Robert Colvile and Sam Dumitriu.

Q41 The Chair: Good morning. This is the Industry and Regulators Committee in the House of Lords. We are looking at the situation between regulators and the growth agenda. Our witnesses this morning are Sam Dumitriu, who is head of policy at Britain Remade, and Robert Colvile, who is director of the Centre for Policy Studies. Welcome to you both.

You will be very well aware of the importance that the Government have attached to growth, and they have made it very clear that they expect regulators to assist. In the first instance, could you outline what you think regulation can do to support growth? To what extent is that within the remit or role of regulators? Sam, do you want to go first?

Sam Dumitriu: The most important thing is making sure that regulation is proportionate to the goals of regulation. The way a regulator such as the Environment Agency, Natural England or the Office for Nuclear Regulation can support growth primarily is making sure that they deliver their mandate while placing as small a burden on business as necessary. That means avoiding requirements that, potentially, have very little safety or environmental protection benefits, or, for instance, making sure that their processes are as simple and easy to understand as possible.

There is an important role in making sure that, when they say they will do something, they do it on the timescales that are expected of them. We know that many projects, particularly housing and development, are held up when regulators fail to deliver on that. We have the Building Safety Regulator causing quite a lot of problems in not getting through its backlog fast enough. We also have the system of statutory consultees, where you are meant to get back in a very short time to make planning applications and things like that as smooth as possible, but we know that many regulators miss their deadlines and face little consequence for doing so. Those are the key things.

On what regulation can do to promote growth, in a lot of sectors they are trying something different—sectors such as fintech and crypto have all sorts of complicated things that are brand new. Often, a degree of regulation is useful in showing people that this is something they can invest in and not something that will be shut down overnight or where there will suddenly be a lot more scrutiny. Some stability in that sense is useful. However, the main role for regulators is to make sure that they are not imposing unnecessary burdens on business.

Robert Colvile: My overall statement is that over the last 50 years we have shifted from regulating the economy directly to indirectly—from government owning and controlling things to government trying to set rules and regulations. If you look at growth rates in the UK and the EU, especially compared to America which has taken a slightly different approach, it is not obvious that this has been the most successful approach.

The argument we make is that, with tax and spend—the traditional method of controlling the economy—Governments have a pretty good track record of monitoring what they are doing. It is not always perfect but they know what their taxes are, how much revenue they are collecting and how it works. Regulation is exactly equivalent to taxation. You are imposing costs and burdens on business and consumers potentially for incredibly good reasons, but there is obviously a cost and a burden to it. We have been very bad about monitoring and measuring the impacts and predicting the impacts of that.

In fact, we have often seen regulation as something that is cost-free. There is a dynamic—and it is not just with regulators and Ministers—that, if there is something bad in the headlines, the instant reaction is, “Okay, we will do some regulating on this”. The impulse is protective, but there is not a system set up. There is no good mechanism of working out whether what we are doing is proportionate, as Sam says, or whether it will have unintended consequences.

The regulators play an important role in growth, but my first principle is Asclepius’s first do no harm. And I do not think we are very good at first do no harm.

Q42 The Chair: Do you think it is clear what the Government mean and want regulators to do? Do you think that the Government’s agenda is sufficiently open and clear for regulators, particularly in the areas that growth can be expected to be delivered?

Robert Colvile: I do not think it is. I do not think that it is the Government’s fault; it is how things are set up.

I was writing this down on the way here. In the last few years I can think of five or six different areas where Government have said, “This is a priority for us”, but the regulatory agenda and the regulatory establishment has cut across it. One example, after Brexit, is the Government saying that novel foods are a huge priority for us and an area we can steal an advantage on. But the regulator just was not equipped to do that—it did not have the resources and it did not make it a priority.

Rishi Sunak’s Government spent a huge amount of capital and time trying to lure Andreessen Horowitz to invest in the UK and set up its headquarters here. It wanted to do things with crypto, but no one had told the regulators that the Government were keen for people to do stuff with crypto. It ended up leaving the UK in a huff because there were no rules for it to operate against.

Similarly, Boris Johnson—Lord Udny-Lister will remember this—stood on the stage at the Conservative Party conference and said, “We have a wonderful idea for 25-year fixed-rate mortgages”, which is something that the Centre for Policy Studies came up with. The idea was that you do not need the 2% stress test on the mortgage if the mortgage never rises in value. Given that interest rates were basically zero, the idea was to try to lock in cheap money for the consumer in the way that quite a lot of companies were locking in cheap money. That was the flagship policy of the Government. It was in their manifesto. The Bank of England had responsibility for regulation of the mortgage market and said, “No. Our job is financial stability. That is our mandate. Our job is not to increase home ownership. It is nice if home ownership increases, but that is not what we are legally mandated to do”.

Similarly, Ebbsfleet was meant to be the big new town—it was a real big government project. Natural England essentially grabbed a large chunk of what was meant to be the town centre on quite weak grounds to expand the Site of Special Scientific Interest there. If you read the papers from it and the responses, Natural England literally says, “We have been told by the courts that we are forbidden from considering the growth implications of our decisions”. On that logic, if someone finds a spider in the middle of Oxford Street, Oxford Street becomes a nature reserve.

That is one of the problems here. Everyone blames the regulators, and it is easy to blame the regulators, but we have set up quite a lot of regulatory systems that are based around absolutes: absolute protection and absolute safety—there are lots more I could cite—often stemming from wording in a directive or law. For example, in the case of nutrient neutrality there was a line that there has to be no reasonable scientific doubt as to the lack of adverse effects from building something. That is a standard that no project can ever meet.

The regulators are hamstrung. The Government say they want growth. They send emails to the regulators saying, “We want you to deliver growth”. They fire the heads of regulators because they are not doing enough to deliver growth. Fundamentally, they do not give the regulators the resources they need or change the fundamental rules that the regulators were established under so that they follow those things. The regulators end up being buffeted by instructions from the latest Minister to move in a particular direction, but they are anchored very firmly in their statutory obligations.

The Chair: We will come on to some of those points later in greater detail. Sam, did you want to add anything at this stage?

Sam Dumitriu: One of the big issues is that politicians are often very keen to add new regulatory obligations and often new duties upon regulators are seen as a very sensible thing to do. As a result, regulators have to navigate an awful lot of different competing demands. Some of those demands have a lot more weight than others—they are typically the reason the regulator was set up—and, as a result, they can end up drifting from their objectives. Governments typically do not apply that same level of scrutiny and they do not—I do not want to say interfere—grip some of the issues where, for instance, habitats regulations encourage them to do things and basically do not give them much alternative to not doing things that are very harmful for the Government’s growth agenda without the Government saying, “We are going to change those rules. It would allow us to build a new town by a high-speed rail station”, for instance.

The Chair: We will draw out some of those points later. I will first bring in our colleague online, Lord Udny-Lister.

Q43 Lord Udny-Lister: Mine is a two-part question. Have the Government considered the different impacts regulators can have on growth in the short term compared to the long term? What I mean by that is, for example, by setting prices low at the expense of encouraging investment in the future. The second part of the same question—you touched on this a bit—is about regulatory predictability and stability to ensure growth. Is there a real downside in regulations if it upsets that stability?

Sam Dumitriu: We definitely see cases where price controls, particularly in water but also in networks for energy, have been probably too restrictive over time and, as a result, we have failed to identify that a lot of demand is coming that we had not quite foreseen.

When you think about something such as a network monopoly, where you will be engaging in price controls, there are certain advantages to it being run privately. One advantage is you have private sector rigour in how you run the company, so you will try to keep waste to a minimum. The other advantage of being a private company—you see it in lots of different companies—is they take risks in anticipation of future demand. A network monopoly might say, “We think there will be loads of energy demand coming online from AI, so we want to build to meet that demand so that Britain gets all the benefits of that and we can make a nice return”. However, the regulator’s incentive is to avoid overinvestment and keep bills down in the short term. You end up in a situation where you do not necessarily get that outcome, which is a shame, and basically means you lose one of the key benefits of privatisation.

There is another problem, which is that the level of activity that goes into these price controls is enormous. We have spoken before about environmental impact assessments 30 times longer than the work of Shakespeare—but, if anything, they are relatively brief compared to the amount of paperwork that goes into persuading a regulator that they should give you this rate of return rather than that rate of return. These are enormous documents and, unlike the environmental impact assessments, the people paid to produce the numbers that go into the documents are paid a heck of a lot more. Enormous amounts of investment go into these things, and it does not seem like a great use of resources to me.

You also have another challenge with this, in that there is a game where the network monopoly’s incentive is to make the case for as much investment as possible. At the same time, they have very little incentive to deliver it in full because, if they can find ways of not delivering it in total, they can pass some of the savings on to consumers, but a lot of it can go to their own bottom line. Again, that is not an ideal situation.

Robert Colvile: I agree with a lot of that, especially when you are talking about price controls. The nature of the regulated asset-based model, which is what we operate in most of the privatised utilities, is that you want to maximise the size of your asset base. If your rate of return is capped, the larger the asset base the more return you make.

