What follows is very much a ‘first cut’ of something currently in hand, but given all the noise about the subject area I thought it might be helpful. If I’m seriously off course here, I’d like to know about it!
I’m getting lost in what the argument is about today’s railway ‘costs’ and whether it is an argument or simply confusion. I decided to look into this after hearing Sir David Higgins at last week’s Transport Select Committee and comparing and contrasting with McNulty’s observation a few minutes prior.
The McNulty investigation had no brief to consider whether today’s railway met passenger needs, or how it was going to cope with future growth or many of the other considerations that are weighing on people within the industry. It was primarily concerned with examining the pre-existing opinion that subsidy was much higher than in British Rail days so there must be huge wastage that could be stripped out to reduce government subsidy. The outcome was a report that thought there was about a billion pounds to save (over and above savings initiatives already in hand between ORR and Network Rail), and that was on the basis there was then (2009-10) about £4bn of subsidy disappearing into the industry, with the industry itself generating about £6.5bn cash. Government support therefore represented between 35 and 40 per cent of total industry costs.
My first port of call was the BR reports and accounts, documents that require a lot of effort from which to extract basic management information from a mass of financial accounting, changing financial policies and a huge amount of noise caused by restructuring and various asset sales.
I only concentrated on the six years up to the start of 1994. I have used CPI indexes to relate costs to today, but feel consumer indexes may slightly underplay cost inflation in an engineering business.
My reading of the situation is (at today’s prices) BR investment had run for some years at a little more than £1 billion a year, rising to more than £2 billion just prior to privatization changes. I would expect an organization with £12bn of assets in 1994 (today’s prices) to require at least £600m a year ‘investment’ to stay still, so it feels as though real, though modest, improvements were being made to the asset base, though whether they were enough is doubtful bearing in mind several generations where asset improvement was clearly lagging behind needs.
The level of actual investment today is hard to gauge, and here is why. Network rail are the only people whose accounts are transparent, but they are only responsible for land, property and the track and signalling infrastructure that sit upon it. Investment in that area is roundly £1.5bn a year today, more than the BR investment until its final year, but for only a proportion of the asset base. The remaining assets, primarily rolling stock but perhaps including lots of services leased or bought in, such as telecommunications, now belong to other people. These other people spend capital as part of their business and the costs of the capital may or may not hit the railway accounts as simple alterations to revenue expenditure. If we assumed that annual spend on rolling stock leasing were the same now as BR spend on rolling stock production each year, on average, then the industry’s current rolling stock costs of (say) £2bn a year must also contain a huge capital investment element. Of course, this is not quite how it works because the RosCos also do maintenance and have financing costs. However there is no doubt that average fleet age has reduced and that perhaps a minimum of £1bn, on average, could be regarded as legitimate capital expenditure, to facilitate historic comparison. Consider also the huge investment by contractors in modern plant, none of which features in railway accounts, but may be covered by increased maintenance and renewal costs (or does the investment mean these costs are actually lower, as economic theory would like us to accept). We don’t know and it isn’t clear, though with a lot of effort no doubt we can establish a proxy.
Feels to me as though current capital spend throughout the wider industry may be of order of £2.75bn to £3bn a year. This is more than double what BR had been spending for a very long time and in a business known to have over-aged and under-maintained assets, and whose growth has already doubled and seems likely to continue to grow quickly, is surely a good thing? We might aim to do better, but for heaven’s sake don’t impede it. Arguably we need to do more.
Incidentally when the expression ‘investment’ is used above, this needs clarification. Ideally the expression would mean capital invested in something new or better to produce a return. Mere renewal is not investment (though it may be termed capital) but an operating cost. BR for various reasons came to present some costs as investment that were really renewal, which slightly muddies things. Network Rail, as you might expect, spend twice as much on renewal as they do on enhancements (about £2.75bn a year) and this is hard to compare with BR, though on things I would expect to be comparable, BR engineering maintenance (excluding rolling stock) was about £1bn. We know track renewal has gone up significantly (according to Higgins), but for comparable NR costs to have gone up by more than 2½ times seems a lot. I think part of this is BR counting renewal as investment but the rest feels like a fruitful area of more research.
