecurity. 5% The most obvious is that we cannot

4% easily distinguish involuntary falls in income from voluntary ones. To give Percentage difference 3% one example, consider a highly paid management consultant who decides 2% to leave that position to instead teach a meditation class. This person would 1% receive a much lower income (though

0% would presumably be enriched in some 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 other, non-monetary ways). In this Lower income risk over higher income analysis, such a person would be counted Figure 4: Comparison of male and female volatility, deciles 3–10, aged 20–64 years as suffering an income loss in the same way as someone who is made redundant 14% and is unable to find work, yet naturally the concern we would show for the 13% former situation is nothing like that we would show for the latter. The spiritually- enriched former consultant does not lack 12% for substantial freedoms, following Sen’s capabilities approach, in the same way as 11% the out-of-work individual (Sen, 1999). The more common instances of 10% voluntary income loss are likely to be those relating to work/life balance and

Chance of two-decile or more drop changing careers. This includes those who 9% reduce their involvement in paid work to undertake study or training, or to care for 8% 2001 2003 2005 2007 2009 2011 2013 2015 children, elderly parents or relatives with Female Male disabilities. However, all such decisions are made in circumstances that are beyond and a higher income group (deciles 7–10), distribution shows much greater volatility our own control to some degree, and there we see that there is a significant divergence of income for younger age groups. The are clearly instances where such situations within the male population. While female rate does rise from age 50 onwards, as should not be understood as being fully income volatility is relatively similar would be expected due to people choosing voluntary. This includes someone between the higher and lower income to retire, though redundancies and health- choosing to retrain because they anticipate groups (and both are higher than for the related withdrawals from work can pose fewer job opportunities in their current general population in Figure 1), the lower difficult challenges for people at this stage industry, or those who have people close than average volatility of males overall of their lives also. The volatility of incomes to them in need of their ongoing obscures the fact that low-income males for those aged 65 and over is included assistance. have very high volatility of income here for comparison, and is considerably Nonetheless, it is important to note relatively speaking. It is the very stable lower than that for any other age group. that an income loss which is regarded as incomes of high-income males that drags voluntary might still be a matter of social the aggregate level below that of females. Limitations concern. Retraining and parental leave are Figure 6 shows the distribution of There are a number of limitations to inevitable occurrences in society, and it is volatility by age. Over the period between this approach, which can be roughly a matter of societal priorities as to whether 2001 and 2014 financial risk was grouped under two headings: those that more should be done to counterbalance experienced in an uneven U-shape likely overstate the degree of individual the income loss that typically accompanies between young and old. The overall income volatility or its actual impact these situations.

Page 48 – Policy Quarterly – Volume 13, Issue 4 – November 2017 There are further factors to take into Figure 5: Comparisons of male and female volaility by income groups, deciles 3–6 and 7–10, aged 20–64 years account that might overstate the impact of income volatility. As our data is 20.% individual, it does not take into account the picture at the household level, which is 18.% more important when it comes to the well-being of those who are part of a 16.% resource-sharing economic unit. Specifically, if a loss of income suffered by 14.% one individual leads to another income earner within the household increasing 12.% their work hours, then any impact on household consumption is likely to be 10.% Chance of two-decile or more drop offset to some degree. Unlike many international studies, 8.% this data is longitudinal only in a short- term sense. We are observing in essence a 6.% large number of two-year snapshots over 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 the 14 years of available data, and so Females 7-10 Males 7-10 Females 3-6 Males 3-6 cannot assess whether a fall in income Figure 6: Average income volatility by age group, 2001–2014 from year one to year two is followed by a bounce back of some sort in year three. 16% Furthermore, before making any judgements on whether an observed 14% income loss leads to a decrease in 12% household consumption, we ought to consider what options are available to 10% households to ‘smooth’ deviations in income. For instance, households could 8% (in some combination) resort to savings or draw upon other assets, borrow to 6% cover a shortfall, or they may have access

Chance of two-decile or more drop 4% to insurance. This is discussed in greater detail below. 2%

Factors understating volatility 0% 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+ Although we do not see a clear secular trend in income volatility over the period of 2001 to 2014, both the level of volatility Zealand is also understated by not projects or relationships with clients, and and its concentration among income including absolute decreases in income often utilises online platforms (Bughin groups and between genders presents an that do not result in someone changing and Mischke, 2016). interesting picture. There are a number of income brackets. For instance, someone at Month-to-month volatility poses a limitations to this analysis which might the top end of decile 5 could suffer a fall in different set of challenges to a substantial understate the level of income volatility income of $5,000, but still remain in the shock to income from one year to the among New Zealanders. same decile bracket. next, which might be associated with a Focusing on downward movement It should also be noted that this period of unemployment or serious between income deciles, and excluding analysis, along with most international illness. A single major shock might be shifts of only one decile, means that this studies, is focused on inter-year volatility. smoothed out through drawing upon analysis excludes deciles 1 and 2. This A growing international literature is savings for those fortunate enough to approach means that we do not capture focused on intra-year volatility, most have them (for a transitory shock), or falls in income experienced by the bottom typically through measuring variances in require a challenging but manageable 20% of income earners, who do not have month-to-month incomes (Bania and adjustment in lifestyle (for a permanent two income deciles below them, but for Leete, 2009; Hills, 2015). This is often shock). However, month-to-month whom any significant decrease in income closely associated with the ‘gig economy’, variations in income require a constant would be a considerable shock. The which involves non-traditional process of budgeting, exhausting and overall volatility of incomes in New employment around short-term tasks, restoring financial buffers and the greater

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 49 Income Volatility in New Zealand likelihood of an ongoing reliance on debt, connected to it in a number of ways. The misfortune. Yet it is important to bear in possibly at high interest rates. Informal rising or increased levels of income mind that upward relative movements assistance from friends and family inequality in many advanced economies between income deciles is the necessary members is another source of buffers for over recent decades has led to concerns corollary of downward shifts. In a similar unstable short-term incomes, though not about intergenerational mobility: that is, sense, the work presented here on income everyone has this option available. whether the concentration of rewards at volatility demonstrates that what is often Furthermore, an individual living off the top of the income distribution is presented as a positive story of income an income that is both low and variable is leading to the development of more rigid mobility also includes a significant level of likely to face an ever-shifting entitlement social classes, where one’s chances in life downward shifts in income in absolute to social assistance. This is important, first are more closely tied to the socio- terms, not simply relative shifts. of all, because some entitlements require economic position of one’s parents. In This underscores an important point: that workers are in a job for a certain this context, income mobility is often seen a society marked by highly unstable period of time before they are available as an indicator of equality of opportunity, incomes is likely to appear as a reasonably (i.e. sick leave, parental leave). Having an though this is less true of (short-term) mobile society over the shorter run; adequate knowledge of one’s entitlement measures of intra-generational mobility however, over the longer run it is plausible to assistance, and the time and complexity than longer-term intergenerational that a society that does not deal with associated with claiming it, can also pose studies (Corak, 2013). individual risk in an effective way will see initial disparities in life chances reinforced. As John Hills has detailed, the advantages ... income mobility is seen as offering of being born to affluent parents are not limited to the early years of life, but are a more nuanced understanding of the reinforced throughout the life cycle, in the form of assistance with higher education, distribution of income in society than a help with housing costs or the deposit for static picture of the shares that different a house purchase, inheritance and other intra-family transfers of wealth (Hills, groups receive at a certain point in time. 2015). All things being equal, these advantages are likely to mean that their beneficiaries will deal with economic risk much better than those without such considerable barriers for some people. In At the same time, income mobility is advantages – a proposition which is September 2017, TV3’s The Nation seen as offering a more nuanced consistent with the higher concentration reported that, according an Official understanding of the distribution of of income volatility among the lower- Information Act request, difficulties income in society than a static picture of middle class. around claiming benefits meant that the shares that different groups receive at As noted, downward shifts between approximately $200 million each year was a certain point in time. To put it another income deciles must by necessity coincide not received by families who were entitled way, the level of income inequality over with upward shifts by others in society. to receive it.4 one’s lifetime is likely to be less than at a This raises the issue of the extent to which single point in time. The argument here is we can weigh these prospects against one Discussion that the implications of a less equal another. It could be argued that individuals There are two main areas arising out of distribution of income might be different are to some degree compensated for the this analysis that this article shall address. depending on the degree to which there is greater risk of downward shifts by the The first is to explore the position of movement between income groups over corresponding chance of upward income volatility in relation to the more time (Barker, 1996). For instance, a New movement.5 For instance, Carter, Mok well-established work on income mobility. Zealand Treasury paper notes that ‘There and Le (2014) found that in 2002 New The second is to put these findings in a is change in incomes between one year Zealanders had an equal chance (6%) of broader context of economic and social and the next, with over 60 percent of the their income increasing or decreasing by trends, and offer some tentative answers as population changing income decile group $20,000 or more in the following year. to why income volatility in New Zealand ... Only 22 percent stay in the same income It makes sense to take this offsetting looks the way it does. decile group eight years later’ (Carter, upward mobility into account; however, Income mobility – the relative or Mok and Le, 2014). there is good reason to view the risk of absolute movements in an individual’s In some sense, income volatility downward movements as a serious income over time – has become a topic of presents the flip side of this story. At the concern, notwithstanding any considerable academic interest in recent superficial level of mental images, the corresponding chance of upward years. This has been parallel to work on level of income volatility discussed here movements in the same year. The first is income and wealth inequality, and may naturally evoke a picture of loss and people’s tendency to experience loss

Page 50 – Policy Quarterly – Volume 13, Issue 4 – November 2017 aversion, meaning that people tend to Table 2: Comparison of loss associated with a two-decile drop vs wealth position by decile value avoiding a fall in their position more Decile Two-decile loss Mean net worth ($) Mean financial Mean cash in than they value an equivalent increase assets ($) bank ($) (Kahneman and Tversky, 1979). If we 1 -$23,000 $400 $1,200 consider subjective well-being to be an 2 $3,200 $100 $700 objective of public policy, then this would 3 $12,300 $200 $2,000 tend to favour a lower volatility system. 4 $8,035 $32,000 $600 $3,100 The second reason is that it is 5 $8,590 $68,700 $1,600 $6,100 unquestionably a lot easier to adjust household finances to an increase in 6 $13,410 $124,800 $2,700 $7,900 income than a decrease. Many of our 7 $18,655 $193,600 $3,600 $10,700 outgoings represent long-term, fixed 8 $21,940 $280,600 $6,400 $15,800 commitments (for instance, mortgages 9 $28,110 $428,300 $14,400 $25,200 and rental accommodation, education- 10 $1,289,700 $63,500 $69,100 related expenses, childcare), and having to Wealth data for 2010 from Rashbrooke, Rashbrooke and Molano, 2017 (data tables provided by authors). unravel these obligations can be difficult, Note: calculation of decile midpoints is based on 2014 data. costly or disruptive to social networks and established relationships. We should also underlying trend. Further work on to the financial buffers that New Zealand bear in mind the social and psychological disaggregating the components of income households have available to them. The costs that might be suffered in the course would provide greater insight on this figures for wealth are at a household level, of this transition, if the end result is that a point. However, it should be noted that whereas the income volatility data is for household experiences a greater degree of the major welfare change in this era, the individuals. material deprivation. Working for Families tax package, is not Partnering and forming a household captured by the data used here.6 provides an additional avenue of risk Social and economic context This analysis provides an ex post view sharing (though obviously that is far from It may be surprising to some that New of disruptions to individual incomes. the only reason for doing so). The share of Zealand does not exhibit the same While we are not in a position to directly couples with dependent children who are observable increase in income volatility assess how effectively households can both employed (as opposed to one or that has generally been found in other smooth volatility in individual incomes, neither being employed) has actually risen advanced countries. However, a number some tentative comments can be made steadily in recent decades. This rate was of points should be borne in mind. First, regarding what is known about options below 60% in 2000, and had reached the many of these (particularly US) studies for offsetting income loss. mid-60s by 2014 (though there has been find an increase in volatility that predates Rashbrooke, Rashbrooke and Molano an even larger increase in the years since the scope of this analysis. A number of (2017) provide evidence from as recently 2014, to close to 70%; Statistics New more recent studies have found a relatively as 2010 on the wealth held by New Zealand Zealand, n.d.). This means that more stable trend over the period since 2000 households. Table 2 compares these households have more than one source of (Hardy and Ziliak, 2014). findings with the midpoint income losses labour market income; however, it may Second, labour market changes are represented by a two-income decile fall also represent an increasing share of often highlighted as a causal factor in (as set out in Table 1). Mean net worth is households that need two incomes to get trends in income volatility, and though greater than the (minimum) income loss by. New Zealand has undergone some measured in this analysis for each decile, significant transformations on this front, though this wealth is typically Conclusions and further research these also predate the years covered by concentrated in housing equity. The two This analysis has provided an indirect and LEED. One third of New Zealanders are in right-hand columns show the mean worth tentative look at the incidence of income- non-standard employment. However, the of more liquid assets: cash in the bank and related shocks, and the effectiveness of growth in part-time work largely took financial assets. For deciles 4–7, the the institutional system in effectively place in the late 1980s and 1990s, and the income loss is greater than the mean total dealing with this risk. The level, trend share of self-employed people has been of liquid assets. For decile 8, it is broadly and concentration of income loss present relatively stable (Statistics New Zealand, equal. an uneven picture. It does not appear 2014). Households are, of course, likely to that the aggregate level of volatility has Third, it is possible that there are smooth income losses through some increased since 2001. However, lower different trends for various components combination of budgeting, drawing upon income groups face much greater chances of incomes, as found by Jenkins (2011) assets and borrowing. However, this of a substantial fall in income than higher with regard to data for the United comparison provides some context for the income groups, but are likely to have fewer Kingdom. Such divergences may act to extent of the income loss that is measured financial buffers and less resilience against offset one another, and disguise any in this study of income volatility, relative economic loss. We also see very different

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 51 Income Volatility in New Zealand

levels of risk between age groups and work patterns and automation on the also be much greater than the equivalent fall for higher deciles. between genders, as well as within genders stability of household incomes is likely to 4 http://www.newshub.co.nz/home/election/2017/09/winz- creating-two-classes-of-kiwis-labour.html; http://www. once income is taken into account. be a major concern for the future. radionz.co.nz/news/election-2017/339064/winz-staff- accused-of-withholding-entitlements. There are a good number of areas Finally, although it is difficult to say 5 Thanks to Norman Gemmell for this point. where further research might add to our what level of volatility ought to be 6 Such changes may have had a dynamic effect on other components of individual income, but the evidence on understanding of income volatility in New considered acceptable within society, this suggests that this package had a positive impact on employment (Dalgety et al., 2010). Such changes would Zealand. Assessing the incidence of there are certainly patterns in the most likely be related to labour market income, and therefore economic loss over a longer time period distribution of financial risk that should visible in the data. would give a better idea of how policy be cause for concern. Further work on changes have affected the level of financial how policy changes can provide greater Acknowledgments risk faced by New Zealanders. Matching up economic security to households is My thanks to Cassie Favager, Jeff Brandt, this work with a better understanding of therefore important. Individuals and Jason Tabarias, Michael Macaulay and how and when households can smooth households have diverse aspirations, and Ben Ogilvie for collaboration on the their consumption in the face of income different things about their lives that they 2017 Deloitte/Victoria University ‘state loss (and how this varies between different each hold dear. However, we all have a of the state’ report, which was the basis groups) would provide a much better common interest in making sure that the for this work. Sarah Burr of Deloitte indicator of how this aggregate level of inevitable incidence of misfortune within contributed the data analysis for this volatility translates to outcomes for society does not unduly determine the work. For comments, advice and helpful individuals and families. This includes course of our lives. discussions, thanks also go to: Jonathan intra-family transfers, and the implications Boston, Simon Chapple, Arthur Grimes, 1 Homeownership in the CPI does not include the price of of this for intergenerational mobility. And, land. It therefore understates the price of homeownership. Norman Gemmell, Girol Karacaoglu, as noted above, the volatility of inter-year 2 All figures referring to decile bands are before tax, and refer Max Rashbrooke, Patrick Nolan, Jess to 2014 figures. incomes does not give us a good idea of 3 Decile 3, the lowest decile measured through this approach, Berentson-Shaw, Max Harris, Courtney has conversely very low volatility relative to other lower how incomes are fluctuating over a shorter deciles. It is unclear how much of this is a limitation of the Kemp, Andrea Fromm, Stephen Oakley methodology at the bottom end of the income spectrum. time frame, such as the measures of For instance, an individual in decile 3 faces the possibility and Liya Liyanage, as well as attendees at a monthly volatility that are emerging of falling up to two deciles at most, compared to someone Victoria University School of Government in the fifth decile who can fall up to four deciles, and so on. internationally. The impact of changing The proportional drop for decile 3 falling to decile 1 would ‘Brown Bag’ seminar in September 2017.

