How Government Experts Self-Sabotage

Front matter

EXPERTS SELF-SABOTAGE

CHRISTIANE GERBLINGER

HOW GOVERNMENT EXPERTS SELF-SABOTAGE

Contents

Abbreviations vii Acknowledgements ix List of Tables xi Preface xiii 1. Introduction 1 Locating the phenomenon of rebuffed advice 1 Case studies and book structure 6 The language of policy advice 9 Primary materials and methodology 14 Do the rebuffed know they are a phenomenon? 18 Contribution and rationale 25 2. Strategies of Impersonality: Constructing a Framework for the Rebuffed 27 İntroduction 27 The text 32 The micro-context 38 The macro-context 46 3. Knowing What Not to Know: Advice on South Australia’s Blackout and the Role of Renewable Energy 57 Introduction 57 Discussion of departmental advice 60 The text 77 The micro-context 91 The macro-context 109 4. Excess of Objectivity: Australian Intelligence Assessments of Iraq’s Weapons of Mass Destruction 127 Introduction 127 Discussion of Australian intelligence assessments 131 The text 146

The micro-context 158 The macro-context 171 5. The Language of the Unrebuffed 203 Contemporaneous comparisons 203 Triangulation 213 6. Conclusion 229 Bibliography 245 Index 287

IAEA

Preface

This book is about how a particular group of experts communicates its knowledge. These experts did not come to my attention by way of a single sensational event. Indeed, while fascinating, they mostly fly under the radar of public consciousness. I did not chance on them by fortuitous accident, nor did I discover them through diligent preliminary research. I already knew them because I had worked with them for more than 15 years.

During my time with the Australian Public Service, I worked on national security issues, particularly those related to intelligence; I served as a policy adviser across a range of portfolios, such as counter-proliferation, energy, health and rural policy; and I wrote a seemingly endless stream of speeches for political and policy actors in an economic portfolio. In one way or another, many of these policy settings nudge the case studies examined in this book, yet I did not work on any of them myself. As such, I cannot claim to offer direct ethnographic insights. Spending a significant portion of my working life immersed in the language and culture of this special type of knowledge production meant, however, that I was able to weed out secondary material that could lend external validation to my own experiences inside the administration. But, more than that, my experiences, observations and conversations with a sizeable number of colleagues gave me confidence that being rebuffed was not an isolated incident and that the language of policy advisers was a kind of practised, hemmed-in expression of expertise.

When researching and drafting this book, it quickly became apparent that there was more than one type of language that ended in rebuffal; there were several. I have identified three types here, but there are probably more, in these and other policy settings, each of them tailored to their circumstances. Future researchers are likely to find a trove of typologies xiii xiv from across international public policy and administrative fields. Beyond my group of experts, there may be other areas of expertise that also communicate in ways that end in some form of rejection. Climate change science, the promotion of genetically modified food, vaccine advocacy, skin cancer education and anti-gun campaigning come to mind as immediate candidates. Are they, like the rebuffed expert policy advisers in this book, also implicated in their own rejection? There will be other, less obvious and therefore probably more interesting areas to uncover for researchers across several different academic disciplines, particularly those interested in knowledge production and communication, such as science and technology studies, policy and governance studies, behavioural science, anthropology, sociolinguistics and organisational sociology.

As I finalised this book halfway through 2021, the world was still traversing a global pandemic. Many of the successes associated with stemming the spread of Covid-19 may well have been due to receptive governments listening to cogent experts. But I suggest that any post-pandemic world— if such can be said to lie ahead—is likely to feature similar challenges to those described here. I therefore challenge experts themselves—whether academic or professional—to find instances of “rebuffal” in their own areas of work. These readers can take the rebuffed and use them to reflect on others’ practices or, indeed, their own. But research and professional (self-) reflection imply that observing the rebuffed is more or less the domain of, well, experts. I do not think this is where the rebuffed should stay. I want them to be watched, chided and encouraged by those who are ultimately affected by their language: ‘everyday’ readers or the public. Despite political rhetoric about including the public in policymaking, publics are mostly sidelined or ignored when final decisions are made. But publics do try to become involved, such as through submissions to parliament or policy reviews, or simply by writing letters to their government representatives. For example, I was told a few years ago that, during a nine-month period, one portfolio received nearly 18,000 separate items of correspondence, which predominantly included letters from the public. Another portfolio, which regularly publishes its own research, received more than 1 million unique online page views for one of its pieces, while another annual publication can receive more than 4 million views. These numbers point to the existence of intense interest in matters of governance well beyond the media, business and the academy.