I then looked at revenue costs. In BR days, gross income was falling from around £6bn a year to just over £5bn. I note that internal revenue today appears to be around £6.5bn. One might have expected a lot more given the vast increase in passenger traffic and is presumably explained by the marginal additional people being carried doing so at rather less than the average fare in BR days! We must not overlook loss of some traffics either.
The revenue grants given to BR during the period I’m reviewing were hovering around the £900m mark, on average. Together with the income above, this pretty much matched operating costs, zeroing out (that is presumably one of the reasons grant was set as it was). However, this was not the only grant as large capital grants were also given. They varied hugely, from £150m to £950m, but the average was £450m a year under BR, latterly.
If we take both grants together it seems to me that on a vaguely like-for-like basis BR was being granted about £1.3bn a year. This excludes government grants for PTE support and various EU-inspired grants that might tempt me to up the government figure in BR days to £1.5bn. This is rather less than three times today’s support which has to handle an industry growing very rapidly and modernizing faster than BR, in part to make up for the latter’s long term lack of investment. The difference is that in BR days we know exactly how it was spent, and today things are not so clear.
‘BR always made a loss’, one hears. Yes, but we need to understand what was going on. The operating account was pretty neutral IF one includes revenue grants. By the time accounts hit group level it made a loss, mainly because of the interest payments it had to make against its borrowing. If you look at the numbers above, it is obvious that capital grants only accounted for roundly a half of capital expenditure. BR could draw down on its reserves and pay for some capital investment, but when reserves were depleted it had to borrow from the government. This happened for three of the six years under review and totalled about £4bn. Interest rates were high, then, so annual outgoings increased from £65m to £203m, producing losses of the same kind of order.
Today, Network Rail has to borrow most of the money which, even at today’s low interest rates, creates an enormous (and mounting) debt mountain where even repaying interest cost about £1.5 billion a year. The Regulatory Asset Base (RAB) was £40bn in 2011, largely raised by issuing bonds. There has been no period of time in the past when railway capital debt was not eventually written off by the government as a pragmatic action (it will NEVER be repaid, after all), and I’m inclined to wonder when this will happen again. It would take at least £1bn a year interest costs out of the rail network at a stroke (without even having to make the efficiency improvements)!
The RAB is a menace. The way it is supposed to work is to reflect only enhancement: the addition of capital value. These enhancements are demanded by train operators who identify (with NR) an improvement that is required and for which passengers will be willing to pay. The work is done by NR, using money they borrow, and its added value is added to the RAB. The resulting increase in interest charges makes it way through the various agreements with train operators to increased station and track usage charges, specific to that enhancement. The increase in fare receipts (by carrying newly-attracted passengers or by charging higher fares) is then supposed to cover the increased charges and leave something for the train operator, otherwise it wouldn’t bother. One can kind of see the logic, but over time traffic results become increasingly distanced from individual improvements so it is hard to know how things are going to pan out long term. This is surely, also, not an indefinite process. It is pretty much the same position as the main line railways and London Transport were in between the 1930s and 1950s, where a bond-fuelled debt mountain pretty much put paid to a lot of investment that should have happened because debt was disproportionately high. I’m not sure we should have gone down this route. What of desirable things that might not yield a return, with sufficient certainty, that this system will stop happening?
- I’m persuaded that the railway costs too much (especially projects), but lack of transparency makes it hard to pin down exactly why.
- Comparison with BR not only reveals the situation is not (quite) as bad as suggested, but there are good reasons why some costs have gone up. In particular BR was under-investing in a system thought to be in long term decline while today’s system not only has to make that good but is also expanding rapidly. Higgins can actually see an end to the back-investment issues.
- It is, in particular, very difficult to make meaningful comparisons between investment and operating costs, seriously confusing things.
- The regulatory asset base system seems to add a lot of apparent cost and was hugely smaller in BR days. How can one compare like with like?
- We don’t seem to have a very good methodology for separating out day to day comparators with those relating to system expansion.
- This whole area is so fiendishly difficult it would probably make an excellent subject for a University thesis, or two. If it isn’t got right, it is useless.
- I need to go on an accountancy course…