Friday Reading List – 3 March 2017

As anyone looking to properly understand London’s transport needs and network knows, context, background and best-practice are important. As readers might imagine, behind the scenes here at LR Towers we thus spend a lot of time sharing links and reading around the subjects we cover here.

We also occasionally share links containing good information about transport topics that we know we just don’t have time to cover. We also all, as authors, occasionally write elsewhere on this or tangentially related subjects.

If you’ve got something you feel we should read or include in a future list, don’t forget to email us at [email protected].

28 comments

  1. I find that Treasury article very worrying.
    Some people still can’t tell the difference between a Train & a locomotive!
    That Paris article will probably require careful study, I think

  2. I love to see how bold some European cities are being in trying to tackle air pollution/congestion. Regarding Paris, I didn’t know about car free Sunday either. I also like the idea of closing one bank of the Seine.

    As already mentioned by Greg though I’m sure it’ll require careful study but sadly I fear we won’t see anything as bold here. I’ll continue hoping…

  3. @Greg T – welcome to the new old world of nationalised industry finance. NR now face an identical dialogue to that which BR (and every other nationalised industry) used to face annually. “Bob, Can you squeeze another £100m from property this year”. The only difference is that the Treasury deals directly with NR – never have allowed that in my day. It’s a measure of the centralisation in the Treasury’s hands these days.

  4. Re: GH – I’d be interested to know whether you think that centralisation occured by accident (an unintended consequence of a sequence of changes to the rail industry structure which were made to achieve something else) or design (someone’s intention, either all along or after the first bits of the “accident” scenario had occurred).

  5. @Balthazar – I can really speak only about DfT/Treasury relations, although I see similar patterns elsewhere. As a general rule, up until at least the ’90s, the Treasury was kept at arms’ length by spending departments. Individual policies and schemes were not discussed with them except at the time of the annual PES round, or when exceptional circumstances,such as the renationalisation of LT, arose.

    Departments organised themselves so that there was a career structure for civil servants specialising in the sponsorship and supervision of quangos and nationalised industries. (My own experience started with the New Towns, progressed through the then Nature Conservancy Council, the National Bus Company, and then BR, with spells dealing with local authority transport finance and setting up LT as a new nationalised industry). With the large scale privatisation of state enterprises, theneed for that sort of expertise was removed, or rather dispersed to regulators, and, in the case of rail, to franchisors. Departments retained very little expertise other than in the mechanics of the industry structures.

    Regulators and franchisors were easy prey for the Treasury, being themselves arms length bodies. Combine that with the Blair years when the effective split in government meant that the PM focussed on foreign affairs and the Chancellor on domestic matters, and the Treasury was nicely positioned to intervene at will.

    You see the evidence today when major policy and project announcements emanate from the Treasury not the departments of state.

  6. Treasury says to NR we’re subsiding you less, you make do. NR says they’ll spend the same amount and generate the money elsewhere. Treasury says they’ll take that money NR made successfully. Do I understand that right? It’s like an inter-government version of privatise the profits and nationalise the risks/debt/losses. Fwiw I think this is just inter-government rivalry not a party or policy specific thing to disagree with along party lines.

  7. @Toby, GH: It sounds very much like a political system that collapsed in (fairly) recent past…

  8. The “restaurant in Vauxhall” in the Treasury article appears to be Nando’s Bermondsey…!

  9. London may not have closed off a freeway like those on the banks of the Seine, but the east-west and north-south cycle superhighways come close and the fuss created mirrored that in Paris. And, as in Paris, I would expect that the extra congestion so complained about will have dissipated by the end of the year.

  10. I liked the Mayor’s comment in the CityLab article: “let’s just throw up some cones & see what happens.” Could that happen in London? Should it? If nit, why not?

  11. @Old Buccaneer
    There used to be a time when the spot for the centre of a mini roundabout was determined by putting a tractor tyre down and seeing where the traffic moved it to. These days, though, even doing experimental traffic schemes needs a Traffic Order with prior consultation. So much harder just to ‘put some cones down and see what happens’. It’s also the case that with any significant scheme traffic takes 6-9 months to settle down after implementation and the risk with just putting some cones down is that you get an immediate problem – that with time would go away – and the outcry forces the scheme to be abandoned. A good example is the removal of the roundabout outside Buckingham Palace and the pedestrianisation of one side. In the initial days the outcry – especially from taxi drivers – was immense because of short term congestion while drivers got used to the new layout. Fortunately we were able to sit out the problems and now no-one would dream of reinstating the old scheme.

  12. @quinlet – in addition to the paperwork, it also seems that TfL need to model proposed changes and that seems to take a long time. I’m guessing that traffic models need specialist people to develop them and there is a long list of projects waiting for those limited resources…

    As someone once said “All models are wrong, but some are useful” so I’m not decrying their usefulness but I wonder if there is a misplaced faith in their accuracy? I don’t recall ever seeing analysis of actual outcomes compared to predictions.

    Also, models produce a range of possible outcomes but the general public seems to have difficulty dealing with probability so invariably seize on the worst possible outcome as if this is being predicted with certainty.

