The way to do this project of course is the A419 model. Get contractors to submit a design, then pay them per vehicle that goes through. Like a toll road, but the government pays the toll.
It’ll get built close to budget, and will rarely close for maintenance (because there’s a big incentive to keep it open).
Alternatively, and this would work for most government contracts, award the contract then pay when it’s finished. No extra payments for “spiralling costs”, just a set price paid upon completion.
Most are supposed to be like that already. But the exception is if the specification is changed. And the civil service always, but always, changes the specification leading to those increased charges.
And bidders know that so they can make low ball bids knowing they will get a second bite at some point.
Of course the civil service is aware of this and tries to make sure bids are realistic to qualify for selection, but they face political pressure to make the project look cheap when it is being planned.
Been there. A few jobs ago, was technical support to a few pubsec tenders, we lost multiple to competitors who were bidding below the hardware cost of the project. For comparison: in a commercial project, hardware would typically be a third to half of the total project cost depending on complexity.
That’s why the “per price” thing works better. Private sector companies avoid changing the specification.
We used to have a drawing in software that looked like a cone. Roughly speaking, correcting things at the scope level (broadly, “what are we doing”) is cheap. At analysis, a bit more expensive. At design more expensive still. And at the point of user testing, really expensive. You get it spec right and then hire people.
Government will just start paying for work on software with a spec that isn’t fully thought out. They’ll go around and around in circles changing requirements. Also, of course, there’s no-one yelling at the staff telling them to stop fucking changing the spec. If there’s a budget, it will get spent. Projects can just be running and then stop. Budget gone. Or change of minister and the new bloke decides it’s not a good idea. Because a lot of things are done on a whim. Like HS2 is. Not actually necessary. Boris just fancied doing it.
Yep. That’s how my company got £10B from NHS for a system that didn’t work. We took the blame in the press. But it wasn’t us, it was NHS.
The core problem is incompetent project managers. People who know by experience the problem with scope creep. Ignorant politicians give the job to their favorite DEI hire. People without enough experience to even realize the need for experienced people. From their view, the job looks simple and easy, so they select leader as a political appointment.
Some years ago there was a discussion on here about a failed NHS IT project and it turned out that the IT consultants had been trying to store X-rays and other images in the database, rather than a link to a file location and costs had gone through the roof.
So its not always NHS managers who are at fault, other than for selecting incompetent IT consultants.
(I forget who it was but it was one of the big companies and selection was probably based on the old nobody ever got fired for buying IBM principle.)
They’re not the sharpest guys around, but if you treat contractors as contractors: “you pay to manage the system” this isn’t an issue. If they burn through money storing images inefficiently, they pay the bill.
I’ve worked on various systems where the civil servants just fannied around. Constantly changing their minds, which attracts billable hours.
Project Management 101: If you are making plans for people, include them in your planning. Central planners designed system without talking to the people expected to use it. Implementation was impossible.
I worked for CSC at the time, but in a different department than the health systems people. In a strange coincidence, my little brother worked in the Blythewood, SC, CSC health systems offices, in space leased by SC-DMV.
No no no. The beauty of the A419 model is that it’s really simple and maintenance is built into the incentives.
If the road gets really shitty, people will take the old road, or go via Fairford or Purton, so they have to keep it in good order to get paid. And they optimise efficiently for road works to not close it.
And it gives you government at the bare minimum, doing the two bits of it that is sensible for government to do, which is deciding a) what money should be spent where and b) whether land should be used for a certain thing. You can’t have morons like Grant Shapps and Heidi Alexander actually running the projects. They know fuck all about roads and trains.
You don’t have to know about how Pierre Jaurant makes the wine, or how Aldi get it into a shop. You just have to know that you like the Bordeaux at the price (£5.49, bit of a bargain in my opinion). Aldi and Pierre Jaurant are the experts. And you pay per time you use it. That’s all government should do with almost everything. Schools, hospitals, roads. Sometimes you can’t do it on a “per use” like a school, but just pay per year for a thing to be provided. No employing bloke to fix the carpets, pay for carpet being provided in the school and they figure it out.
