First report – and first test – of the PBA

8 09 2010

We shall soon see who is more powerful in this country, the elected government or the civil service.

The Network for the Post-Bureaucratic Age today publishes its first detailed report on one way we can get better-for-less. This has been put together by some of the UK’s best thinkers on the subject, led by Liam Maxwell, IT specialist and Councillor at the Royal Borough of Windsor and Maidenhead. It presents examples of where their approach has succeeded and a clear plan – a playbook – for implementation. But will government actually be able to put this into action, or will it be blocked?

The report – ‘Better for Less: How to make Government IT deliver savings’  (iBook here)- investigates the quagmire of government IT.

The British government currently spends somewhere between £16 billion and £23 billion on IT every year. The astonishing lack of clarity over expenditure is symptomatic of appalling failures in IT strategy, procurement, and process. This cannot be allowed to continue, especially during a time of spending cuts in frontline services. The annual cost dwarfs some government departments. It is three times the amount we spend on the army, more than the Department for Transport. Worse, it has been designed badly and, unfortunately this time, the process has been built to last. The problems come from ineffective procurement – much of which is waste.

Each year about the same amount of money is spent on the procurement process (the jumping through hoops to secure contracts) as is used to run the Foreign Office. Savings just in the procurement process – without even counting the savings from better IT –  could finance the entire Sure Start programme, they could fund 50% more school building. And even when the form-filling is done only 30% of projects work. Indeed government productivity has actually declined since IT was introduced. At a time when dynamic change is required –  to reduce cost and deliver better services – one of the principle barriers to that change has become government IT.

Liam and his co-authors are dedicated to bringing government into the information age, and have looked in detail about what should be done to deliver government IT more effectively, and at a much lower cost to taxpayers. The paper spells out exactly how government can deliver a better service for less money – a very different proposition to proposing mere ‘cuts’, where less money means poorer service.

The full report is available at, directly here and as an ibook here.

We would really appreciate any comments from anybody who reads the paper: one of the central tenets of the post-bureaucratic age is that knowledge and skills exist within informal networks, not just companies and departments. We recognise this and encourage anyone to comment below, regardless of their political affiliations.

Stephan Shakespeare

Note: Stephan Shakespeare chairs an informal network of people who are interested in the development of policy towards a ‘post-bureaucratic age’, and has written about what this means  here and here



22 responses

8 09 2010

Before delving deeply (which will undoubtedly generate comments – I have one from page 2 already…), will you be putting up a commentable copy somewhere? Feeding back ‘by hand’ from a 69 page flat pdf is v.painful and an effective disincentive ;-(

8 09 2010

Thanks for your comment. What would you recommend as the best place to put up a commentable document. I looked for the one that The Conservative Party used to crowdsource reactions to the budget but could not remember what it was.

Do you have any suggestions?


Ali ( @aliunwin )

9 09 2010
Darren Woods

Hi Ali,

You could pop it into a standard Google Docs account and this would allow real-time collaboration between potentially thousands 😉

8 09 2010

You might want to have a look at Read + Comment which is an excellent commentable platform. And no, it’s nothing to do with me!

8 09 2010

I do hope you’re joking with that last comment. It undermines your credibility just a tad !

Anyway cost saving is a noble aim. I suspect it will be difficult though (and I speak as someone with experience of many of the ideas in your document).

You are onto something with the cost of desktop systems. These, unlike transactional systems are rather simple and similar to each other. I would throw hosting (secure and non-secure) into the mix – each govt. system seems to do it differently at great expense.

I don’t see a lot of rigour in the production of your desktop comparision figures. Security requirements in central government are costly (if we are in any way serious about data protection you should consider whether you need them locally too).
The fair comparison would be between a well chosen private company (perhaps one who works in the defence and security realm) and a similarly sized central government department – not one between based on your relatively few PCs in the office.

When it comes to building systems to do things, the complexities begin…

You seem to make the observation that systems are designed to fit legislation which makes them expensive and difficult. Personal tax/tax credits would be a good and topical example. However to wish for it to be the other way round (simplifying legislation thereby simplifying systems) is simply a wish for simpler and clearer legislating, which has basically nothing to do with IT and is a rather vague if worthwhile ambition.

Quite early in your document you hit the key issue (to my mind)..
>> Thirteen years of profligate spending has not created a super
tanker that can be gradually turned around – rather, the inheritance could be described as forty or fifty tankers, steaming off in different directions.

