So my train was on time, and is on time

This week I received an email from Network Rail (NR) that detailed the punctuality of the major railway companies. Of course there is that perennial joke that comes up with reference to Britain's railways that they are always late and that you are going to be stuck on a platform for hours freezing your arse off like a penguin waiting for the next iceberg to turn up. Well let me just burst this miss-information bubble and say that most trains in this country do arrive on time, indeed as the email title proudly proclaims, 'NINE OUT OF TEN TRAINS ARRIVE ON TIME DURING JANUARY.' Indeed the figures do give a good picture of companies' punctuality. I may have to sell everything own to pay for my ticket, but at least I'll get to my empty room on time.

For NR a punctual train is those arriving within five minutes of the scheduled time for commuter services and within 10 minutes for long-distance services. They take into account all factors, such as whether, vandalism and passing antelope. Therefore the results I received for six companies and are as follows (apologies the data hasn't come out well on the blog):-

Jan 2009/10 ; Jan 2008/09 ; % point change

First Capital Connect; 85.5% ; 93.0% ; -7.5%

Virgin Trains; 77.9% ; 71.8% ; +6.1%

London Midland; 89.3% ; 83.6% ; +5.7%

Southeastern; 84.1% ; 89.2% ; -5.1%

South West Trains; 90.8% ; 95.2% ; -4.4%

London Overground; 88.5% ; 91.7% ; -3.2%

Having seen these types of figures for years I can roughly say the results fluctuate, month on month, but that generally we can take it as a given that on average 85% to 90% trains have been arriving 'on time' for about a year, after years of improvement. The only constant is that Virgin Trains are always the worst performer, but given recent performance, they are improving.

What these results got me thinking about is how the punctuality compares with the performance of railway companies of days gone by. Are the railways doing better or worse than their predecessors? The answer lies in a number of parliamentary reports from the 1890s that detailed how punctual different railway companies' trains were (an example is shown). The first thing to note is that unlike NR's press release, the Parliamentary returns show the percentage of each company's trains that arrived on time within within more detailed time periods (i.e. 5-10 minutes, 10-15 and so on). The only exception is that those trains that arrived within five minutes of the timetabled time, in which case the returns were split by the percentage that arrived within three minutes of schedule and those that arrived between 3 and 5 minutes.

Therefore the Parliamentary committee of the 1890s gives far more information than NR. But why is this so? Why do NR only give me a dribble of information when they would clearly have access to more? For example Virgin operates only long-distance trains, however all I know is that 77.9% trains arrived within 10 minutes. What if all of them arrived on the 59^th second of the 9^th minute? Surely that would make Virgin and NR look bad? More importantly why can't I have more detailed information? I think the reason is, and this is purely speculative, that in an age when the media will jump on anything that makes the railways look bad, any data that may harm the reputation of the Train Operating Companies' or NR is closely guarded. Therefore, if we had more detailed returns and was able to say that not one Virgin Train actually ran to time (OK, I know this isn't true) then this would be a PR disaster for the company even though by NR (and apparently European) measurements they were technically 'on time' as they came in under ten minutes. This said NR reported in March 2009 that British railway punctuality was the highest recorded since 1992, so I can't really complain too much. Yet, I still think that there is a case for saying that NR keep a lot of information very close to their chest that would be interesting. Anyway I have gone off on a tangent...onward!

For my comparison I will use the results of the company I am studying for my PhD, the London and South Western Railway (L&SWR), and those of South West Trains (SWT) who operate in an almost identical region. This also eliminates in large part having to worry about separating the trains arriving under 5 or 10 minutes in my comparison, as both companies operated mainly suburban and commuter services and as such I can classify all trains arriving under five minutes as 'on time.' I will also compare the punctuality of Virgin trains and the London and North Western Railway (L&NWR) and as both ran mainly ran long-distance services, and therefore the cut-off for trains not running to schedule will be ten minutes. The results are as follows (apologies againthe data hasn't come out well on the blog):

Jan 1890 ; Jan 2010 ; % difference

L&NWR/Virgin Trains; 79.2% ; 77.9% ; +1.3%

L&SWR/South West Trains; 76.4% ; 90.8% ; -14.4%

These results while basic, and highly imperfect, do provide interesting reading. Firstly the L&NWR performed only marginally better than Virgin. However if we consider that the L&NWR and Virgin both operated the West Coast Main Line, that had, and has, the highest levels of traffic of any main line, then the problems with running to time become evident (even though there are far fewer trains on it today). Both companies would have had the difficult job of balancing both goods and passenger trains, would have far more infrastructure that could fail and would have had many more customers to carry. It may be argued that Virgin trains are actually facing fewer problems, given there are fewer trains on the line, however as their trains go at faster speed, then this is a factor that actually negates this consideration because the trains are closer together and require more complex scheduling. However it should also be noted that the recent WCML upgrade was behind schedule, over budget and while now complete, is still replete with bugs, but once things get 'ironed out' there is potential for Virgin's punctuality to improved.

Secondly the results of the L&SWR's performance were poor compared to its successor, despite operating similar services. Of course I cannot venture an wholly adequate reason for the difference, however it may be because SWT operate a clock face timetable, whereby trains leave and arrive at the same time each hour, whereas the L&SWR did not until 1912. Also nowadays the trains arriving at 'x' minutes past the hour have the same rolling stock and passenger capacities. These two factors therefore mean that while SWT can cancel trains if they are running too late, simply turning them into the next scheduled one, a late train on the L&SWR would disrupt the entire timetable as the late runner could not become a following train as they would have put the rest of the timetable into confusion and they would have had different types of rolling stock and locomotives attached to them. In addition on the lines SWT operate, there is very little freight, whereas the L&SWR did have a large goods operation. Therefore the L&SWR had more trains on the line and more potential for passenger trains to get stuck behind them because of failure or late running. Thus overall it seems that the SWT has a much simpler operation than its predecessor that gives it far better performance.

More detailed analysis of past and modern railway punctuality would be welcomed, however, with such limited data, the reasons that I can venture for differences between the past and present can only be speculative at this point.