With water in particular, you have the dynamic where the companies have been saying, “We want to invest huge amounts”, and the regulator has been saying, “No, because that would involve bills going up”. As Sam said, it is not always clear that all the things the companies wanted to invest in were value for money and worthwhile and would be delivered on time, but it is also pretty clear to everyone now that the result has been underinvestment compared to what we will need to do. The level of leaks we are prepared to tolerate and the level of drought we are prepared to tolerate has changed dramatically. There are plans in place for water to be turned off in different boroughs of London on a rotating basis. That might have been okay in the 1970s, but we are not so prepared to accept that nowadays.

On the point about regulatory stability and predictability, obviously you want to have a predictable environment. I will say three things on that. First, a stable and predictable environment tends to benefit incumbents. It is very striking—we saw this in the Brexit process—that companies that opposed particular regulations tooth and nail, once they are in place suddenly say, “It would be an awful lot of faff to get rid of them and we have adapted our procedures”. That means that any new entrant into that market would have to spend a huge amount on General Data Protection Regulation (GDPR) compliance or the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) directive, or things like that.

The other thing is that we do not really have regulatory stability. If you look at the surveys of businesses, they always say it is getting worse and that there is always more guidance. Regulators have more people working for them and those people need things to do with their lives. Their jobs are to produce guidance. Talk to a head teacher about how much instruction they get from Whitehall and how many things they have to read through; talk to a tax lawyer about the size of the tax code. We are not in a situation where we are just setting the rules and letting people get on with it. We are in a situation where, in area after area, there is regulatory creep.

In preparation for this, I did a survey on Twitter—completely anecdotal—and asked, “Is anyone out there happy with their regulator?” I got a lot of very unprintable comments. I got one person saying yes, they vaguely did a good job on this one thing at one point.

Where there have been efforts at deregulation and where there have been concerted examples of this, I am not aware of huge amounts of damage from it. We might get on to this later, but the US state of Idaho literally abolished its entire regulatory code in 2019-20 and then reimposed only the regulations that were found to be utterly crucial. Britain will never do that because of our political culture, but I have not seen many reports of mass deaths or food poisoning in Idaho, or complaints that Idaho is undercutting other areas—though there may be complaints that Idaho is undercutting its neighbours.

Look at Australia, British Columbia, Idaho, Arizona, New Zealand in the 1980s and the Reagan stuff in the 1980s. For those deregulatory drives, there is no evidence that anyone has come up with to say, “That was all very discombobulating and as an industry we would have preferred to be in the old ways”, unless they were the people who were making money from having an incumbent position.

The Chair: It might be a bit difficult if you are saying the regulator is great; it may imply that the regulator is not being very regulatory. I think there is a bit of a moot point there.

Q44 Viscount Chandos: Could I pick up one of the things you said in your answer to the first question, Mr Colvile? You said that regulation imposed costs and burdens on business and consumers. What do you think good regulation is for? Surely there are instances where regulation reduces cost on consumers.

Robert Colvile: Absolutely. Taking lead out of petrol imposes a cost on businesses and consumers but I think that we all probably agree that it is quite a good thing. Monopolies regulation is a classic example, where the entire purpose is to lower costs for consumers.

This possibly speaks to some of the things we will discuss later. One of the problems we have as a country is that we do not really have a model of regulation. We do not even know how many regulators there are. We, and others, have tried to come up with lists, and you can probably name about 95% of them. There might be others you have missed out, and there are voluntary regulators of various professions. A key element to this that we do not really think about is the number of people inside companies and other organisations who are doing essentially regulatory jobs. If you speak to anyone in any industry, they will say that the number of people who are sort-of line-managed by the regulator rather than the company has grown significantly.

I think that Sam’s point is a very good one. The purpose of regulation is to reduce harm and increase competition, to effectively keep the economy and society running as smoothly as possible at minimum cost and with minimal interference. It is a vital function. There is a temptation to say we need a bonfire of the regulators. I was talking to someone in education about this. There is a very good reason for having Ofsted. There is a very good reason for having an organisation that is a group of professional people who are good at checking whether schools are working properly, who are not in the Department for Education and answering to politicians, and who have that independence. There are all sorts of areas where that is absolutely the case, such as the General Medical Council. The standard-setting function is a vital one and the invigilation function is a vital one. However, I think that the balance has tilted too far in the country as a whole. By the way, this is not just a UK thing. You can find very similar complaints in a load of other countries.

Q45 Viscount Chandos: Going back to growth, to summarise Dan Susskind’s book on growth, “It’s all about the trade-offs, stupid”. What is your view about the trade-offs that need to be made and how you judge where you strike that balance?

Robert Colvile: Ultimately, that is a subjective thing and it responds to the individual culture. One of the central problems we have is that we have refused to understand that it is about trade-offs. In certain aspects of our regulatory system we have embedded absolutes in a world of trade-offs. Planning and housing, which is an area where Sam and I have done a lot of work, is one of them, and it is incredibly frustrating. The traditional metaphor is a dial: you can set your risk tolerance from one to 10. I think that most people would say that we are now at a two or three and we could nudge it up to four, five or six quite easily.

Sam Dumitriu: To Robert’s point, the way that certain regulators are set up—I do not want to necessarily blame the individuals working for them—and the structure they operate in make it difficult for them to navigate trade-offs in a proportionate way.

For example, we recently had the Fingleton review into nuclear regulation. One of the things that came up was Hinkley Point C’s fish protection measures. Hinkley Point C is a nuclear power station. It will take in a lot of water. A lot of fish will be sucked in with the water that gets taken in. Some fish will die, but it is not a huge number of fish. It is relatively about one small trawler’s worth of fish. It is very important to note that the measures that they are putting in place will cost around £700 million in total to protect these fish. If you look at protected fish alone, they were looking at something like £250,000 per fish—that is for legal protection. I suggest that this is a bad use of money. We would not typically say that that is a good amount of money to spend protecting the life of one fish, even if they have important features for our ecosystems.

There are ways of doing it that are much cheaper and there should be a win-win that could be struck, where the nuclear power station spends a fraction of that money but they save a lot more fish. That almost certainly is possible. One option would be simply to take some of the trawlers out of the fleet for a year or two. That could be a solution. However, those solutions are effectively ruled out by regulation and that prevents us navigating the trade-offs in a proportionate manner. I think that you need a lot more flexibility for things such as compensation and offsetting. You can deliver a much better outcome if you are able to do that.

Viscount Chandos: How do Government change the culture and turn the dial from two or three to five or six?

Robert Colvile: As Sam has said, in some ways it comes down to quite fundamental legal principles. There are things you can do by sending letters to quangos saying, “You need to have growth”. You can make growth a primary duty and I think that we probably should. You can fire the head of a quango pour encourager les autres.

On the point about nature protection in particular, you are butting up against some fundamental legal rules. It is not just fish discos; it is newts and bat surveys for houses and bat tunnels. Look at the wind farms that have to be three-quarters the size they should be because of the risk to particular birds. As Sam says, there is no provision for us to offer a better overall way to protect these animals or to say that we can do a better job at a cheaper cost. You need some fundamental rewiring of the legislation.

The quote from the habitats directive of there being no reasonable scientific doubt as to the lack of adverse effects has now been interpreted by the courts to mean a single extra flushing toilet within a certain radius of some protected woodland means that you cannot build the house or that you have to set up a very complicated system of compensation for it, when actually the things that are polluting the rivers, which is what the regulation is about, are phosphates and fertiliser from agriculture. As Sam said, I do not blame the regulators in quite a lot of these instances. They are following their mandates.

Sam Dumitriu: The most important things for culture are openness and criticism. What we find in general, as a campaign group that can raise some of these issues, is that, often, we will be given information by a company. They say, “Of course, we can never come out and say this, but if you were to look through some regulatory decisions”, at, say, paragraph 12 of page 17 or whatever, “You will find something quite interesting there”. They do not feel confident to raise tangible objections because often they fear that raising those objections might lead to retaliatory action or reputational damage. In some cases their regulatory position is based not just on whether they meet all the requirements of them but on broad organisational competence. Because of that, they are much more reluctant and it becomes very hard to surface some of these problems.

Something that would be very good at shifting this culture is much more information in the public domain, ideally from the companies and the regulators. The regulator should really be saying, “We will give companies prizes if they can identify where we have been disproportionate and challenge us and provide that information”. At the moment, they are not coming forward and it is only through things such as second-hand groups or think tanks that the information ends up coming out. That is a real problem for public debate.

A proposal in the Fingleton review that came out last week that I thought was very interesting was the idea that whenever there was an environmental impact assessment and a mitigation was proposed to deal with some negative environmental impact of a project—provided it is of sufficient size, because you would not want to do it for everything—there should be cost estimates of that. That cost estimate data should be reviewable by organisations such as think tanks and campaign groups and the Civil Service to try to see the cost of all these things. At the moment, unless we get that information—and companies do not always want to reveal these things; they do not want to say that they pay four times more for their environmental impact assessments than their competitors—we have a very poor idea and a poor information environment for challenging and reforming the regulations and making them more proportionate and effective.