Now I suppose this is true – in that there are quite a lot of fairly ‘heavyweight’ projects out there in the public sector. They are all quite different (necessarily so in many cases). [Incidentally in the private sector this happens all the time but it has nothing to do with spending, more a lack of technical and organisational direction. That seems to be baked into our policy making and political process, I’m afraid ! Anyway, I digress…]

Building ‘super tankers’ as you put it have the advantage that they go in one direction at once. For a big and complex department, this is fantastically difficult and expensive to do of course, as shown rather well by the NHS spine and associated systems.

Building a flotilla of tug boats, local systems is better in other ways – each one is simple. With a massive fleet of tug boats each is cheap and quick to build, but can they talk to each other ? Is each one specified slightly differently ? Are they just as expensive to manage, overall ? Do those on board each boat each have their own slightly contradictory view of the world ? Maybe. We’re back to the reason why the Spine project was imagined.

The question in the NHS or any other department is… why do small projects go in different directions ? The answer is that people do. And people – quite often civil servants – dictate what gets built. In many cases they don’t understand what’s required very well, and so they get what they’re given by their chosen supplier. In many cases they fight among one another and have contradictory views of what should happen. I have seen this personally time and time and time again.

As you point out some common data formats will eventually come along, and maybe they will provide the answer to an extent. A map for the tug boats.
Or perhaps somebody will provide a product that will manage common tasks. A tug boat factory. There have been attempts to do this in the past in all sorts of ways (medical records software, to name but one). Sooner or later, somebody will put simple and easy software to handle, say, parking tickets, or planning applications, or council tax payments, online – and any council or department will be able to sign up. If it happens, it’s quite likely that these systems will be invented by somebody _outside_ government. Similar things have already happened in the area of building and managing content only websites (Drupal, Joomla, WordPress) with the result that they are now quite a bit cheaper to build than when a 6 figure sum was needed for a CMS (and a 7 figure sum to get it to work).

Is a common data format really enough on its own ?
For those services that are delivered locally I don’t really agree that local suppliers will stick to a common format purely by the existance of central ‘standards’. It’s hard enough getting one small department to agree on something never mind all of them (damn people, again).
They’ll end up diverging away from each other (that’s what ‘innovation’ means, going back to my point about people diverging) _unless_ there are systems in place to manage those standards.

Unless they are enforced, small differences will make it impossible to (say) transfer details from place to place. That’s no good if you’re moving between tax offices. Would it even make sense to have a different personal record (standardised or not) held separately in each local authority you lived in ? It wouldn’t really. The data to needs be _stored_ centrally (thus _enforcing_ its format). We’re heading back to a super tanker.

As you say, perhaps this information could be opened up to some extent, but don’t expect people to willingly allow this. If ID cards are unpopular just how unpopular would ‘open’ access to their data be ?

Maybe some non-personal government data would perhaps be useful to application developers (transport timetables could be one…) but most interesting stuff government does has to do with people. Many expensive systems that government builds have to do with transactions related to people (which are often fairly complicated).

People don’t like giving their details to any old company. I don’t necessarily want my medical records held by Google in their ‘cloud’, thanks. Or my parking tickets come to that. Or my tax records.

The last thing I’d say, is on principal 3… I think you are close to the mark here. Bear in mind that it requires greater expertise on the government side which is rarely available (so will have to be recruited or drafted in). I guess this is your intelligent customer role. You’re right to highlight it as without this, the whole thing falls over.

I don’t wish to be overly negative, but there is a need to be realistic, too. When it comes to IT, small steps delivered often is generally the best approach. There does need to be an overall vision of where you’re going though, and that’s the fundamental problem always.

8 09 2010
Steve Horgan

For the record, I am an IT professional working in a fairly senior position in a FTSE 100 company.

The paper is very good, both in diagnosis of problems and proposed solutions. However, I don’t think it is useful to dwell on the good things in it, rather on how it can be improved. I think that the identification of the need for government to be an intelligent customer and the need for improvements to the technical skills of government staff are very important, but this is not really pressed enough. The absolute key problem with government IT is government IT capability, represented by its own staff. Some background here: large-scale IT has changed out of all recognition over the last 20 years. For any large organisation what were once relatively isolated large systems with relatively few components and relatively undemanding service levels have turned into a complicated landscape with thousands of systems interconnected with hundreds of thousands of components, with a proportion operating in real time to deliver services directly to customers. Complexity has increased enormously and so the a world where IT could be left to talented amateurs has been replaced with one that requires professionalism and engineering precision. Most large companies have recognised this and have invested heavily in IT capability not only in terms of direct delivery but also in terms of IT architecture and design. Services companies have also invested similarly in order to sell this capability, particularly to organisations that have not made the required investment. Unfortunately, one of these appears to be the British government.

