Commitment Issues

When is a commitment not a commitment? When a regulator accepts one to avoid having to regulate. A note on promises that are not worth the paper they are printed on.

There is a particular frustration reserved for those of us who read the small print for a living, and it attaches to a single, overworked word. Commitment. It is a fine word. It carries the weight of a promise, the suggestion of something a person can be held to. Which is precisely why it should be used with care, and why it is worth noticing when a regulator reaches for it to describe something that can be held to nothing at all.

Let me set the scene, because I think this matters rather more than the niche subject suggests.

A tidy little decision

For most of 2025, Ofcom was reviewing the wholesale price that mobile operators charge to terminate application-to-person text messages. These are the one-time passcodes, the appointment reminders, and the “your parcel is three doors away” notifications that the NHS, the banks and half the high street quietly depend upon. Termination rates had climbed sharply. Ofcom provisionally found that the operators held significant market power, and possessed both the ability and the incentive to keep pushing prices up. It proposed a cap.

Then, in the autumn, three mobile networks and one of the larger MVNOs each handed Ofcom a set of voluntary commitments on their future pricing. And on 31 October 2025, on the strength of those, Ofcom discontinued the review. No cap. No conditions. The announcement told the public these commitments should prevent uncontrolled price rises, that Ofcom would keep a close eye on adherence, and that it could step back in if needed.

It reads well. It is meant to.

And I want to be fair on the substance, because there is a respectable case for restraint, and it is worth stating at its strongest. WhatsApp for Business has been cutting its prices and may yet discipline the market on its own. The proposed cap, pegged to 2020 rates, would have been fought hard on appeal, and the risk of losing was real. And there is a colder institutional logic underneath all of it. Formal regulation is slow, expensive, and appeal-heavy, and it ties up scarce economic and legal resource for years. Faced with a market already feeling some competitive pressure from over-the-top messaging, a regulator can quite rationally prefer a quick, tidy, face-saving outcome to a long and risky fight it might lose. Seen that way, the decision is not sloppiness at all. It is a perfectly intelligible piece of self-interest. No serious person wants a regulator intervening simply to be seen intervening, and I would not have Ofcom regulate for show.

None of which is a criticism yet. It is the case for the defence, and it is a decent one. My difficulty is not with the decision to step back. It is with the currency in which the retreat was paid for.

But I had a question about the nature of these commitments. So I did the unglamorous thing and asked.

The awkward question

Are they, I enquired, undertakings given under the Enterprise Act 2002, the kind a competition authority can take to a court?

The answer came back plainly. They are not.

They are unilateral: each operator wrote its own. They cover on-net termination only, not the interconnect route, not discounts. They were given, on Ofcom’s own account, while the operators expressly maintained that they did not accept the finding of market power in the first place. And there is no statutory mechanism attached to them: no penalty for breach, no injunction, no lever of any kind. The single obligation with a hint of teeth is that the operators will tell Ofcom before they raise prices.

Notification is not enforcement. It’s like telling the fellow in the pub you’re about to give him a Glasgow kiss, and then giving him one anyway. He had fair warning. He is still on the floor.

So let us be clear-eyed about what has actually been accepted here. If one of these operators departs from its commitment in 2027, Ofcom cannot fine it for the breach. It cannot injunct it. It cannot enforce the commitment at all, because there is nothing to enforce: no instrument, no jurisdiction, no consequence. Its only recourse is to dust off the very review it has just discontinued and begin again, from the top.

That is not a commitment in any sense that would survive contact with a courtroom. It is a mood, dressed up as a term. It is, to put it as plainly as the subject deserves, not worth the paper it is printed on.

Let me be fair to the other side of that, because I do not want to overstate it. “Not legally enforceable” is not quite the same as “worthless.” These four operators deal with Ofcom every week of the year, across dozens of open questions, and a firm that is seen to tear up a public promise the moment it becomes inconvenient pays a price in the relationship, even where no formal sanction follows. In a concentrated market, that reputational and relational pull is real, and it is not nothing. The honest position is that these commitments sit somewhere between legally meaningless and quietly influential, and probably closer to the middle than either caricature.

But notice what that concedes, and what it does not. It concedes that the operators may well behave for reasons of their own. It does not make the commitments a remedy. Good behaviour that depends entirely on the goodwill and self-interest of the party being restrained is not a regulatory safeguard, it is a hope. And a regulator is not supposed to discontinue a market investigation, having found market power and a real risk of excessive pricing, in exchange for a hope. It is supposed to leave something behind that bites if the goodwill runs out. Here, when the goodwill runs out, there is nothing behind it at all.

I should stress that this is not a lie by Ofcom. It never called these things undertakings, and, to its credit, it told me the truth the moment I asked. The difficulty is not dishonesty. It is that the reassuring public word, commitment, is doing work in the headline that the underlying legal reality cannot do anywhere else. The market problem has not been solved. It has been re-labelled as solved, and the label rests on a promise the giver can walk away from whenever it suits.

A live demonstration, helpfully provided the day before

If you think I am being unfair, that a warning surely counts for something and these operators can broadly be trusted to keep their word, then consider what one of them had done twenty-four hours earlier.

