Introduction

‘Why does not the government do something about this?’ is an oft-voiced complaint in legislatures, TV debates, opinion pieces, letters to the editor and cyberspace. Indeed, often such sentiments are retrospective: ‘why didn’t the government or any of its agencies do something about this?’, whether it be recognising that a terrorist cell had been operating undetected until it struck, failing to halt the spread of a pandemic, systematically addressing creeping social crises such as domestic violence or acting on warnings about looming economic recession. In principle, we should be able to turn to the policy sciences to help understand the nature of such ‘doing nothing’. Their locus has been broad, ranging from classic government-centric approaches of those such as Dye (2012: 12) who argue that public policy is ‘whatever governments choose to do or not do’, through to Colebatch and Hoppe (2018) who suggest that policy can also be the structured interactions of governance networks, as well as the process of problematising issues. Yet a common denominator in the policy sciences, regardless of our assumptions about what constitutes ‘policy’, is a bias towards the study of policy activity, and a substantial neglect of policy inactivity.

This gap is a significant one. In contemporary liberal democratic societies with 24-h news cycles, the rise of social media and citizen activism, there are strong expectations that governments (as well as public organisations and governance networks, when faced with imperfect, problematic and undesirable social conditions) will take some sort of action—even if only to create markets or support community organisations doing the actual work. Doing nothing in the face of ‘self-evident’ needs is thus easily framed as evidence of governmental negligence, irresponsibility or ineptness. Economists even produce COI (cost of inaction) indices to prove the point (Anand et al. 2012), exemplified in a report from the Organisation for Economic Cooperation and Development (OECD) on the health impact costs of inaction on air pollution (Alberini et al. 2016). And yet, on the other side of the equation ‘doing nothing’ can be considered good policy in particular circumstances. For example, in dealing with intractable controversies, policymakers are told to wait for the conflict to ‘ripen’ before attempting to resolve it (Cantekin 2016) and in dealing with crises, they are advised to avoid knee-jerk responses (Lodge and Hood 2002).

Public policy analysis needs to develop a more robust understanding and more sophisticated evaluation of inaction. This paper provides an initial roadmap for doing so, while acting as a primer for further reflection and research. It briefly examines the relative neglect of policy inaction within the policy sciences. It then provides a working definition and typology of policy inaction. The main part of the paper presents five forms of policy inaction, followed by analysis of core inaction drivers at four policy-making loci: individuals (coping behaviour), public organisations (information pathologies), governments (agenda control and protection) and networks (non-coordination and lack of feasibility). We conclude by digging deeply into the continuing challenges that inaction poses for the policy sciences in the twenty-first century, followed by the identification of methodological strategies for placing inaction back on the agenda of policy scholars.

Policy sciences and the study of inaction

The policy sciences that emerged after the Second World War were focused on societal betterment, none more so than in the work of Lasswell (1956, 1971) and his seminal work identifying the seven stages through which societies should harness knowledge to produce and continually reflect on public policies. In essence, the intellectual focus of the field has always been on the design, implementation and evaluation of policies as purposeful government interventions in social processes that need to be carefully calibrated before being put into practice or changed in the light of experience (DeLeon 1998; Peters 2019). Accordingly, the study of public policy over the past half-century and more has seen thousands of studies describing, explaining and evaluating policy interventions across a multitude of policy areas.

This intellectual agenda remains persuasive to many, significant critical voices notwithstanding (Fischer 2003; Colebatch 2006). Key contemporary understandings of policy processes are also tilted towards an understanding of policy as intervention. For example, studies of policy instruments quite naturally are premised on the idea of governments ‘doing stuff’; they are not programmed to acknowledge the existence of abstinence, delay, waiting and other ‘tools’ of purposeful inaction (Salamon 2002; Hood and Margetts 2007; Howlett 2019; Peters 2019). The field’s latest trend, the study of ‘nudging’ and other forms of ‘behavioural public policy’, fits the mould in its emphasis on the design of micro-interventions conducing individuals towards, e.g. more cautious driving, taking precautionary health measures, and filing tax returns on time to generate desired macro-level policy outcomes (John 2018).

