Thank you, Eightonedee, for pointing out the Green Signals podcast.

There is an excellent book by Tim Harford, called How to Make the World Add Up. It contains lots of advice on how to critically assess published statistics and identify the way statistics are abused in reporting. It came to mind as I was listening to the podcast.

I think I will start with the podcast's rather critical look at the abolition of peak fares in Scotland, which it uses to demolish arguments for reducing fares. It quotes various statistics that are contained within the evaluation report for the no-peak-fares trial.

The analysis was undertaken nine months into the trial. It seems to be measuring recorded travel during the nine months and comparing with what it calls a "counterfactual" - modelled passenger demand assuming the fares change hadn't been made (and based on what happened elsewhere in GB).

Now it would be rather naïve to assume that nine months would be sufficient to record all the demand change triggered by the fare change. Some people can change their travel arrangements rapidly. However, the full demand change occurs over a number of years, through what is termed "churn". People are most likely to change their travel arrangements when moving home or starting a new job. Indeed, the availability and price of transport prevailing may affect decisions on job/home location. Three years has been a typical assumption for assessing transport changes; I have seen evidence that suggests a much longer timeframe may apply sometimes.

I can't see anything in the Scottish report that acknowledges or attempts to correct this (admittedly after an incomplete reading), so is the report underestimating benefits? The report summary has the following rather unclear statement:


I take this to mean making the same journeys by rail as before, but more often. It tends to reinforce the suspicion that only the "easy changes" have been captured and that more behaviour change is still to come. At least the Scottish scheme is continuing, so we will, hopefully, find out in due course.

 
There's an interesting analysis of UK rail fares in a shorter You Tube from the Green Signals team this week - see -https://www.youtube.com/watch?v=lI1O9ymfFSI.

Not particularly optimistic if you are looking for cheaper fares, but the comparison with fare levels in the 1960s in "real terms" is interesting (see section 7 starting at 6-04).

 
From a longstanding Facebook friend's feed


300 miles Philadelphia to Boston ... £278 pound single at today's exchange rate

300 miles Paddington to Penzance ...