To paraphrase Oscar Wilde:

"DaFT  knows the price of everything and the value of nothing."

It's the modern British desease "Can't do that costs too much."

Look what they are trying to do with freight charges.

It's not as simple as accountancy.

Let's take the example of a route where it is desired to run an "intercity" service and a stopping service. In order to allow the IC to run fast, a passing loop is required where the stopper can be overtaken.

Does the cost of that loop belong
a) to the stopping train
b) to the IC train (as the cost of running faster than the average speed of the stopper)
c) some mixture of a and b?

You could also say that the loop would be unnecessary if the frequency of the services was lower, so the cost would rest on the business case for a higher frequency service.

So, not so easy to divide the total cost between local and other service groups. It depends upon your viewpoint.

I think you are right about overseas, but the legal responsibility of the railways (in this country) was always to make the railway safe against both trespass and against the animals in the fields of the time - so the strength or height of the fencing would probably only be varied dependent on the farms it passed through.  In the Railway Clauses Consolidation Act 1845 [1] the requirement is clearly the railways - the point is that in those days trespass by people would be prevented by appropriate signs and simple wire and post fencing quite adequately. 
 
Times have changed and the modern requirement, mainly in towns and cities, for high security fencing is definitely driven by the public's own actions.   A popular misconception appears to be that DC electrification requires better fencing quality on the line generally, indeed this has been used by NR as a costly reason against the extension of the Merseyrail DC network. 

However this does not seem to have been required historically in the third rail area generally, vast rural areas of the former SR just have the normal wire and post fencing - unless the local public can't be trusted.

Paul

[1] This Act drew together the list of 'standard clauses' which would then be cross referenced by enabling Acts for new railways to avoid duplication.  It is interesting because many of the construction requirements are still quoted in stuff like the Crossrail Act!

http://www.railwaysarchive.co.uk/documents/HMG_ActRailwaysClauses1845.pdf

One thing the "bean counters" have never got to grips with is costing networks as a whole rather than their individual components. The problem being that the value of the parts is probably greater than the whole. 

Hence the problems Trowres listed about costing platform loops. Surely the simple way of looking at it is that the cost of the line from A to B is X the revenue for all traffic using the line is Y. Hopefully Y is greater than or  at least equal to X, otherwise the difference is the subsidy. By doing it that way there are two incentives to reduce X and increase Y.

I think a classic study should be the Leen Valley North of Nottingham, where there were three lines runing parallel to Annesley the ex Mid, GN and GC. All three were closed after Beeching except for a stub of the Midland to serve a colliery. Meanwhile north of Annesley BR freight had rationalised the lines and by putting in a North bound spur from the LDEC to the Midland and rearranged the juctions at Kirby to eliminate an ackward level crossing they managed to single serve all the big Nottinghamshire collieries for Merry Go Round Working  to the Trent Valley power stations.

Subsequently the Midland line had to be reopened through to Annesley and reconnected to the Kirby Mansfield Worksop line, which required the reopening of Annesley tunnel, it is now the very succesful Robin Hood line.

Had Beeching looked at the Network rather than individual lines there were lots of places where by either rerouting services or putting in short links it would have been possible  to close stretches of duplicate routes whilst still providing a rail service between A and B.

Of course the biggest  Network  which is World Wide Web  is impossible to cost hence it's great success as the "bean counters" are unable to cost what any one individual should pay for their usage. So the providers compete for customers by providing competing tarriffs. Plus as the  revenue exceeds costs the providers are happy, although as a moderate user of the web I resent subsidising you heavy users.

Which brings me round in a full circle Networks are impossible to cost and more importantly allocate those costs to individual users except, as a whole.

The biggest problem with UK rail privatisation appears to be our use of Net cost tendering for passenger franchises rather than Gross cost as used in Germany, the Netherlands, Sweden and other countries.

I am fascinated - Could you explain this a bit more please?

So let's see if I undertstand this, by example.

The Highland Council is responsible for providing public transport to a certain level on the Hebredian Light Railway from Isle Ornsay to Uig.

Under a gross cost tendered arrangement, a TOC (let's call the company Skye Trains) runs a service of 6 trains a day each way, with two extra short workings from Isle Ornsay to Portree, Southbound in the morning and northbound in the evening. And they charge the Highland Council 4.5 million pounds a year for this.  The Highland Council takes an income of 8.5 million pounds a year from fares, and spends 5.0 million in hiring / capital cost for trains, maintaining the line, and buying materials to do so, with the shortfall of a million being met by advertising revenues, and taxes on local businesses who benefit from the tourists brought in.

If a net cost tendered arrangement had been used, Skye Trains would have run the same service, but taken the income of 8.5 million pounds.  They would have hired trains from a RoSCo - so have had a gross spend of 6.5 million pounds. They would have paid Highland Council 1.5 million pounds (giving them half a million profit), and the Highland Council would have added to that another million from which they would maintain the line.

The Hebredian Light Railway scheme of 1898 proposed a line along the route I have used in my example, together with a branch to Dunvegan and a further line on the island of Lewis and Harris.  The lines were never built, and the sums quoted are fictional wild guesses - but hopefully they add up. Early morning post - not fully awake yet 

Balance Sheet:

Operational cost: 4 million
Hire of trains: 2.5 million
Maintainance and materials for line 2.5 million
Profit for commercial operator: 0.5 million

Income from fares: 8.5 million
Income from taxpayer: 1 million

Edits to add balance sheet and check sums!