References Bania, N. and L. Leete (2009) ‘Monthly household income volatility in the Hacker, J. (2007) ‘The new economic insecurity – and what can be done U.S., 1991/92 vs. 2002/03’, Economics Bulletin, 29 (3), pp.2100-12 about it’, Harvard Law and Policy Review, 1 Barker, G. (1996) Income Distribution in New Zealand, Wellington: Institute Hacker, J. (2008) The Great Risk Shift: the new economic insecurity and of Policy Studies the decline of the American dream, Oxford: Oxford University Press Barr, N. (2001) The Welfare State as Piggy Bank: information risk, Hacker, J. and E. Jacobs (2008) The Rising Instability of American Family uncertainty, and the role of the state, Oxford: Oxford University Press Incomes, 1969–2004: evidence from the panel study of income Boston, J. and S. Chapple (2014) Child Poverty in New Zealand, dynamics, briefing paper 213, Washington: Economic Policy Institute Wellington: Bridget Williams Books Limited Hardy, B. and J. Ziliak (2014) ‘Decomposing trends in income volatility: Bughin, J. and J. Mischke (2006) ‘Exploding myths about the gig the “wild ride” at the top and bottom’, Economic Inquiry, 52 (1), economy’, November, available at http://voxeu.org/article/exploding- pp.459-76 myths-about-gig-economy, accessed 7 October 2017 Hills, J. (2015) Good Times, Bad Times: the welfare myth of us and them, Carter, K., P. Mok and T. Le (2014) Income Mobility in New Zealand: a Bristol: Policy Press descriptive analysis, working paper 14/15, Wellington: New Zealand Jenkins, S. (2011) ‘Has the instability of personal incomes been Treasury increasing?’, National Institute Economic Review, 218 (190), pp. Corak, M. (2013) ‘Income inequality, equality of opportunity, and R33-43 intergenerational mobility’, Journal of Economic Perspectives, 27 (3), Kahneman, D. and A. Tversky (1979) ‘Prospect theory: an analysis of pp.79-102 decision under risk’, Econometrica, 47 (2), pp.263-91 Dalgety, J., R. Dorsett, S. Johnston and P. Spier (2010) Employment Rashbrooke, G., M. Rashbrooke and W. Molano (2017) Wealth Disparities Incentives for Sole Parents: labour market effects of changes to in New Zealand: final report, IGPS working paper 17/02, Wellington: financial incentives and support: technical report, Wellington: Centre Institute for Governance and Policy Studies for Social Research and Evaluation, Inland Revenue and Ministry of Sen, A. (1999) Development as Freedom, Oxford: Oxford University Press Social Development Statistics New Zealand (2014) Flexibility and Security in Employment: Deloitte and Victoria University of Wellington (2017) Fit for the Future: findings from the 2012 Survey of Working Life, Wellington: Statistics boosting resilience in the face of uncertainty: state of the state New New Zealand Zealand 2017, Wellington: Deloitte and Victoria University of Statistics New Zealand (n.d.) Household Labour Force Survey, Infoshare Wellington HLF185AA Dynan, K., D. Elmendorf and D. Sichel (2012) ‘The evolution of household Stock, J. and M. Watson (2002) ‘Has the business cycle changed and income volatility’, B.E. Journal of Economic Analysis and Policy, 12 (2) why?’, NBER Macroeconomics Annual, 17 Gottschalk, P. and R. Moffitt (2009) ‘The rising instability of U.S. earnings’, Journal of Economic Perspectives, 23 (4), pp.3-24

Page 52 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Tim Hazledine, Michael Wang and Kar Yew Lee

Trends and Determinants of Top Pay in the New Zealand this article is available as Hazledine et al. (2017), and includes an analysis of the apparent broad decline in managerial Public and Private productivity in New Zealand, with a case study of the University of Auckland. The first section gives some Sectors, 1995–2014 institutional background to the changing role and remuneration of public sector Introduction CEOs in New Zealand over the past three decades. Section two introduces the This article analyses recent trends and determinants of chief variables used in the study, and section executive (CEO) pay in the New Zealand public sector, three analyses the data. Section four concludes by drawing out implications for and of numbers and pay of senior managers in the sector. policy and future research. An Appendix documents the regression models used to Comparisons are made with the listed company private explain differences and changes in CEO sector. It turns out that both CEO pay growth and numbers pay in the New Zealand public and private sectors. of senior managers in the public sector have lagged behind those in the private sector, while senior manager pay has The recent history of public sector chief executive pay-setting in New Zealand moved ahead. Before the 1984–90 ‘Rogernomics’1 neo- liberal policy revolution in New Zealand The article complements Hazledine 2014 period, and to which the reader is (Hazledine, 2014), most public sector (2016), which was concerned solely with referred for detail on the listed company activities in New Zealand – including the listed private companies over the 1995– sector and its data. A longer version of activities of ‘trading’ entities, such as the Post Office and the Electricity Department Tim Hazledine is a Professor in the Economics Department of the University of Auckland Business – were organised on departmental lines, School. Michael Wang was a Bachelor of Commerce (Hons) student. Kar Yew Lee is currently in the master’s programme in the Economics Department. with each department run by a permanent head or ‘secretary’ reporting to his (rarely

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 53 Trends and Determinants of Top Pay in the New Zealand Public and Private Sectors, 1995–2014

– never? – her)2 central government foundations of Rogernomics. The basic core central government departments, minister. In the late 1980s this situation assumption here was that ‘management’ is which operate in the ‘public service’.4 For was upturned, in particular following the a generic skill, applicable in any 103 other ‘Crown entities’, their boards passing of the State-owned Enterprises organisation or workplace by any smart come up each year with a figure for their Act 1987, the State Sector Act 1988 and person with a good knowledge of CEOs’ pay, but must then ‘consult’ with the Public Finance Act 1989. Bollard principal–agent (‘agency’) theory. For the SSC before implementing these (State et al. (1996) report that, just during example, generic managers with zero Services Commission, 2016). Consultation 1987–88, 24 government departments medical expertise could replace senior has teeth in the case of the tertiary were ‘corporatised’ into state-owned doctors and nurses in the running of education institutes and the district health enterprises (SOEs; see Duncan, 1996), and public hospitals, and should replace them, boards among the Crown entities, whose a similar number fully or partly privatised, because the medical professionals could proposed CEO pay numbers must be these all previously having been trading not be trusted to not exploit their special ‘approved’ by the SSC. Other public sector departments, selling products – airport position (asymmetric knowledge organisations, mainly local authorities services, electricity, telecommunications, advantages) to further their own private and SOEs, do not come under the banking, forestry and so on – on the goals. jurisdiction of the State Services market. A corollary of managerialism was this: Commission for CEO pay setting or other The remaining public sector if public and private sector managers were matters. organisations – which are our primary basically interchangeable, then their In 1997 the government formally abandoned the idea of a strict private/ public CEO pay link, explicitly recognising The commissioner tactfully suggests that skills required and responsibilities shouldered by public sector managers the public sector’s previous participation which are not generic, and which in ... justified the ... similarity between particular are not part of most private sector CEO job descriptions.5 The state private and public sector CEO job skills, services commissioner put it in these but this doesn’t really make sense. terms: Chief executives who lead departments with policy advice roles6 must understand the capacity of the State to effect change in social and interest in this article – had imposed on remuneration should be closely linked, economic conditions. This usually them various quasi-market procedures and driven by market-determined private demands an understanding of social and constraints: limited-term sector managerial pay rates, as apparently policy or economics disciplines. In the appointments for CEOs, and annual it was – or was supposed to be – in the case of some roles, this understanding performance evaluations, with bonuses wake of the 1988 State Sector Act. But if must be evident in a high level of based on success or not at achieving ‘KPIs’ the linking doctrine was applied it must academic qualification. Several chief – key performance indicators – which have been at a sizeable discount. The state executive roles demand substantial intended to match the disciplines services commissioner, in his 1998 annual professional expertise in, for example, supposedly enforced on private sector report, reported that base salaries for the law, science and technology, or boards and their CEOs by the market. CEOs in ‘broadly comparable positions’ accounting … All chief executive roles Many issues are raised by attempts to rose by as much as 130% in the nine years demand an understanding of senior measure success in providing government to 1997 in the private sector, but by just management in a public environment, services in relatively simple formulaic 42% in the public service (State Services and the responsibilities and terms. These are not our concern here, but Commission, 1998, p.16). In real terms, accountabilities of that environment. we note that accountability by means of adjusting for inflation, this means that (State Services Commission, 1998, KPI performance assessment remains private sector CEO pay increased by 81% p.15) central to the monitoring of public sector and public CEO pay by just under 10%.3 CEOs in New Zealand. We need to note here the difference The commissioner tactfully suggests What has changed somewhat, and is between public service and public sector, that the public sector’s previous relevant to our prime interest in CEO pay, and the role of the State Services participation in market-oriented ‘trading’ can be seen as a partial retreat from the Commission (SSC), and its own CEO, the activities might, in 1988, have justified the full-blown managerialist theory that was commissioner. Currently, as the SSC’s assumption of some strong generic one of the most theoretically interesting 2016 Senior Pay Report explains, the similarity between private and public (even, internationally, unique) commission sets CEO pay for 26 of the 29 sector CEO job skills, but this doesn’t

Page 54 – Policy Quarterly – Volume 13, Issue 4 – November 2017 really make sense. The retreat of the state local government 31,400. Switching job titles of these employees were not from commercial activities via employees in SOEs from the central to the required to be disclosed. The new data corporatisation and privatisation was well local government tally (such then being in began to turn up in annual reports from underway by 1988, and well understood essence the wider public sector) would 1995 onwards. to be sweeping in its soon-to-be- not make a big difference to the size The situation for public sector completed extent. discrepancy.7 There must indeed be organisations in New Zealand is somewhat The government in 1997 also considerable ‘circularity’ or different, with pay disclosure apparently a instructed the SSC to strengthen internal interdependence in the setting of the policy matter, though influenced by the assessment of public service CEOs’ terms and conditions of employment of Companies Act 1993. Some public sector performance and the linkage of this to public and near-public sector chief entities do choose to report CEO and pay their remuneration. What the government executives. band data in their annual reports. Others did not tell the SSC how to do was how to So, can we cut through the circularity make this information available to the set benchmarks for the public service and identify exogenous factors State Services Commission, which in turn CEO salaries as a whole, given that the determining, or at least significantly does publish all the CEOs’ pay in what is administratively useful link – albeit at an influencing, public sector CEO pay? Such now called the Senior Pay Report, covering increasing discount – with private sector is our purpose in this article. As well, we public services departments, district salaries for ‘comparable’ jobs had been will examine the situation of the highest health boards, tertiary education abolished. The commission itself apparently came up with the proposal that was adopted: public service and state As for top pay numbers, these are entity CEO pay would be ‘set against remuneration for chief executives working submitted to the SSC in a form in the wider public sector’, this being mainly the local authorities and state- ‘consistent with the Companies Act owned enterprises (State Services 1993’, but only reported publicly in two Commission, 1998, p.18). Such remains the situation to date – aggregations: all public service entities, annual SSC reports right through to 2016 make fairly familiar reading. There is an and all tertiary institutions. almost amusing passage in the 2003 annual report. The (same as quoted above) commissioner writes: paid managers reporting to CEOs. Their institutions, and other statutory Crown In last year’s annual report I drew pay is not set by the State Services entities. attention to my increasing influence Commission, though we could expect As for top pay numbers, these are on State sector remuneration as a some linkage with what ‘the boss’ gets. submitted to the SSC in a form ‘consistent result of the Commissioner’s with the Companies Act 1993’, but only expanding role in concurring with, or Variables and data reported publicly in two aggregations: all advising Boards of Crown entities on, Data on CEOs and numbers of other public service entities, and all tertiary the terms and conditions of employees on high salaries institutions. We have had to write employment of their chief executives We have in New Zealand what may be hundreds of Official Information Act … The process is becoming ‘circular’. uniquely detailed information on top pay. requests to the individual entities and The more I influence the market, the The New Zealand Companies Act 1993 institutions, requesting the pay band more it loses its value as a benchmark. required companies primarily listed on numbers first for 2014, and then (this (State Services Commission, 2003, the New Zealand Stock Exchange (now done separately) the earliest year available. pp.12-13) NZX), and required thereby to submit Responses to our requests have been audited annual financial reports in New generally good, though not universal, with Indeed. The commissioner was Zealand, to in future provide information the earliest such data being for 1995, but probably right to be worried here, even if in those reports not just on the total most for various later years. he seemed unafraid to run the risk of remuneration of their chief executive The $100,000 salary minimum exaggerating his personal importance in (which had not been required before, but remains in place. However, $100,000 in these matters. The ‘wider public sector’ is, was often made available), but also on 1995 is worth almost exactly $150,000 in relatively, not very wide. The 2001 New the numbers of employees earning more 2014, after inflating by the Consumers Zealand census reported that total than $100,000 a year, with these numbers Price Index. Accordingly, to maintain employment in enterprises owned by the disaggregated into bands of $10,000 – comparability with early years, the top pay central government was 224,000, and in i.e. $100–110,000, and so on. Names and cohort analysed is set to begin at a pay of

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 55 Trends and Determinants of Top Pay in the New Zealand Public and Private Sectors, 1995–2014

Table 1: Descriptive Statistics, Public Sector Organisations We have no data on what is included in (All monetary variables observed before Maximum Average Minimum public CEOs’ key performance indicators. 2014 are inflated to 2014 prices with We ended up with a database on 123 the consumer price index) public sector organisations, each observed Total annual budget, $millions 4731.4 362.7 1.2 in 2014 and in one earlier year. The only Remuneration of Chief Executive, $ 840,000 378,492 100,000 variable for which we have complete data Total remuneration of all employees 195.7 13.1 0 is CEO pay, and the smallest sample is for earning more than $150,000, excluding numbers of non-managerial employees. CEO, $millions Total number of employees earning more 765 61 0 Quantitative analysis than $150,000, excluding CEO Table 1 gives maximum, average and Average annual wage of all other 151,500 81,970 55,213 minimum values for six variables reported employees, $ by our public sector organisations. All Number of all other employees 8908 1343 4 monetary values are converted by the Consumers Price Index into 2014 dollars. The numbers in Table 1 will be compared $150,000 in 2014 prices. That is, we do not · the number of senior managers; with the matching private sector figures, use all the information on numbers of · the average pay of senior managers; from Hazledine (2016, Tables 1, 2). employees earning over $100,000, except · the number of other employees. The average annual public sector for 1995. Data for intermediate years are organisation budget was $362.7 million, pro-rated into higher pay bands according Market factors somewhat smaller than the size of the to each year’s CPI, as explained in To the extent that there is a general market average listed company in 2014, which Hazledine (2016). for CEO services, then prices paid for such was $510 million. The biggest-spending We will refer to employees earning services in this market can be naturally public sector organisation was ACC – the more than the equivalent of $150,000 in expected to affect salaries offered to and Accident Compensation Commission – 2014 prices as ‘managers’, which most of accepted by our New Zealand public which worked its way through $4.7 billion them must be, though outside the core sector CEOs. We have noted above the in that year. public service sector lie the district health difficulties in finding plausibly exogenous The highest-paid civil servant CEO boards and the universities, in which local benchmarks for public sector top was – surprisingly – a military man: the organisations doctors and professors will salaries. We note now our belief that, at chief of the defence forces, who is paid often or mostly earn more than $150,000. least for English-speaking executives, the much more than his equivalent in the market for such talent, in both private United States, who in turn receives much Factors determining CEO pay and public sectors, has indeed become less than the three head coaches of the There are a number of factors which ‘thicker’ because more globalised in army, navy and air force football teams.8 might plausibly be proposed as likely to recent decades. It would be interesting Actually, CEO pay in the New Zealand affect the level of remuneration received to have some historical research on the public sector is higher on average than in by the men and women who head our backgrounds of private and public sector the US, and indeed in every OECD listed companies and public sector CEOs in New Zealand. Our expectation country apart from Italy, a fact that we did organisations: (1) internal/organisational is that, 40 or so years ago, most of these not find noted in any of the state service factors; (2) market factors; and (3) would be both local citizens and internal commissioners’ reports we read.9 performance factors. appointments – career or long-service Nevertheless, the highest-paid New employees of the firm or department they Zealand CEO is not at the top of the Internal/organisational factors eventually worked their way to the top of. overall public sector remuneration list: It is reasonable to expect that CEOs with Now, external candidates sourced from three doctors in 2014 each billed the more ‘difficult’ jobs would get paid more. all over the English-speaking world can Auckland District Health Board more But how is difficulty to be measured? The be found leading our large organisations. than $1 million, presumably for contract reports of the State Services Commission We use an annual time trend to at least specialist services. Average public sector often refer to a concept known as ‘job size’, measure, if not explain, long-term shifts CEO pay in 2014 was $378,000, which was of which in 2000 there were five ‘bands’, in the CEO market. just 56% of the listed company CEO measured in something called ‘Hay points’. average remuneration. We assembled data for four variables that Performance factors Total numbers of what we call ‘senior we thought could be plausibly proposed In the private sector there are financial managers’ – people earning more than as components of public sector CEO job metrics for CEO performance, including $150,000 – are actually highest for the size: profitability and, perhaps, total revenues. Auckland District Health Board, which in · the annual budget of the department Public sector entities are not-for-profit and this case is surely largely due to high pay for or other entity; their budgets are politically determined. medical professionals. Average pay for

Page 56 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Table 2: Comparisons of organisational structure: public and private sectors, 2014 Ratio CEO pay to budget or sales Ratio total managers’ pay to Ratio budget or sales to number budget or sales of managers

public sector private sector public sector private sector public sector private sector

AVERAGE: ALL 0.00827 0.00952 0.06604 0.03909 10,530,066 12,414,646

AVERAGE: SIZE>$10,000,000 0.00542 0.00672 0.06131 0.03486 11,291,797 12,798,516

AVERAGE: SIZE>$20,000,000 0.00357 0.00492 0.05683 0.03290 12,111,225 13,337,886

Table 3: Elasticities of employee numbers and top pay: public and private sectors Number of managers Average manager salary Number of other employees CEO pay

With respect to: public sector private sector public sector private sector public sector private sector public sector private sector

