HMRC fails to plan for £10.4bn contract exit... because it's 'too risky' HMRC's eye-wateringly expensive £10.4bn IT deal – "the largest technology contract in government" – could be too risky to change as the department has failed to plan for a new contract model, MPs have warned. The Aspire contract underpins the collection of £500bn tax income for the government. It has also generated £1.2bn in …
The figure is over £350 per taxpayer. There's only 29.3 million UK taxpayers (HMRC).
They have multiple systems that are not interconnected.
When I worked there, I identified 5 systems doing the same thing. All the system owners agrees that there should be only one - theirs!
There is no strategic direction from HMRC, and there is no benefit to CapGemini or Fujitsu in pointing it out so the whole shebang remains a great lumbering inefficient giant.
Add to that changes in tax law that each successive govt introduce and the mess just gets bigger.
AC 'cos I might want to get back on the gravy train in the future...
Just adding to my previous comment:
Given HMRC's submissions to PAC indicated that by taking ASPIRE back in-house savings of up to 25% were deemed feasible, it doesn't seem that taxpayers have been ripped off by ASPIRE.
Aside: Full PAC report published today here: http://www.publications.parliament.uk/pa/cm201415/cmselect/cmpubacc/705/70506.htm
By asking it to model a ludicrously labyrinthine and baroque tax structure. The Hong Kong tax code was 35 pages long and considered unavoidable. The UK tax code is at least 14,000 pages long and a tax lawyers paradise. Don't forget that after privatisation 20-odd years ago that there is a hefty payment to the outsourcers for their services.
AC because I have spent more time writing, implementing, supporting and otherwise fettling some of those systems for profit and pleasure than I care to recall.
"Holy crap and people in the US complain about ours and its "only" 2600 pages long and our population and GDP is more than 5 times greater."
He's probably talking about something like Tolley's Yellow Tax Handbook, i.e. the entire UK tax legislation (not just personal tax but corporate, VAT, etc.) plus commentary plus history.
I think the US equivalent is something like the CCH Standard Federal Tax Reporter - which is 70,000 pages.
Nothing to worry about for HMRC, having Richard Bacon on your case is like being assaulted with a damp paper bag. Not for nothing is he known as "Emperor Nero" and other less kind nicknames (actually, it is for nothing, he's famed for doing nothing, except being among the very highest claimants of MPs expenses).
I thoroughly forsee that this report will have no effect whatsoever.
Let's be generous and allow 400 million for hardware expenditure and software licensing (assuming they won't use FLOSS because you don't get the thoroughly outstanding support that can be received from proprietary vendors such as Oracle and MS).
That leaves 10 billion for development. If we assume an average rate of £60 per hour for people involved in the project and working on an 8 hour a day 240 day a year basis, that seems to me to work out at 86806 man years of development time. I could write quite a nice system on that sort of timescale.
This post has been deleted by its author
This post has been deleted by its author
> particularly when you cost for the re-skilling and the usual reality of the FOSS software being less capable than the commercial competition.
We're talking here principally about back-end software.
This used to be a good argument when Hadoop technologies, MySQL and Linux were rare, known only to a few bearded hippies.
For that last few years these technologies have become the de facto mainstream rather than the exception. That argument no longer holds.
This post has been deleted by its author
it's definitely not the case at HMRC
Hence the discussion. Government are doing it wrong, hold the presses.
No kidding though why isn't a cost-overrun the supplier's problem to eat? Go to your shareholders, tell them what you burned and enjoy the replacement CEO.
Or put a better way: don't make bullshit bids for projects. I'm serious - why do governments do such a shitty job in contract negotiation? If they can't guarantee the cost then they know full well they're lying about it and they shouldn't get it anyway. If they can't produce deliverables passing acceptance tests why do they get paid a penny and why are they allowed to bid on other projects?
This is fairly basic stuff and it really pisses me off that government haven't learned stuff everybody else knew in the 1970's.
"No kidding though why isn't a cost-overrun the supplier's problem to eat? "
Who said anything about cost overruns? Aspire is remarkable that it almost always comes in on time and to budget*. The PAC criticisms in the article are reflections on this: that HMRC grossly underestimated their future IT needs at the outset (which was, to be fair, 10 years ago) and haven't seemed to have learned their lesson, despite being about to re-tender the work *and* take over as Prime.
Precisely the kind of clear-headed, good old-fashioned thinking that has gone the way of moral hazard, personal and public accountability, and the concept of public service (as opposed to the harvesting of massive backhanders), Can't really blame the companies for bellying up to the trough, after all someone has to do it.