A lot of the horror stories about the cost of PFI don’t explain it properly. When someone pays for a new clock and gets charged £200, that pays not only for the clock, and the bloke coming to fit it. It pays for that being maintained for the remainder of the contract. Clock falls off the wall and breaks, the company goes out and fits a new one. And of course, they have to pay all the costs of the maintenance man. His salary, pension, cost of the van to go there.
It’s how all these high speed choo-choos should have been done. OK, there’s economic benefits of a train, maybe (job creation etc). So, offer a per ticket bounty to a company. Build HS2, and we give you £5 a ticket. If no-one will take the deal at the price, clearly, it’s not worth it. We’d have to spend £15/ticket and we’ve calculated that isn’t worth it.
Precisely WB. It’s all about how the contract is structured. Here’s the requirements, here’s how long you have the concession to run it, make what money you can then hand it back in good order.
HS2 was run through the standard government cost-benefit analysis for new transport projects. Even at the hugely low-ball initial estimate, it failed catastrophically (well, it barely washed its face and there were a score of potential projects that would have delivered better outcomes). Despite this it still went ahead. Odd, isn’t it?
It simply doesn’t make any sense because who cares about going a bit quicker Brum to London?
It’s still too long for a commute, so people use those trains for the occasional client meeting or to go see the grandchildren. And if you’re going to a client you don’t care if it’s 82 or 49 minutes. It’s a day out of the office seeing the client. Have a few meetings, lunch, then a bit of handshaking and chatting, then home. At 82 minutes you finish at about 4 and get to the station. If it’s 49 minutes, maybe 4:30.
You work on the train or get home and do a bit of work.
The value of speed is not linear.
If it takes a month that’s awful to the point lots of people refuse to do it.
Take it down to a week (Bristol to London by carriage), people will do it, but it’s a huge pain to do. Lots of coaching inns to pay for
3+ hours. You just need one night away. Much cheaper
Under 3 hours. You can do it in a day, no need for a hotel. But you wouldn’t want to do it daily.
40 minutes. You can do it every day.
But between those times, there isn’t a lot of value added. 3 hrs vs 4 is nice, but who cares that much. The big value leap is taking it below 3, at which point, you can cancel the hotel and be home in your bed. That’s £150+ saved on the journey. 3 hours to 2, again, nicer, but you aren’t going to change use of it. Once you hit 40 minutes, people will start living there to commute daily.
Most of the value of UK train improvement was pre-WW2.
And this fucking obsession with high speed trains ignores the effect that high speed has on going Paris to Bordeaux. It competed with flights. It made going to see your mum in Bordeaux, or nipping down to the summer pad in Arcachon for a long weekend practical. That’s a lot of value. The only places that far away are Newcastle, Edinburgh and Glasgow. And who wants to bask in the weather of those?
Also, there’s essentially nothing between Paris and Bordeaux. Lumo do nonstops to Newcastle, but that’s because they’re not allowed to stop at Peterborough, Doncaster, York …
“A lot of the horror stories about the cost of PFI don’t explain it properly. When someone pays for a new clock and gets charged £200, that pays not only for the clock, and the bloke coming to fit it. It pays for that being maintained for the remainder of the contract. Clock falls off the wall and breaks, the company goes out and fits a new one. And of course, they have to pay all the costs of the maintenance man. His salary, pension, cost of the van to go there.”
But surely the £200 should only be paid once, at the beginning of the PFI contract? It seems to me that the PFI companies get paid over and over again – once at the beginning then multiple times afterwards when fixing their own crap construction.
You’re reminding me of good old Project Orion. Instead of building complex space capsules, just build a low-tech cruise ship. Of good old steel. Propelled by atomic bombs!!
One could thus go on a cruise to Mars instead of Fiji. But alas the Greens managed to ban the explosion of atomic bombs in the atmosphere. They wouldn’t have used too much plutonium because that’s expensive. Just lots of high explosive to compress the lithium deuteride booster.
Of course you may feel that even a little bit of plutonium in the atmosphere is too much. But I don’t think it’s any more dangerous than CO2.
The Orion project reminds me of a gap in my scientific conceptionalisation. That is – how does a nuclear explosion push a spaceship?
