RTM: why should there be a right to prevent it?


The Right To Manage allows the owners of flats to appoint the manager for their apartment block. RTM was introduced in England and Wales in 2002, since when a few thousand blocks have taken it up. As there is no effective legal right to inspect invoices behind the service charges for flats, the power to replace the manager is generally the only available means of preventing unvouched or fraudulent charges.

But: who should be allowed to prevent RTM? That is, which people should collectively or individually be empowered by law to compel others to suffer being defrauded, exploited or overcharged?

Minimum participation rate

RTM works by transferring a block of flats' management rights and obligations to a leaseholder-controlled company, the "RTMco", the leaseholders being the owners of the flats. All UK-registered companies must have at least one member, and so at least one of the flat owners must opt-in to being a member. These members, however many there turn out to be, elect directors, who appoint a manager. Despite the simplicity and clarity of this arrangement, attempts to muddy the waters about it are so frequent that they practically constitute a minor literary genre in their own right (see the link below about the "Agency Canard").

So the minimum number of flat owners who must be members of the RTM company is one. And membership is not particularly burdensome. However, for the RTMco to claim the management rights in the first place requires it to have had no less than half the flats represented in its membership on the day that it made its claim. Consider the following worked example: for a building with sixty flats, an RTMco would start with two or three leaseholders as founding members. They would then recruit at least half of the flat owners, i.e., thirty members, and get them formally signed up as members of the RTMco. The RTMco could then make its claim to manage the building. That takes a few months to kick in, but it will still occur even if most of the RTMco members subsequently resign.

To recap:

  • incorporating an RTMco takes just one flat owner
  • claiming the management of the building requires half the flat owners
  • but retaining the management long-term requires again just one flat owner.

Why should there be a "half the flat owners" hurdle in the middle of the process? If the other flat owners are dissatisfied with the RTMco, they should join it and sack its directors. After all, all the flat owners have a statutory right to be members of the company; I've never come across a case where the directors of the RTMco have refused a valid request for membership. If they did, they open themselves up to being sued. If this is a genuine problem, which is unlikely, then refusal of a valid membership application should simply be criminalised on the same lines as almost all other breaches of company directors' duties are: a strict liability offence punishable by a daily fine.

The only parties who benefit from this fifty percent hurdle are the outgoing managers. The only effective remedy for bad managers under UK leasehold law is replacing those managers. The harder it is to replace them, the easier it is for them to overcharge. Therefore the hurdle should be minimised: reduce the participation threshold from "not fewer than half the flats" to "one". To do otherwise is to incentivise mismanagement.

Analogy: jury duty

Nothing about being in an RTM company really imposes a "duty" as opposed to a right. The dishonest conflation of the right to manage with the obligation to participate in management is the core of "Agency Canard" detailed in the link at the bottom. We don't say "if you get the power to choose who your electrician is, that means you have to train as an electrician and do the work yourself". But that's basically what people try to say about managing blocks of flats: the power to choose who your managing agent is somehow the obligation to do that work yourself. It's piffle.

Nevertheless, someone has to serve as a director of the RTM company. It doesn't have to be a flat owner, though that is probably the norm. It definitely doesn't have to be any specific flat owner, though.

Where this leaves us is with a nebulous and largely spurious sense of obligation hanging in the air when it is contemplated to obtain control over the management of a block of flats.

But fundamentally this is a collective governance issue. No-one seriously claims that people have a right not to serve on juries, or that because some people aren't willing to be members of Parliament or local councils, that no-one else should be allowed to. But that's very similar to arguments made against RTM, and in practice, the fifty percent hurdle is just a disguised restriction on RTM: the less one likes RTM, the higher one wants the hurdle to be. Effectively, one's saying "I don't want to serve on a jury so even people who would be happy to do so shouldn't be allowed to either."


Collective ownership structures for British apartment blocks


For common ownership of blocks of flats in England and Wales, there are only two viable solutions: share-of-freehold, and reformed commonhold.

The two leasehold models

In England and Wales there are two primary patterns of ownership for blocks of flats which occur when each flat is owned by a different person:

  • conventional leasehold (never mutual)
  • share-of-freehold (potentially fully mutual)

Conventional leasehold involves each flat being held on a long lease by its owner, and someone else owning the freehold of the flats and the common parts of the building such as corridors and the roof. This accounts for most of the so-called "private" housing stock.

"Share-of-freehold" involves some or all of the flat leaseholders, and possibly other parties, jointly owning the freehold over the flats and the building's common parts. Where all the leaseholders have a stake in the ownership, and no-one else does, it can be termed a "fully mutual" arrangement. Although the flats are still leasehold flats, most of the pathologies attendant upon leasehold are absent, because the interests of the owners are broadly aligned.

Fully mutual share-of-freehold arragements operatee, in practice, so differently from non-mutual arrangements that it's often misleading to generalise between conventional leasehold and fully mutual share-of-freehold. The incentive structures are barely comparable.

Alternatives to leasehold

There are three "viable" alternatives to leasehold for blocks of flats:

  • commonhold
  • strata title
  • company title

None of them is currently viable in the UK: commonhold requires secondary legislation or unusual private arrangements to make it viable even on the dozen or so sites which currently use it.

Company title is effectively the predecessor of strata title in Australia. The owners of the building form a company to own the freehold of the building, and each share in the company comes with the right to occupy one of the flats, and there is no lease in addition to the terms of membership of the company; all matters concerning the respective rights of each flat "owner" vis-a-vis each other and vis-a-vis the group are determined on the basis of the company's internal rules.

But look at it another way: company title is basically share-of-freehold without actual leases on the flats. The rules that would appear in an English lease instead feature as provisions of the company's articles of association or of regulations made by the company. With one crucial difference: leases are by definition time-limited, whereas there is no time limit on owning shares. So a share in a company title block of flats is not a wasting asset, but a share in a conventional English share-of-freehold company is.

However, while you could set up a company like that today in England, it would likely be difficult to get a mortgage against a share certificate rather than against a property title. The arrangement is simply unfamiliar in the UK and mortgage lenders would likely decide it was in their interests to turn up their nose at it. So company title, while possible in the UK, is likely even rarer than commonhold.

Why not strata title for the UK?

Strata title, which is the norm in Australia and elsewhere, however, is simply not available in the UK. Commonhold and strata title are very similar: they provide for freehold ownership of individual flats when those flats are vertically one on top of the other, and a fully mutual corporate body for owning and managing the common parts.

So what's the difference between strata title and commonhold?

Strata title and commonhold principally differ according to the type of legal body that comprises the owners' corporation. Under commonhold, this corporation operates under the general company law used by hundreds of thousands of non-profit and millions of for-profit enterprises; it is no different in this respect from share-of-freehold or indeed company title. Strata title on the other hand has a specific type of corporate body that operates under laws that only relate to managing blocks of flats.

This gives rise to a much bigger practical difference in administration, in particular, what government bodies must be dealt with, and whether other economic groups share any of the administrative and dispute resolution facilities.

Commonhold fundamentally involves using a UK registered company as the owners' association for the block of flats. As a company, it has to be registered with Companies House and annually confirmed by its directors. In my opinion, this is better than strata title, because for the small cost of complying with company law, the owners of the flats get all the benefits of being first class corporate citizens in the UK: all the facilities for ensuring democratic governance, finding out the identities of other members, decision-making, audits, accounts, electing and sacking directors, and so on, is shared with five million other UK businesses. Companies House, the company registrar, even provides a very swish API that enables people to manage their company data using third party software, to cover situations that Companies House itself can't or won't support.

There are alternate schemes in the UK legal system which allow groups of individuals to manage their affairs collectively: partnerships, registered societies (think Nationwide Building Society, credit unions and housing associations), trusts, associations, etc. None of these bodies of law is kept up-to-date by Parliament the way company law is. When the pandemic hit, all UK companies were already able to conduct valid shareholder and board meetings electronically, due to legislation passed twenty years ago. Associations and trusts that hadn't foreseen this found themselves hamstrung. Commonhold makes the best trade-off available. Having a bespoke set of laws for the management of owners' corporations for blocks of flats guarantees that they'll become second class corporate citizens, and fall out of the slipstream of big business and small business that currently keeps their governing law up to date.