PREFACE

In an environment that values them in word but not necessarily in deed, it should not be difficult for interested publics to understand the advice that goes to governments as the ostensible base (or not) of policy decisions. This book is not a manual for how to make it easier. It is, as Foucault puts it, ‘a challenge directed to what is’ (1991: 84). In challenging what is—no matter who is in power—it also seeks to elevate the critical contribution that courageous policy advisers could make towards helping publics determine what is in their interest. If it succeeds in doing so, it will be up to those who agree to take the next steps.

XV

Introduction

Locating the phenomenon of rebuffed advice

After official policy advice to governments is publicly released, governments are often accused of ignoring or rejecting their experts. Commonly represented as politicisation, this depiction is superficial. Digging deeper, is there something about the official advice itself that makes it easy to ignore? This book asks: does the expert advice of policy officials feature characteristics that invite its government audience to overlook or misread it? To answer this, I critically examine official policy advice and find the language of the rebuffed: expert advisers reluctant to disclose what they know to accommodate political circumstances.

Perhaps one of the more infamous international cases of advice thus rebuffed is the United Kingdom’s mid-1990s bovine spongiform encephalopathy (BSE) crisis, in which Ministry of Agriculture, Fisheries and Food advice to slaughter affected animals was not adequately disclosed by civil service advisers (Jasanoff 1997) and was ignored by their minister (Millstone and van Zwanenberg 2001). In a case involving a district government in Indonesia in 2003, advice related to ecological preservation was ignored because government actors—perhaps not always unreasonably— believe that they understand the solutions better than the researcher (Dharmawan et al. 2017: 409). In the United States in 2018, the White House ignored expert advice from the Department of Health and Human Services that separations of children and parents at the border would do psychological harm (Colby 2018) and violate migrants’ civil and human rights (Shuchart 2018).

In Australia—a Westminster-style jurisdiction from which this book draws its cases—the phenomenon appears to be growing in frequency. Here, we have seen the commissioning of two major reviews of the tax system (in 2008 and 2015), with both largely swept under the carpet and cast as failures (Benson 2018b; Hodgson 2015, respectively). In 2009, advice about the risks surrounding the so-called Home Insulation Program was allegedly ignored, with the program terminated when it led to the deaths of four young installers (Padula 2017; McGhee and McKinnon 2018). The subsequent government called a royal commission into the program, with the new prime minister then ignoring the advice of his government solicitors not to break Cabinet confidentiality by forcing the previous government to release related Cabinet papers (McGhee and McKinnon 2018). In 2018, the Australian Government also rebuffed advice by one of its security agencies to return refugees to the island detention centres from which they had been medically evacuated on completion of treatment on the Australian mainland so it could continue to claim the refugees posed a security threat (Middleton 2019). In 2019, advice recommending allocation of sports funding to regional Australian communities was ignored in favour of the responsible minister’s favoured electorates (ANAO 2020). This is by no means an exhaustive list, but each example is symptomatic of the phenomenon of rebuffal.

This book’s focus is on where the effectiveness of such advice comes to be established: during expert policy advisers’ construction and communication of advice, and its reception and use by government ministers. Stepping through these junctures, I suggest, demonstrates the argumentation of policy advisers is almost exclusively informationfocused, intentionally incomplete or so neutral as to evade stable meaning. While this is at least outwardly due to advisers’ commitment to notions of objectivity and political responsiveness, it also arises from knowing they could be rebuffed, because they fear clarity will damn them or because they recoil from exercising judgement in the face of risk or uncertainty— all of which is cloaked in the language of evidence to maintain an appearance of responsive expertise. This is not to say that all evidence is powerless, but a language of evidence is incomplete and cannot persuade all by itself. It also casts in rather a different light the regular refrain of evidence-based, evidence-informed or evidence-inspired policy among public policy circles.