Size of organisation or firm 0.70 0.71 0.02 0.00 0.85 0.88 0.16 0.30

Number of managers 0.14 0.26

Average manager salary -0.05 0.60 other workers in the public sector was Then we calculate, average and public sector elasticity is just 16%. Given about $82,000, which is much higher than compare the ratio of the total payroll for that size in the private sector is measured the private sector listed company equivalent senior managers to organisation size. We by market revenues, and thus plausibly a figure of around $65,000. The lowest-paid see that the private sector firms operate key performance indicator for CEOs, the non-managerial employees work for the with substantially leaner managerial difference is perhaps not surprising. Ministry of Social Development; the structures, on average. That is, while the The number of senior managers highest for the Civil Aviation Authority, peak of the bureaucratic pyramid is lower reporting to the CEO is linked to the though this number, being larger than in the public sector (lower CEO pay), the latter’s pay, with an elasticity nearly twice $150,000, must be wrong. base tends to be much broader. as large in the private sector. There is the Table 2 reports averaging of three The third column, however, dividing suggestion that private CEO pay is marked ratios which give further insights into the size by the number of managers – which up on the average pay of senior managers, employment structures of public and might be interpretable as managerial an effect not observed in the public sector. private sector organisations. Averages are productivity – shows a much smaller The elasticity of the number of senior calculated first for the full sample, then public/private sector differential, implying managers with respect to size is very successively excluding organisations with that the larger total payroll ratio in the similar across public and private sectors, annual revenue or budgets less than $10 public sector is mostly due not to more perhaps reflecting some bureaucratic million, and less than $20 million, in case managers, but to these being more highly imperative of modern organisations: the overall picture is distorted by the paid, further away from the lower cut-off double the size/hire, 70% more top behaviour of relatively small units. of $150,000 a year. managers. However, independently of the The first column compares ratios of Table 3 reports estimates of ‘elasticities’, size effect there is a very strong secular CEO pay to total organisation size: budget measuring the quantitative impact of trend in senior manager numbers in the for public sector; total sales revenues for differences in a causal variable, such as private sector, which is about two and a private. In organisations of a given size, organisation size, on differences in the half times larger than the trend in the private sector CEOs tend to get paid variable(s) of interest. These estimates public sector. significantly more than the public sector come from successful econometric As for the average senior managerial counterparts – 38% more in the sub- modelling of the data, as reported in the salary: perhaps surprisingly, this shows no sample of larger organisations. The Appendix, Table 5. significant link anywhere with size, but private sector premium could plausibly be The most powerful association is has been trending upwards in the public linked to the fact that private sector firms between organisation size and CEO pay: sector, at more than twice the rate of have to make profits in competition with in the private sector, a doubling of size public sector CEO pay. Overall, we could other firms, and face the risks of goes with a 30% increase in the CEO’s say that the base of the bureaucratic bankruptcy, liquidation and takeover. remuneration, whereas the corresponding pyramid of senior managers has tended to

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 57 Trends and Determinants of Top Pay in the New Zealand Public and Private Sectors, 1995–2014

Table 4: Average percentage size-weighted annual rates of growth CEOs, we may have to look for political CEO pay Size Size-adjusted CEO pay explanations, with these dating from Spending department 1.04 2.59 0.62 before 1995. Policy department 3.84 1.56 3.59 A final note. Politicians are not included in our database, but it is of District Health Board 1.83 4.09 1.18 interest that the pay (excluding University 3.25 3.87 2.63 allowances) of the most senior public Listed companies 5.38 6.60 3.40 sector CEO of all, the prime minister, has increased annually at the fairly modest widen in listed companies, whereas its and the lowest overall size (budget) rate of 2.3% since 1995, to reach $429,000 peak has risen in the public sector. growth, such that size-adjusted CEO pay in 2014.12 The elasticity of the number of non- growth is actually a little higher for the – managerial employees with respect to generally smaller – policy/regulatory Implications and conclusions organisation size is also quite similar departments. Thirty years on from the Rogernomics across sectors, and seems to imply mildly The spending departments have been upheavals of 1986–88, and 20 years from increasing returns to scale. Both sectors – particularly circumspect in rewarding the adjustments made in 1997, is it now but especially public sector organisations their CEOs with pay increases, at just over time for a fresh assessment of top pay in – have, however, been shedding non- 1% per year – similar to real pay growth New Zealand’s public sector, and of the managerial labour at quite substantial for the New Zealand workforce as a organisational structures that go with it? rates, with negative annual trends of whole11 – and less than this after allowing We think it is. However – and perhaps around -6% and -2%. for increases in the size of these surprisingly – it is the private sector which Next, in Table 4, we partition the departments. District health board CEOs has seen a doubling of CEO pay over the public sector into four groups: 29 mainly have also had rather modest pay increases, past two decades, and which has generated ‘spending departments’ (e.g. the in a faster-growing sector than average. the largest growth in numbers of high- Broadcasting Commission, the New University vice chancellors have done paid employees. Nevertheless, even though Zealand Transport Agency); 56 mainly quite well. public sector CEO pay has barely grown, it ‘policy/regulatory departments’ (e.g. the Differing CEO pay paths within could still be too high: does it really make Broadcasting Standards Authority, the different units of the public sector, and sense that the professional chief of our Ministry of Transport); 18 district health between public and private sectors, seem tiny defence forces is paid more than his boards; and the eight universities. As well, at least consistent with the State Services opposite number in Washington, as also we have calculated rates of growth for 107 Commission’s recognition, from 1997, of are the heads of the major government listed companies in the private sector, inherent differences in the skills and departments? And are there signs of from the database used in Hazledine experience needed to effectively run bureaucratic bloat in the relatively high (2016). The distinction between spending- private and public sector organisations; levels of senior employees in the public oriented and policy-oriented departments these are not the same labour markets. sector for given organisation size, and in is informal, but it does have some Overall, some interesting differences the upwards creep of their average pay? empirical bite, as we shall see. We show emerge from the analysis of our data. For 1 So nicknamed after Roger Douglas, the crusading minister of averaged actual annual rates of growth of listed companies in the private sector, we finance in the Labour government elected in 1984. CEO pay and organisation size, calculated can clearly discern a structure of 2 The New Zealand Official Yearbook 1978 does not list the names of the heads of the then 40 government departments, from the first and last (2014) values of unsurprising linkages between CEO pay, but in describing their duties uses ‘he’ and ‘his’ throughout (page 47). The Ministry of Women’s Affairs was set up these variables, and weighted in the firm size, and pay and numbers of senior in 1985, and headed by Mary O’Regan, the first female averaging by the 2014 size of each managers. The big question remaining (in department head. 3 The Reserve Bank’s inflation calculator tells us that the organisation. New Zealand and in other English- Consumers Price Index rose by 29% from Q2 1988 to Q2 The third column in Table 4 shows the speaking countries) is how this structure 1997. 4 The three exceptions are the SSC itself, the GCSB (spy average CEO pay growth adjusted for any has, as it were, been inflated over time, agency) and the Crown Law Office. 5 For broader accounts of the advance and (partial) retreat changes in CEO pay that can be attributed such that CEO pay has on average doubled of ‘new public management’ doctrines in the New Zealand simply to changes in the size of the in 20 years. public sector, see Whitcombe (2008) and Hughes and Smart (2012). department, agency or company, using For public sector CEOs, there is 6 Policy advice is one of three intrinsically public service duties enumerated by the state services commissioner. the estimated elasticities for size effects perhaps nothing exceptional about the The other two are: ‘exercising the authority of the state’ shown in Table 3.10 growth of their pay since 1995, but, cross- (collecting taxes, setting regulatory frameworks); and ‘direct interventions in social and economic conditions’. The general We see that all four public sector sectionally, CEO pay linkages with points made by the commissioner in the paragraph quoted above would seem to apply equally to the management of groups lag well behind the corporate organisation size and other factors are these other duties. private sector in growth of both CEO pay relatively weak or non-existent. Coupled 7 Currently, the 11 surviving SOEs generate annual revenues totalling $5.4 billion, which is about the size of the budget and overall size, with public sector policy/ with the striking fact of little New of Auckland City (https://mcculloch.org.nz/state-owned- regulatory departments or agencies Zealand’s near leadership, internationally, enterprises/). 8 The chief of the New Zealand Defence Force usually receives showing the highest CEO salary growth in how much we pay our state sector an annual remuneration package in the $600–700,000

Page 58 – Policy Quarterly – Volume 13, Issue 4 – November 2017 range. In the fiscal year 2013–14 the incumbent chief left Health and Environment. We thank Simon Chapple for Acknowledgements the job, picking up a substantial severance package on his bringing this interesting information to our attention. way out. The chairman of the joint chiefs of staff in the US 10 So, for example, the adjusted number for spending Kar Yew’s work on this article was currently receives a salary of US$15,583 a month, according departments, 0.62 = 1.04 – 0.16x2.59. to Wikipedia. The three football coaches are paid (not by 11 Hazledine, 2016, p.198. supported by an Auckland University the taxpayer) annual salaries ranging from US$600,000 to 12 Pay of members of Parliament increased at the lower rate of Summer Research Scholarship. Input US$1,600,000. 1.6%. See www.nzlii.org/nz/legis/num_reg/psaad1995461/ 9 OECD, 2013, Table 5.5, p.107. The comparison is made for and www.kiwiblog.co.nz/2014/10/mps_salaries -2.html. from Jonathan Boston and a referee is the pay of the heads of six government departments: in New gratefully acknowledged. Zealand terms Internal Affairs, Finance, Justice, Education,

References Bollard, A., R. Lattimore and B. Silverstone (1996) ‘Introduction’, in B. Hughes, P. and J. Smart (2012) ‘You say you want a revolution … The Silverstone, A. Bollard and R. Lattimore (eds), A Study of Economic next stage of public sector reform in New Zealand’, Policy Quarterly, 8 Reform: the case of New Zealand, contributions to economic analysis (1), pp.3-8 236, Amsterdam; New York: Elsevier OECD (2013) Government at a Glance 2013, Paris: OECD, http://dx.doi. Duncan, I. (1996) ‘Public enterprises’, in B. Silverstone, A. Bollard and R. org/10.1787/gov_glance-2013-en Lattimore (eds), A Study of Economic Reform: the case of New State Services Commission (1998) Annual Report of the State Services Zealand, contributions to economic analysis 236, Amsterdam; New Commission for the year ended 30 June 1998, Wellington: State York: Elsevier Services Commission Hazledine, T. (2014) ‘Rogered by Rogernomics’, in D. Cooke, C. Hill, P. State Services Commission (2003) Annual Report of the State Services Baskett and R. Irwin (eds), Beyond the Free Market: rebuilding a just Commission 2003, Wellington: State Services Commission society in New Zealand, Auckland: Dunmore Publishing State Services Commission (2016) Senior Pay Report, Wellington: New Hazledine, T. (2016) ‘Tip of the iceberg? Organisational foundations of top Zealand Government pay in New Zealand listed companies, 1995–2014’, Australian Journal Whitcombe, J. (2008) ‘Contributions and challenges of “new public of Labour Economics, 19 (3), pp.199-208 management”: New Zealand since 1984’, Policy Quarterly, 4 (3), Hazledine, T., M. Wang and K.Y. Lee (2017) Bureaucratic Bloat? pp.7-13 Determinants, trends and implications of top pay in the New Zealand public and private sectors, Department of Economics working paper, Auckland: University of Auckland Appendix: Econometric modelling of CEO pay Table 5 summarises the results of Table 5: Public and Private Sector regressions for CEO pay Dependent Variable: log(CEOPAY) Ordinary Least Squares regression Dependent Variable: log(CEOPAY) models for CEO pay in public and private Public sector Private sector sectors. The modelling approach follows number of observations 158 205 that developed in Hazledine (2016), and is further explained in Hazledine coefficient t-stat coefficient t-stat et al. (2017). The main variables were constant -9.641 -1.75 -64.922 -9.88 introduced above, in sections two and log(SIZE) 0.157 17.30 0.302 20.66 three. The coefficients of variables entered in natural logarithms can be log(MANAGERS/ MANAGERSF) 0.137 7.59 0.265 8.10 read off as elasticities, showing the log(TOPSALARY/ TOPSALARYF) -0.046 -0.88 0.597 4.78 percentage difference in CEO pay associated with a given percentage YEAR 0.010 3.56 0.034 10.29 difference in the explanatory variable. DHB -0.277 -6.88

Organisation SIZE is also a determinant FIRE 0.205 2.26 of the number of senior MANAGERS and their pay (TOPSALARY), which would RORHIGH 0.113 2.14 result in ‘multicollinearity’ problems if all R-squared 0.684 0.751 three are included as regressors in the adjusted R-squared 0.674 0.744 model. Basically, the variables would get in each other’s way, blurring their individual effects on CEO pay. To deal included to allow for district health board Both public and private sector CEO with this, actual values of MANAGERS (DHB), FIRE (finance, insurance and real pay models have reasonably high R2 and TOPSALARY are divided by estate) effects, as well as the possibility values, given that this is essentially a their ‘forecast’ values from separate that CEOs in particularly profitable listed cross sectional model. regressions on SIZE and the time trend companies (RORHIGH) get paid more. variable, YEAR. ‘Dummy variables’ are See Hazledine et al. (2017) for details.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 59 Todd Krieble and Danijela Tavich

Civics and Citizenship Education in New Zealand a case for an active citizen participates in the life of a community in order to improve conditions for others or to help shape the change? community’s future’ (Adler and Goggin, Introduction 2005, p.241). This includes voting, as well as broader activities such as donating to The 2013 Constitutional Advisory Panel recommendation charity. Civic engagement is a fundamental for a national strategy for civics and citizenship education part of building and maintaining strong democracies (Blakeley, 2016, pp.99-10). in schools, kura (Mäori-medium schools) and communities There is substantial international evidence that democratic polities enjoy provided the opportunity for an important conversation better economic and social outcomes, about building civic knowledge in Aotearoa New Zealand leading to overall higher levels of well- being (Acemoglu and Robinson, 2012). If (Constitutional Advisory Panel, 2013, p.8). This article we accept this and also accept that New Zealand is a well-functioning democracy explores possible next steps for implementing this worth preserving (see World Bank, 2016), recommendation. It is broken up into two parts: a case for then there are three current trends that, unless attended to, will undermine the change, and potential next steps. quality of our democracy. This section builds on these trends with an outline of A case for change citizens interact with communities and existing research about CCE and the This article explores both civics and societies. Knowledge and understanding current conditions for CCE in New citizenship education (collectively referred of both civics and citizenship concepts Zealand. to as CCE). Civics education addresses constitutes broader ‘civic knowledge’ the formal institutions and processes (Bolstad, 2012, p.7). Loss of a common platform for public of civic life, such as voting in elections, CCE is discussed here with a view to discourse while citizenship education addresses improving civic engagement in New Traditional forms of ‘slow’ media, such as how people participate in society and how Zealand. Civic engagement describes ‘how high-quality, investigative, public interest journalism, are subject to declining Todd Krieble is a Principal Economist at the New Zealand Institute of Economic Research and a profitability and funding restrictions. patron of the McGuinness Institute. Danijela Tavich is a recent Master of Public Policy graduate of This has prompted a shift in the market Victoria University of Wellington and a Research Analyst at the McGuinness Institute. towards digital communication through

Page 60 – Policy Quarterly – Volume 13, Issue 4 – November 2017 channels such as social media, which and Pasifika males had the most limited 2014/15 (Ministry of Business, Innovation offer greater potential for the generation knowledge of democracy (ibid., p.9). and Employment, 2016, p.iv). This of advertising revenue (Armitage, 2016) This indicates that a ‘civic empowerment growing diversity enriches New Zealand and favour immediacy and entertainment gap’ exists (see Levinson, 2012, pp.32- (Gault and Krieble, 2016, p.34), but value (Drok and Hermans, 2016, pp.539- 3), which appears to mirror the other almost 60% of recent migrants did not 41). Despite response to these pressures, inequalities in our society (see Marriott vote in the 2011 general election (Statistics such as the online transformation of and Sim, 2014, pp.26-7; Rashbrooke, New Zealand, 2014, p.9). As many public broadcaster Radio New Zealand 2013, pp.1-6). The ICCS study found migrants come from countries with weak and its recent increase in public funding no clear pattern of association between democracies (Economist Intelligence through New Zealand on Air, the decline students’ average knowledge scores and Unit, 2016), it is important that knowledge of slow media remains an issue for citizens their level of interest in social and political barriers to engagement are reduced and seeking to stay informed (Pullar-Strecker, issues, nor their intentions for future civic that individuals are supported to access 2017; Radio New Zealand, 2013). action (Hipkins and Satherley, 2012, the civic institutions and exercise the These problems are exacerbated by an p.31). These results point to a need for rights they are entitled to as members of increase in the diversity of channels for something other than content or more New Zealand society (Palmer and Butler, news on the internet, making it difficult to ‘academic’ civic knowledge in the New 2016, p.12). However, it must be noted identify quality information. While this Zealand curriculum to encourage interest that a limitation of ‘citizenship’ diversity may be positive overall, it and participation. terminology is that it may foster an represents another aspect of the shift away from professional journalism (Gault and Krieble, 2016, p.35). Additionally, Teachers have significant discretion increasing diversity has caused a burgeoning of ‘fake news’: false and over how they teach under the New misleading news stories that can be shared widely online (Allcott and Gentzkow, Zealand Curriculum and Te Marautanga 2017, pp.1-3). A recent survey by the Pew o Aotearoa, which are based on learning Research Center found that 64% of US adults say fake news causes a great deal of areas, principles and values ... confusion about the basic facts of current issues and events (Barthel, Mitchell and Holcomb, 2016). The growing reliance of citizens on Although evidence suggests that youth assumption that only citizens can vote, social media for news is also causing issues are more engaged than we think, this when permanent residents are also eligible such as online ‘filter bubbles’. Because engagement is occurring externally to to enrol and vote in New Zealand social media platforms such as Facebook conventional forms of participation such (Electoral Commission, 2016). This must determine what content users see based as voting, instead focusing on activities be kept in mind when discussing CCE to on their connections and which pages like volunteering (Wood, 2017). The last avoid perpetuating such ideas. they already interact with, individual three general elections have seen falling news feeds can become an echo chamber, electoral enrolment rates in age groups Inconsistency in what constitutes CCE reinforcing personal biases (El-Bermawy, between 18 and 39, despite enrolment Teachers have significant discretion 2016). This does not bode well for civility, technically being compulsory in New over how they teach under the New which centres on seeing your opponent as Zealand (Electoral Commission, 2015, Zealand Curriculum and Te Marautanga a fellow citizen, rather than the ‘other’ or p.45; New Zealand Government, 2017). o Aotearoa, which are based on learning enemy, and emphasises respectful, Further, Mäori and Pasifika voters have areas, principles and values (Ministry of balanced and constructive public consistently lower turnout rates than Education, 2007, pp.7, 37). The flexibility discourse and debate (Rashbrooke, 2017). Päkehä (Statistics New Zealand, 2014, of the New Zealand Curriculum appears p.9). CCE could provide opportunities for to be advantageous but can be challenging Lack of knowledge about and interest in how addressing these issues by demystifying in the case of CCE. The problem is that democracy works the enrolment and voting process, and there is no consistency about what The 2008 International Civic and perhaps even helping eligible students to constitutes CCE; nor is there any explicit Citizenship Education Study (ICCS) enrol. requirement that CCE be taught (Harris, revealed that New Zealand has some of the New Zealand is becoming increasingly 2017, p.246). This approach disadvantages highest and lowest scores internationally diverse, but the high rate of non-voting students who are not taught CCE or who for civic knowledge. No other country in among migrants could indicate possible are taught less comprehensive CCE (ibid., the study had such a wide distribution issues for civic engagement. In 2015/16 pp.246-50). Given that the low civic (Lang, 2010, p.6). At the bottom end, Mäori resident visa approvals were up 21% from knowledge scores in the ICCS study were

