Stephen Muers is Head of Strategy and Market Development at Big Society Capital. This blog post is based on his time as one of IPR's Visiting Policy Fellows while in his previous role as Director of Criminal Justice Policy at the Ministry of Justice.
The best approach to making decisions when information is dispersed and constantly changing is the subject of a central debate in public policy. Core textbooks for public policy students generally set out the arguments for and against a series of frameworks for such decision-making, each of which lead to different policy tools. More fundamentally, political leaders and senior public officials implicitly or explicitly endorse these frameworks to different degrees, shaping their favoured policy solutions. My argument is that many common frameworks fail to take sufficient account of culture and values both as a way of understanding policy change and as an explicit goal.
Tools for decision-making
Four sets of policy tools flowing from four overarching frameworks dominate policy debate.
Market tools, such as privatisation, financial transfers to enable people to purchase in the market and internal markets within public services like the NHS, will be the favoured options if you believe that the price mechanism and free self-interested agents will tend to arrive at the optimal outcome. A different – and arguably more sophisticated – variation of this is based in the Austrian rather than neo-classical school of economics: not attempting to argue that the market will necessarily optimise, but that that the price mechanism is the best proxy we have for summarising and assimilating dispersed information about the preferences of different actors.
Deliberative tools such as citizens’ juries, micro-level community control of services and the creation of new forums for public debate are the priority if you believe the interplay of unrestricted dialogue between individuals and/or groups is best suited to achieving the most desirable overall social outcome.
Rational state tools such as strategic planning, heavily regulated markets and needs-based allocation formulae will prevail if the preferred thesis is that the government is best able to assess and respond to the competing sources of information and social value. This is typically enabled by the use of bureaucratic expertise to assess options, with elected representatives determining normative issues of value and relative priority.
Informal judgements such as clan or family based networks and precedents are less often put forward as a desirable framework to aim for, but frequently posed as an empirically-based description of what actually happens. Such informal decision-making is a major feature in development studies, but is also highly relevant in industrialised societies (for example, studies of the former USSR have argued that the system was only capable of functioning because of its highly developed informal network of patronage, deals and exchange[1]).
Officials developing policy advice for ministers, MPs debating propositions, journalists commenting on policy issues and academics assessing the results spend considerable time and effort debating the merits of these frameworks and the tools they lead to. Is it better to contract out a service or plan it centrally? Should a service be devolved to local communities so they can decide the approach, or centralised to drive efficiency and fairness? Can a national formula allocate money efficiently and effectively? Should the government give people money to purchase something or provide it directly?
The importance of value systems
There is, however, a good case for saying that debating these different tools is missing the point. Anyone who has spent any time wrestling with public policy will have had the experience of changing the policy tools applied to a problem but getting the same results – or applying a tool that worked well in another locality or country and finding it is completely ineffective in a different context. If getting the right framework for understanding how to make decisions and designing tools appropriately doesn’t work, then what should we focus on?
Values and culture are the missing ingredients. Values and culture in this context refer to the set of (usually implicit) norms, ethical standards and habits shared by the actors in the relevant system. The psychological concept of ‘superordinate goals’ is useful here: the overall aims, sometimes at a broad, high level, which are shared by all the actors and which their ways of operating act to reinforce. Such goals do not need to be, and often aren’t, written down or otherwise expressed in a way that all the actors would sign up to – but can nonetheless be deduced from the way they interact.
Experience shows that values and culture are extremely powerful in determining outcomes. In the famous quote (sometimes attributed to Drucker), “culture eats strategy for breakfast”. After consuming strategy, culture then moves on to policy interventions and tools for lunch. Tools and interventions get co-opted and altered in order to serve the goals, values and assumptions of those using them.
International development practitioners are well versed in this phenomenon. Years of attempts to introduce ‘rational’ western-style merit-based bureaucracies, transparent needs-based allocations and fiscal planning in developing countries have, in many cases, seen those tools simply used to continue the extraction of rents for clan and family members. In a contrary example, similar informal networks and patronage-based approaches in China have produced impressive economic growth – one reason for this being that those networks sit within a very powerful set of superordinate goals around national prosperity and prestige. In western countries there are plenty of examples where supposedly market-based or technocratic processes are actually used to reinforce the existing position and values of elites and insiders: procurement rules that are in fact so complex that new entrants find them impossible, ‘open’ recruitment based on a set of criteria that narrowly reflects the status quo, ‘market-based’ solutions that involve little competition and the state retaining all the major risks (the failed London Underground PPP is a good example[2]).