Robert Colvile: One extra thing that is worth pointing out is the role of Parliament. Andrew Tyrie wrote a piece for us pointing out that the relationship is between Government and the regulators directly. Parliamentary scrutiny of regulators, with the exception of committees such as this, is not resourced and it does not have the oomph that it needs. He thinks that you could spend, say, £5 million or £10 million on staffing up Select Committees in Parliament with people qualified to scrutinise. This was primarily for financial regulators, but you can expand the principle.

The Chair: This committee made recommendations on that in one of our reports last year.

Robert Colvile: Excellent.

Q46 Lord Teverson: Hinkley C is a fairly contested issue. A Welsh Government report quoted the number of 180 million fish per annum that were threatened by Hinkley C. I declare a bit of an interest because I was at the Marine Management Organisation when Hinkley C started, and we went through part of the regulatory regime for that.

Could I come back on the interesting point about regulators and businesses, and businesses not being like regulators? One thing that always fascinates me in some of the businesses I am involved in is that they volunteer for a lot more regulation themselves—for instance, B Corp, the Forest Stewardship Council, ISO 50001, the Marine Stewardship Council and Red Tractor.[1] It is almost as if they want even more. Perhaps the regulator is looking at the wrong things rather than the right things—I do not know.

Robert Colvile: With those things, you get a clear trade-off. You get to say, “Look at us. We are a high-standards organisation. We are moral. We are good”. It is almost like a transaction.

We have done surveys on the tax regime, for example. If you ask people, “Do you object to the amount of tax you are paying or to the hassle of doing all the admin and regulation around it?”, a surprising number say that the admin time and effort is as much of a problem as the amount of money that they are paying. The surveys show that businesses, by and large, say that the amount of regulation they have to deal with has been going up and up.

Lord Teverson: I was not disputing that in any way. I am just saying that many voluntarily involve themselves in even more regulation.

Robert Colvile: A friend of mine is at a regulator that tried to introduce a more principle-based approach rather than being prescriptive, and the people it was regulating said, “No, can you just tell us what we need to do? If we try to use initiative and do our own thing, you might sue us down the line”. We do not let people get on with stuff and check it later, because we are so prescriptive, and people have learned anticipatory compliance. They do not want to risk doing something that might upset a regulator.

Lord Teverson: I absolutely agree. One example I know of is that, in industrial relations and some financial management, the consultants that smaller businesses have to use are all around risk aversion—and that can be the problem, rather than the regulations themselves. Sorry, the Chair will tell me off in a minute.

The topic I want to get into was one that you have started to mention, which is priorities in regulators. This committee has looked a number of times, in a number of areas, at the difficulty of regulators having a number of objectives. Sometimes they are prioritised, but not always. Does this need to be sorted out and how would one do that? Tell us a little about that.

My own view is that you could say, on the other hand, life is complicated, and we are all struck by a number of objectives that we are trying to achieve. Business may be the same, I do not know.

Sam Dumitriu: It depends. You definitely seem to have some cases where it is a bit like the monkey’s paw: you want to be absurdly rich and all of a sudden it makes you absurdly rich in a horrible way. If you pursue a single goal without considering any other objectives, often you end up in pretty unpleasant situations.

My first concern is that, when you have so many different duties upon regulators, you give the regulator a bit too much of a free hand. That is not particularly good if you think that, ideally, Parliament and the Government should be setting the priority of the regulator and making sure that it does roughly what they want.

Secondly, many of these duties can create regulatory uncertainty in themselves. Recently, there was a lot of pushing for regulators such as Ofgem to have a net-zero duty. A lot of companies were in favour of this, because they believed that they would be much more likely to have investment plans cleared; that even if they would not necessarily going to save consumers’ bills, they could make a very plausible case that they were necessary to meeting net-zero obligations.

However, the problem is that you can introduce rules and they can have unexpected consequences. You might think that it will lead to more projects being approved on these grounds. It might instead mean that someone brings a legal challenge saying, “When you did this project, you failed to consider the environmental impacts of the concrete used when you were putting this together, and you failed to adequately assess this or that”. As a result, you end up with a situation where a project is delayed in the courts while they try to figure out what the Government meant in the first place and what they wanted the regulator to do.

In general, I would like regulators to have fewer duties and the focus to be proportionate. The growth duty is probably the way you do that, but making sure that the growth duty is tightly defined, with much more of an emphasis on things such as cost-benefit analyses and considering actual economic costs, rather than the sometimes general wording you get where the growth duty is seven duties to do broadly nice things. Anyone can argue that what they are doing is good for growth somehow, which is another big problem with the looseness of these things.

To come back to your earlier point about businesses wanting regulation, one of the big issues is that there is often a gap between the larger businesses and the smaller businesses on this because regulation is often a fixed cost. If you have a compliance team, you can manage most of the costs of regulation relatively straightforwardly. You might need quite a large compliance team and you will have to pay them quite well, but you can manage those costs. However, if you are a much smaller business, it is much more difficult to manage those costs and to learn about the regulation coming down the line.

If you are a large company, you could quite cynically say, “It’s a good thing if we get overregulated because while it will reduce our bottom line because we have to spend a bit more on compliance, it probably means that we are less likely to be undercut by a competitor”. Whether or not that competitor is behaving in such a way that regulation should intervene, I am not certain. The classic case is the General Data Protection Regulation, where the big tech companies saw a greater share of the advertising market after General Data Protection Regulation, even though that was very much meant to be a regulation designed to crack down on some of the practices of those companies, and the EU clearly had an objective of reducing their market power through many other means.

Lord Teverson: That is an important point; thank you.

Robert Colvile: One problem is that we do not have a template for a regulator. Regulators tend to be set up in response to a crisis or a perceived gap, and they tend to be remarkably idiosyncratic in how that happens. The then Secretary of State and the then Permanent Secretary, and the people working on the policy—especially the people recruited to run it at the start—have an enormous amount of scope to define what it does, particularly if the duties of the regulator have been defined quite loosely. With some of the stuff that has been handed to Ofcom or the Competition and Markets Authority, for example, on tech and digital in recent years, there were any number of directions in which they could have gone within that framework.

The result is that we have a very heterodox landscape. Countries such as Australia, the US, New Zealand, Singapore and South Korea have central regulatory definitions. They have bodies at the centre that say what regulators should be doing, with a definition of competition and of best practice, in a way that we do not.

We then make that profusion problem worse because regulators are always having to respond to crises and having new duties added to them. There was a great review by Penny Dash, I think, of the Care Quality Commission. It looked at the last six or seven reviews of a big problem with the NHS and said, “Every time this happens, we do a review. The review makes 70 recommendations, and 50 or 60 of them are implemented. That all costs money and takes up people’s time. There is no evidence that the cumulative weight of all these regulations, and the different directions that we have shoved the machine in as a result, have done anything for patient safety. In fact, in a macro sense, it has not improved patient safety because we have had to divert people into fulfilling all of the different recommendations”. That is not saying that these reviews were not important, but that we are adding scale, scope and complexity.

We did a big project on regulation via primary legislation—not regulators but the raw regulations. There is only one government department that even has a list of all of the regulations that it imposes, which is the Department for Environment, Food and Rural Affairs, because after Brexit it suddenly got a whole load of new responsibilities and had to work out what it was meant to be doing. In housing in particular, there are instances where it is impossible to comply with every regulation. There were multiple overlapping regulations, and you have to choose and work out which ones you will comply with.

The solution we have suggested—I think it is still a good one—is a system of policy charters, where you ask every regulator to very simply set out the risks they are trying to avert, the priorities, the worst-case scenarios that could happen and which they want to prevent, and the other things that they want to prevent, in rough order of priority. Otherwise, when you have 20 priorities, you have no priorities.

There was a fad in the Cameron era for getting departments to do business plans. I do not know if anyone ever looked at them, but they were nonsense. They said, “Our eight core priorities are the following, in addition to which we have 30 subsidiary priorities”, because government departments are not companies. Regulators should not have just one priority, but, at the moment, it is very hard for anyone to work out what the core priorities are and what the core input is.

Lord Teverson: Briefly, as you know, phrases from Whitehall are very important. Is “have regard to growth” strong enough for regulators, or should the growth objective be stronger than that?

Robert Colvile: I think it should be stronger. In fairness, “have regard to growth” has only been in place for a few years and we have not seen quite what it means, but George Osborne’s point about the stability of the graveyard is quite a good one. The natural dynamic of regulation is to inhibit activity rather than encourage it, and we need something that counterbalances that.

Sam Dumitriu: Personally, I would like something that is both stronger and more precise.

To come back to Robert’s point about the complexity of not writing down all your regulations in one place, I would quite like there to be, essentially, lists of buckets that say, “This is the specific market failure we are trying to address”. If you go through an economic sector, it will give you a list of the standard market failures, information asymmetry, negative externalities and so on. You have all of these very well defined problems. If you work out what your regulation is trying to tackle and put it on to those problems, it is much easier to see whether the solution that we are imposing is proportionate. The problem is that, at the moment, a lot of regulators do not necessarily see it as, “Am I addressing this specific market failure?”, but rather, “Am I maximising the good? Am I doing something that is broadly good?”