So commonhold and strata title are very similar in practical outcomes from the perspective of the relationships that flat owners have with each other and with third parties. But they're not the same, and people who say that commonhold exists outside the UK, equating it with strata title or similar, are really muddying the waters. This leads to unfortunate confusions, such as how internal governance works. Many strata jurisdictions require that there be an annual general meeting, but this requirement was disposed of in the UK to help small businesses; the UK model is much more that company members elect (and if necessary sack) directors to run their companies, but the thresholds for requisitioning general meetings to overrule the board unfortunately do not work for companies with more than 20 members where a single member is aggrieved. This could be trivially fixed via secondary legislation.


The most important factor for the welfare of owners of flats is whether the block is owned on a fully mutual basis. Otherwise there is significant risk of conflicting interests.

In England & Wales, the only realistic means of having blocks of flats owned on a fully mutual basis are share-of-freehold and commonhold. There is no chance of strata title being introduced in the UK because commonhold is so similar, and no chance of company title taking off because it's so weird and runs up against a century of land registration and mortgage-lending practice.

Why commonhold can't be mandated on existing blocks


Who would be upset by abolishing leasehold? Actually, some leaseholders who can't afford their share of their freeholds would. And people shouldn't be compelled to pay what they can't afford, so it follows that commonhold should not be made mandatory on existing blocks without some mitigation of this issue.

This blog post is partly an answer to James Creedy's tweet.

Some leaseholders cannot afford to buy their share of the freehold of their block of flats, even if that resulted in permanent ownership and zero ground rent.

Online discourse about long residential leasehold in England and Wales often alights upon abolishing leasehold entirely.

There are unresolved issues with outright, all-at-once abolition, or at least I've never seen any plan which resolves even the issues I'm aware of. So I avoid saying "abolish leasehold". I would go for the much less snappy wording: "phase out the vast majority of residential leasehold". So what do people mean by "abolish leasehold"?

Roughly, what is meant by it, is that all residential long leases should be replaced by freehold ownership of the same space, and that any common parts of buildings or land shared between multiple flats should be owned on a mutual basis by the owners of the flats. The complications of having multiple financially interdependent freeholds vertically stacked on top of each other would be addressed by schemes such as "commonhold" and "strata title".

Now under the existing law, freeholders must be compensated for their legitimate property interests when their leaseholders compulsorily purchase the building's freehold. To the extent that freeholders or managing agents can extract additional income illicitly, above and beyond their legitimate property interests, then indeed those freeholders and agents would be upset by the abolition of leasehold.

But the other group that might be upset comprises some of the existing leaseholders.

Let us suppose that every residential building were eligible for enfranchisement. Under the existing law, not all such buildings qualify, but suppose those restrictions were removed. All leaseholders could then choose to buy the buildings in line with the procedures in the current law: at a time convenient to them, and a price they would agree with the freeholder, appealable to a tribunal.

There are three main components to this price:

  • the future value of each flat
  • the future ground rent
  • any value in the remaining land that's not part of any lease

If the leases are very long, e.g., hundreds of years, then the future value of the flats will be tiny today. If a newly-leased flat is worth £300000 today, but the lease is 250 years long, then the freeholder doesn't get to sell another £300000 lease until the year 2273. So if you bought about £1.50 of government bonds today and cashed them in in 2273, you'd be able to buy that flat. But the longer the lease is, the more ground rent may be due, so to compensate the freeholder for 250 years of £100 ground rent, you might need to pay £4999 on top of the £1.50. The outgoing freeholder could then stick that money in bonds and get an income stream equivalent to what would have arisen from owning the building.

What you'd get for your £5000.50 is a freehold flat (likely a "commonhold unit"), which would have zero ground rent forever. In practical terms it would be £100 cheaper per year, and you could use that £100 saved to gradually pay off the loan you may have needed in order to finance the transaction.

If you were forced by law to accept a swap, of your lease with its £100 a year ground rent, for a debt that cost up to £100 a year to service plus permanent ownership of your flat, you wouldn't have much cause for complaint.

But consider the case where the lease doesn't have very long left to run, e.g. only 50 years. There are fewer years of ground rent to pay, but since the reversion of the flat to the freeholder is closer in the future, much more money must be invested in order to pay for it. To buy a £300000 flat in 2273 might cost only £1.50 today, but you'd have to invest about £26000 today to buy such a flat in 2073. (There is also currently marriage value to be paid on short leases, though this is irrelevant to the wider point).

There's no way that a reduction in ground rent will always offset realistic repayments on a debt of tens of thousands of pounts; after all, the ground rent might well be zero! So some leaseholders in this position would be bankrupt if forced by law to accept a swap of their lease for a ground-rent free commonhold unit plus a loan for the cost of enfranchising it.

Is anyone serious proposing something like this?

The Law Commission discusses it in some detail, from paragraph 5.136 of their report on Commonhold, and it forms part of their Recommendation 12. The report sets out two main models for reform, one of which is the antithesis of abolishing leasehold. It is "Option 1", and entails watering down commonhold to allow long leases and ground rents, avoiding the problem I've outlined above. Option 2 entails compelling flat owners to accept the swap if enough of the other owners in the building want to convert to commonhold.

The assumption in all this is that most but not all of the flat owners are banding together to convert to commonhold at a time of their own choosing, rather than the government making them convert against their will.

In practice, the options are:

  • retain the requirement for unanimity in converting leasehold units to commonhold units (the status quo)
  • abolish the unanimity requirement (Law Commission's "Option 2")
  • abolish the unanimity requirement except in genuine cases of unaffordability (my own preferred option)
  • abolish the requirement for consent entirely ("leasehold abolition" tout court )
  • don't abolish leasehold after all, but instead allow it to persist within a watered-down commonhold regime (the Law Commission's "Option 1")

The first of these options is effectively the position of the Labour Party. To be fair to the Labour Party, they are publicly committed to abolishing leasehold outside the social and affordable housing sectors. I'm not sure what the Conservatives' position is, but philosophically the party has generally been for gradual change, tempered by occasional drastic shock therapy, which may provide some hope.

I think it's important for commonhold not to get associated with compulsion. Nevertheless, realistically, in the generality of cases, it will be in the financial interests of all leaseholders in a block to convert to commonhold, so so long as it's not genuinely financially detrimental, removing veto players from that scenario would probably be justifiable.

All feedback gratefully welcomed.

Survey of Residents Management Companies (updated)


I recently surveyed the UK's resident management companies (RMCs). There is a proper paper write-up here.

You can do this by querying data provided by Companies House, plus a lot of spadework. The code for doing this is currently here.

Roughly speaking, there are about 50,000 RMCs. I took a random sample of 100 of these and manually read through all of their articles of association. It took many hours of unbelievable tedium.

What is an RMC?

RMC stands for one of the following:

  • resident management company
  • residents management company
  • residents' management company

Note that the first option worrying includes the implication that the company might be used to manage the residents, rather than to manage property on behalf of residents.

Anyway, regardless of what the acronym RMC stands for, there's no official definition of the concept, but there is a generally accepted usage of the term, which is that an RMC:

  • is restricted in its activities to the ownership or management of collective residential property in a particular location
  • is restricted in its membership generally to those who own the property so managed

In practice, RMCs tend to be used for one or a combination of the following purposes:

  • owning the freehold of a block of flats jointly between some or all of the leaseholders
  • separating out the management responsibilities for a block of flats into a distinct legal entity, with or without involvement by the owners of the flats
  • doing the same in relation to some combination of blocks of flats and freehold houses on a private estate (often called "fleecehold")

Effectively, an RMC will have some or all of the responsibilities of the freeholder of the relevant premises.

De facto, RMCs are private local government bodies.

Distinguishing RMCs from MAs

On larger sites, in practice, these activities may be outsourced to a managing agent, and some people conflate this body with the management company. Managing agents are normal commercial businesses, that have no self-imposed restriction in their constitution limiting participation or operations to particular sites. Managing agents will typically have dozens of buildings that they manage up and down the country, and have websites and PR departments and so on. RMCs will generally not have their own websites or staff. The bills from an RMC will often come on the letterhead of its Managing Agent, leading to more confusion.

RMCs may be embedded in the property ownership arrangements for a site. For example, an RMC might be a party to a lease of a flat, noted as the owner of a rentcharge in the transfer deed (TP1) of a house. These arrangements can be nearly impossible to change, so it's important that any company so embedded be an RMC specific to that site, and not a general commercial company.

How the RMCs may be found

The code used for finding the identities of the UK's RMCs used a bunch of heuristics: name, SIC codes, legal form, and so on.