At one time, more attention was paid to policy inaction, particularly as a political phenomenon. In the 1960s and 1970s, some political scientists started wondering what power was at work in preventing certain dissatisfaction, grievances and desires among sections of the public, from being converted into public policies and programmes. Bachrach and Baratz (1970) tackled the frustration of anti-poverty campaigns in Baltimore, while Crenson (1971) studied policy inaction in the face of air pollution in Gary, Indiana and East Chicago. These scholars caught the mood of the era by identifying the phenomenon of ‘non-decisions’ and ‘non-issues’ to encapsulate systemic biases such as rules, procedures and values, which filter out grievances and calls for governmental action. Cobb and Elder’s (1971) agenda-building model emphasised the importance of ‘gatekeeping’, demonstrating that certain actors are able to determine what issues or proposals will not reach the political/governmental agenda.

Much of this type of analysis drew on Edelman (1964, 1977) who argued that positive, ‘feel-good’ political language is used as a palliative to mask inequalities and cultivate regime legitimacy. Lukes (1974) consolidated the fruits of the era in his famous ‘three faces of power’: the power to influence political decisions, thus channelling policy action in some rather than other directions; the power to control the political agenda and thus to block issues from achieving priority attention; and the power to frame the public discourse and thus to stop certain social conditions from being recognised as ‘problems’ that necessitate policy interventions (see also Lukes 2005). Many years later, Cobb and Ross (1997) followed with the concept of ‘agenda denial’, where elites use cultural strategies to avoid, marginalise and redefine issues.

These contributions from critical political science temporarily trained the attention of policy analysts to ‘silences’ in policy discourse and ‘inaction’ in policy design as significant phenomena amenable to analysis and evaluation. The momentum of the early 1970s faded, however, and the scholarly mainstream resorted to focusing on programme design, policy instruments, service delivery—the ‘tools of government’—and on the actors and networks driving these activities. Perhaps the ontological and epistemological puzzles raised by notions such as ‘non-decisions’, ‘non-events’, ‘agenda denial’ and ‘silences’ were deemed too hard to handle. Or perhaps the terrain of counterfactual reasoning that opens up as soon as one takes inaction seriously was deemed too treacherous (Polsby 1982; Tetlock and Belkin 1995; Rosenfeld 2011). Neither has the advent of ‘post-positivist’, interpretive policy science (Hajer and Wagenaar 2003; Fischer and Gottweiss 2012) served to systematically re-focus the agenda of the policy sciences to include governments not doing things.

There are some exceptions. Policy stasis and resistance to change certainly runs through the highly influential frameworks of path dependency (Pierson 2005) punctuated equilibrium theory (Baumgartner and Jones 2009) and the advocacy coalitions framework (Sabatier and Jenkins-Smith 1993). Inaction per se is not the analytical focus in any of these frameworks, although some of the insights therein help inform our analysis. De Vries (2010) also provides a substantive and extended examination of one type of agenda gap, i.e. neglect in policy-making. He focuses particularly on the priorities of different generations, and some issues being neglected, only to emerge later as societal values change. Other exceptions are Newman (2005), where the focus is on government retreating from its interventionist traditional roles, allowing itself to be replaced in part by ‘empowered’ or ‘responsibilised’ citizens or by ‘deregulated’ and globalised (quasi-)market mechanisms (Peeters 2013; Strange 1996). That said, the focus in these studies is generally on the gradual remaking of state–civil society relations rather than inaction per se (Moran 2002; Bell and Hindmoor 2014).

The nature of policy inaction

Whether a government is actually (and literally) ‘doing nothing’ can be a point of argument. In the face of a humanitarian emergency overseas, media and political opponents may accuse the government of ‘doing nothing’ to help resolve the crisis, yet government may say it is ‘doing everything’ it realistically can.

How, therefore, can we identify instances of policy inaction in a politically as well as analytically meaningful manner? Clearly, policymakers, governments, organisations and networks are inclined to select which issues are of potential relevance to them, and marginalise or exclude those on which they do not wish to devote their attention or resources. If policymakers inevitably don’t do things they theoretically could do in a particular instance or on an ongoing basis, how do we single out a universe of cases of inaction? We are aided here by Mahoney and Goertz (2004) who articulate the ‘possibility principle’ as a means of choosing negative cases for comparative research. In terms of inaction, this means that for any given social issue, once we discard extreme and unprecedented policy scenarios we are left with a range of possible courses of action that could realistically be taken. ‘Doing nothing’ can be one such possibility, defined in part as inaction in relation to other possible course of action within the realm of possibilities. A simple example of one form of inaction helps illustrate (see Bauer and Knill 2014). If a government dismantles social service levels by a simple process of neglecting to adjust in line with inflation or higher welfare demand, a plausible course of action within the realm of possibilities would be to maintain existing levels by adjusting for inflation/heightened demand. By contrast, an irrelevant form of action where there is no prior evidence of even a remote possibility would be immediate termination of all social services without any warning to citizens.