New Zealand recently embarked on quite an interesting reform to its planning system. It is trying to strip out almost all of the rules that are not related to a very clear information asymmetry or a very clear negative externality, such as spillovers on your neighbours. It is removing quite a lot of minimum standards where it judges that consumers can quite clearly see, for example, that the house is quite small and so they do not need a minimum standard, necessarily, to protect them, and so becoming much more proportionate in that way.

You might say that it would be good if people had bigger houses or whatever, but probably the right intervention is not through regulation but through giving people access to more money or making more land available for development. That approach—saying that this is what the regulatory system we have is for, and our planning system is here to navigate the impacts of new development on your neighbours—suddenly changes quite a lot.

If you were to apply that to our planning system, it would probably mean that you would have to allow a lot more stuff and you would probably change a lot of the wording in that. That is something we should be looking at, trying to see where we can make things much more focused on the problem that we care about and the problem that these regulatory systems have come to solve, rather than trying to tackle everything. That is the sad truth.

Robert Colvile: There are huge swathes of London in which it is essentially now illegal to build, and they are some of the most desirable properties in the country. People are paying millions and millions of pounds for houses in Pimlico, Paddington and Marylebone or whatever, but, for various reasons, they are all now illegal to build. I am not quite sure that has benefited anyone.

The Chair: That is controversial, but we will move on.

Q47 Viscount Chandos: Mr Dumitriu, does your observation that smaller companies understandably find it more difficult to cope with regulation lead you to say that regulation should be set at the level that enables small companies?

Sam Dumitriu: Some regulations, inevitably—and they could be good regulations—will burden small businesses more, but they should still stay in place. There are economies of scale in a safety upgrade to your factory. It is probably a good idea that we have factories that are safe. That will burden small businesses.

The core problem is the nature of fixed and variable costs. Regulation is a fixed cost in general—not always, but many regulations are. Acquiring a permit is a lot cheaper, as a share of your total sales, if you are a very large company. Some countries try to get around this problem by putting different regulatory burdens on differently sized businesses. There is some merit to that, but there is a big problem. The famous thing is that, in France, there are a lot of companies with—I am not sure whether it is 49 or 50 employees—just the right number of employees before a load of regulation comes in. You end up with a situation where there is a massive effective tax rate on hiring one additional person. You have to guard against that. However, there are things such as varying permit fees for smaller business. Do you need the same level of scrutiny for a smaller company as you do for a large company? Probably not. That seems sensible to me.

Robert Colvile: We never published this because the pandemic intervened, but we did an exercise using the same pollsters and polling pool as the Government’s annual small business survey, and what we found was slightly different from that. Essentially, hiring your first employee is tough. Suddenly, there is a lot of stuff you have to deal with. Then we found that people’s worries about regulation seem to increase as they grew, until they got to the point when they had, say, 250 employees. At the point where you can pay people to make it go away, it becomes an annoyance rather than an existential thing. However, there is a problem where, in quite small companies, the senior managers are dealing with this stuff and it is distracting them from their jobs.

It also happens at the higher end of the scale. If you talk to someone in a FTSE 100 company—this is one of the reasons we do not have many listed companies—and you ask, “What are you spending your board meetings doing?”, most of them will say, “We are spending a lot of time talking about how we respond to government regulation, not how we grow the company”. That is not what they want to be doing in an ideal world.

Viscount Chandos: We heard evidence in a previous inquiry from a representative of big pharma, saying that countries with successful life science industries all have a strong regulator. That came from a representative of a big company. But if you look at the drug discovery industry, it is all about start-ups. Guess what? Meeting successful clinical trials is the essence of the business. That may be an extreme version, but should not companies large and small see meeting reasonable regulation as an intrinsic part of the cost of doing business?

Robert Colvile: The time and cost of doing a medical trial is one of the things that has gone up hugely in the last 20 or 30 years. I come from the same village as Martin Landray, who did the trial during Covid for a 10th of the price of a normal medical trial. That was something that they could do, using the NHS. I do not know if anyone remembers that but it was a wonderful thing. His burning ambition in life is to bring down the cost of medical trials and make it easier to do, in the way that he did during the pandemic. It is really hard, and that is partly about regulation and regulators and the burdens that are imposed.

Small companies can develop the drugs but by the time they get to large-scale trials, they need to basically sell themselves or find someone to give them huge amounts of money on the off-chance that it will work.

Sam Dumitriu: America seems to have the best pharmaceutical sector for innovation, but it has a lot of features that are not necessarily desirable. They spend an awfully large amount of money on drugs. Perhaps that promotes innovation, but we do not necessarily want to replicate that to the same extent. They also have a regulator that is very expensive to comply with.

Would they be as innovative and as productive if the regulatory regime was more proportionate and they did not have as many rules? Historically, there are periods of time when they did not have quite the same level of rules and they seemed to be doing pretty well in discovery. I am not sure you can necessarily attribute it as that simple.

In general, the aim should be proportion and looking at whether a certain regulatory requirement is necessary. Once you have figured out what is necessary, figure out how you can deliver those at the lowest possible cost.

Robert Colvile: This comes to the point about trade-offs. The best thing we can do for medical research in this country—sorry, this is a personal passion—is to link up all NHS data and turn the NHS into the world’s greatest engine for clinical trials, but obviously the privacy concerns there are huge. Those are two entirely legitimate concerns that you have to trade off against each other.

Q48 Baroness Harding of Winscombe: I would like to take us back to risk appetite. Robert, you mentioned your view that the current regulatory approach is too risk averse. The Government have said that they think the regulatory approach is too risk-averse. We have also heard, though, that the Government can be very inconsistent in the level of risk they are willing to accept, and it changes when risk taking has negative consequences. What role can and should the Government play in setting regulators’ risk appetite?

Robert Colvile: Government always plays an important role in this because, fundamentally, regulators report to the Government. The problem is that, at the moment, it is a matter of signals and retrospective punishments. This is entirely anecdotal, but a friend of mine worked at the Financial Conduct Authority before the financial crisis and said, “I think there is something wrong with the mortgage market”, and someone went, “Yes, we all know that—just shut up about it.” The incentives within the organisation were not to raise the issues because at that point the message was about growth and tax revenue—"It is all going great”.

Equally, one of the spectacles we have all seen is that something goes wrong with a regulator and then the regulators get called in to a Select Committee and beaten about the head. Somone described it to me like this. If you are a regulator, essentially your primary incentive in life, informing your decision-making, is to avoid coming into rooms such as this one—especially a less civilised and more confrontational version. It is to avoid being the guy who has to explain why they did something that led to something going wrong. That means that the incentive is always against risk-taking because, fundamentally, no one ever gets promoted for the fact that they said yes to this company which went on to be a real success. It is great for the economy that that happened but it is not on your key performance indicators. The thing you get punished for is when you let something go wrong.

I think politicians are quite bad about this, because of their incentives. They have passed huge amounts of responsibility to regulators because there is a form of buck-passing here. When things go wrong, it is, “Of course it was the independent regulator’s fault and I will be having stern words with it”.

There is a key theme to everything I am saying. Personally, I want more risk, a lighter touch and deregulation. I am a free market conservative and Margaret Thatcher founded our think tank, but whatever your position is on this, we need to be transparent. We need to have clear mandates for regulators. We need to have clear scrutiny and a clear understanding of the impact of regulation before we do it.

To go back to what we were saying in the last answer, we need it to be clearly stated by the regulator the risks that they are trying to mitigate and the failures they are trying to mitigate—that this is where the dial should be, and have that signed off by the Minister. Then, if it goes wrong, it has gone wrong in the sunlight, and everyone can see what has gone wrong and say that maybe one dial should have been moved back down here and another dial should have been moved back up there. At the moment, it is all private conversations and private letters, and no one is really paying attention until something goes wrong.

My core argument is basically that we need to pay regulation as much attention as tax and spend, whatever our position is on the merits of more regulation, less regulation or different approaches to it.

Sam Dumitriu: One problem is that the buck often does not stop with the politician making the demand for a regulator to have a better risk appetite. The politician making that demand is often in a different role and does not have to appear in front of the Select Committee as a result, because their replacement comes in or whatever. If you are working in a regulator, it is all very well to have a Minister say, “We should be trying to be much more proactive and have a much greater risk appetite”. You are not backed up to the same extent.

I think it would be worth politicians making more of an effort to make clear what they want from regulators. We saw in nuclear the other week the strategic steer the Government gave. They were very clear that they want the regulators, and the industry as well, to have a greater risk appetite, to prioritise outcomes over process and to change their mindset. They were very clear and very blunt on this. Most of the asks that politicians make of regulators do not have that same specificity and strength. The regulators now can stand behind that and say, “This is the steer we were given. Things might go wrong but we were told that we should have a higher risk appetite because we saw that the benefits were great enough.”

It is also important to recognise that often there are different risks to manage. If you take something like competition, there are essentially two risks. It is a bit like if you are testing someone for a condition. There is the risk that you get a false positive and there is a risk that you get a false negative. In certain areas, we need to work out which risks we worry about more.

I would argue that false positives are more concerning in regulatory matters than false negatives. If you say something is a monopoly when it turns out that it is not a monopoly, that could lead to a company basically being destroyed or being taken apart and losing lots of value. If you identify that something is a monopoly but you fail to act at that point, first, there are opportunities in the future to intervene, and secondly, the side effect is that monopolies often create an incentive for innovative start-ups to come along and completely disrupt the industry. There are fewer negative impacts if you weight it towards worrying a bit less about false negatives and a bit more about false positives.