A chemical rocket pushes a rocket because it throws out backwards burning gases and reaction pushes the rocket in the other direction.
A nuclear explosion in the atmosphere heats the surrounding air and expands it, causing pressure, which pushes the rocket. So, Orion would work in the atmosphere (which is probably why the Greens banned it – it works!)
How does a nuclear explosion in vacuuum push the rocket? A nuclear explosion is a massive emission of heat – what does that heat heat to cause it to expand and push? Is it just the pressure of the heat itself? How does that work? Is it the kenetic pressure of bits of smashed atoms racing away? “Classic Era” scifi sometimes touched on it as being a heat source, and using water to heat to expel under pressure to push the rocket. Mention of “refuelling stops” to take on water.
So, how does disintergrating atoms on their own push stuff without there being a reaction mass to transfer energy to?
One has a pusher which ablates and the resultant plasma pushes against the rocket. The large amount of explosive provides extra mass to help things along.
In George Dyson’s book, he says that the thing which most bothered the designers was if the atomic bomb failed to explode and the rocket was riddled by shrapnel, not pushed by plasma.
In effect you need to keep the bombs sufficiently small, the rocket sufficiently far away and the shock absorbers sufficiently effective to take the impact. That’s the advantage of a small amount of plutonium fissioned by the neutrons from the lithium deuteride.
Niven and Pournelle’s Footfalldescribes an ‘Orion’ craft quite well. Freeman Dyson was one of the lead designers for ‘General Atomics’ of the original Orion project, there’s film of a test model (using conventional explosives). The aim was to deliver 100t payload to Mars in 125 days.
Like a toll road, but the government pays the toll.
See also the A1(M) between Peterborough and the A14. It’s a good model but it’s not always the most appropriate. If the M6(Toll) had been like this, everyone and its mother would use it and the road would be much less good. By tolling motorists directly on one of two route options there are, well, options.
I’ve walked along a former toll road.
The costs of collecting the toll are greater than any net benefit of collecting the toll – except in a very few cases like the M6
I said “net benefit” which is after the direct cost of collection (not mentioned in the newspaper article) and the cost to the motorists of sitting in a queue every day when they could be working or getting a bit more sleep before leaving for work or …
Not to mention the cost of maintaining and repairing the road (the newspaper carefully chooses not to mention that!).
The M6 Toll is paid by drivers who value the ease and time-saving as being greater than the cost.
My local ex-toll road has a little plaque about it and they stopped collecting tolls because it wasn’t worth it (the county council took over responsibility for maintaining the road – I don’t know how they talked the CC into that).
Let me explain this. The company running the A419 gets paid per vehicle. They have sensors recording this. They invoice the government for the tolls. Government pay it. There’s no toll booth, no delays. That’s easy and cheap to do. Government pay this out of taxes.
That’s what I mean by all roads could be tolled. All roads could be run like that. Not the whole charging individual motorists thing. Although with these little devices you can get for autoroutes, it works quite well.
How would that work? How would the postman, the milkman, the bin lorry get to my house, and all the other houses in town? How would *occupiers* get to their houses, how would visitors get to other people’s houses?
I don’t think it would apply to town roads, but to highways. It’s not worth the hassle.
But as far as you are concerned as a motorist, exactly as now. You going up a road gets recorded. The owner of the road sends the government a monthly bill, counting all cars and government repay it.
The way to do this project of course is the A419 model. Get contractors to submit a design, then pay them per vehicle that goes through. Like a toll road, but the government pays the toll.
It’ll get built close to budget, and will rarely close for maintenance (because there’s a big incentive to keep it open).
Alternatively, and this would work for most government contracts, award the contract then pay when it’s finished. No extra payments for “spiralling costs”, just a set price paid upon completion.
Most are supposed to be like that already. But the exception is if the specification is changed. And the civil service always, but always, changes the specification leading to those increased charges.
And bidders know that so they can make low ball bids knowing they will get a second bite at some point.
Of course the civil service is aware of this and tries to make sure bids are realistic to qualify for selection, but they face political pressure to make the project look cheap when it is being planned.
It’s all part of the great game.
