Here's a radical idea: let's scrap the current fare structure and build a new one. Seems there are lots of people out there who don't like present position - it's complicated, often illogical, and allegedly prevents people who don't use the railway nowadays from giving it a try.

What criteria should drive a revision? What are the difficulties?

The basic framework should be that the new system taken over the whole of Scotland should neither increase nor decrease the size of the amount taken in fares. Scottish Ministers will not wish to increase the subsidy agreed in the franchise (so no net reduction in fares) and passengers will not wish to pay more merely as a result of a change in the structure. So it must be revenue-neutral. It must also be a great deal simpler.

Any change will mean that some journeys will cost less, while others will cost more. This unpleasant fact must not be dodged, nor must the increase be fudged by some means, as this will conflict with the need to keep things simple. The pain of increased fares will soon fade: if it is mitigated it will last longer. While the holder of the ScotRail franchise operates over 90% of the services in Scotland three other daytime train operating companies (TOCs) also run here: Virgin Trains East Coast (VTEC) runs services from Berwick (and points south) to Edinburgh, Glasgow, Inverness and Aberdeen. CrossCountry (CC) runs services from Berwick (and points south) to Edinburgh, Glasgow and Aberdeen. Trans-Pennine Express (TPE) runs services from Manchester Airport via Carlisle to Glasgow and Edinburgh. TPE is the only TOC serving Lockerbie. Sleeper services are not covered in this plan - the Sleeper market is sufficiently differentiated from daytime travel for there to be no need for change.

We start with a blank sheet of paper. What system might we model? One of the simplest is the Japanese, helped of course by the fact that inflation there has been zero or negative for so long that fares have remained constant for many years. (As there are no "bullet train" (Shinkansen) equivalent services in Scotland the supplements charged in Japan are ignored in this paper.)

There are two fare bands - "trunk lines" and "rural lines". Some journeys encompass both trunk and rural sections. For journeys up to 400km the trunk fare is roughly linear at ¥16 per km (£1 = ¥150 in round numbers). Thus a 400km journey on trunk lines in Japan costs around £42. Above 400km the price per km tapers, with an 800km journey costing £70 and a 1200km journey costing £90. Journeys on rural lines are inevitably shorter, again with little tapering under 400km. 400km and 800km (the maximum) journeys costing respectively £45 and £75. It will be noted that ticket prices in Japan on rural lines are typically up to 10% higher per km than journeys on trunk routes. This added complexity seems out of place, and it is doubtful whether a similar structure would be popular with passengers in Scotland given that a large proportion of the routes in Scotland would be classed as rural. This paper therefore discards the idea of a dual structure based on the type of line.

The great simplicity of the Japanese system (modified by ignoring rural lines) is that the length of the journey is the only determinant of the ticket price: nothing could be easier to understand. A degree of tapering is sensible as most commercial purchases of any commodity (or goods) expect a discount for quantity, and rail travel ought not to be an exception. Where, and to what degree, a taper should operate is a matter for detailed examination. The points of change should be round numbers of miles (eg. 100 and 400) reflecting the fact that Scotland is much smaller than Japan, and relatively few journeys exceed 400 miles. A suggested ticket cost would be X pence per mile up to 100 miles, 0.85X pence for the next 300 miles (giving an average of 0.89X per mile, being roughly the same proportion as in Japan for 800km journeys). Any excess over 400 miles could be charged at 0.75X pence per mile. There would no longer be any variation between a journey of 100 miles between say Aberdeen to Kirkcaldy on the East Coast Main Line (now £29) and Thurso to Tain on the Far North Line (now £17.40). Clearly there will be problems in introducing a system where the two prices are harmonized - a good reason why the exercise has lain for so long in the "too difficult" box.

Transport Scotland will have the data necessary to determine the value of X such that the change would be revenue-neutral. Carrying out this exercise will deliver valuable insights about where stresses are likely to occur, and will indicate where modifications may be needed. Historically, unlike Japan, ticket prices on rural routes have been lower than on trunk routes. This can be ascribed to the policy of accepting that rural communities are often disadvantaged purely by their remoteness, and seeking to ameliorate this by ensuring that travel to urban locations (for health, education etc.) should be subsidized to a greater degree than travel elsewhere. It would be simple to carry out a desk-top study where rural journeys (which would be easy to define) would cost kX pence per mile, with k chosen appropriately (k = 0.6 in the example given above).

Once a clear and simple structure is agreed for a normal single ticket thought must be given to other types of ticket. A simple factor of 1.5 for First Class would seem sensible. A return ticket would not exist: there would instead be a single ticket for the outward journey and a separate single ticket for the return. (In Japan a 10% discount is allowed if the return ticket is bought at the same time, but I see no good reason to copy this.) The opportunity for the TOC to market special bargain fares (perhaps at Christmas, or in the summer holidays) to particular destinations would permit a discount being offered on return travel. As this would not be routine it would represent a genuine advantage to intending passengers.

Japan offers no discount for advance purchase, and I see no good reason to introduce that layer of complexity. Apart from a price advantage, the only reason for buying an Advance ticket here is that it brings a reserved seat - or it used to. CrossCountry has rendered that advantage nugatory. There is evidence on any long-distance service nowadays that a significant proportion of reserved seats are not taken up, causing confusion and annoyance to turn-up-and-go passengers. Formerly a small charge was made to reserve a seat, and it would be worth exploring whether a return to a modest charge (not more than £1) would be resisted. If there were no longer any incentive to buy a ticket in advance for price reasons, the attraction of a guaranteed seat for an extra £1 might be significant.

TOCs would be denied the flexibility they now have to price seats according to demand as the time to travel shortens. However this adds greatly to the complexity of the pricing system, exactly what this paper is seeking to eliminate. A passenger who pays £250 for a turn-up-and-go 200 mile journey is not delighted to discover that the passenger in the next seat paid only £40. The rail industry suffers reputational damage when this happens: the story is not "how inexpensive an advance purchase seat can be", but "how much we're being ripped off by the railway". Denying a TOC the ability to adjust prices is not an infringement: it's a much-needed simplification. Devising special seasonal bargain fares, as described above, would not be ruled out, but would be seen as "special" rather than routine.