We define policy inaction, therefore, as an instance and/or pattern of non-intervention by individual policymakers, public organisations, governments or policy networks in relation to an issue within and potentially within their jurisdiction and where other plausible potential policy interventions did not take place. There are several advantages to this definition that help us address, at least for heuristic purposes, the methodological challenges of policy inaction. In the first instance, it is comprehensive enough to capture inaction that is not only inadvertent, but also purposeful. While we have some sympathy with the argument that the latter constitutes ‘action’, it is also ‘inaction’ because it constitutes an instance or pattern of policymakers not intervening: not committing resources, not proposing bills, not authorising or undertaking executive action where such interventions could have been undertaken. Indeed, as we will see, by recognising that some forms of inaction are purposeful and even strategic, we get to the heart of much of the power, ideology and political management of policy agendas involved in actively creating a distance between policymakers and those that might expect their involvement.

Second, it highlights the fact that inaction can apply not just to ‘the government’ but also apply to a whole variety of public organisations and policy networks.

Third, the definition also prompts us to focus on inaction that is meaningful in a particular policy context because it emphasises a problem within and potentially the remit of policymakers. We consider ‘jurisdiction’ not simply in the technical sense of formal roles and responsibilities, but also in terms of what is considered to be its sphere of responsibility by politically relevant actors. If, for example, a government does not have the power to prevent a citizen being extradited for trial overseas, and yet there is significant public pressure for it to intervene regardless, then ‘doing nothing’ becomes part of meaningful political debate (about plausible alternatives) in that jurisdiction. One alternative in such a case—under the possibility principle—would be for the government to ‘do something’ such using diplomatic channels to influence appropriate authorities.

Fourth, our definition includes the term ‘potential’ to recognise that what falls within a particular authority’s jurisdiction can be a matter of ambiguity and dispute and may even be a matter for the courts. When the question of who or what is responsible for (in)action becomes a matter of political or legal disputes, we can observe and note that in our analysis. Also, by recognising the existence of grey areas and ‘potential’, we can ensure that such matters are embedded in our analysis.

Finally, our approach is also quite compatible with policy being disaggregated into different elements of action and inaction, along the lines suggested by (among others) Howlett and Cashore (2009). In the area of unemployment policy, for example, a government could be seen ‘actively’ pursing reform of welfare benefits and actively creating (re)training programmes, but ‘inactive’ when it comes to undertaking job creation programmes or refraining from using particular types of policy instruments such as employer subsidies. Inaction may also be evident at different stages of the policy cycle, from agenda setting through to implementation and evaluation. Lipsky (1980), for example, in his classic work on street-level bureaucracy describes continual patterns of non-enforcement of rules, simply because they are often so extensive, complex and contradictory, that service provision would be unable to function without such discretion being used.

Types of inaction

To aid researchers who want to explore the nature, cause and consequences of inaction, Table 1 presents a fivefold typology of policy inaction. Most forms of inaction will be readily placed in one of these categories, although there are several variations of each, as will be discussed. As analysts, we should embrace such complexity and use the ‘art and craft’ of informed judgement (Wildavsky 1987). Cutting across all are insights from the non-decision, mobilisation of bias, agenda setting and policy stasis literature that ‘doing nothing’ happens in particular institutional contexts, where biases are reproduced and potentially critical issues are not considered worthy of attention.

Table 1 A typology of policy inaction
Full size table

We need to preface our typology by recognising that portrayal of inaction as calculated, reluctant or inadvertent is often a contestable issue, rather than a matter of unambiguous fact. We might be tempted, for example, to view organisational biases and blind spots (to emerging issues and threats) as inadvertent. Yet as Bach and Wegrich (2019) argue, the marginalisation of some issues is intentionally boundedly rational for public organisations working in a political context and striving to achieve their goals amid multiple constraints. Similarly, we might consider that a proposed government reform effectively blocked by a powerful lobby group would be ‘reluctant’, but it would also be ‘intentional’ in the sense that there would be a pragmatic rationale (in effect a boundedly rational choice) that inaction would in political terms be the line of least resistance. Framing the motives of inaction is a contestable issue.