Been there. A few jobs ago, was technical support to a few pubsec tenders, we lost multiple to competitors who were bidding below the hardware cost of the project. For comparison: in a commercial project, hardware would typically be a third to half of the total project cost depending on complexity.
That’s why the “per price” thing works better. Private sector companies avoid changing the specification.
We used to have a drawing in software that looked like a cone. Roughly speaking, correcting things at the scope level (broadly, “what are we doing”) is cheap. At analysis, a bit more expensive. At design more expensive still. And at the point of user testing, really expensive. You get it spec right and then hire people.
Government will just start paying for work on software with a spec that isn’t fully thought out. They’ll go around and around in circles changing requirements. Also, of course, there’s no-one yelling at the staff telling them to stop fucking changing the spec. If there’s a budget, it will get spent. Projects can just be running and then stop. Budget gone. Or change of minister and the new bloke decides it’s not a good idea. Because a lot of things are done on a whim. Like HS2 is. Not actually necessary. Boris just fancied doing it.
Yep. That’s how my company got £10B from NHS for a system that didn’t work. We took the blame in the press. But it wasn’t us, it was NHS.
The core problem is incompetent project managers. People who know by experience the problem with scope creep. Ignorant politicians give the job to their favorite DEI hire. People without enough experience to even realize the need for experienced people. From their view, the job looks simple and easy, so they select leader as a political appointment.
Some years ago there was a discussion on here about a failed NHS IT project and it turned out that the IT consultants had been trying to store X-rays and other images in the database, rather than a link to a file location and costs had gone through the roof.
So its not always NHS managers who are at fault, other than for selecting incompetent IT consultants.
(I forget who it was but it was one of the big companies and selection was probably based on the old nobody ever got fired for buying IBM principle.)
They’re not the sharpest guys around, but if you treat contractors as contractors: “you pay to manage the system” this isn’t an issue. If they burn through money storing images inefficiently, they pay the bill.
I’ve worked on various systems where the civil servants just fannied around. Constantly changing their minds, which attracts billable hours.
Nah.
Here’s an excellent post mortem:
https://www.henricodolfing.ch/case-study-1-the-10-billion-it-disaster-at-the-nhs/
Lots of errors. I think most important is
Project Management 101: If you are making plans for people, include them in your planning. Central planners designed system without talking to the people expected to use it. Implementation was impossible.
I worked for CSC at the time, but in a different department than the health systems people. In a strange coincidence, my little brother worked in the Blythewood, SC, CSC health systems offices, in space leased by SC-DMV.
No no no. The beauty of the A419 model is that it’s really simple and maintenance is built into the incentives.
If the road gets really shitty, people will take the old road, or go via Fairford or Purton, so they have to keep it in good order to get paid. And they optimise efficiently for road works to not close it.
And it gives you government at the bare minimum, doing the two bits of it that is sensible for government to do, which is deciding a) what money should be spent where and b) whether land should be used for a certain thing. You can’t have morons like Grant Shapps and Heidi Alexander actually running the projects. They know fuck all about roads and trains.
You don’t have to know about how Pierre Jaurant makes the wine, or how Aldi get it into a shop. You just have to know that you like the Bordeaux at the price (£5.49, bit of a bargain in my opinion). Aldi and Pierre Jaurant are the experts. And you pay per time you use it. That’s all government should do with almost everything. Schools, hospitals, roads. Sometimes you can’t do it on a “per use” like a school, but just pay per year for a thing to be provided. No employing bloke to fix the carpets, pay for carpet being provided in the school and they figure it out.
A lot of the horror stories about the cost of PFI don’t explain it properly. When someone pays for a new clock and gets charged £200, that pays not only for the clock, and the bloke coming to fit it. It pays for that being maintained for the remainder of the contract. Clock falls off the wall and breaks, the company goes out and fits a new one. And of course, they have to pay all the costs of the maintenance man. His salary, pension, cost of the van to go there.
It’s how all these high speed choo-choos should have been done. OK, there’s economic benefits of a train, maybe (job creation etc). So, offer a per ticket bounty to a company. Build HS2, and we give you £5 a ticket. If no-one will take the deal at the price, clearly, it’s not worth it. We’d have to spend £15/ticket and we’ve calculated that isn’t worth it.
