‘Optimism bias’ falls from favour

In November 2005, the UK Department for Transport (DfT) produced an estimate of the cost of converting road traffic signs for speed and distance measurements to metric units. Optimism bias accounted for between 26% and 33% of the total overall cost. Now, the usefulness of optimism bias is being questioned.

The comments by John Frewen-Lord and The Glob following the last article on Metric Views reinforce the view that the UK’s imperial road traffic signs are a handicap, both for road users and for the country as a whole. Yet, for almost forty years, the change to metric signs has fallen in the political ‘too difficult’ box – too difficult to explain to the public, and too difficult to carry out without running a risk of losing popularity (and votes).

“Too difficult” is not a phrase that comes easy to politicians, and the DfT has been required to find alternative reasons for lack of progress with the sign changeover. For many years, the DfT claimed that it would be “confusing” to make the change while a significant number of drivers had received no metric education at school (Hansard, 2002). More recently however, the cost of the changeover has become the focus of its attention, and an estimate was prepared in November 2005 by the Traffic Management Division. This is available on the DfT web site: #mce_temp_url#

At about that time, the Treasury had introduced optimism bias to counter criticisms of the cost escalation of road schemes. It ranged from 45% for a standard scheme to 65% for a non-standard scheme, adding between £175 million to £254 million to the 2005 estimated cost of the sign changeover.

Recently it has become clear that UK costs of civil engineering projects exceed those of similar schemes in continental Europe by a substantial margin, and the Treasury-sponsored body Infrastructure UK (IUK) has been asked to investigate and report.

Its Chairman, James Stewart, said last week that this work had identified a fundamental need to change the attitudes of those procuring and delivering projects. The process where budgets are inflated to allow for things going wrong was flawed.

“For every public sector project, before you even start you add a 50% optimism bias to allow for contingencies,” he said. “That then creates an affordability envelope, and then you get a budget equal to that affordability envelope.

“So it is no great surprise that most projects come in to that price.”

Thank you, Mr Stewart, for casting doubt on that 2005 estimate – UKMA has always said it was far too high.

Stewart said a better model was the London 2012 Olympics, where the contingency fund is managed separately. The Olympic Delivery Authority chairman and former chief executive of Network Rail, John Armitt, agreed.

“Optimism bias is a fundamental failing,” he said. “We did it at Network Rail and all it does is convince people that the money is there to be spent.”

That said, we should not fall into the trap of becoming obsessed with cost. The real issue is political will. The national interest requires a single system of measurement which is understood and used by everyone for all purposes. On the highways, it requires consistent information in a single system of units used by all those involved. This includes pedestrians, motorists, motor manufacturers, those who build and maintain the network, the emergency services, map makers, and so on, as exemplified by the comments of John and The Glob.

Other Commonwealth countries and, closer to home, the Republic of Ireland showed that where there is a will there is a way. Now it is our turn.

(Some information for this article was taken from ‘New Civil Engineer’ of 4 November 2010.)

14 thoughts on “‘Optimism bias’ falls from favour”

  1. Inflating the cost estimate 50% and then coming in just under it is called an “optimism bias”? To this American, it seems more like a pessimism bias, or possibly criminal fraud.

    Obviously, you need engineers who are looking for clever ways to do it effectively, but at the lowest possible cost, NOT inflating the cost so it appears out of reach, or purposefully making it cost more than necessary.


  2. Yes, the inflated cost of converting road signs is a “red herring” (as we say).

    I see some unfortunate parallels between the anti-metric/anti-EU/anti-change attitudes that have been encouraged in the UK and the anti-change/anti-government/anti-social programs attitudes here in the USA. Noteworthy that the shift in both countries took place in the eighties with the advent of a conservative government (Reagan and Thatcher). In some ways both countries have yet to recover from those ante-deluvian dogmatic beliefs.

    Reframing the debate is going to be critical if road signs are ever to be converted in the UK (or metric widely introduced in the USA). The book “The Political Mind” by Dr. George Lakoff (University of California, Berkeley) and “The Political Brain” by Drew Westen are both highly instructive in this regard.


  3. The costs for installing or replacing road signs is something that should be well known to DfT with little uncertainty. They have plenty of experience. What is so fundamentally different about metric signs?

