By Kit Fotheringham, Bristol Doctoral College (University of Bristol)
In common with many developed nations, the public sector in the UK takes up a sizable proportion of economic activity. Therefore, the way the public sector is run is of interest to the public, both as citizens, who expect good administration of public services, and as taxpayers, who contribute to public finances. Of course, not all citizens are taxpayers (especially children and those on low incomes), nor can all taxpayers be regarded as citizens (notably, companies). Some scholars even question whether the balance sheet of a sovereign government with its own currency is comparable to the household pocketbook, positing that government spending contributes to overall money supply. Nevertheless, politicians play to the narrative that public money is the collective property of taxpayers, and seeking to persuade voters that their policies will offer the best protection against further encroachment on the economic interests of individuals.
In this blog post, I explain how the ‘public interest’ in good public administration is transformed into economic interests, predominantly driven by the notion of ‘value for money’. This, in turn, leads to a managerialist ideology in public administration, which demands ever greater efficiency from public services. Latterly, the drive for efficiency savings has piqued the interest of officials in automated or algorithmic decision-making, in the hope that automation can maintain service quality at reduced cost to the public purse.
A managerialist public sector
Public opinion on the amount the government should be spending on public services can often be at odds with the willingness of taxpayers to provide additional funding for the public sector. The cheats’ way of reconciling these demands is to conjure up ‘efficiency savings’. Such fiscal magic usually consists of rhetoric that blames officials for presiding over ‘waste’ in public services. Politicians make promises to replace ‘bloated’ bureaucracies with more ‘nimble’ alternatives, under the charge of ‘entrepreneurial’ managers, who are empowered to ‘cut the fat’. The trick is to persuade the public that delivering public services in this manner truly is in their interest. However, as the Institute of Chartered Accountants in England and Wales has cautioned, if the ‘public interest’ is manipulated too far, or invoked too frequently, there may be growing suspicion that decision-makers may actually be ‘acting in their own interests’ or in the interests of a favoured group or class. Managerialism attempts to provide a convincing narrative that the interests of a small, exclusive class of managers are to the benefit of all.
Managerialist ideas have their roots in the management of factories and industrial workplaces. In the wake of the Industrial Revolution, sociologists like Frederick Taylor posited that incremental improvements could be made to optimise manufacturing processes through so-called ‘scientific’ management of the workforce. These ideas were readily taken up by the dominant capitalists of the day; indeed, Henry Ford’s application of management techniques to mass production gave rise to the phenomenon we now call ‘Fordism’.
For much of the 20th century, the factory was where management remained. According to Thomas Klikauer, it was only once management had been transformed into ‘managerialism’ that its ideas escaped the factory gate. ‘Management’ could be regarded as a dull, humdrum affair. ‘Managerialism’, on the other hand, promises exciting new horizons, supported by an ideology which claims that managerial techniques can, and should, be applied to everything, everywhere. Any interaction can be turned into a process, and every process can be optimised. Through targeted technical interventions, labour costs can be reduced with no impact on quality. Or such is the siren call of managerialism.
The catalyst for the expansion of managerialism into public administration was the neoliberal reorganisation of the state under Margaret Thatcher. After oil crises and inflation in the 1970s had left the public finances in disarray, Margaret Thatcher’s government believed that the state should do less and set about privatising previously nationalised industries, such as water, energy, and telecommunications. Whereas privatisation brought about a shift from bureaucratic management to capitalist management through changes in ownership structures, bureaucracy itself was not so easily done away with in the rump public administration that was left over. For managerialism to succeed in transforming public administration, a new model of delivering public services was needed.
This was achieved through the New Public Management programme. New Public Management sought to remodel public administration to resemble the management structure of private corporations more closely. This transformation began with ‘Next Steps’ agencies, units which were hived-off from their ‘parent’ government departments in Whitehall so that they could focus on delivering public services with little interference from ministers. Managerialist attitudes to public service delivery were further embedded under John Major’s ‘Citizen’s Charter’; citizens were to be treated as ‘customers’ and managers would be evaluated according to customer satisfaction scores. Tony Blair’s ‘Third Way’ approach continued to adopt a managerialist stance to public services; for example, funding new provision through public-private partnerships managed at arms-length.
