Privatisation efficiency?

[quantinghome]

I have found some interesting data on the financing of the UK rail network, including the pre-privatisation period. Dividing inflation-adjusted costs by passenger-km gives a useful metric of how efficient the industry was then and is now. There should have been significant economies of scale as passenger numbers have grown. However, this hasn't happened. Instead we've seen a rather uneven subsidy level from government, with a trend of rising overall costs.

 

[LNW-GW Joint]

That was exactly the finding of the McNulty report in 2011, leading to the search for productivity efficiencies in subsequent franchise contracts and at NR.
That resulted in the DOO policy, which has run into the sand with all the recent industrial action.
Network Rail had its own efficiency targets to meet, which have also been found to be unachievable in CP5.
We have a right dog's dinner of a railway today as a result.
Is it any surprise that fares continue as high as they are?

 

[Chris Butler]

I have found some interesting data on the financing of the UK rail network, including the pre-privatisation period. Dividing costs by passenger-km gives a useful metric of how efficient the industry was then and is now.
View attachment 54925

Are the data nominal or real. If nominal, they are pretty meaningless.

 

[Chris Butler]

That was exactly the finding of the McNulty report in 2011,

Is that really true ? Here's the McNulty costs chart which shows rising real costs post-Hatfield and then falling costs. I think these are total costs, but maybe there's some twist that I'm missing ?

 

[LNW-GW Joint]

I think the point (made in the text prior to that chart) is that passenger km went up 59%, with cost per passenger km stayed largely static (down 3%).
You'd expect a much greater drop in unit costs for a network carrying 59% more passengers with a similar-sized train fleet.
The figures include some large Network Rail costs (enhancements etc) which don't usually figure in TOC franchise subsidies (the NR direct grant policy).
But it's why NR debt is now £50 billion or so, the interest hitting industry annual costs.

This was in the Interim Report:

We consider that there is considerable scope for the overall increase in expenditure to be reversed. Since 1996-97, although rail passenger-km have increased by 59%, there has been little or no improvement in the total cost per passenger-km, which is remarkable in an industry with relatively high fixed costs. Figure 1 below shows that, while there have been reductions of between 1 and 3% per year in train operating costs and infrastructure operating and maintenance expenditures, these have been largely offset by increases in renewals costs (largely driven by increasing volumes) and enhancement expenditures to accommodate the additional demand.

Of course, the costs were mostly for the 2000s and are now 8 years out of date.
The report has a lot to say about industry organisation and incentives, franchise lengths etc.
The new Williams report is going to re-examine all that.

 

[quantinghome]

Are the data nominal or real. If nominal, they are pretty meaningless.

They're real. I have edited the original post to clarify.

 

[edwin_m]

Is that really true ? Here's the McNulty costs chart which shows rising real costs post-Hatfield and then falling costs. I think these are total costs, but maybe there's some twist that I'm missing ?

View attachment 54932

That looks pretty consistent with the OP's graph to me, bearing in mind that it zooms in on a smaller part and displays it in forward rather than reverse time order. Looking at the left hand part of the OP's graph, costs per pass-km have been falling since McNulty, although not by very much.

 

[Dave1987]

People often talk about the tiny savings that were proposed from cutting staff from the 2011 report. The amount of money wasted through the broken franchising model, the amount of money wasted through the constant use of third party contractors for absolutely everything and the sheer amount of bureaucracy associated with all of that is where the most money is burned up in the system. Axing rail staff is the easy way to save money but it isn’t going to address the fundamental wastage in the system and saves tiny amounts compared to the amount that would be saved in solving the big wastage of money.

 

[Dr Hoo]

It is interesting that the first years of the OP’s chart (going back further than the McNulty chart) show subsidy increasing massively in the BR ‘sector’ era. I have to say that this doesn’t really fit with my own recollection and experience of the time. For one thing the dramatic real increases in fares year by year under BR really ought to have held subsidy in check.
On the other hand the last few years appear to be remarkably stable (even, perhaps, slightly improving) despite what we are told about problems all around.
Something just doesn’t feel right.

 

[route:oxford]

I don't doubt that, since 2008, the subsidy from the Goverment for the railways would have been slashed to the bare bones with salary increments limited to 1% per annum had it been a fully nationalised at that point.