Precisely WB. It’s all about how the contract is structured. Here’s the requirements, here’s how long you have the concession to run it, make what money you can then hand it back in good order.
HS2 was run through the standard government cost-benefit analysis for new transport projects. Even at the hugely low-ball initial estimate, it failed catastrophically (well, it barely washed its face and there were a score of potential projects that would have delivered better outcomes). Despite this it still went ahead. Odd, isn’t it?
It simply doesn’t make any sense because who cares about going a bit quicker Brum to London?
It’s still too long for a commute, so people use those trains for the occasional client meeting or to go see the grandchildren. And if you’re going to a client you don’t care if it’s 82 or 49 minutes. It’s a day out of the office seeing the client. Have a few meetings, lunch, then a bit of handshaking and chatting, then home. At 82 minutes you finish at about 4 and get to the station. If it’s 49 minutes, maybe 4:30.
You work on the train or get home and do a bit of work.
The value of speed is not linear.
If it takes a month that’s awful to the point lots of people refuse to do it.
Take it down to a week (Bristol to London by carriage), people will do it, but it’s a huge pain to do. Lots of coaching inns to pay for
3+ hours. You just need one night away. Much cheaper
Under 3 hours. You can do it in a day, no need for a hotel. But you wouldn’t want to do it daily.
40 minutes. You can do it every day.
But between those times, there isn’t a lot of value added. 3 hrs vs 4 is nice, but who cares that much. The big value leap is taking it below 3, at which point, you can cancel the hotel and be home in your bed. That’s £150+ saved on the journey. 3 hours to 2, again, nicer, but you aren’t going to change use of it. Once you hit 40 minutes, people will start living there to commute daily.
Most of the value of UK train improvement was pre-WW2.
And this fucking obsession with high speed trains ignores the effect that high speed has on going Paris to Bordeaux. It competed with flights. It made going to see your mum in Bordeaux, or nipping down to the summer pad in Arcachon for a long weekend practical. That’s a lot of value. The only places that far away are Newcastle, Edinburgh and Glasgow. And who wants to bask in the weather of those?
Also, there’s essentially nothing between Paris and Bordeaux. Lumo do nonstops to Newcastle, but that’s because they’re not allowed to stop at Peterborough, Doncaster, York …
“A lot of the horror stories about the cost of PFI don’t explain it properly. When someone pays for a new clock and gets charged £200, that pays not only for the clock, and the bloke coming to fit it. It pays for that being maintained for the remainder of the contract. Clock falls off the wall and breaks, the company goes out and fits a new one. And of course, they have to pay all the costs of the maintenance man. His salary, pension, cost of the van to go there.”
But surely the £200 should only be paid once, at the beginning of the PFI contract? It seems to me that the PFI companies get paid over and over again – once at the beginning then multiple times afterwards when fixing their own crap construction.
The extra price is for a change. You don’t have a clock in the contract, and you want a clock. It’s now part of the contract.
Lefties do a lot to avoid explaining things candidly. They don’t give a fuck about value. They want jobs brought back in house.
Or just shoot one in 10 of the management of the contracting company if they come in over budget, aka the Chinese Solution.
A trifle off topic.
You’re reminding me of good old Project Orion. Instead of building complex space capsules, just build a low-tech cruise ship. Of good old steel. Propelled by atomic bombs!!
One could thus go on a cruise to Mars instead of Fiji. But alas the Greens managed to ban the explosion of atomic bombs in the atmosphere. They wouldn’t have used too much plutonium because that’s expensive. Just lots of high explosive to compress the lithium deuteride booster.
Of course you may feel that even a little bit of plutonium in the atmosphere is too much. But I don’t think it’s any more dangerous than CO2.
The Orion project reminds me of a gap in my scientific conceptionalisation. That is – how does a nuclear explosion push a spaceship?
A chemical rocket pushes a rocket because it throws out backwards burning gases and reaction pushes the rocket in the other direction.
A nuclear explosion in the atmosphere heats the surrounding air and expands it, causing pressure, which pushes the rocket. So, Orion would work in the atmosphere (which is probably why the Greens banned it – it works!)