On 30 October 2025, the day before it discontinued the A2P review, Ofcom issued a rather less congratulatory statement about Virgin Media O2, the very operator whose commitment it would rely on the following morning. VMO2 had decided to raise its mid-contract prices beyond what many customers might have expected when they signed up, and to apply the increase to existing customers, leaving some of them facing bills materially higher than they had planned for. It coupled this with the offer of a penalty-free exit. And it is important to be fair here: nothing VMO2 did broke Ofcom’s rules. It gave the notice required, offered the exit right required, and stayed within what the regulations permit. VMO2 did what it was allowed to do, and then did precisely as it wished within that. That was its right.

I would only note, in passing, that a penalty-free exit is a thinner protection than it sounds when a good many of those customers are still paying off a handset, and can only leave without penalty if they first clear that balance in full. Fair warning is not the same as a fair escape. But that is a quibble with the design of the remedy, not an accusation against the operator.

Now here is the part that matters, and it is the opposite of a complaint about VMO2’s behaviour. Those price-rise rules carry real force. They are General Conditions, breach of which can, in principle, cost an operator up to ten per cent of its turnover. And even with that lever available, when an operator did something within the rules that Ofcom plainly wished it had not, all the regulator could do was express its disappointment, observe that the move went against the spirit of its rules, and write a firm letter to the mobile companies reminding them to treat customers fairly. Martin Lewis gave O2 and Ofcom both barrels. And that, so far, has been that.

Let that settle for a moment, because it is the entire argument in miniature. If a rule that carries a ten-per-cent-of-turnover sanction still leaves an operator free to act, lawfully, in a way the regulator dislikes and cannot prevent, what on earth do we suppose an unenforceable voluntary commitment is worth in the same hands? The A2P commitments ask us to trust that four operators will, of their own accord, price in a way they would prefer not to, with no lever behind the request at all. VMO2 has just shown us, entirely legally, how much appetite there is for doing the thing the customer would prefer and the regulator cannot compel. That is not a Glasgow kiss thrown in breach of the rules. It is a Glasgow kiss thrown within them, with fair warning given, and the recipient still on the floor.

We have been here before. I know, because I was there.

If this pattern feels familiar, it is because I have quite literally argued the other side of it, and lost.

Some years ago I acted for a challenger network that believed BT, our former state monopolist, and now the parent of one of these very four, had been overcharging for its Payphone Access Charge, the fee for connecting freephone calls from a phone box. Over the years that charge had risen more than sevenfold, with no clear basis in cost orientation to explain it. Back in 2005, BT had given Ofcom “undertakings” to head off regulation, promising the charge would be cost-based. We tried to enforce them. We went to the Competition and Markets Authority and argued, at length, that those 2005 promises were binding undertakings in lieu of a reference under section 154 of the Enterprise Act.

The CMA’s answer, in February 2017, was that they were nothing of the sort. Because no authority had ever formally assessed whether a market investigation reference was even on the table, the promises had never become section 154 undertakings, and so neither the CMA nor Ofcom had any power to enforce them. A promise given to head off regulation, spoken of as though it bound BT to something, turned out in law to bind it to nothing. It was a mood, dressed up as a term. And across the years the label went unchallenged, the charge climbed regardless.

I took a lesson from that, and it was not bitterness. It was this: a promise extracted to avoid regulation, sitting outside the statutory machinery, is worth exactly what the giver decides it is worth on the day. That is not cynicism on my part. It is the settled position of the competition authority, in writing, addressed to me.

The A2P commitments are those 2005 payphone undertakings in a new suit. Same origin, given to sidestep intervention. Same reassuring label. Same complete absence of a lever to pull when the music stops. And tellingly, when I asked what analysis had been done on the risk that lopsided termination rates between the favoured four and the dozens of smaller range-holders might simply be arbitraged, the sort of gap that, in the messaging world, positively invites artificially inflated traffic, I was pointed back to a paragraph about monitoring. The substance went unanswered, because the framing has no answer to give.

The point, plainly

I hold the people at Riverside House in genuine regard. As economic regulators go, we could do a great deal worse, and much of the world does. So let me be careful about the charge I am actually making, because it is not bad faith.

It is this. A regulator’s entire currency is that words mean what they say. It fines a small provider for breaching a General Condition. It expects the promises made to it, and the obligations it imposes, to bind. That authority is corroded, quietly and cumulatively, every time the regulator itself accepts a promise it knows it cannot enforce, from the largest players in the market, and presents it to the public as a solution. The minnow gets the rulebook thrown at it. The whales get to make a commitment that commits them to nothing.

The remedy is not complicated, and it is not “regulate harder.” It is one standard. If a voluntary commitment is solid enough to discontinue a market review, make it a proper, enforceable undertaking, with a consequence attached, so that the public reassurance and the legal reality are, for once, the same document. And if it cannot be made enforceable, then say so with the same plainness Ofcom demands of everyone else: this is a gentleman’s agreement, worth nothing the day a gentleman decides otherwise, and we are trusting to good behaviour.

That would be honest. What we have instead is a decision resting on a promise that is not worth the paper it is printed on, and a great many people, from the NHS to the corner shop, being told the matter is in hand.

It isn’t. And no amount of monitoring will make an unenforceable promise enforceable.

---

Peter Farmer is a regulatory economist. He has spent longer than he cares to admit arguing about whether an undertaking is, in fact, an undertaking.