Have you ever considered a career in public office?
Right, thought not.... but have an upvote on me and let us know when you capture the tender
It actually tells you in the article.
Only certain companies are in a position to tender for such large contracts. All of these companies deal in quantity over quality. The more complex the customer system, the more time they can bill - why would they recommend improvements if it reduces their income?
If the client has 5 systems all doing the same thing and all the owners think it should be done by theirs, then you have inter-departmental rivalries, not to mention huge overlap in responsibilities.
The vendors have a stake in deploying as much kit as possible, because a 4 year support contract plus licenses will cost more than the kit itself (a lot more).
In the end though, what it all boils down to is that there is no ONE person who is in charge and responsible for the delivery (from within HMRC) and all the decisions are made by committee and consist of bean-counters and sales reps without an ounce of real-world experience of deploying large scale enterprise projects.
I will happily give HMRC some free advice: Hire someone from the Internet world who has experience of deploying large scale infrastructure and reconciling existing systems (plus a large dollop of live critical system migration). Put them in charge of ALL IT decisions that could reduce complexity and overlap - regardless of what the department heads say. Last, but not least, pay them a lot of money to do the job properly and the authority to make decisions stick. This way they will save you vastly more than the project is costing us. Please.
"Except for that whole public-sector union thing."
Actually, in the case of the tribal leaders that head each department, it's not a union as such, it's the First Division Association (a title both laughable and arrogant in equal measure). And it's not so much a union as a club for the most useless Oxbridge and public school alumni to scratch each others back. They wouldn't want their most incompetent club members sacked, because, Lord forbid!, that might set a most unpleasant precedent of getting rid of people who were jointly devoid of either fault or use.
Only certain companies can bid because only certain companies can afford to accept the unlimited liabilities, bureaucracy, prevarication and constant change of requirements that goes with bidding for such a contract.
The people from the Internet world are employed and look at what a good job Martha Lane Fox did before she left. The remainder all work for Digital Services and they are doing a great job on the Government portal and various "Catalogue" sites that Govt customer are supposed to buy through these days.
And of course in all the commentary against this article no one has mentioned networks, firewalls, Govt required security lockdowns, the exceptions to those lock downs and all the other stuff that is needed to be in place before you can even start to build the application.
The change in value of Aspire would appear, based on an August 2014 article, a result of the most popular game in the Public Sector, buy A and then introduce change after change, after change, after... ad infinitum.
Finally, I would love to see the Internet world expert explain to the Dept how they can take their solution built on mainframes and written in ancient languages (like Cobol and Fortran) and easily convert them with a few lines of HTML code in to a brand new solution. In reality what they need is someone who stops them changing their mind about the functionality they need, when they need it and why imposing MI5 level encryption rather than the type used by Banks is a bad idea. Those are the people the Public Sector need but they are also the type of people politicians hate because they stop you implementing stupid ideas and policies by telling you how much it is going to cost.
> I always struggle to workout if public sector IT projects are a disaster due to people being on the take, or if it's everyone involved being incompetent.
In my experience the Government asks HMRC to implement a system that does X, Y & Z and we quote a realistic price. That price is invariably way, way above the budget that Government is prepared to give to HMRC and so a long period of wrangling takes place where the scope changes a bit; figures are fudged and eventually the politicians can announce they are getting a system for £X which represents "outstanding value for money and haven't I <insert name of minister looking for good publicity> done a wonderful job". What hasn't changed is that to do the original job, the original amount of money is required. When the system duly arrives at the higher price, there 'shock and horror' from politicians all round.
Why do we dance these dances? Because the politicians control the money and there is no one effectively controlling them.
As my other half was told many times by the grey haired little men in charge of such things when she worked in the public sector as a Projects Monitor (or words to that effect).
"Oh don't worry..it's only tax payers money!"
The folks in charge of these projects really didn't give a toss as the end result, whatever happened, was they would just move onto another new project disaster. No come backs, no questions just another reshuffle.
They really didn't care, the world would carry on and another £10 billion would land in their laps to piss up the wall.
My other half left after a few weeks in disgust.
Years ago I had to do a software evaluation at HMRC. After weighing up the pros and cons of all the products on offer, we came to a conclusion about which one we wanted, but all this had to be written up in a multi-page report and sent off to a committee to be ratified. After 3 weeks of deliberation, we finally got a response. They were unable to approve our recommendation, because the report was submitted in the wrong font.