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 61 Civics and Citizenship Education in New Zealand: a case for change? found to have a strong association with which CCE is taught is important (Wood explicit references to ‘citizenship’: for poorer socio-economic backgrounds, and Milligan, 2016, p.70). For example, example, notions of citizenship are a key CCE becomes one way to equip all in a 2008 study of 52 high schools in part of the ‘future focus’ principle (Harris, students with the knowledge and skills Chicago, Kahne and Sporte found that 2017, p.247; Ministry of Education, 2007, they need to empower themselves (Lang, ‘active citizenship’ approaches were pp.10, 12). Furthermore, CCE could 2010, p.10). most successful in leading to increased help to fulfil three of the five overall key In addition, the quality and civic engagement. Active citizenship competencies of the curriculum: thinking, accessibility of resources available to approaches link learning to real-world relating to others, and participating and support CCE requires attention. The contexts, for example, by following current contributing (Ministry of Education, recent Ministry for Culture and Heritage’s events and allowing students to study 2007, pp.12-13). citizenship education resources survey and think critically about social issues Social studies presents as the existing evaluated the current state of resources that matter to them (Kahne and Sporte, subject most compatible with notions of and used criteria drawn from the 2008, pp.738, 745-56; Wood and Milligan, CCE (Wood and Milligan, 2016, p.66). outcomes of Aitken and Sinnema’s Best 2016, p.70). Other research shows that the Social studies teachers from secondary Evidence Synthesis (BES) in social effects of CCE in community education schools across the country are already sciences, which examined effective for adults are similar to the effects on working together to form ideas for how to pedagogy in social sciences, to assess the school students (Finkel, 2014, pp.170, 178; implement the NCEA ‘personal social quality of a sample of these resources Gastil, 2004, pp.325-6). action’ achievement standards for social studies in a way that can address the gaps in CCE in the curriculum. These The issue of consistency is exacerbated achievement standards were introduced in 2013 and encourage students in years by the fact that existing notions 11–13 to take social action on an issue of of citizenship in the New Zealand their choice, providing an opportunity for active citizenship learning (Massey Curriculum are vague and provide little University, 2016). Personal social action approaches to clear direction for teachers ... date include a teacher who took students to Wellington to learn about the way Parliament works and a teacher whose class visited the Beehive to make a (Aitken and Sinnema, 2008; Tallon, 2016, International evidence also indicates submission at the select committee p.3). The criteria are grouped into five that CCE can address gaps in knowledge hearing on the Healthy Homes Guarantee overarching categories of social sciences inequalities. In a 2016 study looking at the Bill (ibid).1 A recent study by Wood et al. outcomes pertaining to: students’ United States and Belgium, Neundorf, on the implementation of these standards knowledge and understanding of Niemi and Smets found that civic found that, when students were well concepts; students’ skills in using social education can have compensation effects supported and were tackling personally sciences methods and techniques; for missing parental ‘political socialization’ significant issues, the standards were students’ ability to participate, contribute (p.947). Schools were found to be able to valuable for learning about society and and engage in dialogue; students’ compensate for the ‘civic empowerment social issues, as well as for developing civic awareness of personal, cultural and gap’ between young people from and community engagement skills (Wood layered identities; and students’ privileged backgrounds who were more et al., 2017, p.16). dispositions and emotional responses to likely to have access to academic resources, Tikanga ä-iwi, the subject parallel to learning (Aitken and Sinnema, 2008, political news and the public sphere social studies in kura, also appears to be p.37). It was found that although there are generally, and those from impoverished compatible with CCE, with the subtext of many resources available for civics and backgrounds (Levinson, 2012, pp.32-3; tikanga ä-iwi being a strong emphasis on citizenship education, they can be difficult Neundorf, Niemi and Smets, 2016, p.922). the realisation of rangatiratanga through to locate, lack coherence and are of The ICCS study cited earlier also provides active citizenship (Dale, 2016, p.27). varying quality (Tallon, 2016, p.16). evidence to support this finding (Schulz et However, it should be noted that Mäori al., 2010, p.258). conceptions of citizenship are inherently Existing research different from Western perspectives. Existing research also supports the case CCE and the New Zealand Curriculum Consequently, CCE resources and for change in CCE in New Zealand (see CCE is not currently part of the New strategies developed for tikanga ä-iwi Harcourt, Milligan and Wood, 2016; Zealand Curriculum, but the curriculum must align with the specific vision of Mutch, 2013; Wood and Mulligan, 2016). does include areas in which CCE could tikanga ä-iwi (ibid., pp.20-7). The evidence indicates that the way in be incorporated. There are already some

Page 62 – Policy Quarterly – Volume 13, Issue 4 – November 2017 The issue of consistency is exacerbated Figure 1: Three Cs framework for CCE by the fact that existing notions of citizenship in the New Zealand Curriculum are vague and provide little clear direction for teachers in implementing the principles and values consistent with citizenship (Wood and Milligan, 2016, p.67). This is evident in the findings of the final report of the ICCS study, which concluded that, overall, ‘it is somewhat unclear whether Critical there is a consistent view across New Content Zealand schools about what “civics and citizenship education” ought to involve Thinking and what means are effective in developing students’ citizenship competencies’ (Bolstad, 2012, p.32). Given that there is much evidence of the importance of CCE, and literature detailing appropriate pedagogy, it is worth considering why CCE is not more prevalent in New Zealand. It has been Connection suggested elsewhere that the pressure and narrow foci of national standards and NCEA assessment and the top-down emphasis on literacy and numeracy have led to these being the subjects reported on and prioritised at the expense of other areas of the curriculum (Harcourt, Milligan and Wood, 2016, p.xiii; Thrupp CCE in the community no coordination between efforts and few and White 2013, pp.19-20; Wylie and In terms of community CCE, work is resources available for community CCE. Bonne, 2016, p.25). As outlined by already underway, with resources for Furthermore, there is limited funding Harcourt, Milligan and Wood, this has led community CCE developed by the for community education in general, to the marginalisation of social studies, Electoral Commission with input from with cuts to ACE Aotearoa in recent years and can explain the inconsistency of the Adult and Community Education (ACE) having a significant impact (Pollock, current approach to CCE (2016, p.xiii). Aotearoa that provide a template for 2012). Evidence of poor learning progress for community groups to start learning social studies students in years 4–8 in about civics and citizenship (Electoral Potential next steps comparison with other subjects suggests Commission, 2012). Further, information This article has set out three reasons why that such marginalisation is affecting on the 2017 election is available through civics and citizenship education needs learning in this area (National Education the Electoral Commission in 27 languages, attention if New Zealand is to remain a Monitoring Project, 2005, 2009, cited in and English Language Partners NZ has well-functioning democracy: loss of a Wood and Milligan, 2016, p.68). also produced a resource for its work common platform for public discourse; It seems that furthering CCE would with refugees and migrants (Electoral lack of knowledge of and interest in how entail not only addressing issues like Commission, n.d.; English Language democracy works; and inconsistency consistency, but also some reconsideration Partners NZ, 2017). There are numerous in what constitutes CCE. Although we of sector priorities. Such reconsideration other examples, including civics education argue that there is a case for change, the may not require absolute policy trade- workshops run by ACE Aotearoa current state of CCE in the education and offs: existing priorities are not with a focus on prison inmates (ACE community sectors is not one of crisis. fundamentally incompatible with CCE Aotearoa, 2015, p.23), the work of Active Rather, it presents a picture of promising and could be revisited with a view to Citizenship Aotearoa (Wood, 2017), and foundations and existing efforts that reconciling them with CCE. CCE can the Victory Community Centre in Nelson, would benefit from greater support, provide rich content for literacy and a community hub connecting the local strengthening and coordination. This numeracy learning through topics such as school and wider community with a focus section outlines some next steps drawn voting statistics or political speeches (see on building connection and engagement from the McGuinness Institute’s May Fraser, Aitken and Whyte, 2013, pp.171- (Stuart, 2010, p.86). However, this work 2017 CivicsNZ workshop, and is set out 3). will have limited impact while there is under the Three Cs framework.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 63 Civics and Citizenship Education in New Zealand: a case for change?

A framework for CCE Marautanga o Aotearoa and tikanga online CCE hub would make it easier for Figure 1 shows the Three Cs framework ä-iwi in kura. However, it may also be educators and communities to identify for CCE, adapted from Gault and Krieble necessary to reconsider the current and access CCE content and resources (2016). The framework outlines three priorities of the education sector. (Tallon, 2016, pp.19-20). The BES method critical components of CCE which Arguably, subjects like social studies will for assessing the quality of citizenship would address the three trends that are remain second best until these priorities resources could be used to select resources threatening the quality of our democracy. are revised or even reimagined, by, for for inclusion in this hub and to guide the Content refers to information or ‘academic’ example, exploring options for teaching development of future resources (Aitken knowledge about civics and citizenship; literacy and numeracy through social and Sinnema, 2008; Tallon, 2016, p.3). critical thinking refers to the ability to studies in order to ‘elevate’ the subject The idea of CCE hubs is also viable in critically assess and process information; (see Wood and Milligan, 2016, p.71). a community context. For communities and connection refers, firstly, to the Although some NCEA literacy credits that have built or want to build a civics application of information to problems in can already be assessed through social hub, such as the Victory Community everyday life, and secondly to a sense of studies, this could be extended by Centre in Nelson, it may be worth looking belonging to a specific community within offering more credits, exploring options to the Ministry of Education communities society. The importance of connection for aligning the subject with numeracy of learning/kähui ako model, which links and critical thinking in CCE, rather than credits, and recognising the compatibility networks of schools and kura, as a possible just content, draws from evidence of of social studies with literacy, numeracy, foundation for this (Ministry of Education, 2016). The networks within this model could be used to link schools Though no panacea, CCE can serve with their communities in ways which foster a sense of belonging, inclusivity and as a means of securing the democratic learning, with CCE extending into the character of New Zealand, and ensuring community experience (Wood and Milligan, 2016, p.70). Increases in funding that all can be a part of shaping the may be necessary if CCE is to be accessible to adult learners and those outside the future. mainstream education sector. Existing CCE hubs and other community organisations and efforts will be more effective if they are coordinated, effective pedagogy for engendering civic New Zealand Curriculum and University working to build relationships and share engagement (see Aitken and Sinnema, Entrance priorities (NZQA, 2017a, resources and experiences. Coordination 2008; Harcourt, Milligan and Wood, 2016; 2017b). could be facilitated by government or by Mutch 2013; Wood and Milligan, 2016). one of the existing community groups Aligning curriculum objectives working in this space. The development of Content CCE might also be linked to other areas further resources suitable for community Bolster social studies and consider policy of the curriculum beyond social studies, CCE, similar to the Electoral Commission’s trade-offs as in the earlier example of linkages resources, would also be beneficial for Making CCE an explicit learning outcome between mathematics and CCE through supporting community CCE. of the New Zealand Curriculum would topics like voting statistics. Prospective help to address issues of consistency. and practising teachers could be Critical thinking The learning targets for CCE might supported by the Education Council and Active citizenship approaches and knowing then be met within the existing subject teaching colleges to align curriculum where to look for news of social studies. Social studies could be objectives and subject knowledge. This Critical engagement with current affairs bolstered to support teachers to address would help to embed the concepts of is part of active citizenship. In an era CCE with active citizenship approaches civics and citizenship in everyday life of instant social media and the ability and by drawing on existing best practice (Fraser, Aitken and Whyte, 2013, pp.18- of practically anyone to produce and examples of social studies teachers’ work. 20). distribute ‘news’, critical thinking skills are Bolstering social studies might entail essential. This highlights the importance professional development and guidance, Pull together CCE hubs and coordinate of active citizenship approaches to such as advisory support for teachers, as existing efforts CCE, which entail critical thinking and has been available in the past (Aikin, Consistent with the recommendations assessment, as a way to support learners 1995, pp.68-70). Further consideration is of the Ministry for Culture and Heritage to develop these essential critical thinking necessary regarding whether and how citizenship education resources survey, skills and help individuals seek out and CCE might fit with the vision of Te a single point of coordination and recognise reliable news sources.

Page 64 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Role for public interest media support the participation of groups in civic knowledge that have led to the The role of public interest media outlets society that are less engaged and are apparent development of a civic becomes particularly important for under-represented in voting and other empowerment gap in New Zealand, informing public debate and active citizenship-related statistics. Working which, in turn, reinforces many of the citizenship when the capacity to critically partnerships between Mäori, Pasifika and inequalities we see in society today. appraise news and its sources is limited. migrant communities and government We suggest some next steps that come The 2017 budget increase for Radio agencies such as the Ministry of Youth together as a vision of coordination across New Zealand partially recognises the Development, the Ministry of Social a range of actors, such as schools, kura, importance of a trusted public interest Development, the Ministry of Education, communities, government, not-for-profit media outlet. However, rather than being the Department of Internal Affairs and organisations, the media and the private part of the budget for New Zealand on the Electoral Commission may be able to sector. CCE could be made an explicit Air, Radio New Zealand should have a turn civic engagement figures around by learning outcome in the New Zealand stand-alone budget so that budget trade- appealing to the issues that matter most Curriculum. CCE can support existing offs are transparent and its crucial role for to these groups, learning more about the priorities for numeracy and literacy, and democracy is recognised. driving forces behind disengagement, has a natural home in social studies. and co-producing CCE approaches that Consequently, we argue that social studies Connection are relevant and relatable for diverse should be bolstered to facilitate CCE, Civic engagement is more than voting audiences. through changes such as improved, Civic engagement takes many forms. curated access to online resources and There may be value in beginning a Conclusion advisory support for educators. Though national conversation about what we There are strong arguments for no panacea, CCE can serve as a means of mean when we talk about citizenship, safeguarding democracy. Democratic securing the democratic character of New civic engagement and participation. societies are more prosperous, both Zealand, and ensuring that all can be a The above-mentioned evidence that socially and economically. The three part of shaping the future. people may be participating in civics and drivers outlined in this article that threaten citizenship in little-acknowledged ways 1 Here, it is worth considering some safety protocols for to undermine the strength of democracy ensuring the privacy of students when making submissions, indicates that citizenship and engagement in New Zealand are a significant concern: such as the use of school rather than personal details. are not straightforward, static concepts. a loss of a common platform for public There is a need to democratise the discourse; a lack of knowledge about and Acknowledgements definitions of civics and citizenship and to interest in how democracy works; and The authors are very grateful to the staff, co-produce CCE to ensure that it is useful inconsistency in what constitutes CCE. patrons and workshop participants of the and accessible. How CCE can incorporate In response to such threats, we propose McGuinness Institute, and also to Andrea concepts like civility, to help underpin a case for change to the current state of Milligan, Bronwyn Wood, Andrew Butler, political discourse with respect and to CCE. There is evidence to support Geoffrey Palmer, Murray Lucas, Stasi further a sense of connection in New consistent and comprehensive CCE across Turnbull, Steve Watters, Robyn Baker and Zealand, should also be considered here. schools and communities as an effective Gary Hawke. The views expressed here are intervention for preserving and future- those of the authors alone. Engaging the disaffected proofing democracy. Further, CCE can Government should do more to address the distributional problems in

References ACE Aotearoa (2015) Annual report 2015, Adult and Community Armitage, C. (2016) ‘Online ad revenue passes $800m in 2015’, National Education Aotearoa, retrieved from https://www.aceaotearoa.org.nz/ Business Review, 4 March, retrieved from https://www.nbr.co.nz/article/ sites/default/files/ACE-Aotearoa-Annual-Report-2015.pdf online-ad-revenue-passes-800m-2015-ca-185819 Acemoglu, D. and J.A. Robinson (2012) Why Nations Fail: the origins of Barthel, M., A. Mitchell and J. Holcomb (2016) ‘Many Americans believe power, prosperity and poverty, London: Profile Books fake news is sowing confusion’, Pew Research Center journalism and Adler, R.P. and J. Goggin (2005) ‘What do we mean by “civic media, 15 December, retrieved from http://www.journalism. engagement”’?, Journal of Transformative Education, 3 (3), pp.236-53 org/2016/12/15/many-americans-believe-fake-news-is-sowing- Aikin, S. (1995) ‘Primary problems and the New Zealand curriculum confusion/ framework’, New Zealand Annual Review of Education, 4, pp.57-76 Blakeley, R. (2016) Building Community Resilience, Wellington: New Aitken, G., and C. Sinnema (2008) Effective Pedagogy in Social Sciences/ Zealand Society of Local Government Managers Tikanga a- Iwi, Wellington: Ministry of Education Bolstad, R. (2012) Participating and Contributing? The role of school and Allcott, H. and M. Gentzkow (2017) ‘Social media and fake news in the community in supporting civic and citizenship education: New Zealand 2016 election’, Journal of Economic Perspectives, 2 (2), pp.211-36, results from the International Civic and Citizenship Education Study, retrieved from https://web.stanford.edu/~gentzkow/research/ Wellington: Ministry of Education fakenews.pdf

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 65 Civics and Citizenship Education in New Zealand: a case for change?