Value systems as a policy goal
The first implication of value systems’ triumph over policy tools is that it would make sense for voters in a democracy to choose leaders on the basis of their fit with the values and superordinate goals that the elector wishes to see prevail in the country, rather than their detailed policy proposals. In this model, specific policy proposals offered by political leaders are important not so much in themselves but as ways of signalling a deeper set of cultural assumptions. There is, in fact, a reasonable amount of evidence that many voters do operate in this way. One study of the 1983 UK General Election found that many of the major policies in the Labour manifesto, the so-called ‘longest suicide note in history’, were popular. However, polling on more abstract and value-based perceptions of the two major parties (estimations of their ‘leadership’ and ‘patriotism’, for example) pointed much more accurately to the result[3]. Recently, some analyses of the UK’s EU referendum have argued that cultural values (order versus openness) were the strongest predictor of whether someone would vote to remain in or leave the EU, with attitudes to specific issues and also demographic factors emerging more as proxies for those underlying values[4]. Political campaigns often launch policies not because of their substantive merit but to ’send the right message’ about the assumptions on which their candidate operates.
Moving away from elections to decision-making in government, what are the implications of a focus on value systems? One is that it reinforces a long-standing line of argument in public policy research that having the centre determine processes across a dispersed policy system is futile. Outcomes will not be driven by centrally specified approaches, but by how the individual actors (“street-level bureaucrats” in Lipsey’s famous phrase) use and mutate those processes based on their own needs, values and desires. The process in and of itself has little force against street-level norms and values. A linked point is how to use international evidence and learning effectively. We need to move away from “can we implement the policy tool that was successful in country X?” towards “what is it about the system in country X that enabled them to develop an intervention that was so successful in their own context?”[5]
So if detailed process specification, or scouring the world for good ideas to import, won’t work, what should decision-makers be doing? The logic of focusing on value systems implies that they should be doing what they can to create a culture and value system that supports positive change, learning and evolution. If it is inevitable that outcomes are shaped by ‘street-level’ culture and the constant iteration that happens as people interact and revise their plans, are there things that can be done to help that dynamic lead to better outcomes?
Should government promote positive value systems?
First, it is necessary to address two arguments against attempts to change or create value systems. The first is that such attempts are doomed to failure. Culture is so persistent over time, and develops in such a dispersed, organic and unplanned way, that it is hubristic to think that conscious action by government, or indeed any other institution, can change it. There is a two-fold response to this challenge. The first part is to acknowledge that there is some force in this argument. Changing culture is not easy, and certainly not quick. It needs to be done with the grain of what is already there, and by building broad support rather than by top-down decree. The second part, however, is to point to the clear evidence that it can be done. The extreme example is the actions of totalitarian regimes: it is hard to argue that the Bolsheviks didn’t manage, over many years of focused effort, to change the fundamental culture and values of Russia and the other Soviet republics. The experience of Germany after WWII demonstrates the impact of government on culture through a sort of natural experiment: surveys after reunification showed that different cultures and value systems had become internalised by the populations of the two halves of the divided state. However, they also showed significant continuity, and so emphasised that government can only influence culture slowly and in part[6].
The second argument against attempts to change or create culture emerges from this last discussion: such actions are associated with dictatorial regimes and have no place in a liberal democracy. Shouldn’t a democratic government represent and embody the values of its population, rather than trying to engineer them? There are, again, two counter-arguments to this. The first is that a government cannot help influencing the values of the society it leads. Even if not done deliberately, the way government and its leaders behave and the values they implicitly or explicitly endorse will have an impact on society. If this impact is inevitable, there must be a case for consciously considering it and acting accordingly, rather than allowing the impact to occur by accident. The second is that if we believe that values and culture are central to determining social outcomes, and democratic governments are elected at least in part to deliver outcomes, then taking a position that they should not address those factors is to will the ends but not the means, and potentially condemn them to failure.