This is of course subject to false positives (things that look like RMCs but which aren't) and false negatives, which are much harder to find.

There are only three types of entity that should be used as RMCs:

  • private companies limited by guarantee
  • private limited companies (i.e., companies limited by shares)
  • private unlimited companies

And the latter case, unlimited companies, is fairly rare and somewhat risky. Buying a flat in a converted terrace house in Tooting Bec should not put you on the hook for when the bloke upstairs hires his incompetent mates to fix the roof. There are about as many unlimited RMCs as there are commonhold associations, almost exactly.

I don't think that the two main sorts of company (shares vs guarantee) makes much of a difference to outcomes in practice, though happy to be proven wrong.

What's in the data?

The spreadsheet includes columns with the following data, in no particular order:

  • valid RMC?
  • presence of a PSC?
  • whether the RMC was originally captive
  • the company number

I haven't provided links to the companies' records on Companies House, but you can type in the company number into the form here and get that.

The key information is about whether the companies in the sample are/were captive RMCs, and whether they have a PSC.

Companies House bulk data gives us information about which companies have got a "PSC", or "Person with Significant Control". Any company with fewer than four members will have one, so some smaller blocks of flats will have an RMC with a PSC simply because they're tiny. But in other cases, the presence of a PSC generally indicates that the RMC has been captured by a small group of the flat owners, or by the original developer or the managing agent.

Note that some PSCs cannot legally be registered on Companies House in the PSC register, because they are exempt. This generally includes housing associations, which disguises the fact that some RMCs for market-rate property are under the thumb of institutions in the social/affordable space.

Valid RMCs

In practice, the heuristics I used for finding the RMCs can be a bit off. For each of the 100 companies in the sample in the spreadsheet, I checked the articles of association to see if it really was an RMC, i.e., that it had members who were supposed to own flats, and was only set up for the purposes of collective property ownership or management.

Nine of the 100 companies weren't really RMCs. It's therefore likely therefore that about 9% of the 55000 companies I detected aren't really RMCs. It's harder to estimate the false positive rate, so I shan't attempt to.

There were also about 10 companies that hadn't got valid articles of association on Companies Houses, for various spectacular reasons. These should probably be disregarded, so the false positive rate is a notch higher. Anyway, around 90% isn't too bad.

So what does one do to work out if a company is validly an RMC?

One reads the company's most recent articles of association, and sees if there's some provision restricting the company's membership to owners of flats or similar (e.g., a definitions section containing a definition for "dwellingholder" or "unitholder", etc) and a restriction on the scope of the company's activities to limit it to a particular set of premises (often "the Property" or "the Premises" or "the Development"). Then one checks if these definitions are actually correctly used in the articles.

The notion of a company that is just for managing a block of flats is one that shows up in the tax system, and it is telling that HMRC similarly looks to the connection between the company, the property and the management.

All the companies marked as "True" in the "Valid RMC" column in the spreadsheet passed the test above.


Many companies, including practically all RMCs, have a sort of "startup" phase, when a different group of people are members, e.g., the accountants or lawyers or administrators of the firm setting up the company. The articles of the company will provide for the intended individuals to become members after some event, for the company to be "handed over" to them, e.g., by transferring shares, or by admitting new individuals as members.

These legitimate mechanisms can be abused to provide for long-term control by the successors of the founding members. This might be by giving the founding members three votes for every vote that the other members have, or by giving them initial control of the appointment of directors, and so on. I have seen about a dozen different schemes, including, crucially, ones whose presence need not be publicly acknowledged on Companies House, which doubtless compromises DLUHC's plans for RMC governance.

Often, these schemes to create a "captive" RMC are so successful that they give rise to the existence of a declarable PSC for the company. So that is an additional heuristic for detecting captive RMCs, but the primary consideration is the company's articles. As an aside: if you see the words "three votes" appear in company articles, it's likely a company that is both an RMC, and an RMC with these captive provisions.

I've marked the presence of captive RMC arrangements in the "Originally captive" column of the spreadsheet. Naturally, some of these lapse over time, but often the state of affairs cannot be unilaterally ended by leaseholders. It seems to be the case that Members of Parliament can persuade the controllers of the company (e.g., the company's own managing agent) to hand over control of the company to leaseholders or fleeceholders. If this is widespread, it would suggest that managing agents are more worried about across-the-board legislation against this business practice than they are about retaining any individual captive RMC.


Of the 100 companies in the random sample of potential RMCs, 9 aren't actually RMCs at all, and 9 more are of unknown status. That leaves 82 RMCs.

Of these 82 RMCs, 17 are both designed as captive RMCs and still have a PSC, 2 of those would have a PSC anyway due to their small size. 37 of the valid RMCs, 45%, were designed as captive RMCs, thus about 54% of those RMCs designed as captives have since transitioned to resident/leaseholder control.

Within the still-captive RMC subset are Residents Management Companies controlled by Barratts (2), Encore, Redrow (2), Persimmon and Bellway. Some of those companies are quite old.

Of the RMCs that were never supposed to be captive, there are still quite a few that have a PSC, though about half of them have PSCs due to being very small, and several others exhibit share allotment irregularities that may be confounding results too.

What is to be done about commonhold?


Commonhold in the UK is a system of land ownership that can be used as an alternative to leasehold. However, it is currently very seldom used.

Leasehold has a number of problems, and I know my own situation as a leaseholder without these problems makes it much harder to empathise with people in situations that seem fairly extreme and unjust.

But leasehold is a tool that is used to solve some genuine problems, too. The main ones are:

  • managing negative externalities and collective action problems within blocks of flats or on private estates
  • enabling a tax-efficient system of shared equity home loans, known as Shared Ownership

Commonhold is a solution to the first of these, and irrelevant to the second.

To phase out residential leasehold tenure, there must be an alternative, and commonhold is the only available alternative.

There are two points to this longish article. They are to consider:

  • how feasible it would be to mandate the use of commonhold on newly constructed blocks of flats, and
  • the viability of commonhold on existing sites, given the problems of "fleecehold"

Wider availability of commonhold risks the creation of an additional tier in the UK property tenure system, to the detriment of any leaseholders who cannot convert to commonhold. So the feasibility of conversion to commonhold matters. There is also the problem that the assumed benefits of commonhold are not necessarily available on sites that have converted to commonhold from leasehold, about which I shall write much more below.

Under commonhold, there is no time-limit on how long the owners of flats own the flats. It follows that there's no additional party whose interests must be protected, other than those of the neighbouring flat owners. Commonhold requires that the block of flats be owned by a body corporate whose members are the flat owners. In the UK, this body corporate is just a normal non-profit membership organisation registered as a company with the usual company registry, Companies House.

Since the UK has very democratic company law, and a very technologically progressive registry for companies, this is a good thing; in strata title jurisdictions, the main registry is the one for land ownership, not company ownership. Where a collective enterprise is dealing with the UK bureaucracy, experience has shown that companies, whether big multinationals and tiny non-profits and NGOs, are treated as first-class citizens, and other types of body (trusts, non-company charities, mutuals, trades unions) have to use outdated technology and rules. Commonhold associations get the perks of just having to deal with Companies House.

How ready is Commonhold?

There are one or two problems that mean that commonhold isn't ready for the average block of flats yet:

  • in principle, it's hard to recover service charge arrears in a timely fashion
  • it's virtually impossible to have delegated management of less than a whole block of flats (a "section")

Now some blocks of flats are fairly uniform, but in my own case, there are 12 flats, 10 of which have access to the internal corridors and two which don't. It's not fair for the two non-connected flats to have to pay for the cleaning, heating and maintenance of those corridors, nor is it fair for them to have a say in the management of services. This can be achieved under share-of-freehold but not under the current commonhold legislation. There is some difference of opinion about the management of subsections of a block under the current law. I am not going to say the Law Commission is wrong when they say that it's "virtually impossible" to exclude some flat owners from voting on such matters, but I've certainly heard a credible counter-argument.

The more complicated a site is, the more awkward the inflexible governance model becomes. But this is a key trade-off: inflexibility is the key advantage of commonhold over share-of-freehold, as commonhold lacks the flexibility that would permit deliberate or accidental unravelling of the shared freehold arrangements. "sections", along the Law Commission's proposed lines, can currently be implemented under share-of-freehold but not under commonhold. Only larger, more complex sites would require them.