Managerialism experienced a renaissance during the 2010s under David Cameron’s coalition government. In the context of austerity economic framing, Francis Maude was put in charge of public sector reform. Inspired by the managerialist emphasis on technical interventions as a means to achieve greater efficiency, Francis Maude spearheaded ‘digital by default’, the idea that public bodies should interact with their ‘customers’ through digital technologies wherever possible. However, individuals were not to be given a choice of communication medium. For instance, if somebody preferred to speak on the phone or talk to somebody face-to-face, they would instead have to learn to use the digital systems on offer. In order to be fair and legitimate, the decision to restrict offline modes of administration must be taken with public interest considerations in mind; namely that there are reasons for overriding individual preferences in favour of aggregate benefit to the public as a collective. Here, the notion of ‘value for money’ offers managerialists an easy justification, owing to the way in which managerialism allows assertions of ‘value for money’ to readily substitute for the public interest.
Substituting ‘Value for Money’ for the public interest
Throughout the successive waves of managerialist reforms from the 1980s to the present day, there has been an assumption that the economic interests of taxpayers is synonymous with the public interest. This is made explicit in guidance from HM Treasury called Managing Public Money, which is, in essence, the manager’s manual. The principles set out in Managing Public Money determine what is considered appropriate use of public funds. At the outset, Managing Public Money insists that public services should be ‘carried out in the public interest’. Yet, despite using the phrase ‘public interest’ over two dozen times throughout the document, nowhere does guidance attempt to define what the ‘public interest’ entails.
Instead, upon closer inspection, one gets the impression that the ‘public interest’ is used merely as another way of referring to the notion of ‘value for money’. Perhaps this is unsurprising, given the provenance of Managing Public Money from a finance ministry whose primary focus would naturally be on obtaining maximum benefit at least cost. The concept of ‘value for money’ finds legal expression in the National Audit Act 1983. This legislation enshrined ‘economy’, ‘efficiency’, and ‘effectiveness’ as the standards against which public sector activity should be audited. The ‘three Es’, as they are collectively termed, reinforce a managerialist ideal. In theory, by bringing attention to the economy, efficiency, and effectiveness of administrative activity, ‘value for money’ auditing encourages officials to become less procedural (the bureaucratic ideal) and more focused on the ‘bottom line’.
The managerialist concern for the ‘bottom line’ should be easily disregarded as an ideological project to transform the public sector in its own image. The sustainability of public finances presents real challenges to governments of all stripes, regardless of political persuasion. If the 2022 ‘mini-budget’ fiasco of Liz Truss and Kwasi Kwarteng taught us anything, it is that gilt markets are acutely sensitive to unfunded spending commitments, which has a consequential effect on the government’s borrowing costs extending long into the future. As Neil Wilson explains in a blog post about money creation in the UK, the Exchequer meets its daily spending demands by drawing on the Consolidated Fund account at the Bank of England. Whenever insufficient cash has been brought in during the day via taxation, debt is created to meet the shortfall and ensure that the net position of the Consolidated Fund is nil at the end of each day. The Debt Management Office, the part of HM Treasury responsible for the national debt, sells a form of debt known as ‘gilts’ to investors in exchange for cash to fill any remaining negative balance on the Consolidated Fund. Gilts are attractive to investors because they represent a promise to repay the investor their principal, with interest, on a specified future date. But when too many gilts are created, their projected yield loses value relative to other investments, so investors demand a premium on the price of gilts before they will consider parting with their cash. This results in higher borrowing costs that the Treasury will have to repay in the future.
Deficit budgets are, of course, nothing new. But since the ability of the Debt Management Office to sell gilts is constrained by the attractiveness of projected gilt yields in the eyes of capital markets, officials know that they must demonstrate fiscal prudence. Indeed, as the Treasury’s Financial Reporting Manual makes clear, investors actively scrutinise Whitehall’s financial accounts, so profligate or imprudent spending today could influence the willingness of investors to purchase gilts on terms that the government can afford in the future.
Thus, the fear of frightening-off fickle capital markets means that Managing Public Money insists Treasury consent be sought regarding the design of public services. After all, public services incur the expenditure of public funds, and the availability of public funds is determined by the ability of current and future taxpayers to stump up the cash. Following this logic, it seems almost irrefutable that the ‘public interest’ can only be met by satisfying the imagined interests of a mythologised, hyper-rational homo economicus. Given that the public (as citizens) want quality public services at the least cost to themselves (as taxpayers), the mantra of ‘value for money’ rules supreme.