How to promote value systems
The most obvious way to create a culture that supports positive change and learning is for government to send messages through its words and actions that this is desirable. This would include launching and welcoming experiments in policy and practice and being open to learning from failure, to bottom-up innovation and to constant iteration rather than over-specifying plans from the start.[7]
Such an approach can, however, seem rather intangible. The obvious way to give it harder edges is to use incentives to reward service improvement and thereby encourage innovation that delivers such improvement across the system. This is the philosophy behind recent school reform; the changes gave schools autonomy to innovate, and strong incentives to do so, by introducing a rigorous performance regime.
However, performance incentives alone are unlikely to engender the culture of wholesale creative experimentation discussed above, for several reasons:
- There is a long-standing body of literature which argues that it is hard to design incentives which can’t be gamed and that do not lead to distorting behaviour. There is always a risk that hard performance measures produce great innovation in the management of the measures themselves, rather than genuinely improving services across the board. Classic examples include hospitals meeting the four-hour A&E waiting time target by creating other queues elsewhere in the system[8], and schools focusing their efforts on pupils around the borderline of exam targets at the expense of the less able, who were never likely to make it[9].
- Problems that require collaboration between lots of agencies, with the costs and benefits potentially falling asymmetrically, are difficult to address through performance incentives. While it is theoretically possible to design an outcomes framework that pulls all agencies together behind a common goal (the UK model of joint PSA targets, for example) it is hard to do so in a way that isn’t highly complex and bureaucratic. Simple measures of their own performance will tend to have more traction with service managers.
- Measuring and rewarding performance as a way of promoting innovation has the downside of pushing innovation towards current problems that we know how to measure. The most valuable feature of a dispersed system is, in fact, its ability to react quickly to a changing situation, beginning the task of innovating to respond to new challenges before the central authority has even clocked their existence. The makers of Blackberry phones were proud of the way they encouraged employees to innovate and improve keypad mobile phones – but missed the real innovation of moving to touch-screens and disposing of keypads altogether.
- Service outputs, or even outcomes, are not the only objective of public services in a democratic state. Such services also need to operate in a way that is recognised as fair and legitimate, and which promotes trust between citizens and institutions. While trust and legitimacy is, in some cases, measurable, it is much harder to target with performance incentives[10].
Multi-national corporations that operate with a high degree of autonomy amongst their business units face the same challenge around creating a positive culture of innovation and learning. Evidence from examples such as General Electric indicates that performance incentives alone are not sufficient. These companies have tended to find that, within a decentralised environment, the centre needs to provide an infrastructure for collaboration, mutual learning and opportunities that actively pushes people in that direction, in order to avoid narrow and immediate targets being the sole focus for their creativity. Simply incentivising people to hit performance goals doesn’t lead to a culture that is healthy for innovation in the long term[11].
What a successful infrastructure and environment for collaboration and learning looks like will be highly context specific. What General Electric does to engender innovative learning by local business managers will look very different from what the health service or the police might do. Innovation and learning will be needed to discover how best to innovate and learn, and to keep that understanding up to date. While performance incentives alone are not likely to work, there is also a long history of softer collaboration approaches (best practice networks, peer support, joint boards and shared funding, for example) draining large amounts of resources for little discernible effort. Developing better ways to promote a culture of positive change, learning and evolution is a central task for both public service leaders and the researchers who wish to understand and help them.
[1] Russia’s economy of favours, A Ledeneva 1998, CUP
[2] Blunders of our governments, Crew and King 2013, OneWorld
[3] British General Election of 1983 1984, Butler and Kavanagh, Palgrave Macmillan
[4] “It’s not the economy, stupid”, Kaufmann 2016, LSE discussion paper
[5] “What makes governments get great”, Andrews 2013, Harvard Business School discussion paper
[6] “Value Priorities in the United Germany”, Boehnke et al 1993, European Journal of Psychology of Education; “Individual Preferences for Political Redistribution”, Corneo and Gruner 2002, Journal of Public Economics
[7] “Is your policy a dodo”, Muers 2014 Civil Service Quarterly
[8] “Review into the Four Hour Emergency Access Reporting at Nottingham University Hospital Final Report”, Ken et al 2010
[9] E.g. “Quick fixes to climb league tables”, Marley and Mansell, 2007 Times Educational Supplement
[10] “Creating Public Value”, Kelly and Muers 2004, Prime Minister’s Strategy Unit discussion paper
[11] Author discussion with Monia Mtar, University of Bath Business School
Respond