A govt doesn't need to be efficient to exist, therefore a good service is optional.
A corporation needs to be efficient to exist, therefore a good services is desirable.
It is why politicians like to promote public-private BS. Companies are no better at doing things than the government - ever.
Companies can however be given limited resources to fix *ONE* problem....this is not what we see here.
P.
and the correct response to that is for the committee members to be fired, their supervisory minister to be sacked, and so on. If anyone tries to protect them after such colossal and indefensible stupidity, sack them too.
A couple of hundred examples of that actually happening, and maybe some public sector project costs will start to become vaguely comprehensible .....
The PAC's figures are a little too 'convenient'. Yes, it's true that the costs has increased much above what was originally estimated but so has the scope of the systems. When I worked there (hence the AC) I helped put in a system which abolished a load of paperwork and manual processing. So IT costs went up while staff costs went down (no compulsory redundancies I hasten to add); offices were sold (or no longer let) so premises costs came down, etc. These savings appear on a different bottom line and the PAC conveniently ignores this.
This post has been deleted by its author
... the most profitable thing to do is to ask for a year's extension and a 10% budget increase every year.
Now the government know where they stand, the only question is to decide if they want havoc now for ten billion, or in three years time for thirteen because it is certain the software will not be ready 'real soon now'.
HMRC are not alone in poor IT, many other departments and agencies suffer from the same malaise, it all comes down to an inability to make long term plans and determine future requirements. That is driven by the fact that it's very difficult to keep up with politicians and policy, which change far more in government than they do in the private sector.
Truth is that IT in the private sector is in just as much a shambles as government. "Run it like a business" can't be a solution when business has demonstrably failed at something over and over again. Of course there are some services that by their very nature can't, or shouldn't, be privatized. Only the most rabid zealots of the privatization faith won't acknowledge that. The military or police, for example (I know that privatizing the latter is being tried in Britain, despite all the evidence of how FUBAR private policing and prisons have proved to be in some of the more delusional corners of the U.S.). The key, as always, in both the public and private spheres is accountability. So far none of those who have advocated, implemented and then run away from the consequences of these kinds of bad management practices have been held accountable for their misfeasance. If their jobs, and even more importantly, their pensions, were to suddenly be at risk we might find them a bit less reckless in their disregard of their duty. Failed bureaucrats should suffer the same consequences that failed business executives would in an accountable system. High standards of accountability may not be enforceable in a monarchical corporate culture, but it is certainly something that enough public pressure might bring to pass in government. Or not. Because when the only choice voters have is between worse and worser, it's kind of hard to improve public policy.
Here is how government procurement through open public tenders works (anywhere in the world):
- pre-qualified bidders are invited to submit their bids in sealed envelopes;
- the tender committee officials open the envelopes and carefully count what's inside;
- the highest bidder wins!
Not surprised, having worked on a government contract 10 years ago the levels of incompetence were truly staggering, 2 months of Terminal Services not load balancing properly and on my first day managed to Google the issue and tick a box before 10am.
They were paying the Cisco switch guy £1,200 per week to enter 10 MAC addresses per day as the new machines were rolled out to unblock them.
Isn't this near identical to the botched EDS contract of ten or so years ago.
Hmm, that ended well didn't it.
The fact is the only way to do this properly in my opinion is a brand new system designed from the ground up, tested and tested and tested and then run parallel and then rolled out to all locations in one hit. Of course this is close to impossible, highly impractical and doomed to fail.
>Isn't this near identical to the botched EDS contract of ten or so years ago.
Totally different, the problems with the EDS contract played a major part in defining the form of the ASPIRE contract.
>The fact is the only way to do this properly in my opinion is a brand new system designed from the ground up
A big part of the initial ASPIRE contract was to replace systems that EDS had run and run and run. If memory serves me correctly there were packages running on production systems for which the original developer had ceased trading years previously...
In my experience HMRC is probably the most difficult organisation in which to implement change - process or technical. Its a given that the UK tax rules are incredibly complex but most importantly there is a deadly combination of conflicting technical and political factionalism, timidity, poor programme governance and supplier protectionism coupled with scale that acts like a brake on any undertaking. On an individual basis HMRC has some of the best technical architects, analysts and project managers I have met - in private most confess to be pessimistic and disillusioned about their ability to bring about meaningful change. A transition to another consortium post-Aspire may be a step too far. Even quite simple programmes with obvious and achievable benefts are drowned in the swamp when exposed to HMRC delivery conditions.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