The issue about service charge arrears could be solved with a one-clause private members' bill. The sections issue could probably be solved with a one-clause bill plus a schedule.

The Law Commission's report on reforming commonhold runs to about 20 chapters, a third of which are taken up with making it easier to convert to commonhold, and therefore irrelevant to the question of commonhold on newly constructed blocks. The commission's reports deal with the two issues I raise above, and some unrelated matters. It is not a huge amount of work to partition the commission's 121 recommendations into matters which must be addressed before mandating commonhold on newly constructed blocks, versus matters which are "nice to haves".

There's also the question of Shared Ownership schemes under commonhold; it's my understanding that currently commonhold does not permit a flat to be owned under a Shared Ownership scheme, with the implication that commonhold will might exclude social/affordable housing units. It may be that I've slightly misunderstood this, and there's an equally tax-efficient shared equity scheme that is compatible with commonhold as it currently stands.

In summary:

  • commonhold already works.
  • it needs some reform to improve the handling of service charge arrears, and this affects all sites
  • many sites would also benefit from a statutory scheme for "sections" of the property to be treated separately (in the sense that not all the commonhold members would have a say in the governance)
  • there are some other, less important, reforms that would be beneficial

Mandating commonhold on new sites

We are not at the point where commonhold can be made mandatory for new blocks of flats. But it requires only two or three small reforms to make it viable for new blocks.

Therefore, I support a ban on third party freeholders for new blocks of flats. What that means today is that if such a ban were implemented, any new blocks would have to be share-of-freehold (with no third party investors) or commonhold. Once commonhold has been reformed, that ban could be tightened to mandate commonhold across the board for new blocks.

Until quite recently, it had been assumed by many that there would be a grand bill submitted to Parliament which would include an implementation of the Law Commission's recommendations on commonhold. I never thought that was likely. The Law Commission's proposals go much further than the minimal changes that would be required to make commonhold mandatory for newly built blocks. The Powers That Be seemed to be going for a "big bang" reform, but I no longer think this is the case.

If the measures can be broken down into smaller pieces, then it would be theoretically possible to reform commonhold to the point where it could be mandated. No political party is currently proposing either such a breakdown or such a mandate. Yet.

Aside: cooperative model for commonhold?

Some people want to introduce the one-member-one-vote system of cooperatives into commonhold or into Right To Manage companies. Under the current law, the cooperative model is allowed for share-of-freehold, but completely impossible for commonhold and RTM companies, both of which are strictly required to be companies limited by guarantee, and must additionally be run on a one-flat-one-vote basis.

It is, however, perfectly legal to run a share-of-freehold on a one-member-one-vote basis, due to the protean flexibility of share-of-freehold.

I'm a big supporter of cooperatives. For those unfamiliar with how they work in the UK, they are similar to companies, but have to be for-profit, and come with a statutory lobster trap that protects the little guy from being outvoted by larger investors.

I think the for-profit character is a red herring; the problem is that it will just strike too many people as unfair that someone who owns two flats in a building should get the same voting strength as someone who owns only one. I don't think there's any political will whatsoever to allow such arrangements within commonhold or RTM. As companies limited by guarantee, with prescribed constitutions, commonhold associations and RTM companies already have adequate protection for the little guy.

One possible exception to the above: any one voting member ought to be able to force a commonhold or RTM company to hold an AGM; the current threshold is that five percent of the members must act jointly, which means more than one member if there are more than twenty flats.

Commonhold on existing sites

This is a real can of worms, due to what's known as "fleecehold", the growing phenomenon of private estates.

We must step back and take a broader and historical view. What is happening is that the relationships between dwelling owners are coming more and more to be regulated by a different type of law: before the 1980s, it was mainly property law, the law of landlord and tenant. Increasingly, however, company law is playing a larger role, for better and for worse: commonhold relies heavily on company law, as do tripartite leases and so-called "di Marco clauses". This shift is driven both by industry and by government-initiated reforms opposed by the sector. The roots of the current reform situation go back to the 1980s, when the Thatcher government reformed leasehold law and began the move to implement commonhold, which was the baby of the Lord Chancellor, Lord Mackay of Clashfern; the commonhold project was delayed by Thatcher's deposition and two elections, but taken up by the second Blair government and enacted in 2002.

The 1980s government crackdown undermined the ability of freeholders to profit from the provision of services. A very brief account of this is presented in a deleted article on Flat Living, which claims that tripartite leases arose as an attempt to offload management powers that were no longer going to be so profitable after the accountability measures imposed by this legislation. Thus began the plague of tripartite leases and monocratic Residents Management Companies.

A tripartite lease is one where there are three or more parties. In addition to the freeholder and the leaseholder, there will be one or more management companies, called RMCs ("residents' management companies"). The term "RMC" has no agreed definition, though there has recently been an attempt to specify one under building safety legislation. Tripartite leases will set out the respective roles of the parties, with the management functions under the lease exercisable by the RMC rather than the freeholder. In practice, this made it easier for developers to sell on the freehold and the management rights separately: freeholds are a fixed income investment, management rights not so much.

The management company party to a lease might have any governance arrangement: it might be wholly independent for-profit company over which the leaseholders have control, or a firmly leaseholder-controlled entity, or something in between. There's also no reason that a management company might be restricted to managing leasehold property, and instead might manage a wider estate. Some management companies permit or indeed require leaseholders to be members.

The constitutions of management companies ("articles of association") are available at Companies House, and I've read literally hundreds of them, as they're all different. RTM companies, by contrast, are required to have the articles of association in identical form, and this is effectively true for commonhold associations as well. Resident Management Companies, on the other hand, exhibit a wide variety of constitutional arrangements with respect to how unitholders (the leasehold or freehold owners of dwellings managed) relate to the company and each other:

  • unitholders might be required to be members of the company
  • unitholders might be required to follow the rules of the company, or pay money to it
  • unitholders might be allowed the decisive vote on all matters, or some matters, or none
  • there might be private contractual arrangements overriding the publicly acknowledged arrangements

Now some of this clearly overlaps with the terms of a typical lease, which require leaseholders to pay money and follow the block regulations and so. But it also overlaps with local government, particularly all that stuff about voting and collective management of land. If a body can tax you and enact bye-laws, how different is it from a government? This is basically the thesis of Professor Hazel Easthope who studies strata title in Australia.

Tripartite leases have the effect of compelling leaseholders, freeholders and management companies to enter into a mesh of additional relationships under company law, alongside those they have under property law. Despite its numerous flaws, landlord-tenant law does at least require that service charges be reasonable and reasonably incurred, but an equivalent fee exacted under company law need not be. A lease might provide that the leaseholder agree to join the management company, and the company articles might provide that accepting the transfer of a housing unit constitutes agreeing to join the company, indeed it will often deny the company directors any right to refuse membership to a leaseholder.

Where the leases do not permit the management company to extract money from the leaseholder in the way it desires, there may be a route under company law; this sort of arrangement was upheld by the courts in the Morshead Mansions Ltd v Di Marco case. The management company in the Di Marco case had given up using its rights to exact service charges under the leases, and instead relied on a clause it introduced into the company constitution allowing it to charge its members (which included Mr Di Marco and the other leaseholders) for the purposes of furthering the company objects, which were of course to manage the building.

Where this leaves us is that we have quasi local governments in the form of Residents Management Companies, with very diverse constitutional arrangements, all the way from democratic to monocratic. Membership of such companies can be unavoidable for leaseholders in a block of flats, and the membership may be shared with neighbours in other blocks of flats or of freehold houses. The company might manage both blocks of flats and wider shared areas, which might or might not be accessible to the general public. Some of this property might be municipalisable, and adopted to be maintained at public expense by the local authority. But it's often that local authority which signed off the arrangement in the first place in return for concessions in a section 106 agreement, and the local authority would rather not be maintaining gardens and soft landscaped areas and so on when others are obliged to pay already. Getting local authorities to maintain, at public expense, the parts of the property that are not even accessible to the general public is a political and moral non-starter.

These sorts of estates have proliferated since the Commonhold and Leasehold Reform Act 2002, which introduced commonhold and the Right To Manage. The property sector has successfully sidestepped the act by fighting to overturn parts of it in the courts(!) and by laying out estates such that neither Right To Manage or commonhold is effective in ensuring that flat owners have control over their collective property or service charges.