Constitutional Advisory Panel (2013) New Zealand’s Constitution: a report Marriott, L. and D. Sim (2014) Indicators of Inequality for Ma-ori and on a conversation, Wellington: New Zealand Government Pacific People, working paper in public finance 09/2014, Wellington: Dale, H. (2016) ‘Te whanaketanga o te wähanga ako o te Tikanga a- Iwi: Victoria Business School mai i te kore, ki te wheiao, ki te ao märama (The development of the Massey University (2016) ‘Teaching teens about the political system and Tikanga a Iwi learning area: from nothingness, to half-light, to the full social activism’, Scoop, 23 November, http://www.scoop.co.nz/stories/ light of day)’, in M. Harcourt, A. Milligan and B. Wood (eds), Teaching ED1611/S00112/teaching-teens-about-the-political-system-social- Social Studies for Critical, Active Citizenship in Aotearoa New Zealand, activism.htm Wellington: NZCER Press Ministry of Business, Innovation and Employment (2016) Migration Trends Drok, N. and L. Hermans (2016) ‘Is there a future for slow journalism?’, 2015/2016, Wellington: Ministry of Business, Innovation and Journalism Practice, 10 (4), pp.539-54 Employment Economist Intelligence Unit (2016) Democracy Index 2016: revenge of the Ministry of Education (2007) The New Zealand Curriculum, Wellington: ‘deplorables’, London: Economist Intelligence Unit Ministry of Education El-Bermawy, M.M. (2016) ‘Your filter bubble is destroying democracy’, Ministry of Education (2016) Community of Learning: guide for schools and Wired, 18 November, retrieved from https://www.wired.com/2016/11/ kura, Wellington: Ministry of Education filter-bubble-destroying-democracy/ Mutch, C. (2013) ‘What does a decade of research reveal about the state Electoral Commission (n.d.) ‘Information in other languages’, retrieved from of citizenship education in New Zealand?’, New Zealand Journal of http://www.elections.org.nz/resources-learning/information-other- Education Studies, 48 (2), pp.51-68 languages National Education Monitoring Project (2005) Social Studies: assessment Electoral Commission (2012) Your Voice, Your Choice: learning units, results 2005, retrieved from http://nemp.otago.ac.nz/forum_ retrieved from http://www.elections.org.nz/sites/default/files/Your_ comment/2005_reports.htm#sstud Voice_Your_Choice_Learning_Units.pdf National Education Monitoring Project (2009) Social Studies: assessment Electoral Commission (2015) Report of the Electoral Commission on the results 2009, retrieved from http://nemp.otago.ac.nz/social_ 2014 General Election, Wellington: Electoral Commission studies/2009/index.htm Electoral Commission (2016) ‘Who can and can’t enrol?’, 1 August, Neundorf, A., R.G. Niemi and K. Smets (2016) ‘The compensation effect retrieved from http://www.elections.org.nz/voters/enrol-check-or- of civic education on political engagement: how civics classes make up update-now/who-can-and-cant-enrol for missing parental socialization’, Political Behaviour, 38, pp.921-49 English Language Partners NZ (2017) ‘Topic worksheets’, retrieved from New Zealand Government (2017) ‘Enrol to vote’, https://www.govt.nz/ http://englishlanguage.org.nz/topic-worksheets browse/engaging-with-government/enrol-and-vote-in-an-election/ Finkel, S.E. (2014) ‘The impact of adult civic education programmes in enrol-to-vote/ developing democracies’, Public Administration and Development, 34, NZQA (2017a) ‘Literacy and numeracy assessment standards’, retrieved pp.196-81 from http://www.nzqa.govt.nz/qualifications-standards/qualifications/ Fraser, D., V. Aitken and B. Whyte (2013) Connecting Curriculum, Linking ncea/subjects/literacy-and-numeracy/level-1-requirements/lit-num- Learning, Wellington: NZCER Press subjects/ Gastil, J. (2004) ‘Adult civic education through the national issues forums: NZQA (2017b) ‘Approved subjects for University Entrance’, retrieved from developing democratic habits and dispositions through public http://www.nzqa.govt.nz/qualifications-standards/awards/university- deliberation’, Adult Education Quarterly, 54 (4), pp.308-28 entrance/approved-subjects/ Gault, P. and T. Krieble (2016) ‘Do citizens and communities have the Palmer, G. and A. Butler (2016) A Constitution for Aotearoa New Zealand, news and information they need and want in a digital age?’, Policy Wellington: Victoria University Press Quarterly, 12 (2), pp.31-6 Pollock, K. (2012) ‘Adult education: challenging times, 1980s to 2010s’, Harcourt, M., A. Milligan and B. Wood (2016) Teaching Social Studies for Te Ara – the Encyclopedia of New Zealand, http://www.TeAra.govt.nz/ Critical, Active Citizenship in Aotearoa New Zealand, Wellington: en/adult-education/page-4 NZCER Press Pullar-Strecker, T. (2017) ‘Budget 2017: Radio NZ funding freeze comes to Harris, M. (2017) The New Zealand Project, Wellington: Bridget Williams an end’, Stuff, 25 May, http://www.stuff.co.nz/business/ Books industries/92976168/budget-2017-radio-nz-funding-freeze-comes-to- Hipkins, R. and P. Satherley (2012) New Zealand Students’ Intentions an-end Towards Participation in Democratic Processes: New Zealand results Radio New Zealand (2013) ‘International recognition for online innovation’, from the International Civics and Citizenship Education Study, media release, 1 August, retrieved from http://www.radionz.co.nz/ Wellington: Ministry of Education media/55 Kahne, J.E. and S.E. Sporte (2008) ‘Developing citizens: the impact of Rashbrooke, M. (2013) Inequality: a New Zealand crisis, Wellington: civic learning opportunities on students’ commitment to civic Bridget Williams Books participation’, American Educational Research Journal, 45 (3), Rashbrooke, M. (2017) ‘The indulgence of anger: NZ philosopher Jeremy pp.738-66 Waldron on why politics needs more civility’, The Spinoff, 12 August, Lang, K. (2010) What Do New Zealand Students Understand about Civic https://thespinoff.co.nz/politics/12-08-2017/the-indulgence-of-anger- Knowledge and Citizenship?, Wellington: Ministry of Education nz-philosopher-jeremy-waldron-on-why-politics-needs-more-civility/ Levinson, M. (2012) No Citizen Left Behind, Cambridge, MA: Harvard University Press

Page 66 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Schulz, W., J. Ainley, J. Fraillon, D. Kerr and B. Losito (2010) ICCS 2009 Wood, B. (2017) ‘Young Kiwis more engaged as citizens than we think’, International Report: civic knowledge, attitudes, and engagement anong Newsroom, 22 May, https://www.newsroom.co.nz/@future- secondary students in 38 countries, Amsterdam: IEA learning/2017/05/22/29782/young-kiwis-more-engaged-than-we-think Statistics New Zealand (2014) Non-voters in 2008 and 2011 General Wood, B.E. and A. Milligan (2016) ‘Citizenship education in New Zealand Elections: findings from New Zealand General Social Survey, policy and practice’, Policy Quarterly, 12 (3), pp.65-73 Wellington: Statistics New Zealand Wood, B., R. Taylor, R. Atkins and M. Johnston (2017) Creating Active Stuart, D. (2010) Paths of Victory: Victory village (Victory primary school Citizens: interpreting, implementing, and assessing ‘personal social and Victory community health centre) – a case study, Wellington: action’ in NCEA social studies, Wellington: Teaching and Learning Families Commission Research Initiative Tallon, R. (2016) Citizenship Education Resources Survey, Wellington: World Bank (2016) The Worldwide Governance Indicators, 2016 update, Ministry for Culture and Heritage retrieved from http://info.worldbank.org/governance/wgi/#reports Thrupp, M. and M. White (2013) Research, Analysis and Insight into Wylie, C. and L. Bonne (2016) Secondary Schools in 2015: findings from National Standards (RAINS) Project Final Report: national standards the NZCER national survey, Wellington: NZCER and the damage done, report commissioned by NZEI, Hamilton: Wilf Malcolm Institute of Educational Research

Master of e-Government BETTER GOVERNMENT THROUGH TECHNOLOGY Gain the knowledge and TRIMESTER TRIMESTER TRIMESTER competence to understand, analyse, design and ONE TWO THREE manage transformational EGOV 501 Managing Service EGOV 502 e-Government, EGOV 503 Managing ICT- e-government initiatives. Transformation (Core course) Public Sector Reform and Good enabled Forms of Public The managerial aspect and Governance (Core course) Engagement The challenges, Connect with experts and issues of ICT-enabled service The theories, principles, conditions and requirements a global community of transformation in the public models and strategies for using for managing the use of ICTs to classmates sector. transformational government achieve effective forms of public initiatives to establish good engagement. GOVT 518 Comparative Public Your courses are designed governance, drawing on Management (Core course) EGOV 521 Research Project by experts in government international and New Zealand Comparative public management Applications of theoretical, experience. and ICT and taught by and public policy, with emphasis conceptual, analytical, practical leading academics. Be part on contemporary developments EGOV 520 Introduction to and research methodological of a school that attracts in Asia–Pacific and beyond. Research in the Public Sector knowledge to the design An introduction to approaches, and conduct of an applied not only local professionals 2 x 15 point electives from ethics and methods of doing research project in the area of but a talented group of EGOV 500–599, MMIM 500–599, research in the public sector. transformational e-government. MMPM 500–599, MAPP 500–599 international students. This course is co-taught with MMIM 513 Managing IT-related GOVT 561. Change FURTHER MMIM 510 Information Systems The impact of information INFORMATION: Management (Core course) technology on an organisation The principles and practices and issues concerned with IT Academic enquiries of managing the information related change. e [email protected] systems function within an 1 x 15 point elective from EGOV organisation. Enrolment enquiries 500–599, MMIM 500–599, e [email protected] 1 x 15 point elective from EGOV MMPM 500–599, MAPP 500–599 p +64-4-463 6599 500–599, MMIM 500–599, MMPM 500–599, MAPP 500–599

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 67 Belinda Storey and Ilan Noy

Insuring Property under Climate Change Climate change will increasingly create severe risks for In the best of all possible worlds New Zealand would face far fewer risks from New Zealand’s coastal housing stock. Even a small amount climate change. Sound science would of sea level rise will substantially exacerbate the costs of communicate future risk, which would be understood by all decision makers. These flooding and storm surges (Parliamentary Commissioner decision makers would make socially optimal decisions and would coordinate for the Environment, 2015). Under the Intergovernmental across all levels and parts of government. Panel on Climate Change’s (IPCC) three mitigation scenarios, Existing homeowners would take risk into account when making housing investment global average sea levels are likely to rise by between 28cm decisions. Home buyers would take risk into account when purchasing, and this and 73cm by 2100 (above the 1986–2005 average). Under the would affect both what they purchase IPCC’s high emissions scenario the sea level is likely to rise by and how much they are willing to pay. Developers would take future and current between 52cm and 98cm by 2100 (IPCC, 2013). Only collapse climate risk into account in siting and of parts of the Antarctic ice sheet, if triggered, could cause the designing developments. Insurers would pool the residual risk across individuals, sea level to rise substantially above these ranges. Some regions and would obtain affordable reinsurance in international markets. Councils would in New Zealand (including the main urban centres) have high credibly commit to an adaptive decision- enough quality geographic data to infer the number of homes making approach for land use and building decisions and would continue to at risk. In those regions, there are over 43,000 homes within adjust this approach as the climate, and 1.5m of the present average spring high tide and over 8,000 sea level, change. Alas, we do not live in this Panglossian dream. within 50cm (Bell, Paulik and Wadwha, 2015). In reality, there is limited information about how climate risks will change over

Belinda Storey is a doctoral student in economics at Victoria University of Wellington. Ilan Noy holds time and the information that is available the inaugural Chair in the Economics of Disasters at Victoria University of Wellington. is often not accessible to the layperson.

Page 68 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Even with good information, people often been passed down by past generations such as earthquakes and floods that are make poor decisions under different types (Awatere, 2008). The assets which will be often excluded elsewhere. of uncertainty: they overreact to small exposed to climate change decades hence Other institutions are also used to threats, easily forget previously observed are being built now, so the far-reaching manage risk. The New Zealand Coastal loss, and exhibit optimism bias when risks consequences of current decisions should Policy Statement 2010, established under are high. They often discount future be reflected in current policy (Stephens, the Resource Management Act (RMA), is events heavily and do not plan well for Bell and Lawrence, 2017). the overarching planning document them.1 These same obstacles and barriers Here, we focus on how escalating regarding New Zealand’s coasts along faced by homeowners and renters also coastal housing risks induced by sea level with part 2 matters in the RMA. Each afflict, of course, policymakers, and the rise, coastal storms and extreme regional council is required to prepare a employees of private sector entities precipitation will affect the future regional coastal policy statement which involved in residential housing, such as availability of insurance in New Zealand. outlines how, among other matters, the banks and insurers. Insurance does not reduce risk. Instead, regional council addresses management When coastal disasters occur, insurance allows one party (the insured) of natural hazards (including climate homeowners experience significant losses to transfer some of its risk to another change) and must give effect to the and displacement. Some may be forced to permanently leave their community after a single and sudden-onset disaster like a Climate change may make the storm surge, flash flood or landslide, or following a series of smaller events that calculation of actuarially precise accumulate to large losses (Moftakhari et premiums more difficult as changes to al., 2017). Property developers and existing homeowners may seek to block hazard frequency and intensity render the transmission of information about risk to potential home buyers. Local and historical data less relevant. central government may face high costs from protective measures and continued provision of infrastructure when abandoning housing may be more party (the insurer) through a financial national policy. The New Zealand Coastal efficient. Local authorities, and their contract which compensates the insurer Policy Statement states that a pre- insurers, may find themselves holding for the receipt of the transferred risk. cautionary approach should be adopted unexpected liabilities if future courts rule Evidence from international markets to resource use and development in that councils are liable for resource suggests that when a risk is perceived to coastal environments wherever the effects consents provided to homes threatened have increased and become unprofitable of activities ‘are uncertain, unknown, or by climate change. All this may happen, to transfer, insurance companies classify little understood, but potentially and almost surely will. affected areas as ‘uninsurable’ and significantly adverse’ (policy 3). It Climate change will render some withdraw insurance altogether. The New highlights climate change as a particularly currently inhabited locations Zealand government plays a major role in important source of vulnerability for uninhabitable. This transition could well the provision of some natural disaster coastal resources and outlines policies for be costly for individuals and their insurance through the Earthquake identifying and managing coastal hazards, communities. It is not yet known how Commission (EQC). Several other including the effects of climate change many locations will face this transition features differentiate the New Zealand (policies 24–27) (Department of over the next few decades. Many of the residential property insurance sector from Conservation, 2010). Local councils are costs considered in this article will be those in other countries. First, New responsible for managing risk insofar as accrued in the more distant future. The Zealand has mostly international insurers they construct infrastructure and make traditional use of government discount in the local market. Second, disaster planning decisions which affect exposure rates might portray such costs several insurance take-up is unusually high to coastal hazards. decades from now as innocuous. Heavily (probably around 90-95%), because it is discounting the future losses from past automatically attached to fire insurance. Issues with pricing climate-sensitive and present actions is generally Third, prudential regulation of New insurance inappropriate, however, especially when Zealand insurers is quite recent and does Climate change may make the calculation those harmed will be uncompensated not constrain the product categories in of actuarially precise premiums more (Cowen and Parfit, 1992; Stern, 2015). which an insurer may operate (Dean, difficult as changes to hazard frequency Moreover, discounting is incongruous 2010). Fourth, the New Zealand insurance and intensity render historical data with the emphasis within mätauranga industry has historically offered all-perils less relevant. In recent decades most of Mäori on safeguarding treasures that have coverage, including major natural hazards the increase in global coastal risk has

