Some basic UK rail statistics
Before looking at sets of rail statistics, the following points should be noted, all is not as it seems:
| Tendency to reduce information available to the public |
From the earliest days of railways, it has been a
government requirement for rail companies to furnish basic
data. This tendency increased during late Victorian
times and was put on a firm basis by the Railway
Companies
(Accounts and Returns) Act 1911, which required
substantial data to be provided pro forma; the objective
was that data would be
comparable across companies and from one year to
another. This provided a very stable long running source, but made it
hard to compare with data provided before that act came
into force. Data was originally consolidated by the Board of Trade, later the Ministry of Transport. On Nationalization the responsibility fell to the British Transport Commission who published very detailed and helpful statistical rail information. In 1963 responsibility passed to the British Railways Board, who initially provided statistical data to the former BTC standard, but whose quality of reporting deteriorated considerably from the early 1980s when data was reduced, and quality less consistent. Privatization created a data hiatus where data quality and quantity much diminished. The Strategic Rail Authority made a valiant attempt to improve this position and provide consistent data. Responsibility now lies with the Office of Rail Regulation. |
| Lack of data consistency | One might have thought that for basic
railway data there would be a consistent dataset
allowing one to make fair comparison of one year
with another. Put such thoughts out of your
mind. The first problem is that for a number of reasons decisions are taken periodically to change the way data is collected, presented or interpreted either to meet immediate business needs or to answer to some externally set objective. This is all well and good, but unless what has happened is fully explained then it is either impossible to make valid comparisons across the data change, and if the data change is not explained (as it often is not) then analysis will be flawed. Unfortunately, data is often viewed uncritically, and some analysis is indeed flawed as a result. The second problem is that those putting the data together over the last twenty years or so do not always have the interest and experience of the old office clerks and will cheerfully churn out huge sets of data without understanding how to relate it to past information and sometimes without having much knowledge or experience of what it is they are collecting and presenting. This leads to the production of data which is at best ambiguous and because it is not understood it can also be inconsistent within itself. There are often no rules, so different people will happily attack a simple problem in different ways. Asking how many Underground stations there are (and relating it to historical data) is but one example of a trivial problem that will waste many researcher's time owing to data being presented without consistency. If you think basic rules are not essential, a little thought about how to measure something as simple as distances between stations is a fruitful area for reflection. What chance quoting correct network size? A third problem is data aggregation, where data that would be perfectly accurate in the context of one company or operator becomes defective when added to that of another, mainly because it is apt to produce double counting. Not asking 'what is a passenger journey' is a particular area calculated to trip up the unwary who uses data without understanding the complications and which will produce inaccurate research as a result. |
| Data changes identified in table below | 1947-54 Season and ordinary ticket split
estimated as basis upon which data presented in
reports altered from journeys excluding journeys
of London Underground origin to those including
such journeys. 1948 Rail nationalization - reallocation of lines between national rail and London Transport alters statistical presentation. 1955 Railway Strike. 1956 Oil Restrictions. 1969 Series disrupted. This appears to have been because certain sections of BR line were transferred to London Transport and it was desired to make statistics comparable. 1974-6 There were several disruptions to the data series, in part caused by altering assumptions about season ticket usage from 540 journeys a year to 480. In reality, of course, this would have happened gradually over (say) 15 years. 1980 Series disrupted, apparently owing to including of West Midlands travelcards. 1984 Series disrupted as year changed from Jan-Dec to Apr-Mar. 1984 report was for 15 months and data reworked on 12 month basis. In addition, data no longer available about how many journeys took place only from stations operated by BR (the surviving stats represent journeys on BR that started from a BR or an LUL station). After 1996 it is not clear what the statistics represent (whether journeys originating from main line station or main line and LUL or a PTE station). There is some double counting (under five per cent) owing to aggregating train operator numbers and double counting some journeys, on the other hand open access operator numbers are omitted altogether. It is not known how journeys on LUL trains are counted when running over Network Rail tracks. ORR confirm figures quoted exclude journeys on light rail systems like Newcastle and Manchester Metros, DLR etc. |
| How we can help | Fifth Dimension is a knowledge-based consultancy and can provide a range of useful services ranging from data collection, presentation and interpretation, and checking of datasets for sense and inconsistencies that then need a clearly researched and written explanation. |
Back to Resources