Each form of inaction is outlined briefly below and as Table 1 suggests, each can be understood in terms of potential ‘drivers’. Our rationale here is not to offer testable hypotheses in line with some recent debates on causation within the social sciences. Rather, it is to highlight what we consider to be important plausible courses of explanation for each form of inaction (see Daigneault and Béland 2014; Dowding 2016; Hay 2017). As Parsons (2007) argues, ‘first cut’ explanations direct our gaze to what we want to know more about, while recognising that further and more detailed methods and analysis are less important at this initial stage.

Type I: Calculated inaction

Inaction can be deliberate, strategic and tactical. There may be a danger of rushing in before an issue has sufficiently matured. At times it may be better for decision makers to wait for more evidence, while buying some time to gauge how ‘hot ‘the issue is liable to be. Decision makers may also try to protect core goals and minimise risks to these goals. Risk calculations and risk management are a fact of institutional life in all areas of public policy-making (Althaus 2008).

Type II: Ideological inaction

Ideology and values can shape purposeful inaction. Ideological stances about the role of the state versus other mechanisms of public problem-solving may play a significant role in limiting the scope of and sympathy for governmental and public intervention. The same applies to ideas about where to draw the line between the private sphere of individual liberty that should remain exempt from public regulation and matters affecting the public interest where government action is appropriate and needed (Pesch 2005; Peeters 2013).

Type III: Imposed inaction

Inaction can also stem from the realisation that political powers, institutional realities and checks and balances are such that ‘taking action’ is not feasible. A lack of legal or financial levers to act may be blunted by powerful blocking coalitions. Bureau–political stalemates between departments and agencies within the executive branch can stymie a government’s actions. The institutional architectures of some political systems—separation of powers, federal, and multiparty coalition systems—bear greater risks of policy deadlock than those of others. Furthermore, specific institutional devices such as corrective referendums or the decision-making procedures of some international organisations also create ample opportunities for effective veto-playing (Cameron 2000; Tsebelis 2002).

Type IV: Reluctant inaction

Inaction may occur when viable policy options, tools and resources to address a particular issue are simply not available. Think of the logistical, military and/or diplomatic constraints of taking effective humanitarian action in complex emergencies. Likewise, the ups and downs of the economic and fiscal cycles sometimes create perfect storms where economic conditions require public sectors to do a lot while their fiscal positions are bad and worsening because of these very conditions (Offe 1984). Going further, it may be that policymakers have few or no practical levers at their disposal to actually address the problem in hand, hence leading to reluctantly taking a step back from concerted action.

Type V: Inadvertent inaction

Inaction can spring from policymakers’ cognitive processes in coping with the manifold, voluminous and ambiguous data and information they are routinely exposed to. This phenomenon is well recognised in cognitive psychology, neuroscience and behavioural economics. Human processing of complex information or conflicting signals proceeds by reliance upon intuitive ‘shortcuts’ that may result in disproportionate information processing (Jones and Baumgartner 2005). One possible consequence of this is that a discrepant event, condition or trend may not be considered important enough to warrant serious attention. This leads to ‘not seeing’ particular phenomena because they have not happened frequently enough to be familiar, or because decision makers’ pre-existing beliefs and assumptions about the phenomenon work against any substantial reappraisal of that phenomenon as a problem (Kahneman et al. 1982). Not just individual policymakers but entire organisations (Seibel 1996; Bach and Wegrich 2019) and institutional cultures (Douglas and Wildavsky 1982; Schwarz and Thompson 1990) can display these problem-negating cognitive biases.

Drivers of inaction

If these are the most manifestations of policy inaction, what are their most prominent drivers? We explore this question and suggest key mechanisms operating at four analytically distinct but in practice overlapping loci of policy-making: individuals; organisations; governments; and networks. We surmise that particular types of mechanisms conduce inaction in each of these realms and that policy inaction is most likely to occur and persevere when these loci and mechanisms overlap and reinforce one another.