The mechanism is as follows: an estate is laid out and developed, with shared areas (car parks, bike sheds, bin sheds, gardens, soft landscaped areas, the corridors and roofs of blocks of flats, etc). As part of this, the developer concludes an agreement with the local authority not not adopt the shared areas. In any case, local authorities would not adopt a lot of this land, e.g., car parks or the common parts inside blocks of flats. Anyone buying a house or flat in the estate is compelled to join the management company under the terms of a tripartite lease or an equivalent provision in the freehold transfer deed. The management company will usually have provision for different members to pay different amounts depending on what access they may or may not have to various "sections" of the common areas, and may or may not have provisions for different members to vote on matters affecting only those sections. This can lead to a situation, for example, where owners of houses have the majority of the votes determining the size of the sinking fund for a block of flats that none of them occupies.

How do commonhold and RTM address this?

Well, RTM now covers only the block of flats and any property that is exclusively used by the owners of the flats but not by the wider estate, which may be a problem. Commonhold completely extinguishes the tripartite leases, but on such estates that's only one of the two relationships that the flat owners will have with the management company: they likely would remain involuntary members, and thus subject to the management company's rules and charges along Di Marco lines, or they'd lose their membership, and with it their vote on wider estate matters. Similarly, the management company might well lose the revenue from the former leaseholders, which may bankrupt it if has a structural deficit without the income from the flat owners. The situation could put leaseholders in flats at serious odds with their neighbours.

Commonhold for an entire estate is a non-starter, because when a commonhold is dissolved, the commonhold association becomes the owner of all the housing units, including the detached houses, which is not a situation that house buyers will likely enter into knowingly.

There is a provision in the 2002 Act, s3(1)(d), allowing the government to give management companies a veto on whether a block of flats can convert to commonhold. Expect this veto to be enacted once leaseholders start converting to commonhold on private estates. I raised this issue at a 2022 meeting of the All Party Parliamentary Group on Leasehold and Commonhold Reform, and publicly the minister wouldn't address it, but at least the issue is understood in government. I raised the matter publicly with Michael Gove when he was reappointed, and similarly he dodged it.

The widespread rollout of private estates and the transition to company law has created "facts on the ground" which impede the adoption of commonhold. Estates are laid out in a manner which entangles blocks of flats with nearby dwellings, and commonhold, which does address the fundamental "wasting asset" issue of leasehold, does not interoperate well with private estate management. This likely affects millions of home owners.

What commonhold gives you is ownership of a flat without a time limit or a third-party freeholder. It is too widely assumed that this brings management, charges and regulations under the control of the flat owners, but sadly that presently will only apply in relation to the block of flats, but not the car park, gardens, bin stores and so on. Effectively the property sector has had a twenty year head start to vitiate the benefits of commmonhold for many sites.

What is to be done?

In addition to reforming commonhold and mandating it for new blocks, there has to be mandatory democratisation of residents management companies and estate management. Effectively, fleecehold needs to be banned.

Who really is the NFSP?


Please note, this blog post has been significantly extended and updated since it was first circulated.

I have recently discovered that the members of the subpostmaster's trade association were not actually permitted to vote in the association's affairs: the National Federation of Subpostmasters' members are required to nominate a board-approved intermediary to exercise their vote on their behalf! Now in pratice the board has been letting subpostmasters appoint themselves as their representatives, but in principle, the directors have a veto over who can exercise the subpostmaster's vote.

This is a long post, but there are recaps and summaries along the way.

Where I'm coming from is a rather idealistic and romantic view of membership organisations. I think it's great that in the UK, moreso than in many other places, and as part of a very long tradition, strangers from different families can get together, organise themselves and solve their problems amongst themselves, without troubling the government for money or having to ask much in the way of permission from outsiders. Dissenting religious groups could for centuries get together and employ their priests, without having to ask the officially Anglican government for permission. Nowadays we have societies, associations, clubs, trades unions, building societies, social enterprises and all manner of other private collective enterprises. I really don't like it when the proper balance between a group's management and its wider membership gets out of kilter. And if the little guys can't call a meeting and make their case that the group should think again, it never, ever, ends well.

But it's not clear that the NFSP is really a membership organisation in that ideal sense.

The public inquiry into the Post Office scandal has recently taken evidence from a former NFSP general secretary, concerning that body's role in the Post Office scandal. So it's appropriate now to subject NFSP to a bit more scrutiny.


The NFSP used to be a trades union. but about ten years ago, the it was decertified and struck off the list of trades unions. This deprived it of its quasi-corporate status, and restored it to being purely an unincorporated association. Lacking corporate status complicated the enforcement of contracts and ownership of property, so around 2015, NFSP negotiated a funding deal with the Post Office(!), and incorporated as a company limited by guarantee.

Mark Baker, formerly of the NFSP and latterly of the Communication Workers Union, gave evidence to Parliament to the effect that

"Also the formation of the company arrangements of NFSP ltd should be called into question as such a company should be owned and controlled by its members as detailed in their Articles of Association of NFSP Lrd. I am still in contact with many Postmasters who thought they were such company members but they have all confirmed that they have never been asked to legally take on company membership nor have they ever voted at an AGM or appointed the Directors which is the normal practice for members of a company limited by guarantee. Democracy within NFSP Ltd does not appear to exist. The free membership of the NFSP that the Post Office claims they pay for appears to be nothing more than a sham."

(I think Baker is accidentally referring to a subsidiary company here; the names are confusing, a point to which I shall return ...)

Companies all have public constitutions, called the articles of association, available on Companies House. Due to my interest in the leasehold scandal I have read the articles of association of several hundred companies, normally those which are used to allow (or not allow) owners of dwellings to have a say in the collective management of shared facilities. So I know most of the tricks used for pretending that members have power when they don't. (Side track: DLUHC launched a consultation about this yesterday, 1 December 2022).

Do regular NFSP members have a vote?

Section 284(4) of the Companies Act 2006 provides that company members' votes are to be counted in accordance with the company's articles. Unless the articles say otherwise, it's one share one vote if the company has shares, and one member one vote if it hasn't a share capital (as is the case for NFSP).

So for a company limited by guarantee like NFSP, any restrictions on voting rights will be found in the company's articles, and will therefore be a matter of public record. From my experience of Residents Management Companies, I've found that the basic mechanisms whereby members' votes might be restricted are things like:

  • clauses that give the company's founders' appointees a supermajority of the votes
  • clauses that give certain members a veto
  • classes of membership that carry no votes at all
  • "friends of" and "supporter" membership schemes that don't amount to formal membership of the company at all

And there indeed legitimate and arguable grounds for using some of these, particularly on a time-limited basis. There's even a company that weights members' votes according to how many internet domains they've registered. I myself am director of a company that allows upgradable, non-voting membership for registered supporters. But we don't go round pretending those individuals are full members.

What matters is how the company holds itself out. If it holds itself out as a bona fide membership organisation, then the members should be able to vote, particularly in elections for the board of directors. Anyone who is liable for the debts of the company should get a vote. Anyone paying for voting membership ought to get a vote, and a vote that's usable.

According to its articles of association, NFSP members aren't supposed to be able to directly vote, even though they are liable for the company's debts. Now there is a £1 limitation on that liability, but obviously the amount is beside the point. The gory details of how the NFSP has rigged this up are in article 8 of their articles of association (filed on 10 Sep 2015). They provide that

  • "All Members shall exercise their membership rights and fulfil their membership responsibilities through one or more Organisational Representative(s)." (a8.2)
  • "The Council may not accept, or require the replacement of an individual as an Organisational Representative provided an alternative is possible. If an alternative is not possible the Council may exercise constitutional rights in respect of the Member directly." (a8.10)
  • "Voting rights under the Articles may be exercised only by Organisational Representatives." (a8.11)
  • "To become a Member [...] a sole trader [...] must [...] nominate [...] one Organisational Representative [...]" (a9.2)

I've never seen anything like this. If you don't want people to have voting rights in a company, you can say so. For companies with share capital, it's routine to have non-voting shares, which often come with better dividends, and there's a nice wizard on the Companies House website to help you write the legalese that you have to publish to describe the situation. For companies without share capital, like NFSP, you can write it in plain English. You don't need a convoluted scheme like NFSP's Organisational Representatives unless you're ashamed of the arrangements and don't want to own up to them. I find it difficult to believe that NFSP's peculiar voting arrangements obtained when it was still a trades union; they might have their roots in an attempt to ensure that unincorporated businesses are equitably represented in voting weights, solving a problem like the relative voting strength of a two-person partnership that runs one sub post office, versus a sole trader who operates two sub post offices, and so. But nonetheless, you don't need to give the board control over members' votes to fix this.