It also needs to be part of the calculation of what we want to avoid. We are probably happy to lose a few false positives around safety, to say, “This practice is probably safe but we will crack down on it anyway because we think there is a reasonable risk”. It is managing that. It is not simple as there being a dial where you say, “Maximum risk”, because often there are two risks that you have to go against and it is, “Which one do you care more about?”

Robert Colvile: One specific example that I should have mentioned is the regulation of the tech sector and the Online Safety Act. We are simultaneously saying, “We want to have an amazing, dynamic tech sector that is the heart of our growth agenda, we want to reserve freedom of speech, and that if anyone says anything nasty on your platforms, we will fine you 10% of global turnover.” Understandably, it is the 10% of global turnover that people’s attention concentrates on. Now you have people saying, “They are being excessive in their regime for producing online content”. Yes, because we have established a regime that pushes it in that direction. No one came into it with that intention but it was the obvious consequence of where the balance of risk lay.

The Chair: It is the impact of the risk as well, is it not? It is not just a percentage; it is the actual impact and what happens if something does go wrong in a very significant way.

Robert Colvile: Yes, absolutely. There is a difference with regulation in cases where catastrophic failure is a risk because it is inevitably going to be more of an issue. In the tech space in particular, there is a problem that it is effectively impossible to police everything that everyone is saying, which is why our regime is designed to be about not the individual content decisions but the architecture of moderation behind that. Even so, the incentives are still very clearly tilted in a particular way.

Baroness Harding of Winscombe: In defence of the Online Safety Act, harm to children is something where there would be very limited tolerance for false positives. I think we have to be very careful in that space and I would hate for this committee to be somehow thinking that online safety does not have catastrophic risk. Having campaigned in the child online safety space, it just triggered me slightly.

Robert Colvile: I completely accept that. At the same time, yes, we can have that as our objective but that implies a whole host of other aspects of how we police the internet that are quite unpalatable. Fundamentally, it is a horrible thing to say but it is impossible to keep everyone safe from everything at all times and we do need to accept that.

I was one of the people who wrote the Conservative manifesto in 2019, and I was looking back at all the other stuff they did. On tech policy, you could absolutely tell which department had written each section because there was a bit from the Cabinet Office about, “We will make it easier for you to get your driving licences online”, there was a bit from the Department for Digital, Culture, Media and Sport saying, “We will do lots of start-ups”, and there was a bit from the Home Office saying, “Nothing bad will ever happen on the internet ever, and terrorism and child pornography will be absolutely expunged from it”. Balancing those tensions is very hard.

Baroness Harding of Winscombe: You beautifully describe the problem. I think I have heard that Government need to be more transparent about what their actual risk appetite is, and regulators need to be clearer and, again, more transparent about how they are interpreting that steer. What else, if anything, can Government and regulators do to make sure that there is greater consistency when you hit upon events?

Robert Colvile: Sam, you can come in as well. The point about trying to measure the impacts more accurately and having some centralised standards for regulators is important. Having time limits and reviews is also very important, as is trying to remember that guidance is not just guidance. People treat it as the word of God and we should measure it in that fashion.

As I said, whatever your position is, we need to take this stuff seriously and understand that people are completely reshaping their companies. On the tech stuff, there is the strategic market status thing that they have introduced. Regulatory decisions made in the UK are, for better or worse, absorbing vast amounts of time at the highest levels of US tech companies, partly because they are worried that other people will do the same. That is, at the moment, butting up against our idea that we want to be a great place where everyone comes and starts their tech companies, and bases their tech companies. At the moment, no one will train a Large Language Model AI here, for example, because the laws around copyright in the UK are uncertain. Whichever way you fall on that, whichever decision we make, we should probably make that decision quite soon, rather than letting it float on.

Sam Dumitriu: It is worth thinking about the kind of decisions you want regulators to consider, and the kind of decisions you want politicians to consider. For instance, if I am the Competition and Markets Authority, I am considering a merger and I think that on the balance of evidence, this specific merger might have a negative impact on competition in this specific market I am looking at on the balance of evidence. That might be the case; however, we know that many start-ups rely on the ability to sell up to a larger company as a reason why they can both attract investment from venture capitalists, and dedicate their time to a project that probably has a greater than 60% or 70% chance of failure.

That is probably not something you want the Competition and Markets Authority to consider every time it makes a decision on a merger. When you are setting the standards for which mergers should be screened and what the balance of probability is, would we go for high confidence, would we put the burden of proof on the merger to prove, with a very good level of evidence, that this merger is not going to harm competition and it should go ahead? These sorts of questions should be for politicians to tackle, looking at the thing as a whole.

Part of the problem is that often we have not done that phase. We get to the phase where the regulator is making the decision on that specific merger, and it says, “I looked at the evidence, I looked at all of the processes, and I think, on the balance of everything I was meant to look at, that I have made the right decision”. That is because there has almost been a dereliction of duty to consider these things. Then they will criticise the regulator for making some absurd decision, but you were the legislator who put the rules in place. Unless you fix those, you will have a lot of big problems.

Baroness Harding of Winscombe: That is helpful, thank you.

Q49 Baroness Valentine: My question is about the action plan the Government have published. The Government brought together their current policies on regulation in the action plan they published in March and the progress update published in October. To what extent do the plan and the update provide a coherent strategy for ensuring that regulators drive growth?

Sam Dumitriu: I will focus on the areas where I have probably the most expertise. First, on the stuff they wrote about planning reform and changing environmental protections, part of the issue is that the Government have a very clear vision of what they want to do, and they make very strong and persuasive arguments to this effect. However, what we have seen is that the legislation coming through does not necessarily match up to the rhetoric in solving some of these problems. That is one concern I have. They have promised a big game on planning reform but the scale of reform is quite incremental at this point. I suspect that they will not see how incremental it is and how they have not reached the standard that they should have until quite late down the line.

The other thing that is interesting in the action plan and is worth discussing is the target set around administrative costs, because it is very important to distinguish between administrative costs, compliance costs, and total economic costs. Administrative costs are important. We should make it as simple as possible to comply with regulations in the forms you fill in, the paperwork you have to keep track of and so on. It is perfectly worthy to have that as your objective. However, administrative costs are not the thing that we need to worry about with regulation. It is the overall cost of compliance.

The administrative cost of the second staircase rule, for example, is not particularly great. You know that you have to build your building with an extra staircase. However, what you lose in floor space and the cost of building that staircase is very substantial. The real thing that we should be addressing is reducing those compliance costs.

Robert Colvile: There is a wider impact in that case of people not bothering to put in planning permission for tall buildings because it is taking nine months, 12 months or 18 months for the regulator to approve them.

Sam Dumitriu: Often people say, “We need to reduce compliance costs”, but then you go to the third thing and say that the simplest regulation for compliance and administrative costs is a straightforward ban. We do not necessarily want to ban lots of economic activity. We have to be careful that this does not prompt people towards more severe regulatory requirements where you do not even have to fill out the form, and you do not have that flexibility to do something provided you comply with X, Y, and Z.

I think the Government’s focus should be on shifting as much of that concern as possible, for reform, on to overall compliance cost. Correct me if I am wrong but I think that when the Prime Minister initially announced it, he said “compliance” rather than “administrative” costs.

Robert Colvile: He did, yes.

Sam Dumitriu: I suspect there were quite a few civil servants who busily said that while 25% of administrative costs is not that hard a target, 25% of compliance costs would require much more serious reform. I would argue that we should be looking at that, given the poor state of economic growth. Targeting admin is fine, but there are limits to this “red tape challenge” sort of stuff and we need to think more seriously about it. Do our regulations have very high compliance costs and are they ruling out economic activity overall?

Baroness Valentine: On your point about planning and the rhetoric not matching what is coming through, what do you think is causing that?

Sam Dumitriu: For instance, let us take the issue of environmental protection, which has been probably the most controversial and most debated part of the Planning Bill Part 3. The Government have the correct vision for what planning should be for major infrastructure projects and big housing developments, which is that they take a strategic look. Fish protection measures at a nuclear power station, for example, might not require you to necessarily protect all of the fish at that exact power station. They would give you some flexibility. The idea is that instead you could protect fish in some other way where you get much more bang for your buck. Then you get a win for nature but a lot of the bureaucracy, impact assessments and cost of designing and implementing these compensation measures is eliminated. That would be a good change.

However, these systems as currently proposed do not bring that in. For very many projects, you are reliant on Natural England putting these plans in place, when it does not necessarily have the capacity to do that, and it can only put the plans in place if they meet X, Y, Z conditions, where X, Y, Z conditions are quite hard to meet, particularly because the Habitats Regulations sit directly underneath. Particularly around species, that becomes very complicated to comply with. You have them talking a strong game about ending bat tunnels, fish discos and so on, but the infrastructure lawyers I have talked to are very sceptical that it will have that impact on the ground, as it is drafted currently. I do not think you are necessarily seeing a huge impact there.