Individual-driven inaction

In his constraints model of policy decision-making, Janis (1989) posited that individuals (and groups) employ cognitive, affiliative and self-serving rules, designed to keep in check the inner conflicts they experience, their sense of self-esteem and the relations they maintain with significant others. These rules serve their equanimity in the sense of keeping the stress they experience in their roles of decision makers, at levels they can tolerate. But, as Janis and countless other decision theorists have since demonstrated time and again, these coping behaviours can come at the price of a capacity for reality testing and can indeed immobilise them in the face of mounting warning signals about imminent threats to important values and interests. In earlier work, Janis and Mann (1977) identified several patterns of coping with the pressures of responsibility under which policymakers labour that may conduce them towards inaction:

  • Unconflicted adherence to the status quo, by selective attention to information about past, present or future conditions and selective interpretation and forgetting of information that conflicts with their benign interpretation of the status quo;

  • Shifting responsibility (‘buck passing’) for taking a decision or acting on a signal to other people, departments or organisations;

  • Bolstering decisions already taken in the past by rationalising away the need to reconsider them;

  • Procrastination, i.e. continued indecision while searching for more information, engaging in further deliberation or determining to defer the making of a decision.

These coping mechanisms make for observable patterns of behaviour that indicate individual policymakers’ propensities for non-acting or for choosing not to change pre-existing policies. For potential answers to the follow-up question of what drives this behaviour—particularly in individuals whose role descriptions and self-conceptions propel them to act as ‘decision makers’ who direct, lead act—we can also turn to Janis. Use of his pioneering work—including his best-known work on ‘groupthink’ (Janis 1982)—in policy studies has since been eclipsed by the predominantly cognitive- and neuroscience-driven ‘behavioural economics’ that is now in vogue. Yet anyone who has ever been close to real-life policymakers and crisis managers realises that it is not just bounded rationality, ‘framing effects’ and cognitive processes that drive policymakers. Their motivational drives, their emotions, their psychological needs and hungers, their self-categorisations and group affiliations have been demonstrated to be at least equally pivotal drivers for them to act—and, importantly here, not act—in the ways that they do (Haslam et al. 2011).

Public organisation-driven inaction

Gathering, receiving, interpreting, creating, communicating and disseminating information is at the heart of the work that public organisations do. How they process information determines how they act—and whether they act. They require information to help inform their direction and operations, as well as to rationalise pathways not taken and issues not addressed (Bach and Wegrich 2019).

While in some senses public organisations may seem to confirm classic Weberian notions of bureaucratic inevitability and dominance, they do not necessarily equate with ideal-type efficiency and effectiveness in information handling (Hood 1998). Variations in public organisations in terms of structures, processes and openness to external environments can lead to variations in ‘sensemaking’ capacity where there is a struggle to rationalise what is ‘out there’ and line up with what is ‘in here’ (Weick 1995: 70). Issues that plausibly could or should be within the frame of public organisations to act upon can linger on the fringes of their attention spans. As Hood (1998), Wilensky (2015) and Bach and Wegrich (2019) argue, each configuration or organisational characteristic brings its own ‘Achilles heel’. Wilensky (2015), for example, famously demonstrated that hierarchy is conducive to concealment and misrepresentation of relevant issues; centralisation can produce out-of-touch and overloaded leaders who do not have enough information, interest or capacity to reliably assess what is relevant, and specialisation cultivates a culture of turf wars and lack of information sharing.

An organisation may also be so focused on particular priorities that it blinds itself to other issues—in effect forgetting how to see, interpret and act upon them (Vetzberger 1998; Stark 2019). A study by Zegart (2007) of the CIA prior to the 9/11 attacks found an organisation unable to acknowledge and act on threats from Al Qaeda because it was still stuck in a Cold War mindset. Hood’s (1998) analysis of the role of cultural archetypes (hierarchist, egalitarian, individualist, fatalist) in governance shows that systems dominated by the latter are likely to display passivity and inertia stemming from a self-negating mindset. In countries such as Pakistan, for example, fatalistic beliefs have proved a formidable barrier to the enhancement of road safety, especially participation in health-promoting and injury prevention behaviours (Kayani et al. 2012).