Thank goodness for long term franchises and guaranteed funding models.

 

[edwin_m]

It is interesting that the first years of the OP’s chart (going back further than the McNulty chart) show subsidy increasing massively in the BR ‘sector’ era.

I think that's down to the amount spent on the privatisation process, which got under way fairly quickly after then 1992 election. Before that was probably a result of the early 90s recession.

 

[mrmartin]

I think the point (made in the text prior to that chart) is that passenger km went up 59%, with cost per passenger km stayed largely static (down 3%).
You'd expect a much greater drop in unit costs for a network carrying 59% more passengers with a similar-sized train fleet.
The figures include some large Network Rail costs (enhancements etc) which don't usually figure in TOC franchise subsidies (the NR direct grant policy).
But it's why NR debt is now £50 billion or so, the interest hitting industry annual costs.

This was in the Interim Report:


Of course, the costs were mostly for the 2000s and are now 8 years out of date.
The report has a lot to say about industry organisation and incentives, franchise lengths etc.
The new Williams report is going to re-examine all that.

Would you? The marginal costs of improvements is absolutely enormous. Resignalling/electrifying + acquiring new rolling stock to take a line from (say) 4tph to 6tph costs an absolute fortune to deliver not much more passenger-km.

I'd be much more interested to see the cost of ToC spend + NR opex, which is a better comparison IMO. I'm pretty sure the trend would have went down significantly.

 

[big all]

no one was interested in buying the railways
they had to double the subsidy in the lead up to privitisation to get any real interest
with most off the additional subsidy pre 1994 would be the cost off setting up all the seperate elements for sale

 

[Chris Butler]

I think the point (made in the text prior to that chart) is that passenger km went up 59%, with cost per passenger km stayed largely static (down 3%).
You'd expect a much greater drop in unit costs for a network carrying 59% more passengers with a similar-sized train fleet.
The figures include some large Network Rail costs (enhancements etc) which don't usually figure in TOC franchise subsidies (the NR direct grant policy).
But it's why NR debt is now £50 billion or so, the interest hitting industry annual costs.

My point was, and still is, that the OP said costs had an upward trend. The data shows more a number of phases, to which McNulty referred. In the (then) latest phase (since 2003 and the Hatfield splurge) there is a definite downward trend, especially if enhancements are excluded. The unit costs are down about 25% in 6 years and Operating plus Maintenance have nearly halved. That's very different from a rising trend.

The question about fixed costs is valid, but not as inexplicable as it might first seem. How much it can be explained by a backlog of maintenance/replacements as the nationalised railways, having been managed for decline, needed to be managed for growth, I am not sure, but I guess that is some of the story.

I am no apologist for Network Rail, but I thought McNulty was a big waste of time, mainly because it was so superficial.

The point about interest is very valid when looking at the accounts, but I don't think it is included in the McNulty 'expenditure' chart.

 

[Chris Butler]

The marginal costs of improvements is absolutely enormous. Resignalling/electrifying + acquiring new rolling stock to take a line from (say) 4tph to 6tph costs an absolute fortune to deliver not much more passenger-km.

I'd be much more interested to see the cost of ToC spend + NR opex, which is a better comparison IMO. I'm pretty sure the trend would have went down significantly.

I had a similar reaction, namely that marginal costs are low when there is spare capacity, but not so low when additional capacity is needed. That is doubly true when the existing infrastructure was being run down with insufficient maintenance and replacement to keep it 'in a steady state'. In that case the previous costs didn't really cover the long term cost of the infrastructure.

The chart I posted earlier (from McNulty) does show improvements separately. The effect isn't as big as one might imagine because enhancements are a small proportion of total spend (10%'ish).

The rolling stock acquisition wouldn't really show up directly in these numbers since the ROSCOs effectively spread the costs over the fleet lease life (after adding the cost of financing).

 

[quantinghome]

There’s certainly plenty of room for discussion about the various phases of spending and what they represent. However, the chart does nail one issue, which is the view that British Rail was terribly inefficient until private companies took over and sorted it out. There’s just no evidence for it.