Ditto with some of the planning reform stuff. There is a lot of talk about making it much easier to do certain things, but there are limits to what you can do via changing the National Planning Policy Framework. This is one of those things that we talked about earlier, where everything has “due regard to”, pretty much anything can become a decision where an expert is weighing up every different competing concern. It is a process where there is a lot of political pressure to prioritise certain concerns. Also, because there are so many different competing concerns, it is very possible that you could have a legal challenge questioning a planning decision where they have approved something on the intention of the government policy.

For instance, there was a time a few years back where the Government said that you have a permitted development right to build an extra storey on your property, building up. They tried to define it so that it did not cover every single property, but they wanted it to be a clear tick-box exercise. Councils had a lot of say in it, and very few people took up that additional freedom because it turned out not to be a freedom that they could actually use. In those cases, reforming the planning system is very hard and you need very strong, clear legislative wording. That is not what we currently have because there is a desire to keep the system as one that involves planners, to a very large extent, making their own judgment calls.

The other problem is there have been a lot of concerns around whether you can fix some of the environmental legislative problems that we have. We see very clear examples of waste and inefficiency, things like a 50,000-page environmental impact assessments. The planning application for the Lower Thames Crossing will have cost £300 million to produce, more than Norway spent on its longest road toll, I think about twice as much. You have all of these problems that they know are there. They are concerned that if they attempt to fix them in a more robust way, there would be issues with the EU trade agreement, for instance, around whether that would count as non-regression. We are in a quite a tricky position there.

We go back to the point of risk aversion. Very few government lawyers will encourage you to take a risk on these sorts of things, even if there is a decent chance that it will turn out that the plans can go through as intended, because of the cost of them saying, “We can do this”, and then, down the line, the EU putting up a big fight and there being a big international trade dispute. They are not going to take those risks. As a result, these measures have not been quite as radical as we would have hoped.

Baroness Valentine: To check I have understood, what are you recommending that the Government could do differently, just be less risk-averse on that example you just gave?

Sam Dumitriu: On infrastructure, if you were to take all of the recommendations from the Fingleton review around the habitats regulations, which go further than what is in Part 3, and you were to implement them in full, as suggested—this is, by the way, a document that was produced with one of the panellists being an infrastructure lawyer. There is at least a legal case for doing this. There are well paid, intelligent lawyers who have looked at this and said, “This is possible”. If you were to do that, that would move the dial.

When it comes to the planning system, I think it would be good to put the principle of proportionality in the planning system, and also to have a much clearer statement of, “This is what you can do and this is what you cannot do”, rather than the current approach, which gives enormous discretion. Essentially—and I think the Government have started to use this phrasing—make the planning system as rules-based as possible.

Rules is quite an interesting point because there is a trade-off in regulation. You can have prescriptive rules, “You do this”, or, “You do not do this”, where it is clear what you can and cannot do, or you can have a system where it is very much down to discretion, where you make your application and try to persuade the regulator that what you want to do is proportionate, in line with the regulations, and gives regard to all of these different competing considerations. In some areas you want that because you want the innovation, but in a lot of areas—and I would say planning is one—an inflexible regulator that say yes to a lot of things is probably better than a very flexible regulator that, in practice, does not say yes to much.

Robert Colvile: You asked about the action plan originally. The action plan is best understood as part of a continuum of efforts from Governments dating back to the Blair-Brown era and long before. They all say that they want the same thing, which is simpler, quicker, less burdensome and more predictable regulation. Huge amounts in the action plan are almost word for word what the last Government were doing. It is the stuff I have been talking about, trying estimate the costs upfront.

One of the issues we have is that, essentially, people make decisions and then they calculate the impacts at the last second to justify the decisions that have already been made. There has been a huge amount of thinking and work on how you fix that. Do you try to embed the regulatory work further? Do you beef up the Regulatory Policy Committee? Do you copy what America does, where there is essentially an entirely separate organisation outside the department that marks the impact of the regulations, and then can throw them back at the department and say, “No, we are not going to do this”? There has been a lot of thinking about this.

Sadly, what I think has happened with the action plan is what always happens with these things, which is that it starts off very strong because it has been written at the centre and the people at the centre want growth, dynamism and opportunity, and then it gets out to the departments and they push back. The classic example of this is the Treasury exempting itself from the regulatory targets under the Cameron Government, which meant that huge swathes of economic activity were not covered.

What we have is a set of good things. If you look at all of the things that it says it will do, they are all good things, and they will hopefully save the amount of money they say they will save. However, it is kind of a laundry list at this point now, especially in the update. It is not a fundamental transformation of the way Government do business and think about regulation.

That is partly because Ministers want be Ministers. No one ever got promoted in the private office for saying, “That is a brilliant idea, Minister. However, we just need to go and check whether the numbers add up on those.” They get promoted for saying, “That is a brilliant idea, Minister”. Essentially, regulatory reform is in large part an attempt to force politicians to limit their own powers and follow structures that they do not particularly like. Understandably, that often does not happen.

One thing I did object to, though—it shows the incredibly slapdash quality of so much of the work in this space. Sam has referenced second staircases. We were able to show that the estimates for the cost of that were an order of magnitude out because there was no estimate in the impact assessment for the lost floor space, which was the main concern in industry, because they just said that you can build the building taller or wider. Anyone who has worked on planning knows that you cannot build a taller or wider because this is England and we do not let people do that.

The example I always give is the Markets in Financial Instruments Directive (MiFID) II. If you ask people the cost of that, they would say billions, not a full percentage of GDP but a really big deal. If you look at the original impact assessment, the estimated cost of MiFID II is something in the order of £120 million a year, except it is not because they put the minus sign in the wrong place and they forgot the “million” mark. Technically speaking, MiFID II is meant to give the UK economy a benefit of about £120 a year. I cannot remember the exact figure. However, no one cared enough to spot that and correct it until we came along 10 years later and went, “Hang on a second”, which I think is symptomatic.

In this case, on the action plan—I wrote about this for the Times—what was notable to me was when Keir Starmer announced it, as Sam said, there was a shift from compliance costs to admin costs, which is a massive downgrading. It is cutting it to about one-third of the original ambition.

Sam Dumitriu: A bit less, I believe.

Robert Colvile: The estimates are all so vague. There was still an estimate that if we go by this study of Holland or America from 20 years ago and compare it to the UK now, this could mean that we will save a huge amount of money. Then, when the update came out, that huge amount of money had been cut to something much smaller. I was a bit puzzled by this, so I looked at what they had done. Essentially, they had taken the estimates in the small business survey. What they do for the small business survey is call up a random sampling of companies and ask to speak to the senior person who does compliance. To the people who stay on the phone, they say, “Just off the top of your head, how many hours a week or a month do you and your colleagues spend doing compliance stuff?” The person says, “I do not know, about eight hours”, and they write it down.

What the Government did for the estimate of this cost was take the answer from this survey, multiply it by the number of companies in the country, and then multiply that by the average wage. That is how they came up with the target for what they are saving. Even by government standards, this is an appallingly bad number, for four main reasons.

First, it is a random survey and it is not a very big survey. Secondly, there is no allocation for different types of companies, different sectors or different impacts. It is a raw figure. Thirdly, it is taking the average wage when the survey itself says that the people you are talking to are overwhelmingly senior. They are at director level or board level. If it is a small company, they are the owner of the company. These tasks are not being carried out by people on the average wage; they are being carried out by people on many multiples of the average wage.

The final thing is that the survey—I do not have the figures to hand—has a ludicrously small floor. It says, “How many hours a month do you and your colleagues spend on this stuff? Nought to one, one to two, two to five, five to 10, 10 to 20?” Apologies, this is completely off the top of my head. There are large companies in the tech sector, biotech, and finance—the heavy regulated, big engines of the economy—that are spending decades of time on compliance every month. They have teams of dozens of people working on this stuff, even a number of people who will be following the proceedings of this committee, writing it down, and, when your report comes out, going through it. There is an absurdly long tail of large companies, which are a very small part of this sample, spending huge amounts of time on compliance, which is completely not captured.

Again, I am glad that they want to cut the hassles with compliance, but this is a ludicrously bad estimate and it infuriates me that no one seems to care enough. No one seemed to stop and think, “Hang on a second, the number has gone down by a quarter. Where are we getting it from?” It was obviously just, “We want to make this announcement and we need a number”. There is no punishment mechanism. No civil servants lost their job for this. No Minister has been called to the floor of the House of Commons and told, “What on earth are you doing? This is ridiculous, Rachel Reeves. Pull your socks up.” It enrages me because, again, it speaks to the fundamental thing: we do not take this seriously. You cannot tackle what you cannot measure and we are not measuring it.

The Chair: Noted, but can I just say that we will be running out of time? We have to keep things a bit tight.

Robert Colvile: Sorry. Will do.

Q50 Baroness Valentine: Can I ask for a quick answer to this question? To join up the points you are making, I think there is a Regulatory Innovation Office in the action plan. Your contention that we need to scrutinise regulation better seems, in some way, to be muddled up with this thought of a Regulatory Innovation Office. I do not understand that connection because it seems to me that office is having a look at AI, “Wouldn’t it be fun to use more AI?” and stuff. It does not seem to me to be the same thing, and yet the two seem to be confused in the way that they are referred to in the action plan.