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 69 Insuring Property under Climate Change come from increased exposure through undergoes an extratropical transition that event for a number of years, or where the urbanisation and economic development results in a large storm surge in Auckland average household tenure at that location (McGranahan, Balk and Anderson, 2007). or Tauranga, where many exposed homes is short, the demand for catastrophic However, as the sea level rises at faster and businesses are located. insurance is likely to be subdued. rates, and storms intensify, changes in Accurate estimates of future coastal International studies have found that hazards are expected to replace changes hazards are further limited by deep the demand for insurance cover for in exposure as the primary driver of uncertainties around polar ice sheet catastrophic events is more price sensitive escalating coastal risks (Prudential response and future global emission than for insurance for non-catastrophic Regulation Authority, 2015). pathways, the high collection cost of up- events (Botzen and van den Bergh, 2012; As the global sea level rises, king tides, to-date high-resolution land elevation Grace, Klein and Kleindorfer, 2004). storm surges and waves will reach further and asset data sets, and the limited When insurance premiums for low- inland. Research commissioned by the precision of risk models for extreme probability events rise in price, individuals Parliamentary Commissioner for the events and their actuarial counterparts. may stop insuring or may under-insure, Environment has demonstrated that even The price New Zealand pays for even when those premiums are subsidised modest increases in the sea level will reinsurance in global financial markets is below their true actuarial cost (Dixon at dramatically reduce the return period of an important determinant of the retail el., 2006; Petrolia, Landry and Coble, major events. For example, with a 10cm cost of insurance. If international 2013). Those with lower incomes are somewhat less likely to buy catastrophic insurance than those with higher incomes International experience suggests that (Grace, Klein and Kleindorfer, 2004). in the absence of an EQC-type scheme, The combined effect of price and income means that as insurance premiums most homeowners do not insure against consume a greater proportion of disposable income, homeowners are less natural hazards and governments are likely to retain insurance (Landry and compelled by public pressure to provide Jahan-Parvar, 2011) and demand for insurance falls faster for lower-probability, ad hoc assistance high-impact risk (i.e. the risk from catastrophic events) than it does for higher-probability, low-impact risk (i.e. sea level rise the return period for a one- reinsurance markets harden – currently frequent nuisance events) (Grace, Klein in-100-year storm surge in Wellington they are facing historically low costs of and Kleindorfer, 2004). In other words, if will reduce by a factor of five to a return financing – the reinsurance premiums the AEP (annual exceedance probability) period of one in 20 years. Once the sea paid by EQC and New Zealand’s private of an event doubles from 0.5% to 1% (i.e. level has risen by 30cm, a one-in-100-year insurers could significantly rise. Since if a one-in-200-year event becomes a one- event is expected to become an annual New Zealand catastrophic risk is in-100-year event), insurance demand is event in Christchurch and Wellington. uncorrelated with other markets, likely to fall by a greater amount than if The global sea level is projected to rise by reinsurers are more likely to raise prices the AEP doubles from 2% to 4% (i.e. if a between 17 and 38cm by 2065 (IPCC, than to leave the New Zealand market one-in-50-year event becomes a one-in- 2013). altogether, but reinsurers may ultimately 25-year event). The resolution of climate models is withdraw cover for some perils, such as The Household Economic Survey such that it is difficult to predict even large storm surges or earthquakes. conducted by Statistics New Zealand events such as tropical cyclones (Roberts, suggests that average household Colle and Korfe, 2017; Walsh et al., 2015). Demand for residential insurance expenditure on building insurance has Nevertheless, most climate models predict Even risk-averse individuals tend to almost doubled since the first Canterbury that while the number of tropical cyclones underestimate risk, particularly low- earthquake in 2010, from $540 in 2010 to in the South Pacific will reduce slightly, probability, high-impact risk (Kousky, $1,051 in 2015 (New Zealand Treasury, the proportion of tropical cyclones that Michel-Kerjan and Raschky, 2017; 2017). This includes the 2012 increase in reach category 4 and 5 will increase and Kunreuther and Pauly, 2004; McClelland, EQC premium rates from 5 cents to 15 the path of tropical cyclones will move Schulze and Coursey, 1993). Demand for cents per $100 cover, but not the poleward (Holland and Bruyere, 2014; insurance often increases immediately November 2017 increase to 20 cents. As Kossin, Emanuel and Vecchi, 2014; following a catastrophic event (Browne insurance premiums take up a larger Munich Re, 2016; Ramsay, Camargo and and Hoyt, 2000; Michel-Kerjan and proportion of disposable income, more Daehyun, 2014; Woodruff, Irish and Kousky, 2010), but returns to modest low-income households are likely to Camargo, 2013). This could increase the levels within a few years (Gallagher, 2014). under-insure or allow their insurance probability that a tropical cyclone In locations that have not experienced an policies to lapse.

Page 70 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Homeowners may not purchase we’ll have to deal with that then’ (Radio demands this would place on EQC and on insurance if they believe that they will be New Zealand, 2017). The government its interactions with the private insurance compensated by government. also offered a $700,000 support package sector. Retreat by private insurers from International experience suggests that in to approximately 100 affected businesses particular locations could increase the the absence of an EQC-type scheme, most (Morton, 2017), thereby creating a unfunded fiscal risk to the Crown homeowners do not insure against natural disincentive for businesses to purchase associated with private property in natural hazards and governments are compelled business continuity insurance for future disasters, should the Crown elect to by public pressure to provide ad hoc events. provide relief to uninsured homeowners assistance (Kousky, Michel-Kerjan and (ibid.). Raschky, 2017; Kunreuther and Michel- Public supply of residential insurance Applying a standard (flat) EQC Kerjan, 2014). This may encourage The government believes it is important premium price nationwide helps spread homeowners to avoid insurance, thereby to help private property owners avoid the risk faced in more hazardous locations increasing the future fiscal risk for ‘socially unacceptable distress and loss across all policy holders. This makes government if it is induced to provide in the event of a natural disaster’ (New catastrophe insurance affordable for those compensation (Raschky and Weck- Zealand Treasury, 2015). EQC has helped who are most exposed and helps ensure Hannemann, 2007; Raschky et al., 2013). ensure greater insurance penetration in high insurance penetration. In doing so, Following the Canterbury earthquakes, the government did not assist homeowners who had not purchased insurance except EQC does not protect against erosion where it was compelled to do so for property that was red-zoned.2 Any public caused by slow-onset events or rising objection to that decision was muted, as seas, but the courts may hold EQC liable the number of people who reportedly did not have insurance was very small. Public for land loss caused by storms – many objection and government response after a climate change-related disaster may be of which will be more destructive with very different if the number of uninsured sea level rise and changes in extreme properties is higher where insurance had previously been prohibitively priced for precipitation. affected communities, or if the event is perceived to be at least partially a consequence of government failure. New Zealand so that homeowners have a however, it also mutes the price signal Case study: Edgecumbe and Cyclone Debbie much higher take-up rate of catastrophe which otherwise may discourage or shape In general, in the immediate aftermath insurance than is the case in other development in more hazardous locations. of natural disasters public officials countries (New Zealand Treasury, 2015). The nature of EQC land cover is face intense pressure to offer support, EQC protects private residential currently being reconsidered as part of the particularly to those households and property and contents from damage by review of the Earthquake Commission businesses that for whatever reason earthquake, volcanic eruption, Act 1993. There is, apparently, no current are not insured (Boston and Lawrence, hydrothermal activity, landslip, tsunami, proposal to remove EQC’s land cover for 2017). Public support for ad hoc or fire caused by natural disaster. EQC storms or floods (ibid.).3 assistance is bolstered by media coverage land cover extends the range of perils to In May 2017 the New Zealand of natural disasters and can reinforce include storm and flood hazards, but government announced an increase to expectations that relief will be offered in excludes coastal erosion. EQC does not EQC’s premiums (also known as the EQC similar situations in the future (Seifert cover damage to residential structures or levy), effective from 1 November 2017 et al., 2013). For example, immediately contents from storm or floods (or coastal (EQC, 2017). This increase was designed following the Edgecumbe floods caused erosion). to ensure that the premiums paid by by extratropical Cyclone Debbie in EQC premiums are collected by homeowners reflected EQC’s long-term 2017, the prime minister, Bill English, private insurance companies and are costs, including expected future losses acknowledged that ‘there’s probably going embedded within residential insurance from natural hazards. Historically, 85% of to be people there who aren’t insured or policies that include fire insurance. EQC’s historical land claims have been for whom it’s had a huge impact’ and Consequently, if private insurers withdraw less than $20,000 (New Zealand Treasury, suggested that public assistance would be from certain markets, homeowners would 2015). Recent analysis commissioned by offered to households to work through need to apply directly to EQC for cover. It EQC assumes that the expected losses for ‘their immediate issues and then if the is not clear how many homeowners would EQC’s land cover – for all perils – will long-term one is lack of insurance then seek to do so and what administrative remain less than 10% of EQC’s total

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 71 Insuring Property under Climate Change average annual loss (New Zealand to differentiate is likely to increase as the 1993 Earthquake Commission Act, Treasury, 2017). cost of estimating individual risk falls commercial properties are insured only Except for severe liquefaction and with higher-resolution climate models, by the private sector insurers. Without major landslides, land damage from advanced data aggregation and analysis the involvement of the public insurer geological hazards can usually be (‘big data’) and the expansion of geodata (EQC), the response of commercial policy remediated: volcanic ash can be removed such as LiDAR. Policy discrimination in premiums to the Canterbury earthquakes and buckled earth can be levelled. EQC New Zealand is uncommon and has was more dramatic. The loss of cover for does not protect against erosion caused by historically taken the form of higher earthquake-prone buildings was more slow-onset events or rising seas, but the excesses rather than higher premiums, but acute in the non-residential sector. It is courts may hold EQC liable for land loss this could change. also likely that some of the other factors caused by storms – many of which will be that might be dampening the expected more destructive with sea level rise and Impact of insurance retreat on mortgage increases in premiums on individual changes in extreme precipitation. When availability and cost residential properties do not apply in the land disappears with storm surges or flash Insurance is a requirement for residential commercial sector, where the full explicit floods, full compensation may be required. mortgages in New Zealand and failing to and implicit risk is borne by the insurers. Given the rising value of coastal and maintain insurance can trigger ‘technical’ For commercial properties, insurance riparian land, EQC’s exposure could default. The possibility of default is plays an additional, socially beneficial role by allowing entrepreneurs and small and medium-size businesses to transfer some As insurance retreats from particular of the risks they incur, thus facilitating locations, house prices in those more investment in future activities and growth. The vast majority of businesses in areas are likely to be affected and New Zealand are small, so this risk- infrastructure investments may be more transfer role for insurance is potentially very important (though as yet difficult to justify. unquantified). Similarly, business continuity insurance contracts are often tied to property insurance, so the potential withdrawal of commercial property become orders of magnitude greater than exacerbated by maturity mismatches insurance will impose cascading barriers historical averages. between residential insurance and on the operations of small and medium- mortgages. While mortgages are often size commercial entities. Private supply of residential insurance granted with repayment periods spanning Insurance covers risks for which there is decades, insurance contracts are renewed Looking ahead significant uncertainty. As such, insurers annually. Insurers are thus able to leave The landscape for insurance demand and will retreat from coastal and riparian an insurance market within 12 months, supply is already changing. Insurers are locations once risks are sufficiently while it may be a decade or more before gradually moving away from the New probable. Insurers may retreat from a lenders’ loans mature. As a consequence, Zealand practice of all-peril policies. They coastal or riparian area of New Zealand in the future, bankers may lend to owners are offering different types of maximum following a climate event in that location of coastal property less often, require cover for different hazards, even though or in another New Zealand location. more equity as collateral, or offer shorter this adds complexity to insurance Alternatively, they may retreat after their mortgage terms (Lawrence et al., 2016). contracts. This may lower demand for experience in another country convinces Even now, and despite their rules requiring disaster insurance from homeowners. A them that risk profiles have changed mortgagors to insure, the general absence requirement that insurers provide a non- because of sea level rise or other climatic of compliance checks means that banks technical explanation of the key elements changes. Insurance retreat from coastal do not currently know whether particular of the insurance policy on a single page and riparian locations could increase the properties they mortgage remain insured could support homeowners’ decision unfunded fiscal risk faced by the Crown beyond the first year of ownership. making. Reforms of consumer finance and decrease house prices as mortgages in the United States following the global become unavailable (or more costly). Commercial insurance financial crisis, such as the Credit CARD Insurers may be more willing to The same problems and trends that we Act of 2009, provide an example of this continue to provide insurance to high- described for insurance for residential type of requirement. A breakdown of the risk areas if they decide to discriminate buildings and contents4 also apply to cost component for each category of risk between these areas and lower-risk ones in commercial ones; though commercial – for example, how much of an insurance the policies they offer (in premium prices, cover responds to these pressures premium price covers the risk of fire in excesses or in policy wording). Pressure even faster. In New Zealand, since the as opposed to the risk of floods at that

Page 72 – Policy Quarterly – Volume 13, Issue 4 – November 2017 location – could strengthen the signalling and supported economic growth, will coal mine, and inform decisions to effect of insurance. From experience with become less available. As insurance gradually withdraw residential properties other policies – for example, mandatory retreats from particular locations, house from those locations most at risk from pricing of plastic bags – we know that prices in those areas are likely to be climate change. even very modest pricing signals can have affected and infrastructure investments 1 Economists, with their talent for esoteric terminology, call a material impact on behaviour. may be more difficult to justify. Since this myopia ‘hyperbolic discounting’. As hazards become more likely, and insurance companies are unlikely to 2 Quake Outcasts v the minister of Canterbury earthquake recovery [2017], NZCA 332 (1 August 2017). particularly for homes experiencing commit to long-term insurance contracts, 3 EQC land liability is the smallest of: the area of the insured land that is damaged; the minimum-sized area allowed for repeat events, insurers are also transferring local and central governments will need to use of a residential site under the relevant district plan; the some risk back onto homeowners by commission their own analysis of the value of the 4,000m2 closest to the dwelling (New Zealand Treasury, 2015). requiring very high excesses. Insurers are potential timing and scale of insurance 4 Note: since private insurers do not provide cover for land already withdrawing from some retreat from locations affected by climate damage, and EQC does not provide cover for commercial properties, no insurance is available for land damage on individual flood- and earthquake-prone change hazards. commercial properties. properties. Policymakers are limited in Furthermore, the New Zealand their ability to determine the extent of this government’s fiscal exposure will expand Acknowledgments problem, as private insurers do not as the government faces ever greater This article has been informed by a disclose their commercial decisions pressure to respond to climate disasters facilitated dialogue run by Motu Economic regarding insurance availability. As with financial assistance and other post- and Public Policy Research as part of the climate hazards escalate, these trends will event recovery support programmes. Deep South National Science Challenge accelerate. In some jurisdictions, the Creation of innovative funding impacts and implications programme. insurance regulator collects information instruments that support pre-event The inclusion of the ideas in this document from insurers on insurance coverage (Plitt adaptation measures that enhance societal does not imply any recommendation, and Maldonado, 2012). Aggregated data resilience and reduce risk could lessen consensus or endorsement by dialogue from insurers on key contractual terms, these claims on central and local participants or presenters, their affiliated including premium pricing and excess government when disasters occur (Boston organisations or the programme funder. levels, and expected and actual losses, as and Lawrence, 2017). All opinions, errors and omissions are the well as notification of withdrawal of Since climate change will render some authors’ own. coverage from particular locations, could currently inhabitable locations be very useful for policymakers in uninhabitable, policy interventions in the assessing emerging risks. insurance market can only achieve so Insurance, which has long contributed much. Projections of probable insurance to the financial security of homeowners retreat could serve as the canary in the

References Awatere, S. (2008) ‘The price of mauri: exploring the validity of welfare Dixon, L., N. Clancy, S.A. Seabury and A. Overton (2006) The National economics when seeking to measure mätauranga Mäori’, PhD thesis, Flood Insurance Program’s Market Penetration Rate: estimates and University of Waikato policy implications, Santa Monica: RAND Corporation Bell, R.G., P. Paulik and S. Wadwha (2015) National and Regional Risk EQC (2017) ‘Budget announcement: EQC levy to increase’, 25 May, Exposure in Low-lying Coastal Areas, report prepared for the https://www.eqc.govt.nz/news/budget-announcement-eqc-levy-to- Parliamentary Commissioner for the Environment increase Boston, J. and J. Lawrence (2017) The Case for New Climate Change Gallagher, J. (2014) ‘Learning about an infrequent event: evidence from Adaptation Funding Instruments, Wellington: Institute for Governance flood insurance take-up in the United States’, American Economic and Policy Studies and New Zealand Climate Change Research Journal: Applied Economics, 6 (3), pp.206-33 Institute, Victoria University of Wellington Grace, M.F., R.W. Klein and P.R. Kleindorfer (2004) ‘Homeowners Botzen, W.J.W. and J. van den Bergh (2012) ‘Risk attitudes to low- insurance with bundled catastrophe coverage’, Journal of Risk and probability climate change risks: WTP for flood insurance’, Journal of Insurance, 71 (3), pp.351-79 Economic Behaviour and Organization, 82 (1), p.151-66 Holland, G. and C.L. Bruyere (2014) ‘Recent intense hurricane response to Browne, M.J. and R.E. Hoyt (2000) ‘The demand for flood insurance: global climate change’, Climate Dynamics, 42 (3-4), pp.617-27 empirical evidence’, Journal of Risk and Uncertainty, 20 (3), pp.291- IPCC (2013) ‘Sea level change’, in T.F. Stocker et al. (eds), Climate Change 306 2013: the physical science basis, Cambridge and New York: Cambridge Cowen, T. and D. Parfit (1992) ‘Against the social discount rate’, in P. University Press Laslett and J.S. Fishkin (eds), Justice Between Age Groups and Kossin, J.P., K.A. Emanuel and G.A. Vecchi (2014) ‘The poleward Generations, New Haven: Yale University Press migration of the location of tropical cyclone maximum intensity’, Dean, R. (2010) ‘Bringing financial stability legislation to the insurance Nature, 509 (7500), pp.349 industry: the Insurance (Prudential Supervision) Act 2010’, Reserve Kousky, C., E.O. Michel-Kerjan and P.A. Raschky (2017) ‘Does federal Bank of New Zealand Bulletin, 73, pp.19-28 disaster assistance crowd out flood insurance?’, Journal of Department of Conservation (2010) New Zealand Coastal Policy Statement Environmental Economics and Management 2010, Wellington: New Zealand Government

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 73 Insuring Property under Climate Change