How does a nuclear explosion in vacuuum push the rocket? A nuclear explosion is a massive emission of heat – what does that heat heat to cause it to expand and push? Is it just the pressure of the heat itself? How does that work? Is it the kenetic pressure of bits of smashed atoms racing away? “Classic Era” scifi sometimes touched on it as being a heat source, and using water to heat to expel under pressure to push the rocket. Mention of “refuelling stops” to take on water.
So, how does disintergrating atoms on their own push stuff without there being a reaction mass to transfer energy to?
One has a pusher which ablates and the resultant plasma pushes against the rocket. The large amount of explosive provides extra mass to help things along.
In George Dyson’s book, he says that the thing which most bothered the designers was if the atomic bomb failed to explode and the rocket was riddled by shrapnel, not pushed by plasma.
In effect you need to keep the bombs sufficiently small, the rocket sufficiently far away and the shock absorbers sufficiently effective to take the impact. That’s the advantage of a small amount of plutonium fissioned by the neutrons from the lithium deuteride.
Niven and Pournelle’s Footfall describes an ‘Orion’ craft quite well. Freeman Dyson was one of the lead designers for ‘General Atomics’ of the original Orion project, there’s film of a test model (using conventional explosives). The aim was to deliver 100t payload to Mars in 125 days.
See also the A1(M) between Peterborough and the A14. It’s a good model but it’s not always the most appropriate. If the M6(Toll) had been like this, everyone and its mother would use it and the road would be much less good. By tolling motorists directly on one of two route options there are, well, options.
That’s different. The M6 toll road is more of a luxury. The old A419/A417 was slow road.
But really, all roads could be tolled, and include various pricing based on demand.
I’ve walked along a former toll road.
The costs of collecting the toll are greater than any net benefit of collecting the toll – except in a very few cases like the M6
You can’t say something like that on the blog of a man from Bath:
https://web.archive.org/web/20220201221312/https://www.somersetlive.co.uk/news/local-news/gobsmacking-sum-money-baths-toll-6578291
I said “net benefit” which is after the direct cost of collection (not mentioned in the newspaper article) and the cost to the motorists of sitting in a queue every day when they could be working or getting a bit more sleep before leaving for work or …
Not to mention the cost of maintaining and repairing the road (the newspaper carefully chooses not to mention that!).
The M6 Toll is paid by drivers who value the ease and time-saving as being greater than the cost.
My local ex-toll road has a little plaque about it and they stopped collecting tolls because it wasn’t worth it (the county council took over responsibility for maintaining the road – I don’t know how they talked the CC into that).
Let me explain this. The company running the A419 gets paid per vehicle. They have sensors recording this. They invoice the government for the tolls. Government pay it. There’s no toll booth, no delays. That’s easy and cheap to do. Government pay this out of taxes.
That’s what I mean by all roads could be tolled. All roads could be run like that. Not the whole charging individual motorists thing. Although with these little devices you can get for autoroutes, it works quite well.
That is not a toll road – the driver does not pay a toll to use it.
.
How would that work? How would the postman, the milkman, the bin lorry get to my house, and all the other houses in town? How would *occupiers* get to their houses, how would visitors get to other people’s houses?
I think that it’s called a “season ticket”. Visitors walk from the municipally-owned car park or the bus stop.
I don’t think it would apply to town roads, but to highways. It’s not worth the hassle.
But as far as you are concerned as a motorist, exactly as now. You going up a road gets recorded. The owner of the road sends the government a monthly bill, counting all cars and government repay it.
There’s a specific term for that: shadow tolling
Every day is a school day. There’s quite a few listed on Wikipedia.
‘Angin’s too good for ’em.
Apparently, the planning documents run to 660000 pages. An oustanding example of an endemic problem.
£3.1bn spent and no work done. A tribute to Britain.
It’s amazing how expensive not building things has become. Soon we won’t be able to afford to not build anything.
A good 25% of that is legal fees. The harder things are made (by the legal profession) the more the lawyers profit.
Gawd, y’all are backwards. We quit using artists for this stuff decades ago.
Game, here’s another impression of the Lower Thames Crossing and HS2.
That would be good. If they collide, all of Essex, part of Kent and half of London would disappear.