The Regulatory Mindset in Civil Service Management
We should seek to promote excellence, not just avoid incompetence.
Regulation is one of the core pieces of the government’s toolbox. When civil servants (or ministers) want to solve a problem, they often turn to regulation. But I think this has a knock-on effect when thinking about what to do about managing the civil service internally. I’d like to coin this as the ‘regulatory mindset’ in contrast to the service mindset’.*
What is the regulatory mindset™? Where does it come from?
The ability of the government to set rules that dictate how other people and organisations act sets it apart from private businesses. Where businesses begin to acquire this power, people tend to start arguing they are monopolies or have inappropriate government-like powers - look at the control social media has over news media, or platforms like the App Store have over app developers. When Facebook and its associated apps went down earlier this week, some people argued that infrastructure like this should not be in the hands of one company.
Regulation of society has its analogue internally to an organisation. Management can set rules that workers must follow, or else they face consequences, from withholding a raise or bonus to being fired. These rules are set above the individual team they apply to, and apply to all employees seeking to carry out a particular activity. This type of rule setting happens in private business too, but I think it’s over-used within the civil service. The instinctive decision to manage by setting rules is what I want to call the regulatory mindset.
There are many examples of regulatory-style management rules in civil service policymaking. Almost every piece of government legislation or consultation comes with an impact assessment, which has a template and format to follow. This must follow the Green Book approach to policy appraisal, which business cases for spending approval must also follow. Some rules are literally written into regulation, like the Public Sector Equality Duty. Others relate to the specific ways that departments must arrange their recruitment or their pay scales, dictated from on-high by the Cabinet Office or Treasury.
The Government Digital Service (GDS) has some of the most service and user-oriented people in the civil service. But perversely, it seems to be one of the areas with the highest level of internal regulation. Literally every pound of digital spend, even when it doesn’t touch the public, must be approved by the central GDS. This is in stark contrast to the amount of spend in other areas which is delegated to departments. In order to get this approval even for a simple tool, there are a huge number of bureaucratic hoops to jump through in order to demonstrate in detail how the service meets the extensive government standard.
There are undoubtedly many people in government with a service mindset or customer focus. A large part of government is dedicated to managing public services like healthcare, welfare and tax administration. But the people dedicated to service delivery are rarely those who climb to the top of the civil service, nor are they a majority within the policy profession. Civil service reform is led out of the Cabinet Office, which doesn’t deliver any major public services. Significant power is also held by the Treasury, which also doesn’t deliver major services. So I think it’s safe to assume that this service mindset is less prevalent in those who hold the most influence in the management of the civil service, especially policymaking.
There is an alternative diagnosis of the problem. It could just be inertia. For years the question of how to manage the civil service more effectively has been more rules, top-down dictats and centralised reforms. Perhaps the idea just rarely occurs to reformers that their job could be about providing services and solutions to workaday civil servants. If that’s the case, maybe putting a name on it can make this default assumption explicit.
What’s wrong with the regulatory mindset?
Regulation and rules have a role in the management of any organisation. Rules are most effective where something absolutely must or must not happen. Enforcing rules about safety prevents accidents and upholding rules about accounting prevents internal fraud. They are also essential for maintaining standards across multiple versions of the same product. A Starbucks coffee tastes the same in any branch because their employees follow rules from a company handbook about how to make each particular drink.
The rules I described above for policymaking or digital projects generally encourage things that any good policymaker should do. We should consider the impact on people from groups with particular needs or interests, and assess the expected outcome of a policy using best practice. Plus, we are setting rules for the whole country and spending public money, so it is right that there is a high standard. But forcing everyone through a painful approval and assurance process seems a brute force way of ensuring standards are maintained, and definitely puts people off exploring innovative solutions like digital projects unless unavoidable.
No organisation can run entirely on strict rules, especially an organisation broadly in the ‘knowledge economy’ business of producing ideas rather than products. The first problem is that what employees produce is not standardised. Different problems require different policy solutions, even if there are often commonalities. The bigger problem is that knowledge workers tend to have more autonomy than service or manufacturing workers. If you try to impose rules, you’ll often fail because they’ll subvert, ignore or half-arse them. This happens to service workers, as in the book Street Level Bureaucracy which describes the ways frontline public sector workers like police and social workers bend the formal rules to get stuff done. But it happens even more to knowledge workers, who tend to be harder to pin down partly due to just having more social power, but also because their outputs are difficult to measure, so imposing targets is close to impossible.
These types of rules provide no incentive for colleagues to follow them properly. When you want to get something done, the rules tend to get in the way of that rather than support that. This is fine to the extent that the rules are in place to stop people doing stupid things, but they fail to the extent that they get in the way of doing sensible things as well. What we want is people to build best practices into their daily working lives, so that the rules preventing people from doing stupid things only come in on the rare instances when that happens. Most civil servants very genuinely want to do the right thing, and need to be assisted rather than prevented from doing it. The best rules are those that don’t need to be policed.
What’s the alternative?
Businesses, in contrast to government, rarely use regulation-style activities to get what they want. Their main way of interacting with outside individuals or organisations is to persuade those people to be their customers and pay for products or services. Rather than telling people how it’s going to be, they seek to provide the best product possible so that customers will want to take up the company’s offer of their own free will. If they don’t produce things people want to buy, they won’t survive very long. Effective business leaders tend to have a pretty strong service or customer focus.
As should be obvious from my discussion above, I don’t think all regulation-style management rules should go. Clearly there should be some standard applied to ensure that government spending proposals aren’t a waste of taxpayers money, for example. The problem comes from when preventing poor outcomes is the limit of the attempt to improve standards (and they aren’t even very good at avoiding disasters).
The opposite approach stems from the service mindset and focuses on what can be done to provide colleagues with the most effective ways of doing their job. This would require central reformers like the Cabinet Office to stop seeing line departments and their employees as opponents who want to impose all sorts of bonkers ideas on the country. Instead they would treat line departments as customers who genuinely want to do the best for the country, but don’t always have the capability to make that happen.
In practical terms, this may still involve setting standards and guidance, but it wouldn’t use those as a stick to beat people with. Instead, these standards would be the starting point for building a comprehensive set of tools that policymakers would actively want to use. Rather than telling teams they have to recruit in a certain way, aim to provide the most effective recruitment process possible, so that people use it because it gets them really great team members, not because they have to use it. The goal of central civil service reform teams should be to design systems so good, they don’t have to enforce their use.
There are some examples of this in government, but they tend to be limited in scope. Groups like the Policy Lab produce some useful guidance and toolkits for how policymakers can do effective work. The main problem with these, as I mentioned previously, is that they focus on an idealised academic policymaking process, rather than improving the things policymakers do on a daily basis. They also therefore tend to be slightly too high-level. The goal should be to minimise the amount of low-value thinking policymakers have to do about how to organise themselves and their teams, so they can spend as much time as possible just focusing on analysing problems, proposing and testing solutions.
*Scott Alexander opened his recent review of Scout Mindset by mocking the proliferation of “mindsets” in modern discourse. Unfortunately, I’d already thought of the name for this post.