Kunreuther, H. and E.O Michel-Kerjan (2014) ‘Economics of natural Prudential Regulation Authority (2015) The Impact of Climate Change on catastrophe risk insurance’, in M. Machina and K. Viscusi (eds), the UK Insurance Sector: a climate change adaptation report by the Handbook of the Economics of Risk and Uncertainty, Oxford; Prudential Regulation Authority, London: Bank of England Amsterdam: North Holland Radio New Zealand (2017) Interview with Prime Minister Bill English, Kunreuther, H. and M. Pauly (2004) ‘Neglecting disaster: why don’t people Checkpoint, 7 April insure against large losses?’, Journal of Risk and Uncertainty, 28 (1), Ramsay, H.A., S.J. Camargo and K. Daehyun (2012) ‘Cluster analysis of pp.5-21 tropical cyclone tracks in the southern hemisphere’, Climate Dynamics, Landry, C.E. and M.R. Jahan-Parvar (2011) ‘Flood insurance coverage in 39 (3-4), pp.897-917 the coastal zone’, Journal of Risk and Insurance, 78 (2), pp.361-88 Raschky, P.A., R. Schwarze, M. Schwindt and F. Zahn (2013) ‘Uncertainty Lawrence, J., P. Blackett, N. Cradock-Henry, S. Flood, A. Greenaway and of government relief and the crowding out of flood insurance’, A. Dunningham (2016) Climate Change Impacts and Implications for Environmental Resource Economics, 54 (2), pp.179-200 New Zealand to 2100: synthesis report RA4: enhancing capacity and Raschky, P.A. and H. Weck-Hannemann (2007) ‘Charity hazard: a real increasing coordination to support decision making hazard to natural disaster insurance?’, Environmental Hazards, 7 (4), McClelland, G., W. Schulze and D. Coursey (1993) ‘Insurance for pp.321-9 low-probability hazards: a bimodal response to uncertainty’, in C. Reed, A.J., M. Mann, K.A. Emanuel, N. Lin, B.P. Horton, A.C. Kemp and Camerer and H. Kunreuther (eds), Making Decisions About Liability and J.P. Donnelly (2015) ‘Increased threat of tropical cyclones and coastal Insurance, Springer Netherlands flooding to New York during the anthropogenic era’, Proceedings of the McGranahan, G., D. Balk and B. Anderson (2007) ‘The rising tide: National Academy of Sciences, 112 (41), pp.12610-15 assessing the risks of climate change and human settlements in low Roberts, K.J., B.A. Colle and N. Korfe (2017) ‘Impact of simulated elevation coastal zones’, Environment and Urbanization, 19 (1), twenty-first-century changes in extratropical cyclones on coastal pp.17-37 flooding at the Battery, New York City’, Journal of Applied Meteorology Miche-Kerjan, E.O. and C. Kousky (2010) ‘Come rain or shine: evidence on and Climatology, 56 (2), pp.415-32 flood insurance purchases in Florida’, Journal of Risk and Insurance, 77 Seifert, I., W.J.W. Botzen, H. Kreibich and J.C. Aerts (2013) ‘Influence of (2), pp.369-97 flood risk characteristics on flood insurance demand: a comparison Moftakhari, H.R., A. AghaKouchak, B.F. Sanders and R.A. Matthew (2017) between Germany and the Netherlands’, Natural Hazards and Earth ‘Cumulative hazard: the case of nuisance flooding’, Earth’s Future, 5 System Sciences, 13, pp.1691-705 (2), pp.214-23 Stephens, S.A., R.G. Bell and J. Lawrence (2017) ‘Applying principles of Morton, J. (2017) ‘Heartbeat: flood-hit Edgecumbe battles on’, New uncertainty within coastal hazard assessments to better support Zealand Herald, 15 August coastal adaptation’, Journal of Marine Science and Engineering, 5 (40), Munich Re (2016) Natural Catastrophes 2015, Munich Re pp.1-20 NatCatSERVICE Stern, N. (2015) Why Are We Waiting? The logic, urgency, and promise of New Zealand Treasury (2015) New Zealand’s Future Natural Disaster tackling climate change, Cambridge, Mass: MIT Press Insurance Scheme: proposed changes to the EQC Act 1993, 15 July Storey, B., I. Noy, W. Townsend, S. Kerr, R. Salmon, D. Middleton, O. New Zealand Treasury (2017) ‘Aide memoire: EQC Review: update on Filippova, V. James and S. Owen (2017) Insurance, Housing and expected costs of potential future EQC schemes’, 2 May Climate Adaptation: current knowledge and future research, Motu note Parliamentary Commissioner for the Environment (2015) Preparing New 17, Wellington: Motu Economic and Public Policy Research Zealand for Rising Seas: certainty and uncertainty, Wellington: Walsh, K.J.E., J.L. McBride, P.J. Klotzbach, S. Balachandran, S.J. Parliamentary Commissioner for the Environment Camargo, G. Holland, T.R. Knutson, J.P. Kossin, T. Lee, A. Sobel and Petrolia, D.R., C.E. Landry and K.H. Coble (2013) ‘Risk preferences, risk M. Sugi (2015) ‘Tropical cyclones and climate change’, WIREs Climate perceptions and flood insurance’, Land Economics, 89 (2), pp.227-45 Change, 7 (1), pp.65-89 Plitt, S. and D. Maldonado (2012) ‘When constitutional challenges to state Woodruff, J.D., J.L. Irish and S.J. Camargo (2013) ‘Coastal flooding by cancellation moratoriums enacted after catastrophic hurricanes fail: a tropical cyclones and sea-level rise’, Nature, 504 (7478), pp.44-52 call for a new federal insurance program’, Brigham Young University of Journal of Public Law, 27, pp.41-96

Festive greetings from the School of Government The School of Government would We wish you all a happy and restful like to extend our sincere thanks festive season and look forward to and good wishes to all those who working with you all again in 2018. had contact with the School during 2017, with particular School of acknowledgement of our 2017 Government graduands and prize-winners.

Page 74 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Graduate Pathway Students: Three Contributions The School of Government at Victoria University with the Productivity Commission, and exploring of Wellington each year accepts a small number covenanted land and sustainability for the Waipä of promising undergraduates into its Professional District Council. We are very proud of our excellent Masters programmes. They undertake an Graduate Pathway students and look forward to accelerated programme that involves a for-credit their graduation this December. short internship. In this issue of Policy Quarterly We sincerely thank our internship sponsors from - we feature short articles by three students – 2016/2017 - Waipa- District Council, Akina Alice Denne, Matthew Macfarlane and Danijela Foundation, New Zealand Productivity Commission, Tavich – based on their internship research. The the State Services Commission and the Ministry of students had highly varied experiences ranging Justice. from supporting a social enterprise in its work with Dr Barbara Allen, Graduate Pathway Coordinator contractors to government, researching productivity

Alice Denne Lessons from an Internship at Waipa- District Council motivations and incentives Creating change through policy interventions relies most now all but 7.5% of the district’s land has been cleared. It is home to 73 nationally often on changing individuals’ behaviour. To create effective threatened species, including kahikatea, change, it is important for policymakers to understand the tawa and podocarp forest, känuka shrub and species of bat and gecko (Kessels & attitudes and motivations of the people most affected. I Associates Ltd, 2013). As a local authority the district council learned how important this is while spending my summer must work to prevent further destruction interning at the Waipä District Council in Waikato as part of the heritage that remains. The Resource Management Act 1991 (RMA) requires of Victoria University’s Master of Public Policy graduate territorial authorities to protect ‘areas of significant indigenous vegetation and pathway programme. significant habitats of indigenous fauna’, Isolated from the rest of the world, New reduced the dense native bush that while preventing inappropriate land use, Zealand is home to much indigenous covered most of the country just 200 years subdivision and destruction (s6(c); Waipä natural heritage. The rise of agriculture ago. The Waipä district is not exempt District Council, 2010). The challenge is and suburban living has dramatically from this trend and, as a farming region, that the most vulnerable land and many threatened ecosystems are found on private Alice Denne is a recent Master of Public Policy graduate of Victoria University of Wellington. property, outside council authority. In

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 75 Lessons from an Internship at Waipa-: District Council motivations and incentives order to meet biodiversity goals, the efforts be divided into two camps: those who plant the land, their ecological benefit will of landowners to protect and maintain the farmed for money and economic gains, and be minimal. For this reason, not only are quality of the environment that surrounds those who farmed for the lifestyle, driven by the covenants important, but it is also them is essential. Limited in their ability to stewardship values to care for the important for landowners to have a monitor or enforce regulatory measures, environment (Scarlett, 2004; Sullivan et al., stewardship attitude towards their land the councils’ role is then to encourage 1996; Taylor, Cocklin and Brown, 2011; and be motivated to look after it. Hence, landowners to protect their own natural Maybery, Crase and Gullifer, 2005). the current policy is not doing enough to heritage sites. These motivations were consistent change the attitude and behaviour of Natural heritage covenants are with what I found in the Waipä district: landowners unwilling to protect natural promoted as an effective way to ensure the four of the six landowners were protecting heritage and threatened species. ongoing existence of threatened species their natural heritage because they cared Some interviewees talked about on private land otherwise vulnerable to about the environment. They talked about neighbours who saw covenants as ‘black destruction. Covenants prevent both being motivated by the knowledge that marks on the landscape’ and thought that current and any future title holders from they were leaving something behind for maximising the agricultural potential of removing the heritage feature and are future generations and felt a responsibility land was always more important than administered by either the Queen to protect what was left. protecting bush. It is these attitudes that need changing to prevent the further destruction of the district’s natural This study taught me the importance of heritage. Awarding subdivision entitlements to those unwilling to knowing the people whose behaviour maintain their heritage sites brings few you are seeking to change and benefits to the region, and at the cost of undermining the district’s rural growth understanding their motivations. strategy. The strategy aims to reduce ad hoc rural development created by these unplanned residential lots scattered Elizabeth II National Trust or the council. For the environmentally motivated between towns. In the Waipä, landowners are incentivised farmers, the subdivision entitlement was a As it stands, the current incentive to use covenants by being offered a reward for the work they were already offered by the Waipä District Council subdivision entitlement in exchange. doing on their property, not an incentivising reaches the low-hanging fruit, mostly These can be used on another section of factor to apply for covenant status. The those individuals already motivated by their land or sold to another property legal protection a covenant provides is not stewardship values to restore the natural within the district. This creates a financial enough to ensure ongoing protection of environment in their care. The financial benefit to compensate for the loss of threatened species; many sites require incentive is not great enough to overcome potential economic value the land had fencing, riparian margin planting, and the loss of property rights or compensate prior to covenanting. ongoing weeding and pest control. Before a for the ongoing work that must be done To better meet the requirements of the covenant is created, landowners are made by farmers. It is not easy to see the RMA, the council sought to evaluate the aware of the management required and are degeneration of natural environments effectiveness of this scheme. I wanted to able to make a decision on whether this that happens gradually over generations. know how people were responding. Was ongoing cost is worthwhile. Even with a The council should increase awareness of this incentive attractive to landowners? covenant, without the effort, time and the issue of biodiversity loss in the district. Were farmers interested in covenanting financial investment of landownersIt needs to inform people of what they can their land, which often comes with costly maintaining their sites, the preservation of look for as indications of this on their management plans on top of loss of natural heritage would be lessened by own land and highlight what changes they property rights, for the economic benefit uncontrolled weeds and pests. should be concerned about in their of a subdivision? Or was there another, The two farmers who were motivated waterways, soils and visible landscape. more important motivator for seeking primarily by the economic value of the After awareness is increased, the covenant status? entitlement claimed that their sites strategies the council uses to reduce the My exploratory study was centred on required very little, if any, maintenance. effects and prevent further damage can be interviews with landowners who already Neither was motivated to protect the better communicated. I suggest that there had covenants on their properties. environment in their care. should be greater awareness of the Altogether, I conducted seven exploratory As they provide ongoing legal biodiversity corridor strategies and interviews in early 2017, six with landowners protection, natural heritage covenants are protection for significant natural areas, and one with a local property valuer. Existing essential for the long-term preservation of and why these are important priorities for literature suggested that the farmers and biodiversity. But without the effort of the district. Support can be built through their motivations, broadly speaking, could individuals to weed, control pests and profiling the environmental work already

Page 76 – Policy Quarterly – Volume 13, Issue 4 – November 2017 being undertaken by role models and capacity, skill and time. These can be assumptions and then design policies that champions in the community. Greater reduced by profiling external funders, miss the mark because of this. Policy is benefits will be achieved if social norms community restoration projects, and about people: it is about their values and regarding land use and the environment increasing education about what motives and the way they respond to can change. landowners can do to restore natural incentives, sanctions and barriers. Yet changing attitudes alone will be heritage on their own properties (Steg and Policymakers do not know everything and ineffective when significant barriers still Vlek, 2008; Taylor, Cocklin and Brown, must often make decisions based on exist. I recommend that the council tries 2011). limited information. But it is important to to better understand the specific barriers This study taught me the importance engage with those most affected by the faced by Waipä residents who may intend of knowing the people whose behaviour policy and develop a better understanding to act pro-environmentally but are limited you are seeking to change and of their motivations and barriers to in their capacity to do so. I found that understanding their motivations. It is easy change. barriers are both financial and related to for policymakers to make inaccurate

References Kessels & Associates Ltd (2013) Significant Natural Areas of the Waipa- Sullivan, S., E. McCann, R. De Young and D. Erickson (1996) ‘Farmers District: terrestrial and wetland ecosystems, technical report 2013/16, attitudes about farming and the environment’, Journal of Agricultural Hamilton: Waikato Regional Council and Environmental Ethics, 9 (2), pp.123-43, doi:10.1007/ Maybery, D., L. Crase and C. Gullifer (2005) ‘Categorising farming values BF03055298 as economic, conservation and lifestyle’, Journal of Economic Taylor, A., C. Cocklin and R. Brown (2011) ‘Fostering environmental Psychology, 26 (1), pp.59-72 champions: a process to build their capacity to drive change’, Journal Queen Elizabeth II National Trust (2011) ‘About covenanting’, http://www. of Environmental Management, 98 (1), pp.84-97, doi:10.1016/j. openspace.org.nz/Site/About_covenanting/default.aspx jenvman.2011.12.00 Scarlett, L. (2004) ‘Citizen stewardship and the environment’, Society, 41 Waipä District Council (2010) Environment Strategy, retrieved from http:// (6), pp.7-11, doi:10.1007/BF02688225 www.waipadc.govt.nz/our-council/Waipa2050/Documents/ Steg, L. and C. Vlek (2008) ‘Encouraging pro-environmental behaviour: an Environment%20Strategy%202010.pdf integrative review and research agenda’, Journal of Environmental Psychology, 29 (3), pp.309-17, doi:10.1016/j.jenvp.2008.10.004

Matthew Macfarlane Two Key Lessons from a Public Sector Internship accountability in public sector contracting and client-level data - I served as a research intern for the during the summer of 2016-17. Akina, which - Akina Foundation through the School of is dedicated to fostering social enterprise in Government’s graduate pathway programme New Zealand, was contracted by the Ministry

Matthew Macfarlane is a graduate of the Masters in Public Policy programme at Victoria University of Wellington.

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 77 Two Key Lessons from a Public Sector Internship: accountability in public sector contracting and client-level data of Social Development (MSD) to The following reflections focus on two specifically in a contract; their other facilitate and manage the co-design and specific issues raised by the new policy: accountabilities are only those set out by implementation of its Acceleration for first, the implications of the new form of their individual organisation (e.g. in codes Results programme (Äkina Foundation, contracting for accountability; and of conduct) or their professional 2017). This programme seeks to transition second, the privacy issues that have arisen standards. the providers with MSD contracts from as a result of the desire of policymakers, in This is a problem because public output-based contracts (i.e. purchase the interests of more effective resources are expended by organisations agreements) to results-based contracts interventions, to acquire greater client- not covered by the State Sector Act. Aside (i.e. contracts which specify desired level data from social service providers via from each organisation’s professional changes in recipients’ quality of life the new contracting regime. practices, and what is outlined in their according to prescribed measures). The respective contracts, non-governmental new contracts are intended to assist MSD The impact on accountability organizations (NGOs) are not obliged to in reaching its Better Public Services When government programmes are act like public sector agencies. For (Ministry of Social Development, 2015) contracted out, the tasks to be undertaken instance, they are not responsible for targets by measuring (and incentivising) are carefully defined at the outset so that providing free and frank advice to provider performance based on client both parties have matching expectations. ministers; they are not obliged to serve the government’s wider objectives, such as stewardship; and they are not required to ... one of the risks of contracting out the contribute to the objectives of public agencies with which they do not have provision of publicly funded services via contracts. Hence, contracting out presents tightly defined contracts is that NGOs a structural risk in the design and implementation of public policy in New focus narrowly on the terms of the Zealand. One example in the social investment space is that there is no contract, potentially at the expense of mechanism for NGOs to account for the wider public goals. benefits which accrue to departments they are not contracted with. For example, an NGO contracted to the Ministry of Social Development has no incentive to create savings for the Department of Corrections outcomes, rather than directly purchasing This creates a client–provider relationship other than what may be stipulated in a services from providers. dynamic, and enables providers to be specific contract. By contrast, a public During my involvement the policy held to account for their performance. servant working in MSD is ultimately programme was at a pilot implementation These features have an impact on both accountable to cabinet (in hierarchical stage. That is to say, there was a small the design and the implementation of the terms) and therefore has a responsibility sample of social service providers working policy, because the relationship of a client to consider the wider public interest. with MSD and Äkina to understand the to a provider is different from that of a Structural risks necessitate structural goals of the policy, identify barriers to department or ministry to a minister. solutions. One way forward would be to change from the perspective of both MSD Public servants juggle difficult,place an additional responsibility on and providers, and iteratively design some sometimes conflicting accountabilities – public servants to ensure that contractors prototype contracts. Approximately 20 to their line managers and chief executives, provide ongoing feedback on how their providers were involved across the four to the government of the day and the activities might contribute to wider workshops I attended, and about 15 MSD individual minister, and to their governmental objectives. This implies staff from regional and central offices. stakeholders – alongside responsibilities strong relational contracting, with regular My role for Äkina was to observe the to be good stewards of resources and to interactions between those tasked with processes and context specific to the new maintain political neutrality. Public commissioning and those responsible for contracting policy, and to conduct an servants are offered tools to navigate these delivering publicly funded services. A key independent literature review. From these accountabilities: the State Sector Act, the goal must be to ensure that departments I prepared recommendations that code of conduct for the state services, and cannot contract away their responsibilities highlighted international practices and other resources, written and unwritten, to be good stewards of public resources. potential risks. The project also involved including constitutional conventions. The some rapid processing of data from accountabilities for non-government Individual client-level data workshops and surveys to assess provider entities working on government Individual client-level data is, as defined readiness, mood, and satisfaction with the programmes are different. The bulk of by MSD, ‘information about a client that early implementation stages of the policy. their accountabilities are outlined typically doesn’t change over time and