Robert Colvile: There is the classic task of checking whether regulatory estimates are fit for purpose, which the Regulatory Policy Committee does. Another example of how seriously we take this is the Employment Rights Bill, where the Regulatory Policy Committee said, “These numbers are nonsense. You have not calculated all sorts of things. Can you go back and do it again?” and the Government said no. They have not. Keir Starmer was giving a speech, “We will cut the burden of regulation”, even as they were refusing to estimate the cost of their flagship piece of employment regulation. Again, you can think this is a good piece of regulation but at least be honest about the cost. Be honest about the impacts.

Then there is a separate issue. There is an overlap in things like the RegData approach that the Mercatus Centre has produced. Rather than trying to go through a list of laws by hand, you feed the whole thing into a Large Language Model and say, “Look for all of the phrases like ‘have regard to’, ‘must’, or ‘should’”, that kind of command language. That should spit out a list of obligations on companies, which is good.

You are right. How we regulate innovation is a fascinating thing, and we did a paper on this. Our contention was, effectively—and I am speeding through because we are short on time—that there are two different types of innovation, born free and born captive. There is stuff where people come up with something and no one has ever thought of it before. Crypto would be a great example. People go off and do that, and then the law catches up. That works quite well and generally it does not destroy the economy.

There is also born captive, which is where people are trying to do things within the existing regulatory space. That is very hard, especially where things cross. If you think about it, if you are using a drone to irrigate crops, which regulator is responsible for that? You can make a theoretical case for about four or five. If you are coming up with a financial product that is pensions but also insurance but also a savings product, which financial regulator do you go to? Who writes the rules on that?

In those situations, where things cross boundaries or you are asking people to carve out exemptions within existing spaces, we fall down. In one example, the person who had come up with a product literally said to the regulators, “Can one of you pick a regulator that it should be and let us do this?” They were like, “No, you go off and do it, and then we will sue you if we do not like what you do”. That is not how we should be regulating.

That is an interesting area to think about, and it is something that comes up in the Fingleton review. It is not just the regulator and the thing; it is the fact that there are huge areas of activity in our economy and our life that are different things at the same time.

The Chair: Can we move on to Viscount Trenchard? We will have to be brief in answers, if you will.

Q51 Viscount Trenchard: We are already moving on to the territory of my question, relating to whether the action plan and its commitment to cut administrative costs for business by 25% by the end of this Parliament is indeed sensible, realistic and achievable. You have already started talking about that but I would like to extend it a bit more.

I know very well that if you look at the executive committee of regulated businesses, the people who do most of the talking are not the people who are developing the business, who have fantastic ideas to sell a new product and to make lots of money; they are the compliance officer, the finance director and the in-house lawyer, and most of the talk is about, “Does this comply with regulation? How do we avoid getting into trouble with the Financial Conduct Authority?” or, if you are dual regulated, the Prudential Regulation Authority.

I do not want to trespass on to Lord Best’s territory but in our inquiry into the building sector, we were told that building approvals are taking much too long because we do not have resources. How can the Government on the one hand say that you have to cut by 25% the administrative costs of regulation and on the other hand say that it is unacceptable and that you need more people in regulation? There is a huge conflict here.

Sam Dumitriu: I will address this on planning, which is an area I know a lot about. You often hear people say the problem that we have is we do not have enough planners, they cannot get through things quickly enough, and that causes a big delay. Recently I visited a place where there was a planning application out. It was going to be 14 flats and it was in what was called an opportunity area in London. In theory, it is the place where we want to build in London, where we have decided, “This is the best place to build”. It had a planning application that was 1,250 pages long, with 40 to 50 different validation studies. I think it is perfectly consistent to say that you could cut admin costs and get through stuff much more quickly if you were to say, “Do we need those 50 studies?” Instead say, “Look, do not do those”. Then, all of a sudden, you will find that the people whose job it is to check whether those studies were done properly have a lot more time on their hands.

We should be thinking about whether these things are proportionate to the task. We visited another building in the same area that was built in, I think, 1937, the same height with the same number of flats, give or take a couple. It was approved within three weeks and rubber-stamped a few days later. It had a three-page planning application with some drawings and some very clear rules. It is perfectly possible to cut admin costs and to have regulators with a lot more capacity.

Robert Colvile: Based on historical precedent, especially with the shrunken definition, it is possible to cut admin costs by that much. Doing so can be beneficial not just for the companies, but for the regulators. A huge amount of planners’ time is spent doing very basic, entry-level tasks. In financial services the head count at the regulators has gone up, but it is not a fun job for them. Planning in particular has huge staffing issues. Everyone says we need more planners; we probably have more planners, it is just that we make them do a pretty miserable job. We make them print out PDFs and transfer them through multiple incompatible computer systems. Planners want to be planning and designing, not signing off roof extensions and side returns. Basic principles like, “If someone else on the street has done it already, it is probably okay”, can cut through quite a lot of tension.

Q52 Lord Best: I will take a different example from the world of planning and housing, which is the Building Safety Regulator. You have had a look at this. Is it not the case that the number of people to do the application checking and then to issue the approvals was so small compared with the task? We can say we should reduce the task, but not many people think that Grenfell did not reveal a need for more regulation and more action. Is it always the case that you can get away with fewer staff by cutting the number of duties that you give them to perform?

Sam Dumitriu: There is an element of Parkinson’s law, I think it is, where when you expand the number of people to scrutinise an industry, whether that is building safety or pretty much any sector, you get to a point where they scrutinise more things. There is not always a direct correlation of the sort that this is a well-staffed regulator; therefore it will give you a decision on time and do a satisfactory job. When you have all of those staff, you have to justify their existence in certain ways.

The Building Safety Regulator had a pretty unique set-up by international standards. There was not a very clear idea of what would be required and we did not do a good workforce plan to know how well we needed to staff that, based on the requirements we were asking for. As a result, you have ended up with a situation where it has not been able to get through things very quickly.

There is a genuine question about whether the structure lends itself to doing this in an efficient way and whether it would have been better if we had kept the previous system but funded it much better, with a few very specific regulatory changes to tackle things that we know were not working, which would have been an alternative model, or whether something would have been a standard-setting thing, given existing capabilities.

We talked a bit about the importance of regulatory stability before. When you set up an entirely new regulator, you tend to encounter problems that you have not foreseen. In these situations, is it better to take existing forms of regulation, existing processes, funding and resources?

Lord Best: I think we can agree that there are various other factors, but are you saying that there is no case for an increase in the admin costs or resources going to the Building Safety Regulator for its staffing? Would you never concede that a regulator could need more staff?

Sam Dumitriu: My understanding is that when we talk about admin costs reduction, we are not talking about reducing the amount spent by regulators but the cost to businesses.

Robert Colvile: It is the cost to business. If things have been identified as national priorities, such as novel foods or Medicines and Healthcare products Regulatory Agency (MHRA) medical devices, and if there is a customer service element, you want to have enough staff to make sure that the customer service element can be done quickly and efficiently.

The problem with the Building Safety Regulator was not just the staffing. It was essentially approving every single planning application, as if no one had ever put in a planning application before. There is an excellent report—I know we are short on time so I am not going to quote much of it, or any of it—by Maxwell Marlow for The YIMBY Initiative [2], which had MPs from three or four different parties endorsing it, setting out all sorts of ways in which the Building Safety Regulator could and should have been done better. There was a complete failure to understand the economic impact and the impact on industry of what would happen. It has been absolutely disastrous and it has left things more unsafe. More people are in older, less safe buildings because we have stopped building new ones.

Lord Best: There is another excellent report on its way from this committee on the same subject.

On how you speed things along, there is this idea of having paid for fast lanes, which is criticised on the grounds that it would help the big boys, who have compliance departments and so on, to whizz along, but the small fry SMEs would be disadvantaged by not being able to pay or finding it difficult to pay the extra levy. Do you have any sympathy for the idea of people being able to pay for speed?

Sam Dumitriu: Yes, because you will see in general that the cost of a regulatory delay is much greater than the cost of paying for a fast lane. In the US, for instance, you can get a much faster Food and Drug Administration decision if you are a drug company and you contribute to some other cause. This is how they get loads of drugs for the developing world. If you are doing this, it suggests that there is a high value that people place on shaving a few months off. If you are a start-up, for instance, you will have a very limited runway and you will be burning through cash at a very fast rate. If you can spend that money and have a product out the door before you have burnt through all your cash, that is a very useful thing. In housing, you have a very high cost of capital relative to other things because you are typically very leveraged when you have bought the land and so on. The ability to generate that return suggests that it is probably worth it.

We need to be very careful to make sure that if you pay a regulator for a service, it is very transparent on how it is charging you, it can provide you with an estimate saying, “We spent this amount of regulatory time looking at this. We are doing this, this, and this” in an itemised format, and you can hold it to account on whether it is doing its job. My concern is that where people can increase planning fees, or you pay for the regulator’s time to do certain things, they are not necessarily getting a service that they might get in, say, a competitive market, if they were dealing with another private company.