It is difficult to imagine that civil servants have not noticed this is how their advice is constructed and used. Indeed, Boswell and Corbett (2015) describe a cynicism and even stoicism among policy elites that suggest an intense awareness of, and frustration with, the way policy is made. Yet, the status quo largely prevails, which implies that maintaining this style of advising makes them participants in their own rebuffal. With this, they have essentially institutionalised the provision of unproductive advice. One might go so far as to suggest they are complicit in their own marginalisation, as if advice were being written with an underlying invitation to ‘rebuff me if you need to’. This does not simply contribute to the erosion of their integrity as policy experts, it also weakens policy advice as a tool with which to legitimately govern democracy. That is, if public institutions, whose role it is to furnish governments with thorough, unmistakable and relevant information, are not doing so, there is no record and, thus, no accountability for tracing how political decisions are made. The result is effectively the same as concealment—that is, despite proceeding in relatively plain view, advice is articulated in a language that tries its best to blend into the background. Advice of a sort is being given, ‘facts’ are being provided and evaluative and accountability measures are being observed, but meaning and views are invisible and cannot be critically interpreted. What one ends up with is policy advice as a political token or plaything, symbolising facts and evidence but uninterpretable by most—and, therefore, paradoxically, infinitely interpretable.

My central argument—that the language of such advice evades stable meaning and diminishes the democratic right of citizens to scrutinise the work of government—is on the grandiose side, certainly. If we want to determine whether expert advisers are holding back their knowledge, we need to be able to see the policy stories behind some of the examples above. But gaining a detailed understanding of the daily workings of organisations, particularly how their employees communicate their knowledge to each other and their superiors, can be a ‘problematic enterprise’ (Alvesson 2003: 13). This is perhaps especially the case in relation to policy advisers inside civil services. As Wagenaar suggests, ‘we know surprisingly little about what the work of public administrators entails’ (2004: 643). Rhodes concurs, observing they ‘have been studied rarely (2005: 5). Wagenaar continues that, on the one hand, critical appraisals are ‘remarkably silent on the actual activities of administrators’, while on the other, relevant textbooks ‘largely ignore the sociology of administrative work’ (2004: 643).

To be sure, there are mountains of scholarly work on policymaking and there are, indeed, some studies that provide insights into how policy advisers in service to governments communicate in some of their day-to-day dealings (Heclo and Wildavsky 1974; Mackie 2015; Stevens 2011). There is also research, largely in the interpretive tradition, that has examined the communications of policy advisers, but this broadly occurs at a theoretical level (Schmidt 2008), where policy and political communication are grouped and not viewed as separate entities (Hajer 1993), or to demonstrate how policy analysts (that is, analysts in the field of policy studies) should argue when making their case (Fischer 2003). When the impact of expert advice on government decision-making is considered, the focus is generally on scientific advisers (Boswell 2009; Cairney and Oliver 2017; Moore 2017; Wynne 1989). Overall, scholarly writing on policy tends to be quite targeted, such as on policy advising during the digital governance era (Marando and Craft 2017) or the public profiles of senior civil servants (Grube 2014a, 2015), or on the degrees to which evidence can or should be incorporated into policy (Head and Di Francesco 2019; Howlett and Wellstead 2011; Nair and Howlett 2017; Sanderson 2009).

All in all, gaining a detailed understanding of how official policy workers routinely communicate their knowledge to governments has largely been an under-researched area, at least in the English-speaking world (De Vries et al. 2010; Eriksson 2016; Hoppe and Jeliazkova 2006). This relative absence of study is due partly to the difficulty of accessing government policy advisers (Williams 2010), and to the fact that most of their advice to governments remains confidential. But it also stems from the fact that most of their advice to governments is being increasingly withdrawn from public accessibility, with Goldfarb (2009: 58) arguing ‘the government has claimed more privacy for its actions but has provided the citizenry with less. While Goldfarb writes in the US context, a similar situation can be observed in Australia, where requests for government information are on the rise and compliance with those requests by government departments is down (Australian Government 2019: 14).