What NFSP membership amounts to formally is a written agreement to obey the NFSP's regulations, an agreement to accept limited liability for the NFSP's debts, the various non-voting powers available to members under the Companies Act, such as zero fees for requesting certain information from the company, and the notional power to nominate a board-approved Organisational Representative to vote in one's own place. Oh, and the right to sue the board for discriminating between members, pursuant to section 994 of the Companies Act ...

It's not clear that most people who had been members of the previous NFSP structure before incorporation as a limited company ever actually formally joined the company, or they'd have been aware of the indirect voting scheme and complained about it. Someone should demand access to the NFSP's membership register to see if anyone other than the subscribers ever managed to join.

But my point here is not really how many members the NFSP has on its books, but the fact that the NFSP's constitution seems to say that members can only vote through intermediaries; that might work when everyone is content, but when push comes to shove, members could find it extraordinarily difficult to exercise their rights. That's the sense in which it's a "sham membership organisation". It's obviously not a sham organisation, it's there on Companies House and actually provides unusually transparent publications, e.g., on its finances, that go beyond the legal minimum requirements. But I suspect it's not really a membership organisation in practice.

There remains the question how much the Post Office influenced these arrangements ...

Why does NFSP's membership model matter?

One would tend to assume that a company limited by guarantee is not for the profit of its own members. This isn't formally the case. You can have a for-profit company limited by guarantee, and a not-for-profit company limited by shares, like St John's College, Durham.

The National Federation of Subpostmasters can apparently hand its assets over to its own members in expectation of the organisation being wound up. It therefore matters who the members actually are. A lot of those assets seem to have come from Post Office Limited, a company wholly owned by the British government. Since the net assets of NFSP were £3,627,700 in its 2021 accounts, that would amount to about £250,000 each if it only the directors were members. I suspect that the membership is much broader than that.

But we don't actually know how many members NFSP really has, and it would be useful for someone to request access to the NFSP's membership register to find out.

Who are the NFSP's members?

NFSP's unusual constitution requires that members nominate an Organisational Representative to exercise their voting rights in their stead. You can find the constitution on Companies House, in the very first document filed for the company, in September 2015. In principle, you can't now join NFSP without nominating an Organisational Representative, as we shall see.

NFSP is now incorporated as a company limited by guarantee. Purportedly, in 2015, all the members of the predecessor unincorporated association were to have been admitted to membership of the new company. This is in article 9.1 of the new NFSP's constitution. It's not clear whether this kind of arrangement really works legally without each of those transferring members formally agreeing to take on some liability for the new entity and agree to be bound by its rules, provide an up-to-date service address, and so on. Some former or serving subpostmasters have repeatedly questioned the membership arrangements.

(NFSP's articles may be found here or as its first filing on Companies House.)

New subpostmasters joining NFSP post-2015 are required under article 9.2 to nominate one of these "Organisational Representatives" as a condition of admission. This shows that the automatic transfer of members is a bit of a fudge: all members, auto-enrolled or otherwise, must nominate an Organisational Representative in order to vote, but only in the case of new members admitted post-2015 is this stated to be a requirement of admission. That may leave the auto-enrolled members without a duly appointed Organisational Representative.

Anecdotally, it would seem that subpostmasters may not have been asked to agree in writing to be transferred to the new NFSP in 2015, or to nominate an Organisational Representative subsequently. NFSP's membership application paperwork makes it absolutely explicit that new applicants are agreeing to abide by the NFSP's articles of association. But it doesn't mention the requirement under article 9.2.2 to nominate an Organisational Representative. Either that requirement is improperly being waived, or it is being fulfilled separately from the membership application form, or no-one is actually really being registered as a new member of NFSP. Of course, it's not in the NFSP's interests to draw attention to this requirement, so maybe if it's being fulfilled at all, it's by correspondence after the membership form has been submitted.

In any case, article 16 suggests that the identities of Organisational Representatives of Members must be stored together with the Register of Members. So NFSP really is supposed to be recording this data.

There is a form, which I've linked, for appointing an Organisational Representative, on the NFSP's website. The question is whether that form is really being used.

To summarise so far: it's not obvious which persons have been properly registered as members of the NFSP. And it's important to understand that, according to the NFSP constitution, these questions are linked, as new members are required to nominate their Organisational Representative as a condition of joining the company.

Association or Foundation model?

There are two basic models for the membership of non-profit organisations, which in the case of one special type of non-profit organisation (CIOs) have even got semi-official names: the Foundation model and the Association model; sometimes also called "small membership" and "large membership". The Association model is where a large group of members elect a smaller committee to be in charge, and can hold that committee to account. The Foundation model is where the only members are the committee. One is a self-selecting democracy, the other is a self-selecting oligarchy. But these two models are just two points on a spectrum, and an organisation may combine elements of both.

The committee in either of these models has power day-to-day. The question is, is there a wider group with a final say, above the committee, or not? Both models are perfectly legitimate; they're just different.

What is not legitimate is holding yourself out as following the Association model, whilst actually operating on the Foundation model.

There'll now be a little digression about how trivial mistakes by accountants can sometimes shed light on the nature of membership, and they'll crop up later on, too ...

I became interested in this problem a few years ago in relation to residents' associations (RAs): the RA for my ward, Trumpington Residents Association, was clearly set up on the Association model, and indeed was both a limited company and a registered charity, but its annual accounts stated explicitly that it was the Foundation model: supposedly, the only members were the board of directors/trustees. I pointed this out at a meeting, wondering if I myself were really a member, in the sense of the Companies Act, of the Trumpington Residents Association; the alternative was that the statement in the accounts was right and the only members were the board of directors, and that the elections weren't really binding and the people who thought they were members, and paying for the privilege, weren't really members --- what I'd sometimes call a "sham membership organisation". There was a little headscratching, but they quickly confirmed that the organisation was operating on the Association model and made sure the accountants didn't misrepresent this in future years. It had simply been a harmless mistake by the accountants, including boilerplate language more appropriate to a Foundation-style organisation.

(By contrast, the residents association for my own development, Trumpington Meadows Community, operated on the Foundation model but held itself out as a bona fide association: it simply was not prepared to permit outsiders to have any influence, any ability to hold the committee to account, and indeed TMC did not maintain a register of members. Given that it also held itself out as a conduit of communication with the developers and managing agents, this caused considerable tension; it had a strange resemblence to the NFSP in this regard.)

To summarise: the committee managing an organisation might be accountable to a wider group of members, or it might not. It is illegitimate, in my view, for there to be too great a divergence between theory and practice: if the rules say the committee is accountable to a wider membership, that should be true in practice too. That is why I described NFSP as a "sham membership organisation": when push comes to shove, the board can refuse to accept a member's choice of who should exercise his/her vote.

Proxy voting

These Organisational Representatives sound very much like proxy voters. The right to vote by proxy on company business is provided by section 324 of the Companies Act. The NFSP's articles of association go to some length to try to circumvent the possibility that a member might appoint someone other than an Organisational Representative to this role, article 50 providing that "It is expected (but not required) that [...] Organisational Representatives, rather than the Members they represent will exercise any power to appoint proxies".

It may be the case that every Member since 2015 has always been able to appoint himself, or a preferred individual, as his Organisational Representative (and I say "him" for brevity here, though the public statistics show there are lots of female subpostmasters, as you'd expect). But when push comes to shove, the NFSP board could easily nobble the votes of hostile members.

What I suspect is going on is as follows: the NFSP did indeed register as members the people who had been members of the previous unincorporated association, on or shortly after the NFSP turned itself into a company. It probably has been registering those who've subsequently joined as well. The register of members is held by the NFSP and anyone wishing to access it must give ten days' notice. This allows companies to correct an "out of date" register before disclosure.

What I'm much less confident of is the possibility that the NFSP has ensured that all members even have a duly appointed Organisational Representative.

But, the register of members' Organisational Representatives is an NFSP-specific thing, rather than a statutory requirement, and so is not subject to a right of access, and it is doubtless also restricted by GDPR.