As with individual policymakers, organisations too can succumb to motivated reasoning—‘seeing what they want to see’—in the furtherance of organisational goals or rationalisation of core beliefs (Weick 1995). The consequence is that a range of plausible and pertinent issues get filtered out that do not aid the furtherance of the leadership’s priority goals. As Bach and Wegrich (2019) argue, it is not possible to scientifically separate blind spots and biases that lead to selective information processing, because ‘bias’ has normative implications of departure from a desired optimum. One implication is that inaction on the part of public organisations is necessary for them to fulfil goals, but judgments on the blurred boundaries between inaction as necessity and inaction as pathology are generally easier to make in hindsight, once there is greater clarity on whether ‘doing’ nothing has been a success or failure.

Government-driven inaction

High aspirations, popular mandates and governments ‘doing stuff’ are embedded in the very fabric of liberal democracies. In practice, however, the actual business of governing necessitates continual, dogged policy inaction, and the eschewing of multiple plausible alternatives (Baumgartner and Jones 2015). Ideologically driven inaction can be the raison d’être for some governments who seek to promote a particular relationship between the state and civil society, where the former secedes as much as possible from continual and active involvement in the lives of citizens (Fried 2007). Policies of abstention (or termination, deregulation, liberalisation) have been at the heart of neoliberal party manifestos and government programmes throughout and beyond the OECD during much of the 1980s–2000s. To a considerable extent its ideological appeal has withstood even the force global financial crisis which according to many observers occurred precisely as a result of the hands-off, minimal regulation stances that governments had adopted as a result of their neoliberal commitments (North 2016).

Importantly, ideologically driven inaction can occur on all shades of the political spectrum. The Liberal prime minister and founding father of the Dutch constitution, Jan-Rudolph Thorbecke, took the principled view that government has no role to play in shaping the arts, and made sure the party and the governments he led came nowhere near doing so, and his successors have maintained this stance and extended it to areas like media policy (Winsemius 1999). Likewise, since 1840 and continuing into the present day, Christian political parties in the Netherlands have successfully fought for and defended what they call ‘freedom of education’—in effect restraining the Dutch state from asserting control over both the organisation and content of education, thus leaving proponents of religious and other worldviews to set up schools that reflect their philosophical commitments (and compel the state to fund them if they meet basic quality standards) (Van Bijsterveld 2010).

In addition to purposeful governmental inaction, the pragmatic demands of governing means that inaction is also a coping mechanism. Governments have limited attention spans and do not have the time or space to provide equal and sustained attention to all issues (Downs 1972) and consequently deprioritising issues is a fact of life (Jones and Baumgartner 2005; Baumgartner and Jones 2015). Also, no government or policymaker can ever win all the policy debates they are involved in at any given point in time—and so they slice policy conundrums into manageable bits and pace their attempts to get things done, meaning some aspects get addressed first and others are not acted upon until the stars align.

Furthermore, in a world of immense complexity and uncertainty about how (or whether) issues will escalate and with what support or media attention, ‘doing nothing’ can be the product of governments deciding to wait in the hope that the problem will dissipate or disappear, or more favourable conditions for addressing it will emerge. Doing nothing can be intended to prevent inflaming contentious issues and ride out what they perceive to be fleeting media storms and temporarily inflated public concerns by refusing to ‘overreact’.

Whatever pathway governments forge, and whatever plausible policy alternatives are cast to the margins of agendas, both involve an element of risk (Althaus 2008). Governments routinely deem certain risks and costs acceptable in the pursuit of policies that deliver on ‘core promises’ made during electoral campaigns or otherwise suit the core values and interests of major constituencies. Conversely, they are unwilling to absorb the risks associated with programmes or reforms that address ‘non-core’ promises, eat into their political capital, or cater to less essential constituencies—and thus choose not to act (Vis 2010).

Of course, purposeful and/or pragmatic attempts to control agendas and legitimate inaction in the face of calls to go down plausible, state-centric policy routes, do not always work out as intended. As indicated for example, by analyses of ‘warning signs’ of the 2007–2008 global financial crisis, emerging risk in sub-prime markets and a bubble in asset prices, were not acted upon because of a pervasive ideology that market mechanisms had ‘delivered’ economic growth, healthy profits, low inflation (AUTHOR). Hence, inaction on emerging risks was legitimised by an ideology and mindsets that risks were symptomatic of the natural operation of markets, and remediable through market correction.