Page 78 – Policy Quarterly – Volume 13, Issue 4 – November 2017 is identifiable by its nature’ (Ministry of or at least were willing to trade privacy for Concluding remarks Social Development, 2017). This includes the promise of ‘better’ outcomes. Even so, My experience as an intern brought to the information such as names, ages and these providers were unwilling to support fore critical issues of governance. One of locations, as well as what services clients the government’s proposals. This raises these relates to the risks associated with are accessing. Here I briefly outline the questions about whether it is right governments contracting out the design key controversies that have become the ethically for providers to adopt a of public policy. Another concerns the subject of media attention (Brunton, 2017; paternalistic approach to their clients, and practice of contracting with NGOs for Kirk, 2017) and a report from the privacy conversely whether it is justified for the the provision of services. In both cases, commissioner (Privacy Commissioner, government to use its coercive powers and important issues of accountability arise. 2017). These arose during my internship funding mechanisms to compel More specifically, one of the risks of as the collection of individual client-level organisations to release this required contracting out the provision of publicly data was (at that time) to be included in information. funded services via tightly defined every new contract. The opportunity for The willingness of some clients to contracts is that NGOs focus narrowly on first-hand exploration of the issues was sacrifice their privacy for a better quality the terms of the contract, potentially at personally exciting. service aligns with the view that privacy is the expense of wider public goals. Another Concerns surround both privacy as a luxury good (Trepte and Reinecke, 2011; concern is that government departments something intrinsically valuable, and the Angwin, 2014) that can be traded away for are unable to fulfil their responsibilities to potential consequences of a client’s something individual clients value more be good stewards of public resources and unwillingness to consent to providing highly. The question of whether people, their capacity to pursue wider government personal data – such as reduced trust in especially vulnerable people reliant upon objectives is diminished. At the same providers or choosing not to use services social services, should be compelled to time, it will be important in an age of they would benefit from, thereby incurring exchange privacy for the services they big data and the unequal distribution of harm. In the workshops I attended need raises significant ethical issues. Much personal privacy and material resources providers were almost uniformly opposed depends on the value placed on individual that governmental contracts with NGOs to the collection of individual client-level privacy and the extent to which less do not impose unjustified demands for data. The privacy commissioner reports privacy is likely to enhance the personal information. that anecdotally around half of providers achievement of other important societal have clients who would be unwilling to goals. For those who value privacy highly, release their personal data. there will need to be powerful reasons for However, some providers also reported trading even modest amounts of it away. that their clients were largely unconcerned,

References Äkina Foundation (2017) ‘Acceleration for results’, http://akina.org.nz/ Ministry of Social Development (2015) Community Investment Strategy, programmes/acceleration-for-results, accessed 4 September https://www.msd.govt.nz/documents/about-msd-and-our-work/ Angwin, J. (2014) ‘Has privacy become a luxury good?’, New York Times, work-programmes/community-investment-strategy/community- https://www.nytimes.com/2014/03/04/opinion/has-privacy-become-a- investment-strategy.pdf luxury-good.html, accessed 28 June 2017 Ministry of Social Development (2017) Building Financial Capability Beeck Center (2016) Funding for Results: how governments can pay for Services Guidelines, https://www.msd.govt.nz/documents/what-we- outcomes, http://beeckcenter.georgetown.edu/wp-content/ can-do/providers/building-financial-capability/building-financial- uploads/2016/06/Funding-for-Results_BeeckCenter.pd, accessed 6 capability-services-guidelines.pdf, accessed 28 June February 2017 Privacy Commissioner (2017) Inquiry into the Ministry of Social Brunton, T. (2017) ‘Budget advisory refuses to send client information to Development’s Collection of Individual Client-level Data from NGOs, government’, Stuff, http://www.stuff.co.nz/timaru-herald/ https://www.privacy.org.nz/assets/Files/Reports/2017-04-04-Inquiry- news/90011360/budget-advisory-refuses-to-send-client-information- into-MSD-collection-of-individual-client-level-data-OPC-report.pdf, to-government, accessed 28 June accessed 28 June Kirk, S. (2017) ‘Government backs down over collecting individual’ data Trepte, S. and L. Reinecke (eds) (2011) Privacy Online: perspectives on until security confirmed’, Stuff, http://www.stuff.co.nz/national/ privacy and self-disclosure in the social web, Springer eBooks politics/92925264/government-backs-down-over-collecting- individuals-data-until-security-confirmed, accessed 28 June

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 79 Danijela Tavich New Zealand Productivity Commission Internship social sector productivity

For my internship I prepared a short paper on social adjustment, which ensures that quality changes are considered when measuring sector productivity for the Productivity Commission. The productivity, to ensure a fair picture of objective was to consider the implications of introducing performance and minimise the promotion of productivity improvements at the productivity measurement in the social sector, given some expense of quality (Hanushek and Ettema, of the complexities of observing outputs and outcomes 2015). Consequently, to be able to apply standard productivity measures to social for certain sector tasks. To do this I selected one typology, sector tasks we must be able to observe some defined input, and the output or James Q. Wilson’s matrix of government tasks in Gregory outcome, and we must be able to quality- (1995b), and attempted to apply this to a set of tasks within adjust the result. one organisation in the social sector, namely the Ministry Theory Wilson’s typology differentiates among of Social Development (MSD). The tasks were drawn from types of tasks by the observability of MSD’s annual report for 2015/16. outputs and outcomes, and identifies four categories, as shown in Figure 1. Productivity in the social sector of the volume of output to the volume Production tasks have observable outputs The social sector is complex, covering of inputs (Gemmell, Nolan and Scobie, and outcomes; procedural ones have a variety of activities, including 2017). Productivity is a useful concept in observable outputs but outcomes that health, education and welfare services. the social sector because delivering more are difficult to observe; craft ones have Productivity measures the capacity of or improved services with the same inputs observable outcomes but outputs are an economy, industry or organisation to (or the same services with fewer inputs) difficult to observe; and it is difficult to produce goods and services (outputs) using can potentially enhance well-being, all else observe both outputs and outcomes of inputs such as labour and capital (such as being equal (Conway, 2016). coping tasks (Gregory, 1995a). machinery, computer software and land). An important component of Wilson’s typology, and this analysis, It is a quantitative measure of the ratio productivity measurement is quality employ a broad definition of outputs as the work that organisations carry out and Danijela Tavich is a recent Master of Public Policy graduate of Victoria University of Wellington and a the things (goods and services) produced, Research Analyst at the McGuinness Institute and outcomes as the effects of this work

Page 80 – Policy Quarterly – Volume 13, Issue 4 – November 2017 Figure 1: The James Q. Wilson matrix Figure 2: Tasks of MSD, adapted from James Q. Wilson Observable Outcomes Observable Outcomes YES NO YES NO Production Procedural

Administrating Data, anaytics and

S YE S YE Production Procedural Income Support evidence services/ to Seniors policy advice vable Outputs vable Outputs Craft Coping Obser Obser Improving Care and NO NO Craft Coping employment and protection social outcomes services support

Source: Tavich 2017 Source: Gregory, 1995b, pp.172-3 on communities and society at large can cause significant issues for attributing In some instances, a classification that (Gregory and Lonti, 2008). However, outcomes to a single organisation by would allow a task to straddle the lines outputs and outcomes can be further making it hard to unpick the causes and between quadrants might be most broken down according to the level at effects of contributing efforts (Gregory, appropriate. For example, certain tasks which we wish to observe them, a point 1995b). As craft tasks have some can have sub-tasks that are more or less which will be returned to later. observable outcomes, the effects of co- observable. For instance, fraud prevention production are slightly less significant for is a sub-task of administering income Practice measurement, as there is some observable support to seniors (Ministry of Social Performance measurement practice end which can be linked with production Development, 2015). Fraud prevention in the social sector is governed by the technologies (technical knowledge). requires more discretion on the part of requirements of the Public Finance Act However, in the case of coping tasks this officials and thus is more difficult to 1989, which, in contrast to Wilson’s matter can be challenging (ibid.). observe and might not fit within the typology, considers all organisations Nevertheless, much measurement already production quadrant. Nonetheless, other to be based on production-type tasks, exists in the social sector, including for tasks fit relatively neatly within the with observable outputs and outcomes tasks which have complex, long-term typology, with care and protection (Gregory, 1995b; Treasury, 2005). For goals, such as social work (see Ministry of services (social work) a clear coping task. many years the government has tried Social Development, 2016). to shift performance measurement in One-size-fits-all approach not sufficient the social sector to an outcomes focus Key findings Despite the limitations, applying the (Destremau and Wilson, 2016, p.33). Limits of the typology typology to MSD highlighted that much However, it can be difficult to observe There are some clear limitations to task diversity exists within a single social (and, consequently, to measure) some Wilson’s typology when applying it to sector organisation, as shown in Figure outcomes, particularly those that only practice. Indeed, Wilson himself once 2. This indicates that a one-size-fits- become evident in the long term (Alford warned that the typology should be used all approach would be inappropriate 1993; Productivity Commission, 2015). ‘with caution’ (1989, p.159). Significantly, for productivity measures in the sector. Further, co-production can cause the attempt to categorise many complex Inevitably, the measurement of a coping issues for measuring some social sector tasks into a small number of categories task like social work will require a tasks. Co-production entails the – in this case four – is unduly restrictive different approach from that required contribution of people or organisations (Lonti and Gregory, 2007). The exercise of by a production task like processing external to the producing organisation applying Wilson’s approach to the services payments (Gregory, 1995a). In these (such as the target group being regulated, undertaken by MSD (see Figures 2 and more complex cases, moving away from or other public actors) to accomplish 3) illustrated this, and the lines between standard productivity measures and objectives (Alford, 1993). Co-production quadrants in Figure 2 have been dashed towards productivity-type measures, is usually required for more complex to represent that the classifications are not such as cost-effectiveness (i.e. measuring tasks, namely craft and coping tasks, and definitive. the relationship between inputs and final

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 81 New Zealand Productivity Commission Internship: social sector productivity

Figure 3 The reliance of tasks on co-production There is a need for greater clarity about what we mean when we refer to PRODUCTION PROCEDURAL outputs and outcomes, as using the same words interchangeably can create confusion, not least when exploring SELF-PRODUCTION performance measurement issues. Certainly, it will be easier to observe a Less need for learning capacity higher-level overview of an output, and (taller hierachies) an intermediate outcome, than it would be to observe outputs and outcomes according to Wilson’s definitions, and the CRAFT COPING former set of definitions are more consistent with those employed in the CO-PRODUCTION Public Finance Act 1989. Nevertheless, which level of measurement is appropriate Greater need for learning capacity depends on the objective – what it is that (flatter hierachies) we are trying to achieve or learn by measuring the outputs and outcomes of a certain task – as different goals will require Source: Gregory, 1995a, p.174 different levels of detail. Approaching measurement in this way will help to outcomes) and more innovative bespoke suggests), or at a higher-level overview of avoid ‘hitting the target and missing the measures or proxies, might be most this work, such as the number of hours point’ (Bevan and Hood, 2006, p.421). appropriate. worked or the number of children seen in the case of a social worker (see Laking, Conclusions Defining levels of measurement 2008). Similarly, outcomes can be broken Wilson’s typology is a useful tool that The way we define outputs and outcomes down into intermediate outcomes and raises interesting questions about social affects how we might go about measuring ultimate outcomes, the former being sector productivity measurement. them. Wilson’s typology employs a broad more observable, shorter-term goals and In many cases standard productivity definition of outputs and outcomes, the latter being the final desired effect concepts are compatible with social but these can be further broken down of the task, which could take years to sector tasks. In others, however, we may according to the level at which we wish to eventuate and is more consistent with need to apply more innovative, bespoke observe them. Outputs can be observed Wilson’s definitions (Coglianese 2012; methods to capture performance. Above either at the level of the specific day-to- Gregory 1995a). all, it is imperative to define outputs and day activities of individuals (as Wilson outcomes clearly and avoid a one-size-

Table 1 Compatibility of selected tasks with productivity concepts Task type Inputs (labour and capital) Outputs Outcomes Ability to measure quality change

• Production – • Capital (money, computers, • Monetary transfers • Ability of seniors to • Using outputs or Administering income buildings etc.) • Entitlement eligibility maintain independence outcomes support to seniors • Labour intensive assessments and social participation

• Procedural – Data, • Capital (money, computers, • Advice delivered to • Unobservable – • Using outputs analytics, evidence buildings etc.) Minister outcome attribution and policy advice • Labour intensive issues, impacts of work uncertain

• Craft – Improving • Capital (money, computers, • Unobservable – much • Clients moving closer • Using outcomes employment and buildings etc.) discretion by officials, to independence social outcomes • Labour intensive difficult to prescribe (away from benefit support • Attributes of clients etc. outputs dependency)

• Coping – Care and • Capital (money, computers, • Unobservable – much • Unobservable – • Difficult protection services buildings etc.) discretion by officials, attribution problems, • Labour intensive difficult to prescribe impacts of work • Attributes of clients etc. outputs uncertain

Page 82 – Policy Quarterly – Volume 13, Issue 4 – November 2017 fits-all approach. Plainly, the social sector Acknowledgement Patrick Nolan for his time in supervising is too complex to be fully captured by My internship was a great experience. My my work. The views expressed here are my Wilson’s typology, which consequently, as thanks to the Productivity Commission own. he warned, must be used ‘with caution’. for the opportunity, and especially to Dr

References Alford, J. (1993) ‘Towards a new public management model: beyond Hanushek, E.A. and E. Ettema (2015) Defining Productivity in Education: managerialism and its critics’, Australian Journal of Public issues and illustrations, Stanford: Hoover Institution, Stanford Administration, 52 (2), pp.135-48 University Benington, J. (2012) ‘Leadership, governance and public value in an age Hughes, O.E. (2012) Public Management and Administration: an of austerity’, retrieved from, http://www.treasury.govt.nz/publications/ introduction, 4th edn, New York: Palgrave Macmillan media-speeches/guestlectures/pdfs/tgls-benington-slides.pdf Laking, R. (2008) ‘New Zealand public management in action: a case Bevan, G. and C. Hood (2006) ‘Have targets improved performance in the study of organisational performance’, International Public Management English NHS?, BMJ, 332, pp.419-22 Review, 9 (1), pp.76-93 Coglianese, C. (2012) Measuring Regulatory Performance: evaluating the Lonti, Z. and R. Gregory (2007) ‘Accountability or countability? impact of regulation and regulatory policy, expert paper 1, Paris: OECD Performance measurement in the New Zealand public service, Conway, P. (2016) Achieving New Zealand’s Productivity Potential, 1992-2002’, Australian Journal of Public Administration, 66 (4), Wellington: New Zealand Productivity Commission pp.468-4 Destremau, K. and P. Wilson (2016) Defining Social Investment, Kiwi-style, Ministry of Social Development (2015) Strategic Intentions 2015–2019, NZIER public discussion paper, Wellington: NZIER Wellington: Ministry of Social Development Gemmell, N., P. Nolan and G. Scobie (2017) Public Sector Productivity: Ministry of Social Development (2016) Annual Report 2015/2016, quality adjusting sector-level data on New Zealand schools, Wellington: Wellington: Ministry of Social Development New Zealand Productivity Commission Productivity Commisssion (2015) More Effective Social Services, Gregory, R. (1995a) ‘Accountability, responsibility and corruption: Wellington: New Zealand Productivity Commission managing the “public production process”’, in J. Boston, The State Tavich, D. (2017) Social Sector Productivity: a task perspective, Under Contract, Wellington: Bridget Williams Books Wellington: New Zealand Productivity Commission Gregory, R. (1995b) ‘The peculiar tasks of public management: toward Treasury (2005) A Guide to the Public Finance Act, Wellington: New conceptual discrimination’, Australian Journal of Public Administration, Zealand Treasury 54 (2), pp.171-83 Wilson, J.Q. (1989) Bureaucracy: what government agencies do and why Gregory, R. and Z. Lonti (2008) ‘Chasing shadows? Performance they do it, New York: Basic Books measurement of policy advice in New Zealand government departments’, Public Administration, 86 (3), pp.837-56

Policy Quarterly – Volume 13, Issue 4 – November 2017 – Page 83 ENROL NOW FOR 2018.

Study at one of the ADVANCE world’s leading business schools BETTER Victoria Business School holds the triple GOVERNMENT crown of international Gain a qualification in e-government, public management accreditations. or public policy from Victoria—New Zealand’s leading education and development provider in public services. Master of Public Management: Upgrade your personal skills and competencies as a manager and improve your public management practices and your impact. Master of Public Policy: Develop your skills and knowledge in policy analysis, development and evaluation in public and non-government sectors. Master of e-Government: Learn how to successfully manage complex technology-based initiatives in the public sector. Flexible learning options—study full time or continue to work while you study. Courses offered in Wellington.

victoria.ac.nz/sog | [email protected]

FC0367 - VBS Adverts May Updated - SOG Tri 2_A4_v3.indd 1 1/11/17 1:33 PM