Of course, the officials of public administrations and public or civil services serving executive governments are regularly reviewed or on view—for example, in royal commissions and parliamentary inquiries, during audits or parliamentary hearings where they account for outputs like policy advice. Yet most of these appearances can be characterised more by confining information than by its clear presentation. Although some inquiry processes have yielded insights into culture, becoming more cognisant of the way policy advice to governments is argued by policy workers at the coalface, as it were, does not appear to be an easily achievable task. Indeed, it is practically a given that government advisers advise impenetrably. But the popular narrative of impulsive governments ignoring the well-reasoned expertise of their policy workers only scratches the surface. Much more is at work here, because the language of official policy advisers enables certain affordances, serving as a kind of springboard from which to launch political manoeuvres. That is, it leaves political actors free to not just ignore but also reinterpret the offerings of their policy advisers as greenlighting any and all policy directions because those offerings are expressed in ways that render meaning malleable.

Given the central position of public accountability in democratic governments, one might reason that understanding how policy advisers to such governments go about the business of communicating their policy expertise should not be an arduous or prohibitive process. Further, in an ever more contestable policy environment in which citizens themselves are assigned only restricted roles in the decision-making process, they should be able to observe better than they do now how expert policy knowledge is produced and argued on their behalf. They should be able to know why and how some advice succeeds in convincing its political audience. They might also reasonably ask why and how it does not, particularly when one of its aims is to present the best available evidence. Indeed, instances of policy advice failing to convince may be particularly germane to those very citizens whose interests are thought to be at least nominally represented in that advice. If citizens cannot do any of these things, control of the government is not in the hands of the governed (Gruber 1987: 1). Further, if policy advice is being constructed in ways that curb this control, advisers risk appearing complicit in neglecting the democratic needs of citizens.

When public institutions furnish governments with such advice, interested publics are effectively excluded from gaining accurate impressions of how decisions are made on their behalf. Following Dewey, much as others have in staking out claims for more open, cooperative forms of deliberation (Forester 2012; Minteer 2005; Sanderson 2009; Wagenaar 2004), I argue that only in being able to judge and assess policy rationales can they assemble a picture of what is in their interest. Yet, far from helping assemble that picture, the language of the rebuffed seems designed to break it up.

Case studies and book structure

This book examines three Australian case studies, which are drawn from three streams of written knowledge production across the Australian Public Service (APS): the economy, energy and the environment, and national security. They have been chosen due to their originally confidential nature but eventual public availability by means of freedom of information (FOI) requests and a parliamentary inquiry. Examining advice destined predominantly for ministers’ eyes makes it possible to observe, as closely as publicly possible and with some qualifications in mind, how policy advisers communicate when executing one of their core roles: supporting the government of the day to deliver its policy agendas and priorities. The first case study—on policy advice regarding the taxation of investment properties given in the leadup to an election—is discussed as a kind of abridged blueprint to help us construct a framework with which to interrogate the language of the rebuffed and the conditions under which it is produced. Two further and more exhaustive case studies drawn from distinctly different policy areas across the past two decades are then juxtaposed using this framework: advice about a 2016 statewide electricity blackout event blamed on renewable energy and Australia’s intelligence assessments in the leadup to the 2003 Iraq war. Each will form one chapter, with chapters on the blackout and Iraq making up a major component of this book, and each is divided into smaller sections dealing with aspects of my framework.

A brief note on the selection of these cases is in order. Although the simple fact of their availability (the first two by FOL, the third by parliamentary review) played a role in their selection, my ultimate choices were based on the fact they brought with them the visibility of policy advice and political reaction. This was essential given it would be difficult, if not impossible, to be rebuffed without first offering something. By seeing both offer and retort, it is possible to distinguish a performative pattern in which advice has the repeated perlocutionary effect of either irritating or polarising its political audience. Seeing both offer and retort, therefore, makes it possible to identify that advice has been rebuffed, rather than having simply been considered but set aside. Moreover, to understand how advice is rebuffed as an act in which advisers participate, I chose advice that grappled with uncomfortable knowledge to observe how policy advisers communicated when they anticipated the likelihood of a negative political reaction. This, I hope to show by way of my three case studies, is the environment in which rebuffed government policy advice is generated. To test my findings, Chapter 5 discusses cases of politically accepted advice as the language of the “unrebuffed”.