Habermas (1984) famously identified the disconnect between the experiences and concerns of ordinary citizens in their day-to-day ‘life worlds’ and the institutional logic of perceiving, categorising and appreciating policy problems in the ‘system world’ of governments. Acknowledging the continued existence of poverty, family violence and discrimination in the otherwise successful and ‘progressive’ north-western European welfare states that reside at the top the OECD rankings, requires for instance, that their policymakers actively look beyond the reassuring picture painted by those rankings and the high-level statistics about unemployment and fiscal positions. The contemporary populist revolt in many Western polities rides partly on its critique of the ‘elitism’ of a political and administrative establishment that downgrades, ignores and thus not acts upon the real needs, fears and losses of the common people (Rooduijn 2015). The 2018/2019 ‘yellow jackets protests in France are a case in point, as is at least part of the pro-Brexit movement on the UK, for example among embattled working class voters in the North of England who feel their plight to have been ignored by distant elites as Westminster championing globalisation and the influx of cheap labour it enabled (Hobolt 2016).

Network-driven inaction

From social welfare reforms and urban redesigns, to mega-infrastructure projects and the fight against drug trafficking, working across boundaries has become the norm for almost any significant contemporary policy initiative (O’ Flynn et al. 2014). It is increasingly difficult to find policy problems and a range of viable policy solutions that do not in some ways straddle traditional geographical, institutional, sectoral and jurisdictional boundaries. In response, joined-up government, networked and collaborative governance have become the new aspiration (Head 2014). Much can be achieved by cross-silo working, not least because it can involve the pooling of knowledge, expertise and resources (Carey and Crammond 2015). And yet research suggests that in practice coordination structures, networks and collaborations are complex creatures whose ability to ‘do stuff’ is contingent upon many factors coming together to enable trust and momentum (Ansell and Gash 2008; Emerson and Nabatchi 2015; Laegreid et al. 2014) to occur. It follows that when such fortuitous confluence of preconditions and relations does not occur, such structures of interactive governance may become arenas in which miscommunication, mistrust, interagency politicking and other forms of centrifugal behaviour produce policy stalemates or non-realisation of shared service delivery and co-production ambitions (Klijn and Koppenjan 2016; Bach and Wegrich 2019).

The mere fact that network structures are often designed to overcome collective action problems does not mean they succeed in doing so. Paradoxically, both the design of networks themselves and the behaviour of their participants in networks can actively contribute to their non-delivering on their purpose. When labouring under adverse circumstances or when not managed with a view towards their becoming productive, networks in fact reproduce and succumb to the very coordination problems they were built to address.

Investigating inaction: moving forward

In the 1960s and 1970s, Schattschneider (1960), Edelman (1964, 1977), Bachrach and Baratz (1970), Lukes (1974) and Cobb and Elder (1977) sought to redirect our attention towards ways in which initial post-war, public policy and planning aspirations in the USA (and presumably other western countries) had been partly blunted by systemic power biases, agenda control, symbolic gestures and continued exclusion of minorities and marginalised groups. Today, several decades into the era of neo-liberalism and globalisation there is firm evidence that social inequalities have risen rather than declined, not just in terms of key economic, health, safety and well-being indicators (World Inequality Lab 2018), but also regarding access to and participation in the democratic process (Bovens and Wille 2017). Arguably therefore, there is an urgent need to revive the power-critical agenda of the policy sciences.

If, as their founding fathers intended, the policy sciences are to remain a relevant ‘critical friend’ to democratic governance they need to balance their propensity to engage with policymakers and institutions that are ‘acting’ with examining when, how, why and to what effect policies are not-proposed, not-designed, not-enacted, not-implemented. In public policy, inaction is just as real a phenomenon as action—at least to those who are prepared to acknowledge its existence. Policy scientists need to examine how in the policy processes avenues of problem inquiry get closed off, the range of policy options that are considered get narrowed, soft and critical voices in public policy conversations go unheard, and non-acting gets institutionalised and legitimised. To be able to do so, they need undiminished awareness of the fundamentally political character of their enterprise (Botterill and Fenna 2019). To a limited extent, contemporary policy analysis continues to be influenced by such impulses as it seeks to scrutinise new forms of late-modern state activism such as ‘evidence-based policy’, ‘behavioural’ public administration ‘new public governance’ and ‘collaborative innovation’ (e.g. Brandsen et al. 2015; Wegrich 2019; Feitsma 2018) that tend to assume a benign and activist state.

