Ofcom recently released “Statement: Reviews of Call Termination Markets and End-to-End Connectivity Condition”, a consultation we had a lot to input to.
One of our suggestions was the introduction of a charge per call attempt, yes, whether the call was successful or not. In the current paradigm all providers in a call’s chain of custody incur costs at every stage but only generate revenue when the call is answered. This creates an asymmetry, particularly when there are heavy flows of calls to invalid numbers or calls which only remain connected a few seconds, i.e. spam and scam calls. This asymmetry causes two effects:
- Providers have limited economic incentive to block such traffic as the majority of the cost is offloaded to their competitors. It seems some only do things when there is a profit in it – morals and the good of the market be damned.
- Spammers have every economic incentive to pump as much crap as they can towards our consumers because they only pay when they get a win. The invalid numbers, the astute customers who terminate calls immediately cost them nothing-to-little. The old granny who keeps a call open while they rob her of her life savings, becomes the exceptional call which has a more meaningful cost, albeit many orders of magnitude less than her life savings.
Our suggestion was that by allowing terminating networks to impose a cost per call attempt, however small, the incentives would be reversed. Operators would still receive at least some income even for calls they blocked, and more importantly, the spammier and scammier a call profile was, the greater the proportion this charge would be making the business of originating nuisance calls harder. For ordinary business and consumer traffic, this cost would likely be inconsequential however and, given the framework within which it would be set, would have a compensatory reduction in the per minute costs. That being so, average duration calls would cost the same, and long duration calls would cost less; only short duration and unanswered calls would cost more.
What isn’t to like about that? Ofcom’s response was, politely, fascinating:
“Likewise, we do not agree with Simwood’s suggestion that we should recalibrate our cost models to introduce “per call attempt fees” for calls that are not answered. We recognise the rationale behind Simwood’s argument that introducing a per call attempt fee could in theory help to reduce the incidence of spam calls. However, this would be a major change of approach to wholesale charging and implementing it would be costly as it would require significant changes to providers’ billing systems. Determining an appropriate fee would also require extensive modelling, though for reasons set out above it would likely be low as we would only expect the combination of any such fee plus WCT charges to cover the incremental cost of termination. So, we consider that moving to such an approach would be disproportionate.”
The bold has been added by us in relation to points we’d like to explore now.
There is little theoretical about price elasticity of demand. It is proven in every business across the land every day, not just on the sell-side but also on the buy side as record youth unemployment and rising general unemployment demonstrate following lunatic rises to the cost of employing people – a rant for another day. “In theory” feels to me as an attempt to diminish the effectiveness of this suggestion so what follows next doesn’t sound so unreasonable.
They then say this would be a major and expensive change to billing systems. Seriously? Here, we record every CDR whether a call was successful or not. Those which have an answered duration of greater than zero get billed, others are just archived. Amending that to put them all in the billing queue would be a single-digit number of characters change in code, take less than 2 minutes, including running our many unit tests, and could be deployed the same day. Even providers who are adding CDRs to a database and pulling out those to later bill would simply need to remove the “> 0” from the associated SQL statement. Meanwhile, Ofcom continues to advocate for Dirty Origin Surcharges, something we will opine on separately as we had some suggestions there too (TLDR: bin the fecking things), which were far and away the most complicated change to billing systems I recall in 30 years. So complicated in fact that a number of carriers continue to output incoherent rate sheets manifestly exhibiting human-hands in their preparation. But that’s ok, because there was a gravy train attached to it and only the third world suffers. Spending a few minutes to save some domestic grannies from being defrauded? Oooh, no, that’s too hard and expensive.
They then suggest any per call charge would be low because they’d have to vary the per minute element to compensate. That is our point earlier – only short-duration crappy traffic would see an increase in cost – average traffic would cost the same and high quality traffic would see a saving. Only the spammers should be adversely affected.
In conclusion, price elasticity is theoretical, it would cost too much to implement, and it wouldn’t cost the spammers much. That is the definition of “disproportionate” folks! Meanwhile, a regime which costs much much more to implement and administer, encourages spammers to ratchet up their levels of fraud (i.e. spoof UK numbers to avoid surcharges), does nothing to help protect the public, but affords a gravy train for those operators low enough to implement it, is deemed fair and proportionate.
If logic and reason wasn’t enough, we also made the small point that preventing telecom networks being a criminal vector can be interpreted as the will of Parliament. Who cares though? Not the Regulator.
If you’d care to read our consultation submission and Ofcom’s response they are linked here.