Chapter 2 briefly details my first case study on the Australian Treasury's advice about income tax deductions facilitated by investing in property (known as ‘negative gearing’). Provided to the government in 2016, this advice followed an early election announcement by the opposition Australian Labor Party (ALP) of a policy to restrict how homeowning property investors could minimise their income tax burden. The conservative Coalition government called this the most ‘destructive policy ever proposed’, which would lead to ‘mum and dad investors’ being forced out of the housing investment market. Emphasising the financial mayhem this would wreak, the government also projected the ALP” policy change would reap a mere $600 million over four years. Two years later, in 2018, an FOI request by a news outlet seeking Treasury's 2016 advice on the issue found the agency had confirmed the ALP’s claims that its proposed reforms would ‘increase revenue in the long run between $3.4 and $3.9 billion a year’ and ‘the limit on negative gearing for established property would not differentiate between more or less wealthy investors’. At first glance, Treasury’s advice was ignored. By carefully avoiding the political context and practising ‘strategies of impersonality (Porter 1995: 229), policy advisers constructed their advice in a way that would protect them from key political contexts. The political retort to this advice was the treasurer entrenching the government's existing policy further.

Given the gap in research on the confidential communications of public policy advisers, I construct a new framework from this short case study with which to trace the conditions under which policy advisers construct and communicate their advice, and how government ministers receive and use that advice. Briefly, this framework will unpeel three layers that encase policy advice: the rhetoric of its text, the organisational aspects of its micro-context and the international comparisons suggested by its macro-context—each of them revealing an aspect of policy language. For example, by closely analysing the text itself, I will observe the rhetorical construction of advice and its relationship to time and timing, context, framing and self-perception. The micro-context focuses on the institutional, legislative, bureaucratic and cultural effects and burdens on advisers’ language, drawing much of its supporting literature from organisational and administrative studies. The macro-context connects to a much larger sphere of ideas around bureaucracy, objectivity, rationality and evidence, both in Australia and internationally. By interrogating case studies with the help of my tripartite framework, I hope to demonstrate that policy language is more than words on a page; it is a complex arrangement of tracks from which one can deduce the settings under which this expert community weakens its own credibility.

The second case study, detailed in Chapter 3, chronicles advice released under FOI by the departments of the Prime Minister and Cabinet and Environment and Energy after a statewide power blackout in South Australia in late 2016. To accommodate misleading government rhetoric about wind energy as a béte noire and cause of the blackout, this advice sidestepped any mention of renewable energy and climate change, thereby expunging discord and uncertainty (Rayner 2012) and giving the appearance of certainty and solid evidence. It also gave the impression of being both objective and responsive to the government by frequently issuing damage status updates, but little else. My framework here makes possible some critical discoveries. For example, one finds a widespread anticipatory compliance—a dynamic that, because government actors did not ask policy experts about deeper arguments and rationales, saw policy experts not offer them or discuss them among themselves (at least on paper). Policy advisers became entirely transactional, functioning as vendors of rudimentary information, perhaps because the only advice they could provide objectively and responsively was on the status of storm damage. Official advice was intentionally ambiguous, deliberately one-dimensional and unknowing even while political hyperbole was at its most strident; it spoke a language of ‘knowing what not to know

Chapter 4 discusses my third case study, which considers Australian intelligence assessments of Iraq’s weapons of mass destruction (WMD) between 2002 and 2003. The assessments by Australia’s then Office of National Assessments (ONA) and Defence Intelligence Organisation (DIO) represent political responsiveness and detached neutrality when they are pursued separately. Within the text, the DIO presented as highly objective in its repeated acknowledgements of uncertainty and gaps in the available evidence, and revealed a lack of consensus, offering an almost textbook study of how to present evidence impartially. John Howard, Australia’s prime minister at the time, turned this element of doubt into strength in that it helped him demerit mounting arguments that absolute proof was needed before invading Iraq. The ONA, on the other hand, became increasingly responsive by falling into line with the dominant political discourse. Even while neither was used by the government and both were later blamed as failures of intelligence, their inconsistency and ‘excess of objectivity’ (Sarewitz 2004: 388) channelled an uncertainty that contributed to a highly effective political strategy.