  13. @Reynolds953 – it’s standard procedure for those supervising modelling work to check predictions against past outcomes; indeed, that is the way in which most models are constructed, you know.

  14. @Graham H – TfL may check outcomes against traffic modelling predictions but I don’t recall them ever making this public, either coming clean that predictions were way out or blowing their trumpet that predictions were accurate.

    It is common for scientific and engineering data to have a confidence level but I don’t recall seeing this type of information being provided for traffic modelling.

    However I do have a degree of sympathy for TfL in that providing lots and lots of detail about modelling will just invite interminable questions.

  15. @Reynolds953 – I can assure you that, having managed many teams of professional transport modellers for my sins, backcasting is one of the first things we do whenever a new model is built or recalibrated. I suspect that TfL, like most professional transport planning organisations, assumes either that the public aren’t interested or that everyone would get bogged down in an “amateur night” discussion about assumptions and the like. You have only to look at the average output of crayonistas (and the BML2 proponents in particular) to see their point.

  16. @Graham H – I can certainly understand the point about “amateur night”. Last week I went to a meeting at my local borough and TfL were there to talk about a traffic scheme. A resident wanted to talk about traffic modelling in obsessive detail and even from my layperson’s perspective (although I have an engineering background) I knew the resident was missing the point completely.

    On the other hand, it can be frustrating to be presented with a “computer says no” response.

  17. @Reynolds953 – of course! The “Scientists have invented a black box” syndrome. Combine that with the “elite expert scum” syndrome and we are all probably lost. What is necessary is for projects and analyses to be explained simply and clearly but that is hard work and few technical departments have the right skilled people to do that. (I have made something of a living doing this since BR privatisation, but have met relatively few competitors…).No glib answer, I’m afraid.

  18. My closest experience was traffic modelling was witnessing animations of the A23 in Brixton town centre before the pavements were widened and junctions altered. It was all very well, but I noted that all of the busses in the model departed the string of bus stops without a single one changing lane (to extricate itself from terminators in front still unloading) while all the cars in the remaining other lane sailed serenely by. I therefore took it with a large pinch of salt, and wondered with some awe what it takes to produce the models. They are certainly certainly useful, but it doesn’t seem to take much for a component detail like the idea that buses don’t change lanes to throw it all askew. Perhaps these models are more sophisticated now, though I would not have wanted a more realistic rendering to have thwarted the removal of a lane and widening of the pavements.

  19. @NickBxn – I don’t know what model was used in this instance, but the functionality of certain pedestrian modelling suites – Legion, for example – where pedestrian flow can be modified to deal with random obstructions, suggests that micro-simulation of lane changing shouldn’t be too difficult these days.

  20. In my experience the smaller scale models tend to be the most accurate and useful. More ambitious models are tend to be less reliable because of the more complex assumptions that need to be made. Graham H will probably be familiar with the RHTM (Regional Highway Traffic Model) fiasco of the 1980s. And the DfT traffic forecasts for England have been consistently wildly over-optimistic for the last 20 years. One early forecast predicted that there would be one HGV for every man, woman and child in the UK by 2020. Backcasting is one issue but clearly not enough sanity checking goes on. Even today, DfT forecasts an increase in car ownership in London of about 40% by 2030 (based on currently low levels of car ownership against household income in inner London) without any consideration of the fact that there is nowhere to put them at night, let alone space on the roads to drive them during the day.

  21. @quinlet – I would agree with all that and add that too many models -Railsys is a good example – are effectively “told” the required answer by selecting the inputs accordingly. Thus, for Railsys, the particular layout and service is calibrated against services which are “like” the test case. This undermines the value of the model in my view.

  22. quinlet /GH
    Given the harrowing rail modelling gets from DfT ( & others) resulting in [Carefull, could’ve been a full snip. LBM] the under-invested partial re-opening of the Waverley route, how come road models get away with the opposite gross fault & so often?

  23. @GregT – one of the great secrets in DfT is the matter of scheme selection. If you don’t look for something, you won’t find it.

  24. @Greg
    History has a part to play as traffic forecasts until the eighties tended to under forecast rather than over forecast, whereas rail forecasts have a history of overestimation as well. The Treasury may also have a hand in this, for example, with the optimism factor being introduced to down play benefits.

  25. TfL has a document which explains its London Transportation Studies
    Model (LTS)
    which models the behaviour of drivers, passengers, cyclists and pedestrians. Inter-related with LTS are four other models:

  26. The London Land-Use and Transport Interaction Model (LonLUTI), which predicts the use of land for different human activities depending on Government policies and transport investment
  27. Railplan Public Transport Assignment Model, which predicts the public transport mode (eg. rail, underground, bus) and route that a person chooses to get to their destination, as well as the associated crowding impacts.
  28. The Highway Assignment Models (HAMs) which contains five models covering the whole of London, to predict the routes that drivers choose and the associated congestion and delay impacts on London’s roads.
  29. The London Regional Demand Model (LoRDM) which is still in development.
  30. @LBM – I don’t know about LoRDM, but I don’t think the others are guilty of the sin I mention – selecting the inputs in accordance with what you want the outcome to be.

Comments are closed.