Robert Colvile: In an ideal world, everything would zoom through and everyone would get treated equally. You can completely understand why people pay for speed. It does advantage some firms over others, but at least it gets some decisions made. One of the worst things that we have at the moment, especially on planning, is a unofficial system where planners will say, “We are not going to get around to this one. We do not want to not hit our targets. Could you ask for an extension at your end? Otherwise, we will turn it down.” That is the worst of all possible worlds.

Q53 Baroness Nichols of Selby: In this ever-changing world—that is probably where I want to start—how do you think the wider changes in data technology and AI are changing how companies and regulators operate, and what opportunities do they provide to improve their efficiency or effectiveness?

Sam Dumitriu: In theory, a lot of what you do in regulatory work could be greatly sped up by AI and the ability to analyse lots of different documents. In things like planning, the AI should be able to give you roughly where the regulator would approve or disapprove of a certain thing, so that you would have that certainty without having to go through the whole process and have that back and forth, for instance.

However, we need to be careful. Regulation seems to have become more complicated with the advent of the word processor. It also becomes much easier to increase regulatory burdens in the same way. In planning, we have noticed that a lot of organisations have used AI to make it much easier to object to a planning application. You could even draft certain forms that will imply that it would be easy to bring a judicial review or something like that. We have to be very careful about how people can use these technologies.

The most important thing is the benefits that things like AI can have in making our lives better and bringing new products to market. There, we need regulators to be much more nimble and willing to approve things, and that is very difficult when you are an existing regulator that has a very set brief.

Robert talked about people who are captive to the regulator, who are working in a certain industry but may want to do something that cuts across three different regulators. One idea here is to create a super regulator. You could either go through your existing regulatory channels or, if you are a particularly innovative business that is hard to pigeonhole, where we do not know which bucket to put you in—maybe you are doing a bit of fintech but also some very innovative things with data, or an innovative thing here—and we cannot manage that very well, do it quickly, and approve what you want to do, you could go to this super regulator, which would act as a general, economy-wide sandbox. It would say, “We will indemnify you against various regulatory risks. We will be your regulator and will approve whether or not you do things.” That could potentially be a very exciting route for people who want to, say, combine AI with law or AI with medicine in ways that legal regulators or medical regulators are not particularly well suited to.

Robert Colvile: In theory, AI should be an immensely helpful in regulation and deregulation because it is the kind of problem that Large Language Models were set up for: very long-winded and complicated, but also quite simple. It is just a list of sentences and instructions that can be boiled down. However, as Sam said, the risk is that you end up with AI generating vast amounts of verbiage on both sides, and the whole thing becoming incomprehensible to humans. I already referred to the RegData approach, using innovative ways to scan the existing corpus of legislation. That is really interesting.

More broadly, what is happening within the tech sector illustrates quite how slow moving a lot of our regulatory structures are. The Competition and Markets Authority seems to be engaged in trying to regulate the tech world as it was about four years ago when we were talking about the FAANGs [3], and now you have OpenAI, Anthropic and the Chinese companies. You have this incredibly disruptive space and the growth of data centres and all sorts of other stuff that we are not in the mix for, partly because tech companies are quite nervous about our regulatory environment.

The Chair: We have three more people who want to ask questions, and a very limited amount of time. Can I ask for very direct questions and very direct answers, please?

Q54 Lord Udny-Lister: My question is something you have touched upon. Do government departments have the right tools and information to hold a regulator to account? The Government have said they will strengthen regulators’ accountability, publish key performance indicators, and expect sponsor departments to do more reviews. Does any of this help a department review a regulator?

Sam Dumitriu: In general, that would be helpful but it would not be sufficient. The key things are the examples that you only get when you have an industry-wide review or someone slips something to a think tank because they do not necessarily want to reveal it themselves. Getting that information out should be the key thing.

One idea I will suggest is the concept of a regulatory red team, essentially people inside regulators whose job is to challenge internally the regulatory decisions, provide that feedback, and publish as much information as possible so that politicians, legislators and people in government can all see exactly what is going on and where those points of challenge are. It works very well in private sector organisations for lots of different functions; I think it would work well for regulation.

Robert Colvile: I think the key performance indicator is accountability. It is a good agenda. Internal challenge is vital. Speaking to people who work in regulators, in some ways having a direct channel to the Secretary of State is quite useful because it allows you to hear and work out directly what their priorities are rather than having that filtered through different layers. However, just to repeat myself, there is also a role for Parliament. Parliamentary scrutiny of regulators has not been what it could be, and I welcome everything this committee is doing in that direction.

Lord Udny-Lister: I have a quick supplementary, if I could. Is it clear what good looks like for a regulator?

Robert Colvile: Not really. What is the incentive? Is it to do what the Secretary of State wants and make them happy? Is it to act as a consumer champion? Is it to protect the public? There are multiple conflicting things and, as I referred to before, every regulator has wildly different set of priorities and standards. For quite a lot of regulators, not being noticed is probably what good looks like.

Q55 Viscount Chandos: Regulators have said that they need legislative change to follow the new directions. Do you think that is fair or is that them looking for an excuse? A lot of regulators have their hands tied effectively. If you are the Office for Nuclear Regulation, you are very much tied by health and safety regulations and legislation developed a very long time ago, which might not be particularly appropriate to what exactly you are trying to do. If you are an environmental regulator, what you are doing is very much shaped by what the Habitats Regulations that we have put into law say.

One thing that would help here is sunset clauses. Parliamentary time is limited and getting primary legislation to knock some bad regulation on its head, when there is not necessarily a direct vehicle for that, is sometimes very difficult. It would be good to at least remove some of that friction if you want to get rid of a rule that has not worked. The problem is that at the moment we have a lot of rules that have not worked, but there will not be time to fix them.

Robert Colvile: One of the challenges is that if you are trying to reform the regulatory apparatus, it will be impossible to go quango by quango, regulator by regulator, because there are so many of them and they are doing so many different things. Equally, if you try to come up with one big beautiful Bill that has a lot of reform efforts, because all these quangos are different things doing different things, the worry is that you break things into individual regulators that you did not anticipate. You probably need to do some big, central things rather than go one by one, but you also to be conscious of the sensitivities.

Q56 Lord Teverson: The Government have made their intent clear: they want to look at merging regulators—the nuclear industry is an example you just gave—bringing responsibilities back into departments, or abolishing them altogether. Quickly, what are your views on that? Is that just rearranging the deckchairs on the Titanic? Does it mean lots of distraction without action, or would it get over this issue of multiple regulators for individual businesses?

Sam Dumitriu: In certain cases, bringing stuff within departments, where there are very clear reasons to do so, makes sense. Part of the problem is that you seem to have a situation, for instance, with Natural England where the Government want it to do one thing but Natural England is not doing that thing, and their hope is they can change the culture within. It seems to me that it would be much more effective if they had direct control, and that is what you see in lots of different countries. Not everything has to be completely independent in the way that certain things need to be independent.

I am genuinely sceptical of mergers. I find that they do not necessarily solve many problems in regulation. Often there will be three regulators dealing with the same topic and they all have slightly different standards of evidence. For instance, if you are getting permits for a nuclear power station, you will have something from the Environment Agency and something from Natural England. They might have a slightly different standard of evidence. Something like the commission model that that the Fingleton review suggests, where you essentially have one regulator sitting over who can engage and deal with these disputes, might make a lot of sense, and that could be a model applied to different regulatory areas.

Lord Teverson: Is there a good example, quickly, of where that has happened?

Sam Dumitriu: Not in the UK, but that commission approach is used in a lot of different places.

Robert Colvile: Theoretically, it is a good idea to rationalise and bring functions back. In practice, it results in significant disruption. I think I saw reported that they have only three full-time equivalent people working on this, which someone worked out was 186 quangos per person per year, or something insane. That suggests it might be a slightly Potemkin process. If you look at what is happening in NHS England, which is obviously the largest regulator, that is turning into a bit of disaster. The chance of it hitting its target for April 2027 is pretty much zero, people would say in the sector, and that has had a knock-on impact on the state of the NHS.

There is another example—it is not quite regulators—with planning. The Government have said that their core priority is to deliver planning reform. They have also embarked on a process of combining councils into unitary authorities, which means combining planning authorities. We have a local plan-based planning system. Loads of councils are not producing local plans because they do not know what the councils will be, they do not know who will be writing the plans and they do not know who will be in charge. If you are a planning officer, the incentives are not to write a local plan that says, “We will build loads of houses”, because your new boss will probably be someone who says, “Let us build the lowest possible number of houses that we can”. It is inevitably disruptive but you need to do it at some point, I suppose.

The Chair: We have gone through quite a few issues this morning. Thank you both very much for your evidence, and I will close the public part of this session. Thank you.

[1] B Corp is shorthand for Certified Benefit Corporation, which is a for-profit company verified to meet high standards of social and environmental performance, transparency and accountability. ISO 50001 is the international standard for energy management systems issued by the International Organization for Standardization.

[2] YIMBY stands for Yes In My Back Yard and is used for supporting increased housing development, in contrast to NIMBY or Not In My Back Yard.

[3] FAANG is an acronym for five leading U.S. tech companies: Facebook (Meta), Amazon, Apple, Netflix, and Google (Alphabet).