At the same time, it would seem important to develop this needed focus on policy inaction as a political phenomenon unencumbered by heavy-handed assumptions about the existence of a coherent, conspiratorial and self-evidently controlling power elite. The study of inaction should be an even-handed intellectual enterprise, not a political crusade. Going down this path will lead us to come to terms with complex and sometimes paradoxical phenomena: the weight of silences amidst all the talk in public policy conversations; the power of looking away amidst all the policy analysis that is routinely performed in and around government; the selectivity of the emotions—indignation, enthusiasm, fear, hope—that drive policy agendas. To set some direction for the efforts we hope our colleagues will agree might be needed, we close this essay by proposing three pivotal areas of inquiry.

Capturing inaction

This involves three major challenges: (a) mustering the epistemological versatility (in particular acknowledging the knowability of ‘non-events’ and the salience of counterfactual reasoning in doing so, Lustick 2011); (b) developing fit-for-purpose methodological tools, for example to notice absences in policy networks, silences in policy discourse and non-decisions in cabinet behaviour (e.g. Yanow 1992); and (c) applying them to policy fields to ‘observe inaction’ at work. Carefully constructed comparative case designs to track differences in the presence/absence, timing and thrust of public policy responses in different jurisdictions to emergent social, ecological, medical, technological seem particularly promising. A good example of such work employing different methodologies to be emulated includes studies of the notable—and consequential—differences between what could be called active-responsive versus reluctant–abstentionist policy approaches to the occurrence of Aids (Perrow and Guillen 1990; Bovens et al. 2001), sea-level rises (OECD 2019), cybercrime (Fafinski 2011) and complex emergencies (Albala-Bertrand 2010).

Explaining inaction

Many explanatory avenues can be derived from the putative drivers of inaction in policymakers, organisations, governments and networks presented above. That entire section can be read as a bundle of propositions awaiting systematic study, and we shall not belabour them here. In addition, however, a pivotal cluster of researchable issues centres around the question of when, how and why policy inaction begins and ends. Just as they do with regard to the programmes that they do adopt and implement, governments and networks can be forced to consider terminating their policies of inaction. When do signals about the growing costs and risks of inaction become so strong that hitherto blindsided policymakers and unresponsive institutions start paying attention to issues they could not or would not deal with previously? To what extent do the same actors and mechanisms—such as pro-reform and anti-termination coalitions, negative feedback and escalation of commitment play a part in causing policies of laissez-faire to give way to policies of intervention, for example. Historical–comparative studies of how and why poverty, housing, the use of alcohol and various forms of recreational drug use became subjects of state regulation fit this mould (De Swaan 1990). Also, why were some states and non-state actors noticeably earlier than others in stopping to treat the internet as a fundamentally benign self-governing space? (Hofmann et al. 2017). Conversely, when and why do governments decide to ‘offload’ policies they have long pursued, i.e. through termination, deregulation and privatisation of social security schemes? (Breen and Doyle 2013).

Evaluating inaction

From the preceding pages, readers may have formed the impression that inaction is more often than not a negative phenomenon triggered by the tricks our brains play upon us, insidious pathologies of organisational life or the flawed functioning of political institutions and processes. This certainly has not been our intent. There is, we think, no a priori reason to value action over inaction. As we noted, inaction is sometimes purposefully pursued and enacted, and for a variety of reasons. In that sense alone, it cannot be treated as not just a product of inadvertence and dysfunction. However, what is really needed is systematic evaluative research. Just as policy-as-action can be assessed as a success, failure or somewhere in-between in view of programmatic, process and political criteria (AUTHOR), so can policy-as-inaction once we have managed to unearth it. Again, comparative studies of active and inactive responses to similar policy challenges, e.g. whether or not to vaccinate teenage girls against cervical cancer, or mandatory or non-mandatory approaches to evacuating at-risk areas in natural disaster (McCaffrey et al. 2015).

We hope that by pursuing these three pathways of inquiry, policy scholars will begin to open up questions about the flip side of its standard object of analysis (policy-as-inaction) that all too often remain un-asked and un-investigated. The knowledge they will gain and disseminate as a result may sometimes be awkward but it is, we believe, potentially important in our ongoing quest to keep speaking truth to and about power. And who ever said that life as policy analyst was meant to be easy?