The fact that a member can appoint a proxy other than an Organisational Representative is a chink in the armour of this scheme; I imagine very few succeed in doing so in practice, and that there are no general meetings at which such member-appointed proxies could vote. After all, under the Companies Act, to call a general meeting requires a requisition supported by five percent of those entitled to vote at the meeting, so that would involve dozens or hundreds of members each persuading their board-approved Organisational Representatives to sign the paperwork. An a general meeting is the only way to hold the board to account.

Effectively, the NFSP constitution creates by the back door a non-voting membership class. Of course it may be that the board has always allowed members to appoint themselves as their Organisational Representatives, but it still amounts to the board having a veto on members' voting rights.

What do the accounts say?

The accounts of the old NFSP are available for a few of its final years, at the National Archives. The figures around 2012 note membership income of £893194 from 7168 members, implying a subscription fee of £125 per annum.

When it was deregistered as a trades union, the NFSP could have swiftly incorporated a company limited by guarantee, with members paying the same subscription fees to the new company. That's not what seems to have happened; instead a Grant Framework Agreement between the NFSP and the Post Office provided that the Post Office would top up NFSP's subscription income to £1,500,000 beyond what was raised in member subscriptions, and the subscription fee was effectively reduced to zero. Some weeks after the initial Agreement was concluded, NFSP incorporated, and registered its articles of association; article 12.8 provides "subscription policy may take account of funding arrangements between NFSP and POL.", which is quite the understatement.

Before the funding agreement with the Post Office, the NFSP was fiscally accountable to its members; if they didn't like what the NFSP board was doing, they could vote with their wallets as well as their feet. Of course, much of that income was really down to the fact that the NFSP had a monopoly on representing subpostmasters to the Post Office. By tying itself to the Post Office's purse strings, the NFSP untied itself from its own membership.

Accounting for members' funds

That £1,500,000 figure comes from Schedule 1(II) of the Grant Framework Agreement.

Is this figure borne out by the accounts? NFSP to its credit provides nice searchable PDFs of its more important Companies House filings, and for the 2021 accounts it does so with additional detail. NFSP turned over about £1.8 million in 2021, slightly above the guaranteed grant income. I tried comparing the version of the accounts filed at Companies House with the more detailed figures on NFSP's website, as the documents seemed to be from different time periods (the list of directors had changed).

When the light came it on almost blinded me.

It was another of those harmless accountant's errors. All limited companies in the UK must publish their balance sheet, which shows that the net assets of the company match the members' funds. In company law, "member" and "shareholder" are fairly interchangeable; if a company has shares, then the members are all the shareholders and only those shareholders. If a company doesn't have shares, then its members are just members; we would only call them shareholders very loosely. In a company with shares, the equity line on the balance sheet might be called something like "Shareholders' funds". For a company without shares, it might be "Members' funds" or similar.

Yet NFSP balance sheet says "Shareholders' funds".

Why do the accountants think of the members as shareholders? Are they entitled to those funds?

From the point of view of accounting, it doesn't make any difference whether the equity is described as belonging to shareholders or members, but it got me thinking. We tend to think of the difference between companies limited by shares and companies limited by guarantee as the difference between "for profit" and "not for profit". The shareholders / members must contribute funds if the company hasn't got enough money when it is being wound up, up to a defined limit. But before winding up, while the company is still in business, it can pay dividends to shareholders / members out of its accumulated profits, but not out of money they've invested. The reason is that the invested money has to be kept in case it's needed to repay creditors. This is a basic quid pro quo in return for being able to limit the members' liability to creditors. Companies limited by guarantee are not allowed to distribute profits to non-members. But what about distributing profits to members? They can do so; however, a very large proportion of companies limited by guarantee choose not to do that, and instead their constitutions provide something like the following:

"The income and property of the Association, whencesoever derived, shall be applied solely towards the promotion of the objects of the Association as set forth in these Articles of Association, and no part thereof shall be paid or transferred, directly or indirectly, by way of dividend, bonus, or otherwise howsoever by way of profit, to the members of the Association."

There are thousands of organisations that include some variation on this formula in their governing documents: charities, non-profit associations, every block of flats' Right To Manage company, even some Royal Charters have it. Having this provision or something equivalent is a statutory requirement, for limited companies that omit "Limited" from their official name. As we know, the official name of NFSP is "National Federation of Subpostmasters". Not "National Federation of Subpostmasters Limited".

The requirement in regulation 3(3)(b) linked above relates to the behaviour of the company while it is in business, not when it is being wound up. Winding up is covered by regulation 3(3)(c), which requires that "all the assets that would otherwise be available to its members generally to be transferred on its winding up either [...] to another body with objects similar to its own, or [...] to another body the objects of which are the promotion of charity and anything incidental or conducive thereto [...]".

So I looked in NFSP's articles again, and sure enough, there was a variation on the "No profit distribution to members" formula in article 5, and I say "variation" advisedly. But when I found the provisions as to winding up, in article 61, that warning light became an alarm bell. For on winding up, NFSP's articles provide that the assets can indeed be given to other organisations fitting the rubric of "similar objects" or "charitable or community purposes", but also "distribution among the Members on a reasonable basis determined by the Council"!

So that disqualifies NFSP from entitlement to the benefit of the section 59 exemption from using "Limited" in their name, which has stood as a provision of English company law since time so ancient the memory of the Internet runneth not to the contrary. These sections of the Companies Act go all the way back to an origin in the Companies Act 1867, section 23. If anyone has a digital copy of that, please do send it me.

A closer look at NFSP's article 5 shows there's a loophole: "No part of the income and property of NFSP may be paid or transferred directly or indirectly by way of dividend, bonus or otherwise by way of profit to any Member (subject to Article 61)." [italics added, natch] Some variation!

This means that NFSP's constitution is, at best, self-inconsistent. The previous article (4) says NFSP's "income and property shall be applied towards the promotion of its objects in pursuit of its authorised activities", but those objects are drawn very widely indeeed.

Cui bono?

If people see a limited company without "Limited" or "Ltd" or "Plc" or their Welsh equivalents in its name, they're entitled to expect that it's not ultimately being run for its members' benefit or allowed to give the Members its assets if they decide to wind it up. Now there is a benefit of membership that the NFSP has probably failed to tell its members about!

There are no prizes for slipping one past Companies House; they don't check the paperwork submitted. But NFSP should really change its articles, or change its name. NFSP is a for-profit company, and it should bear the name of a for profit company, included the "Limited".

Large membership organisations that are sitting on piles of cash come under pressure to demutualise, and release the accumulated cash pile for the benefit of the current members, who therefore might have an incentive to vote for winding up. This might be a reason why the members are impeded from voting. In any case, the question who is a really a member of NFSP, and the question can NFSP members really appoint Organisational Representatives in defiance of the board, are questions with a potentially significant price tag attached. And that price has basically been paid out of Post Office funds.


As a matter of public policy, the Post Office should not be handing what's effectively public money to an organisation which holds itself out as not being for its own members' benefit when in principle it is. Nor should public funds be doled out to an organisation with such peculiar governance arrangements. I shall be writing to my MP on this question promptly.

What NFSP can do to fix this is either of the following:

  • change its name, to add "Limited", and remove the impression that its assets cannot be distributed to members
  • or remove the inconsistency in its articles, and actually unambiguously prohibit distribution of assets to members

It should also remove the absurd paraphernalia about Organisational Representatives. Either:

  • reclassify the members as non-voting (good luck with that if there are a lot of them)
  • remove the concept Organisational Representatives and let members vote directly

It should be more explicit, when asking people to join, about what the constitutional terms of the membership offer actually are.

Productivity Shearing in Voluntary Organisations


There's a recurring problem I've observed in voluntary organisations: potential volunteers have different levels of ability and experience, across different areas. There is a key area that affects the cohesiveness and effectiveness of a voluntary organisation: if volunteers have to interact with each other, how well do they do so?

The sorts of interactions I'm concerned with are really things like these:

  • organising when people are available for meetings
  • selling tickets for events
  • circulating news
  • tracking ongoing activities

For some things, it doesn't matter whether everyone uses the same system, but for others it does:

I do some litter-picking around my neighbourhood occasionally. There is a local group I'm part of which organises litter-picking, but which is largely focused on other issues; anyone can just turn up, group or no group, with a glove and a shopping bag and pick up some litter; you largely do not need to worry about how good anyone else is at litter picking: your gloves and bags don't need to be compatible with their gloves and bags.