My penultimate chapter will discuss some contemporaneous Australian examples of unrebuffed policy advice: amendments to Australian citizenship, marriage equality and the legal basis for military action in Iraq. Here, I consider whether the policy language conveyed features associated with successful advisers, such as openness to adaptation, a culture of curiosity and comfort with uncertainty (Luetjens and ’t Hart 2019: 28). Is this the language of the unrebuffed? Or were additional, or entirely different, factors at play here? These examples, which follow a document-led approach, are juxtaposed with two short comparisons using a different methodology to test my findings. Following Rhodess (2018) suggestion to take up ‘bricolage as part of a menu of tools when undertaking qualitative research, I will briefly consider interviews I conducted with Australian policy advisers reflecting on the introduction of a consumption tax in the late 1990s, as well as Dutch policy actors who worked on prison reforms throughout much of the 2000s. These show us another style of policy advising with which to test my hypothesis. A conclusion will consider how both the public and the public interest are excluded by the type of policy advice examined in this book. Envisioned by pragmatist John Dewey (1946: 208, 218) as denoting either oligarchy or soliloquy, such advice helps normalise policy advisory environments in which civil servants neglect the democratic need of the governed to understand why they are governed in particular ways and not others.

The language of policy advice

Let us now consider why I have chosen language as the locus from which to study rebuffed policy advice. Policy language is not easy to pin down. Like the rhetoric of politics and expertise, the language of policy advisers seeks legitimation by being agreed to. In a political context, the rhetoric of government leaders is said to be both central to ‘the practice of modern government itself’ and integral to ‘the way we are governed’ (Grube 2013: 3), while ‘establishing oneself as an expert entails a judgment from an audience that is contingent on a rhetorical effort’ (Hartelius 2008: 3). Although the rhetorics of politics and of expertise contrive to be noticed, the language of policy advice appears to aspire to authority through dullness. Rather than provoking inspiration or engagement, its flatness seems to discourage this. This does not mean policy language is less premeditated than political rhetoric, nor that it contains only univocal content. As Majone (1989: 1) has suggested, policy is ‘made of language’ and its contents can be understood as expressions of organisational realities (Alvesson and Karreman 2000), such as institutional and subcultural constraints, expectations and circumstances.

Underpinning this view is an understanding of language as ‘more than reflectlingl what we take to be reality’ (Fischer and Gottweis 2012: 8)— not simply as ‘assertions about the world of “things” (Shapiro 1981: 26), but also as a particular kind of argumentation determined by one’s world view. How policy advice is argued and argued about is at the crux of any endeavour to examine its influence—or lack thereof. The idea that language is a construction that has much to tell us about how its users view the world also sheds light, in the institutional settings I will observe, on how its users think their political audience would like to view the world. As Michael Shapiro continues: ‘Insofar as we do not invent language or meanings in our typical speech, we end up buying into a model of political relations in almost everything we say without making a prior, deliberative evaluation of the purchasing decision’ (1981: 231).

Following Shapiro, Frank Fischer (2003: 43) concludes that, in the world of policy, ‘given that the languages of politics inscribe the meanings of a policy problem, public policy is not expressed in words, it is literally “constructed” through the language(s) in which it is described’.

David Farmers (1995: 19) work on the language of public administration offers the view that we are condemned to see the world through the eyeglasses of a language. We cannot escape from the lens of language. We can see the world only in terms of some conceptual system, some perspective. Public administration is no different in this respect.

There are thus powerfully influential, premade grooves along which the language of policy advisers within public administrations comes to be conceived of, delivered and heard. In addition to the grooves and frames provided by politics, policy language is also shaped by the educational backgrounds of its speakers and by the organisational enculturation (Srivastava and Goldberg 2017) that takes place while working inside government and policy institutions. These may not be as separate as they sound. When policy analysts or advisers come to public policy employment, they are often equipped with tertiary qualifications from the social sciences and will usually already speak at least a rudimentary version of the language of evidence or rationality (Dryzek 1993; Fischer 2003; McCloskey 1994).’ When they enter the work environment, the language that is spoken and rewarded there takes great pains to express itself as neutrally and free of value or emotion as possible, underpinned by the vernacular of social science methodologies (Porter 1995; Stone 1997), favouring the epistemological status of statements that sound like science over those that sound like value judgements (Farmer 1995). This depiction makes policy workers sound like robots. They are not, and we will see some instances to the contrary. However, the pervasive influence of precepts and organisational ethos on policy knowledge and construction should not be underestimated.