Without a competent architecture and design function it is almost impossible to avoid being fleeced by IT suppliers and services companies, because buyers simply lack the technical skills to negotiate a decent deal. From my observation, government IT projects delivered by outside suppliers cost roughly ten times more than they should, largely for this reason. While I am not positing particularly bad behaviour on the part of IT businesses that sell to government, salesmen are only human and an ill-informed customer that lacks the skills to dissect a plausible PowerPoint presentation that glosses over key technical difficulties is always going to get a bad deal. Worse still, without technical skills at a high level most organisations will lack the capability even to understand their own IT landscape, much less organise it in a beneficial way. Now, as I write that I am aware that those outside of professional IT, or who don’t understand the sheer bloody complexity of modern systems at a large scale will think that I am exaggerating. I am not. There is a selection of quite expensive IT tools whose primary function is just to make a complex IT landscape understandable. However, they are useless without highly skilled staff to operate them.

The technical proposals in the document are broadly sensible, but without a serious government IT architecture and design function that employs properly-qualified people then there will be little capability to deliver them in any coherent way. Please note, I am not talking about some ivory-tower IT Strategy function, but quite large numbers of senior IT professionals to manage overall architecture and with a controlling interest in the technical design and project processes, operating within an overall IT target architecture. For comparison, my company employs a group of more than 400 in this role, amid many thousands of other IT staff, who also design strategic solutions at the highest level in addition to their other responsibilities.

If government cannot match the skills of systems integrators then they will continue to get poorly-designed and overpriced solutions from them.

If government cannot match the skills of the commercial sector then the disparity in IT delivery between the private sector and government will only continue to grow.

The solution is not open source versus bespoke development, it is to employ people who actually have the capability to strategise, architect, design and then deliver the solution.

9 09 2010
Darren Woods

I think I slightly agree, Steven…

The architecture and low-level service logic that make up large scale systems will always remain complex. So, just as we did with OO code, maybe we need to abstract the logic that’s important to end users, empowering them and making them agile and self self-sufficient. Business people know their business. All they need are intuitive graphical tools to manage their processes, and not care about the fine grained logic of the integration or transaction layers (Managed by the experts in those fields).

If the architecture is abstracted and simplified to incorporate a BPM layer, based around open standards, we can lift the relevant business logic out of the complex world and get the tools into the Business people’s hands so they can manage their own processes locally, enabled and supported by an overarching best practice.

The BPM system used by each organisation could be different, be-spoke or open source, as long as they support the standards.

Working in a FTSE 100, I suspect you may have a conflict of interests when it comes to the open source and ‘freemium’ business models? 😉

9 09 2010
Steve Horgan

I entirely agree with this. While BPM is not a panacea, nothing ever is, the logic of abstracting business processes to the point that they can be delivered by configured, as opposed to constructed, systems is absolutely sound. Many companies, including my own, are doing this for core business processes, and my company operates on a scale commensurate with that of UK government functions.

However, I will say that getting this right is not easy and the solution must be properly governed, architected and designed, especially in order to interoperate with those systems that are not amenable to the BPM approach, which are primarily database and transaction oriented.

There is not conflict of interest with large companies using BPM. Open source has some difficulties, because set against demanding service levels it requires the employment of specialist third-parties and very considerable in-house skills. Large businesses cannot wait until a back-room community of programmers with whom they have no contractual relationship get around to fixing a fault, and so support has to come from one of the companies that supply this on a commercial basis, supplemented by teams of their own staff. That having been said, there are opportunities, and I was very proud to put in the first large LINUX platform in our company last year. It has proved a tremendous success. It fact the only problem is that it has proved so much of a success that we need to upgrade its capacity much earlier than we thought due to business demand.

9 09 2010

>> The solution is not open source versus bespoke development, it is to employ people who actually have the capability to strategise, architect, design and then deliver the solution.

This is quite correct. And not cheap.

9 09 2010
Steve Horgan

Dead right. I recently saw a job advert for the Lead IT Architect for HMT. This role was for the top IT Professional at one of the great Departments of State, a department where excellence in IT provision is a key requirement for its every activity. However, the salary point was £50k. This is a level of remuneration where any remotely qualified candidate in the private sector would have to take a pay cut. This is a level of remuneration where any remotely qualified candidate in the public sector would long have left for the private sector and double the salary. God knows who they ended up with. Senior IT professionals are in great demand, and have proved largely recession-proof as a result. They enjoy very good salaries and if government wants them then they will have to cough up.

Pay peanuts, get monkeys.

9 09 2010

That is truly staggering. I’d assume you’d need to live in London, too.

I’d fancy a tilt at a job like that – just about ready for something of that size but my god for 50K I wouldn’t go near it. I was earning 35 to 40K a decade ago as a mid-level, mid-twenties programmer in the private sector. I believe Boris gets about the same for trotting out 10 sunday paper columns. What a state this country is in.
I suspect a top IT bod in a comparably sized private company would earn nearly double.