    It is quite ridiculous that they should lump 45% – 65% on top just to allow for uncertainty. The work involved is quite routine. If anything the actual (pre-bias) costs are over already over the top. If it really costs as much as they say to put up or replace a road sign then why do highways authorities deploy so many of them? (they have recently been accused of sign clutter)

    There are other ways in which this estimate is disingenuous. For example there is no allowance for the fact that the true cost of replacing a sign for the purposes of metrication is partly mitigated by normal replacement liability. If say the normal life of a sign is 10 years and it is already 5 years old, the true cost of replacing it is about half the money actually spent – you get half the money back because its next replacement is delayed by 5 years beyond when it would have been replaced anyway.

    There is then the question as to whether the whole approach is wrong. The DfT figures are based on an instant replacement nationwide. Is that really necessary? They didn’t do it that way in Ireland. Distance signs were phased in over 8 – 10 years so that for most them the effective cost was zero. Only the speed limits were changed in a day or so. 10 years may be bit too long but it is possible to significantly reduce the overall cost using this approach.


  4. The Human Rights Act, which is part of British law, is supposed to outlaw discrimination and give all British subjects equal treatment. Why are British subjects like myself who understand and prefer to use a universal measurement system being discriminated against by the Department of Transport who give priority to a parochial measuring system of feet yards and miles? Why is the British media, particularly The Daily Mail, allowed to continue exclusively use the Imperial system of measurement without fear of prosecution? Celebrities like Steven Fry and Jeremy Clarkson, who are very popular and amusing comedians, continually ridicule the metric system with impunity. Perhaps someone with more knowledge of the law could challenge the government on these issues.


  5. In Australia, we changed all of the old pre-metric signs to fully metric signs in a single day. It was all done Sunday 1974 July 1.

    Naturally, to do this took a high quality plan that involved sign design and preparation ready for the big day. Also maintenance levels had been reduced so that signs that would have been replaced in the year or two before changeover were maintained at a safe level rather than at a high performance level – these savings then went to preparing new metric signs – maintenance budgets were sort of shifted forward.

    Also, other road dependent services had time to plan. “Metrication in Australia” wrote:

    “On the roads, all distance and speed signs, traffic regulations, touring information and maps were metric, all new cars since 1974 had kilometre speedometers and odometers, fuel and oil capacities in litres and claimed fuel consumption was in litres per 100 kilometres.”

    The nay-sayers, who claimed carnage on the roads and costs that were far too great were all proved to be wrong. Australians adjusted to the new conditions in a week or so and there has been no pain – only gain – ever since.

    Pat Naughtin
    Geelong, Australia


  6. I would just like to echo Pat Naughtin’s comments regarding the Canadian experience. As I’ve said many times before, Canada changed all its speed limit signs in one night (incredible when you think how big Canada is, and that all roads, including freeways, have speed limit signs spaced at regular intervals – exception, blanket signs on entering a large metropolitan area). The solution used was a stick-on vinyl sign over the existing sign, using the same material that new signs are made from anyway, so the converted signs lasted just as long as they would have anyway, and most were only fully replaced due to other reasons (changed road layouts, accidents, etc).
    There were all the nay-sayers predicting carnage of course, none of which happened. There was no visible change to the accident rate – and back then (1977) few cars in North America had dual marked speedometers. Many new cars had metric-only speedometers, even before the signs were changed.
    I don’t know how many times this story needs repeating to the single-digit IQ policy makers in DfT, but they obviously still don’t get it. Sad.


  7. The reason I pointed out the two books I cited on political discourse and persuasion is that the position of both the former Labour government and the current one has nothing to do with facts or reason. This is why, although John Frewen-Lord is well justified in his frustration with DfT, I believe it is misplaced.

    The only hope I see is for UKMA and other groups and individuals who support completion of metrication to find a way to reframe the debate that finesses the hot button emotional hook of anti-EU sentiments that has been attached to resisting road sign conversion by replacing it with a framing of modernization and competitiveness.

    Ideally, a group of industrialists and other influential folks (politicians, pundits, media folks, entertainers, etc.) would both lobby the current government and create a media campaign (advertisements, articles, appearances on talk shows, etc.) to over time fundamentally change the way Britons think about this issue.