 

[Chris Butler]

There’s certainly plenty of room for discussion about the various phases of spending and what they represent. However, the chart does nail one issue, which is the view that British Rail was terribly inefficient until private companies took over and sorted it out. There’s just no evidence for it.

Nor any clear evidence against it either. This data will never answer, or nail, that question.

If I were arguing the contrary case (which I am not) I would point to the fact that operating, maintenance and rolling stock costs are lower than at privatisation while the service is faster, more frequent and safer. In contrast to those, privately managed, costs, the main cost categories that have increased are the replacement and enhancement costs that are executed, in the main, by the state run part of the railway.

My feeling, having worked for BR, is that it was a well run and pretty efficient organisation, but with capabilities suited to a mission of managing decline and benign neglect of a wide swathe of assets, focusing growth very selectively. That is what it was asked to do and it did it well. I really doubt whether it could ever have generated or managed the growth in traffic seen since privatisation.

 

[quantinghome]

Nor any clear evidence against it either. This data will never answer, or nail, that question.

If I were arguing the contrary case (which I am not) I would point to the fact that operating, maintenance and rolling stock costs are lower than at privatisation while the service is faster, more frequent and safer. In contrast to those, privately managed, costs, the main cost categories that have increased are the replacement and enhancement costs that are executed, in the main, by the state run part of the railway.

It would interesting to see the stats on that. Given it was the spiralling costs on the West Coast Main Line route modernisation which did for Railtrack, I have my doubts that the private sector would be any more capable of delivering infrastructure upgrades. They have an inherently high risk of overrunning where there is significant interface with historical assets.

My feeling, having worked for BR, is that it was a well run and pretty efficient organisation, but with capabilities suited to a mission of managing decline and benign neglect of a wide swathe of assets, focusing growth very selectively. That is what it was asked to do and it did it well. I really doubt whether it could ever have generated or managed the growth in traffic seen since privatisation.

Intercity 125
Network Southeast total route modernisation
Electrification
Sprinterisation
Station and line reopenings

Arguably the private operators and network rail have simply continued where BR left off on these, which isn't that surprising given it was basically the same people running it.

As for growth, I think there are significant underlying changes in society which have driven the shift to rail, rather than privatisation. The end of the government's love affair with roads has made a big impact, as have changing working patterns.

 

[Pigeon]

HSTs...

The nadir of passenger numbers was in the mid-70s. The "Thatcher dip" in the 80s notwithstanding, the trend in passenger numbers has been upwards since long before privatisation was even a possibility - since about the time the HSTs were introduced. So if anything is going to get the credit for "turning around the fortunes of Britain's railways", it makes more sense for it to be the HST than privatisation...

 

[Chris Butler]

... the trend in passenger numbers has been upwards since long before privatisation was even a possibility - since about the time the HSTs were introduced. So if anything is going to get the credit for "turning around the fortunes of Britain's railways", it makes more sense for it to be the HST than privatisation...

I really don't think the facts justify the absolute certainty with which you state your hypothesis.



In addition to the raw numbers, the modal share graph below takes out the 'noise' caused by economic activity and the comparison with other European suggests that there were wider forces (probably including oil prices) causing some of the increase, but that the rise in the UK was more prolonged and steep than in other counties.

 

[deltic]

To look at efficiency - need to consider a large number of factors.

As I understand it rail freight has seen considerable improvements in productivity with freight moved per employee, path, locomotive all increased significantly in recent decades

On the passenger side there has also been a radical improvement in fleet utilisation and the number of passengers per employee has increased

Its on the infrastructure side where productivity has not improved as much which seems in part down to tightening of safety standards, the massive increase in train kilometres meaning more wear and tear coupled with less time to access the network and the need to pay compensation to train operators.

As other people have noted rail does not necessarily have low marginal costs. Adding an extra coach on London Overgound services allowed more passengers to be carried but the revenue generated did not cover the additional carriage leasing costs. This is seen in many parts of the network, carrying more passengers sometimes leads to costs increasing more than revenue.

 

[mrmartin]

Also - in other areas of transport we see this. Heathrow for example charges way higher landing and passenger fees than Southend, despite handling 75x the volume. The bigger the airport the more expensive it is, so you could argue than transport has negative marginal cost savings.