That group also collects and shares information about problems around the neighbourhood. This is where the trouble starts: we could store the information in our heads, or on paper, or on a specific computer, or on a networked computer. Individuals will have their own preferences, and some of these preferences can be very strong, for two reasons:

  1. some systems are more familiar than others, and it costs time to become familiar;
  2. some systems are much more efficient than others, and it costs less time to use the more efficient systems, so long as you're familiar with them.

There is therefore a minimum and maximum level of efficient that each individual is prepared to work at. Some people, largely down to personality traits, are willing to put in a lot of time to acquire familiarity with new systems which might prove more efficient, or which at least seem to be more effective for co-operating with their colleagues. Others less so.

In an organisation where people are getting paid for what they do, the organisation can simply use its resources to train people up on the systems that it wants, and mandate their use. In a voluntary organisation, much more leadership, persuasion and strategy is required.

Now this only matters where the systems used by one volunteer have any impact on the systems used by another volunteer, but that is a very common occurrence.

Therefore, there will be potential tensions about the range of levels of efficiency that individuals are prepared to work with in a situation where:

  • people want to co-operate,
  • but are not being paid to do so,
  • and where there is a need to use compatible systems.

Some individuals' minimum or maximum levels of efficiency won't even overlap. That is to say, there will be no system which everyone is happy using for booking events, because some only want to use Eventbrite and some only want to use cheques and postage stamps.

This is the "shearing" effect: the cohesion of the group is undermined because the requirement for efficient use of technology affects its members differently depending on how comfortable they are with particular systems. If you push people too far outside their comfort zone, they'll lose interest and volunteer for a different group instead.

The choice, then, is to handicap technologically proficient volunteers by making them use systems that may be orders of magnitude less efficient than what they use in other areas of their lives, or encourage a possibly painful learning process to get other volunteers "up to speed", or some combination of the two. One unexpected barrier may be that people like inefficiency because it gives them something to do, and if that means more productive individuals stop volunteering, so be it. None of this is happening in a vacuum: there are plenty of other things people could be doing with their time, and the rest of the world will on balance be getting more productive as time goes by.

There is, then, a particular danger for organisations that rely on volunteer labour. Persuade your volunteers of the strategic importance of investing time in learning the most productive collaboration systems, or perish!

I loved email. It's dead.


I loved email. It's dead.

We should start thinking of email addresses only as attack vectors.

An email address a piece of information which, once disclosed, allows someone or something to communicate with you forever. The consequence of this communication is that you may get interrupted by a notification, and bear the cost of storing, reading, and/or deleting the message; the message also increases the cost of searching through all your other messages.

These costs are small. The number of emails you receive, however, is very large. I have received well over a hundred thousand emails so far. Over time it adds up.

There are many-to-many communication systems which are indexed on other kinds of addresses (such as your phone number, postal address, your Facebook identity, your cryptographic public key, and so on). Email is like a phone number or a postal address: it has the property that "knowledge-is-permission", i.e., if you know the address, you can send data to it. Unlike other knowledge-is-permission addressing systems, or "capabilities" to abuse the computer science lingo, sending an email is almost costless, much less than the smallest unit of any normal currency.

The problem is that sharing your email address is a transitive operation: you are granting the recipient the capability to share the address with whomever he/she/it chooses. It is of course much worse than that: the address might be obtained accidentally or maliciously by a third party with whom you have no relationship, due to error, or the recipient going bankrupt, or a data breach, or being sold. There are some laws against sharing "personal data" without permission, but they're not remotely sufficient and probably not the right tool for the job anyway.

There is a commercial incentive to obtain email addresses from customers. They improve price discrimination, which means that customers collectively have to pay more (though some may pay less). Therefore companies try to force customers to hand over email addresses. You are required to divulge an email address to obtain the product; this is useful because it helps keep you informed as the product is delivered. But then a few weeks or months later, you start getting adverts from the company.

In the time it took me to write the previous paragraph, an advert arrived by email from a company from which I bought some blinds for my flat in December.

But in the time it took me to write that paragraph, I blocked all future emails from them.

What I have done is established a system of individual addresses for each company I and organisation I deal with. When I signed up with Blinds2Go, they got given my email address as mk270-blinds@no.ucant.org. But all I had to type was:

address-tool --retire mk270-blinds

and all future email from them is prevented with a curt "bounce" message, and I never receive a notification or store the message.

Effectively, this amounts to having one email address per interlocutor, with revocation indexed on sender email address.

What we actually need is a distributed store-and-forward messaging system where addresses are not transitive: instead, one would receive an invitation to communicate which could only be used by the recipient and not by third parties. This is vaguely similar to the PGP web of trust, Facebook messages between friends, and so on, but is probably most closely represented by the Scuttlebutt system.

To be continued ...

Experimenting with CompCert


A few weeks ago I experimented with CompCert, a C compiler from INRIA, written largely in Coq, with chunks in OCaml; this allows the Coq parts of CompCert to be formally verified (see below for more on this).

Now I have no need of a guarantee that my compiler is bug-free, but to the extent that translation my code into the subset of C supported by CompCert reduces the bugcount rather than increases it, it's a win. I'm basically using compcert as a lint tool, but it's fun and instructive anyway. The real-world scenario which makes any of this interesting is therefore if you have a C codebase and suspect a bug in your compiler and want to know how hard it would be to maintain that codebase such that it compiled with a compiler believed to be bug free.

For many years I have maintained a codebase of 40K lines of fairly odd C, that implements a computer game I used to run in the 1990s, and which predates modern conveniences that might have been used, such as sqlite, pcre, libevent, reliable IP stacks on NeXTSTEP, ANSI C, free C++ compilers, free Erlang, etc, etc. The code is also unusual in shunning the use of struct, malloc and pointer arithmetic. For almost the last twenty years I've kept it up-to-date with the C toolchains on a number of OSes, as a way of keeping an eye on what the cool kids are breaking.


Firstly, the codebase needs to be able to cope with multiple compilers; gcc and LLVM's clang are close to drop-in replacements for each other from the perspective of the Makefile. Not so, CompCert: -Wall -Werror are not accepted as options by CompCert, as they're effectively on by default. CompCert isn't going to want to know about any code that doesn't pass gcc -Wall -Werror, but there are a few things LLVM thinks it's Ok to warn you about that CompCert is cool with, which feels like LLVM is wasting my time. Getting the build system and revision control happy about parameterisable compiler options has to happen first.

I was forced to do change all the remaining instances of conflation of integer widths. Anyone who's done arithmetic in OCaml will recognised this as one of the house microfascisms of INRIA, but it's a deep issue: a lot of corner cases depend on your installation of the header files and libraries and so on. In my case, function prototypes are culled into a .h file automatically with cproto, which by default changes the width of integers in K&R-style C functions:

void my_function(i)
short i

is output as

void my_function(int i);

which gcc and LLVM tolerate, but CompCert doesn't. There were a couple of other legitimate "Well Don't Do That Then" moments that I won't tax you with. Effectively one's forced to get all the prototypes and headers and includes exactly right. This showed up a bug: a variable which was supposed to be declared extern wasn't, and was separately allocated from the global it was supposed to represent.

The more formal treatment of integer widths also meant fixing a lot of sprintf format strings.

The next thing I had to fix was the idiom

char *messages[] = { "...", "...", "...", NULL };
int x = sizeof(messages) / ...;

CompCert insists on the length of messsages[] being explicitly specified, which means this technique isn't allowed.

The harder stuff was signal() and stdarg; basically, CompCert supports an anaemic subset of C, and doesn't allow stdargs, though it provides the sprintf() clique of functions. Since wrapping sprintf() is about the only thing varargs is used for in C, this turns out not to be a problem, but I originally bet that parts of the codebase were outside the CompCert C dialect and would need to be shunted into libraries.

My own adventure in CompCert land basically amounted to learning new stylistic restrictions in C. Reading around what people have been doing with CompCert I came across a few interesting articles and from this chap I learnt about concolic testing which is another technique I have no use for but am glad to have spent time learning about.

Aliens don't crash land


Today is the anniversary of the Roswell Incident, which is the subject of various conspiracy theories. I don't like conspiracy theories; that way of thinking always tends to involve being selective about whether particular things are plausible.

There's no reason to suppose that life necessarily exists outside our solar system, or that it is impossible for life to exist elsewhere, but that is not the point: we are invited to believe that intelligent creatures from outside our solar system crash landed in Roswell this day sixty-six years ago.

Is plausible that aliens could master interstellar travel, but not the ability to land without crashing?