  8. It is certainly true that a change of thinking is required. It isn’t just the politics though. It’s the way people think about measurement. Any society that mixes metric units with non-metric loses sight of the importance of a coherent sytem. In the media I am constantly hearing, otherwise metric friendly, presenters switching between metres and miles.
    The DfT and their political masters do not seem to understand that measurement units on road signs cannot be left out of the equation. The indefinite postponement of the change back in the 1970s was an early sign of this muddled thinking. I mean what is the point of getting school children to build their education around the metric system and then neglecting to do the same with public signs?


  9. Last weekend my 61 year old cousin made an interesting observation. We were talking about satnav systems, and she was complaining that when her system said ‘Turn Right in 400 yards’, she was always getting it wrong – either turning too early or too late. ‘Why don’t they have some way of measuring yards in your car?’ she complained.
    I suggested that if we switched to kilometres on our roads – like the rest of the world bar the US – the problem would be solved, in that 4/10th of a kilometre on your odometer would be 400 m – which is the alternative to 400 yards that her satnav could meaure.
    She looked at me a bit confused, and obviously didn’t really understand what I was saying. I’m beginning to wonder if that is the real problem behind the resistance to converting our road signs – the general populace simply doesn’t understand the measurement units involved. If so, this is an appalling indictment on our education system – notwithstanding that our children have been metric educated for many decades now.


  10. One point I think is often missed, a way to decimate the cost of re-signing, and bring benefits to the environment. In short, use less signs! I firmly believe that there are currently (in the UK) in the order of 5 times more road traffic signs than is necessary or desirable. One tiny example is coming in on the A2100 from Hastings (junction of A21) into Battle. Every km or so there’s a speed change. 30, then 40, then National (Bannatyne roundabout) , then 40 (Telham), then 30 (Lower Lake). Between the later two, the 40 mph signs are repeated about every 50 metres. All this occurs in less than 5 km and it’s similar elsewhere. Ridiculous, ineffective, wasteful, confusing, dangerous, expensive. Who are the people who wanted this madness? The whole route could be subject to a global 30 mph (just two signs instead of 20+) with a corresponding improvement in safety. Put some responsibility on drivers to know the rules and drive safely at all times. Have an inner town area which is all 20 mph and an outer region area of 30 mph, everywhere else except motorways, 40 mph. Simple, effective, and save hundreds of innocent pedestrian and cyclists lives (per year) in the process.


  11. There is currently a project in East London which involves changing the speed limit on a stretch of the A13 from 40 mph to 50 mph (in conjunction with new speed cameras). Some new “40” mph signs have been put up; and it is obvious that the “4” is on a removable sticker which presumably has a “5” underneath. It will be interesting to see how they handle the old signs when the project is finally implemented. I’ll keep an eye out and let you know.


  12. Interesting to hear John Frewen-Lord’s comments about satnavs and yards. Since my car’s trip odometer displays distances using decimal, rather than base-88 (100 yards is 5/88ths of a mile), I long ago gave up and switched the satnav app on my iPhone to metric, along with the digital dashboard in my car. Problem solved.

    Admittedly I then have to convert speed limits, but there’s only six of those (20-70 miles per hour) with simple metric approximations (30 km/h for residential streets and 50, 65, 80, 95, 110 km/h for the rest) that are easy to remember by rote. This prevents distraction at critical moments while driving on an unfamiliar road as I try to work out how many tenths of a mile on the odometer corresponds to x yards to the next turning. And it’s only what any UK driver has to do when driving their own car on the Continent anyway; even our dumb-it-down, “too difficult to metricate” government doesn’t appear to think we’re too thick to manage that!


  13. Further to my comments in December, the new speed limit on the A13 is now in operation. The strategy for replacing the signs has been to put up the new signs in advance with a sticker showing the old speed limit, which has now been removed. The old signs have either been covered with a new stick-on sign, or completely removed where the new signs had made them redundant. This would seem to be a perfectly feasible approach to take for the entire country, and would enable signs to be put up ahead of time and then revealed on the day.


  14. Another example of inflation of cost estimates for construction projects in the UK emerged last week. The initial estimated cost range for the Forth replacement crossing was £900 million to £1.2 billion. However the Forth Bridge Constructors Consortium (FCBC) has now been chosen as the preferred bidder to build the bridge and connecting roads. The price is £790 million, that is between 12% and 34% below the estimate. A respected observer commented that the winning bid price showed that “Transport Scotland dramatically inflated the price. It did this because it did not want to get it wrong, but it was not sure how much it was going to cost.”


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: