There are a number of reasons why Boris Johnson’s policy for buses in London is wrong. In particular, as he now appears to have acknowledged, his advisers’ costing of £8m for the replacement of London’s fleet of 340 articulated buses by a “21st century Routemaster” was wildly inaccurate.

This was borne in on us when The Guardian newspaper rang last week and asked for a view from TAS on the argument over the figures. We did some quick calculations based on our National Bus Model. It quickly became clear that, even on the most generous of assumptions, Boris’s figures were entirely wrong: I’ll explain the details later on.

But the policy is also wrong for a number of other reasons – and particularly the proposed reintroduction of conductors. The concept of a bus operated by two members of staff was developed in an age when virtually everybody paid cash, and had to be issued with a small piece of card punched in an appropriate place to prove that they had paid. One can certainly say that the Routemaster was the ultimate design of crew-operated bus – and its longevity in service is a tribute to that design – as well as to its hugely expensive 12-year gestation period and very high initial capital cost.

However, it really is from another era. Progressively over the last 20 years, we have entered the world of Travelcards, Freedom Passes and Oyster cards. Even before the huge success of Oyster pre-pay and the introduction of cashless boarding in the inner zone, the proportion of passengers paying their fare in cash was very small indeed.

Indeed, in the later days of Routemaster operation, conductors had virtually nothing to do, except supervise boarding – and all too often rarely moved from the platform: I had more free rides on Routemasters in the last five years of their existence than ever before – not because I avoided paying, but because I was deprived of the opportunity to pay: nobody came for my money.

Thus, adopting a design of bus that would require the reintroduction of conductors is, on the whole, a silly idea: by all means recruit more staff – and use them to supervise, to assist people at busy times, and check tickets: but do not make the running of the service totally dependent on them.

Next we come to the issue of speed. Articulated buses were seen as the solution to the problem of how to replace the Routemaster because, with three doors, their speed of boarding and alighting at stops would be the same as the crew-operated double decker. In fact, arguably, they have proved to be even faster. At the same time, however, they are much safer than buses with open rear platforms: even if you could get approval for a new design of open platform vehicle, they are probably now uninsurable.

At the same time, the higher capacity of an articulated bus means that the same number of places per hour can be provided on fewer vehicles. Thus, on conversion from RML to artic, the peak vehicle requirement on the 38 fell from 50 to 44. On the 73, the saving was three and on the 12 it was six.

I have heard it expressed that the length of the artics causes more congestion. This is nonsense. The idea that 340 buses each seven metres longer than a double decker can actually cause congestion is just plan daft. Especially when the amount of time they spend at stops is lower than a double decker, and their use results in having fewer buses on the road.

Go back to a double decker with its smaller capacity and higher boarding times, and those savings would disappear. The number of buses on the streets would increase again, so contributing more to congestion.  We estimate that to provide the same peak capacity as now on the twelve routes currently equipped with artics would require an additional 217 buses.

Which brings us back to the costs of Boris’s policy. It seems that the £8m figure was based on the number of conductors required being one per bus, paid the same as the staff manning the remaining Routemasters in service on the two ‘heritage routes’, £24,600 a year.

In fact, of course, on these routes, the buses would be running for around 20 hours a day, needing three shifts for each vehicle on each day. Then there’s sickness and holiday cover, spare cover, signing on and signing off time, travelling time and all the other rings and bells associated with crew scheduling. To man a fleet of 341 buses a day running 20 hours would, we estimate, require a workforce of over 2,000, costing some £58m a year.

Then there would be the extra costs associated with restoring the capacity – 217 more buses, 543 more drivers and a similar number of extra conductors. We cost that little lot at another £41m a year, taking the total to just over £100m. Add a 10% premium to the capital costs for a small production run on a non-standard design and you get to £114m – slightly above the figure quoted by TfL.

Sorry, Boris. You’re wrong on this one: the reason that many thousands of articulated buses are operated in cities large and small around the world is because they are very efficient. There are certainly things you could do to improve things on London’s buses, but this is not one of them.