Automation serving the public interest?
So, in the managerialist public administration, the public interest is deemed to be satisfied if ‘value for money’ is achieved. If we recall that managerialism treats all interactions as processes which can be optimised, then the assumption is that each component of a public service is liable for improvement through technical intervention or the application of market logic. For instance, the government’s Digital, Data, and Technology Playbook rehearses the managerialist assertion that ‘innovation brings about benefits such as increased productivity [and] value for money’. Therefore, to realise these benefits in public administration, officials should actively foster ‘an environment which enables innovative and creative solutions’ to come forward.
Against this backdrop, automation enters the frame. As algorithms have become increasingly sophisticated and the cost of computing has reduced, automated decision-making has emerged as an attractive proposition. The calculus of ‘value for money’ means that officials are now almost under an obligation to consider the use of automated technologies, with the presumption being that human decision-making is more costly than operating an algorithm of comparable effectiveness. Procurement guidance reinforces this view, arguing that public bodies should stimulate market competition for innovation in artificial intelligence and other automated technologies because this will ‘drive innovation in delivering public services’ and thereby ‘achieve value for money for taxpayers’. To the extent that ‘value for money’ trumps other considerations, the allure of automation appears irresistible.
The proliferation of automated decision-making throughout the public sector has largely taken place quietly, with few attempts made to check that the use of automation adequately meets the needs and interests of service users. In large part, the fact that there are generally no legal rules which compel officials to consult on the design of public services means that consultation may be deemed a costly distraction. Even where an attempt is made at stakeholder engagement, there are a number of methods for doing so, and a formal public consultation can be avoided.
Sherry Arnstein illustrated the various ways that the ‘public interest’ can be manipulated, or simply ignored, through her ‘ladder of participation’ analogy. Lower rungs of the ladder are colourfully described as ‘window-dressing’, consisting of empty rituals which do little to reassure citizens that their interests will be properly taken into account. Towards the top of the ladder, citizens exercise greater control over the issues and can more effectively direct the allocation of resources to services they have had a hand in designing.
| 8 | Citizen control |
| 7 | Delegated power |
| 6 | Partnership |
| 5 | Placation |
| 4 | Consultation |
| 3 | Informing |
| 2 | Therapy |
| 1 | Manipulation |
Sherry Arnstein used her ‘ladder of participation’ model to show that the ‘traditional power elite’ retain control. Advisory committees are an example of ‘placation’, whereby citizens ‘can be easily outvoted and outfoxed’ by handpicked appointees. An exemplar of this is the Social Security Advisory Committee, whose present (2024) chair, Dr Stephen Brien, was for most of his chairmanship concurrently a senior member of the Legatum Institute, a think tank founded by the investment firm of the same name. Stephen Brien’s biography describes him as providing the ‘blueprint’ for Universal Credit, one of the major welfare benefits which are overseen by the Social Security Advisory Committee.
Meanwhile, there are examples too numerous to mention about how the automation of Universal Credit, and its ‘digital by default’ service offering, has wrested control and autonomy from individuals who rely on Universal Credit as their main source of income. If individual circumstances do not match up to the algorithm, benefits can be reduced automatically. In the worst cases, claimants who otherwise meet all the eligibility criteria get sanctioned, just because their employer paid them a few days early to avoid a weekend or bank holiday. The Child Poverty Action Group took the Department for Work and Pensions to court over such cases where the ‘computer says no’. Whilst they were successful in persuading the court that the claimants had been treated unreasonably, the judges accepted the managerialist narrative that automation is necessary in the ‘public interest’.
Ultimately, in a managerialist public administration where ‘value for money’ effectively substitutes for the ‘public interest’, there is little that individuals can do to refuse the introduction of automation. Yet, despite the hype put out by technology evangelists, there is nothing inevitable about automation. Alternative conceptions of public administration are possible, such as those which prioritise care of the individual at the ‘street-level’ and where citizens are active participants in government, choosing the manner in which public services are offered. If such models produce public services in a manner which better satisfies the interests of citizens we should consider them urgently, before automation ‘locks-in’ a partial, monolithic view of the ‘public interest’ which optimises for cost alone.