I wonder if the problem is that the lead architect can get away without really doing anything since it’s all outsourced ?! (that’s not to say someone excellent couldn’t keep an eye on the suppliers and save a few million here and there).

9 09 2010
Darren Woods

A very good paper! A long needed critique of the status-quo!

There are a couple of points that I think are missed and really should be discussed in the paper… The first is BPM. The new process paradigm in software design, development, and ongoing change management, really does impact massively on the decisions being taken right now regarding the future of ICT and the UK government. If the Government were to adopt a BPM strategy from the ground-up (As opposed to the traditional top-down strategy management), we could see a revolution in shared services and best practices. This methodology would support many of the innovations presented in this paper, such as breaking large programmes into manageable scopes at the local level, driving innovation, and increasing the interoperability of operations across the public sector. The BPM approach also enforces standards such as BPEL, BPMN, SOA, etc… And if you get the people on the ground focused on how they work, they buy-in to, and evolve the processes as a team. With a graphical process and screen designer included in any good BPMS, the end user can even change their process without the need to speak to analysts and developers working for expensive consultants. This empowerment is what BPM is all about – supporting the only constant: Change!

The second subject is licensing. I know the paper mentions Open Source in many places, but maybe there could be an entire section dedicated to the subject of license fees, SaaS, Cloud Apps, and alternative licensing models? We don’t charge for our software, or for licenses to use it, we simply bill the customer for the work that our people do (As a business model, our margins are low, but our resilience is high!).

We have configured process in our BPMS for many UK police forces to support areas such as Briefing and Tasking of Officers, Human Resources and even Stop & Search (Yes, it’s still happening on paper!). These e-forms and processes may be shared, along with our process engine, with other organisations and the processes remain the IPR of the organisation that designs them (the open source code is abstracted from the IPR of the customer via the standards based BPEL specification of the process they design!).

The open source semantic web is a powerful model for the future that could be adopted as policy and forced upon vendors by our government right now. My only concern is that they don’t do it because of the adverse impacts on existing interests. Hopefully they’ll prove me wrong.

10 09 2010
Super Tanker Pilot

Having not yet read the paper but having many colleagues and friends in the IT industry and having righted a supertanking government programme back on track, I can and only agree with the majority of sentiments here. The real situation is massively complex not only from a technology standpoint, but from a commercial perspective. A previous commenter (Steve Horgan) has already elluded to the issue; the reward mechanisms do not exist within the public sector to attract the appropriate talent and human capital toward what could only be described as a challenging task and environment. Coupled with a commercial landscape in which the vendors actually engineer lifecycles into the products and programmes, it becomes a herculean task as a friend described; You’d need someone like Stephen Hawking to dissolve the complexity of the mechanistic IT machine in use by the government. Technology itself will not solve the issues, the wont of the individual in front of the ‘machine’ needs to be willing to take this on…

10 09 2010
pinboard September 10, 2010 —

[…] First report – and first test – of the PBA « The Network for the Post-Bureaucratic Age UK report via @swardley: Better for Less: How to make Government IT deliver savings. #gov20 […]

10 09 2010

[…] Better for Less, How to Make Govt. IT deliver savings #yamAlready some interesting comments on the report page – will add some of my own when I've ploughed through all 69 pages!Government_IT Spending_Review […]

12 09 2010

The term “local” and “localism” is used throughout the document, and yet there is little clue to what this means. Local could mean per council, per department, per development group, etc. And surely a “local” deployment is meanignless. This blog post could be based ona server in Timbuctoo for all you care.

Use of the “Betamax over VHS” argument changes meaning each time it is used.

“zero touch” – puzzling uses throughout the document.

“Open Data – government data must be transparent” – open data is not the same argument as transparency.

“Success will come from a technically literate government where ministers and officials are comfortable with their ability to use technology to do government better. This requires a commitment to learn about – and embrace – technology and the considerable determination that will be required to push through such a cultural change in Whitehall.”
– If you believe the above, we are all fucked.

“access rackspace offerings.” – A Freudian slip?

“Assume that citizens will access on-line public services using a market or ecosystem of accredited third-party identifiers (issued for example by a range of existing online services, credit bureaux, or banks)” – Just say Twitter and Facebook. If you don’t think you can do that, review your arguments.

“Android operating system” – You mean Mobile Phone operating system. A lay person will assume you mean something on a PC like Windows.

ILC approach as described seems to lack testability; I’m sure it should be mentioned.

The term “mash-ups” cannot be dropped in without explanation. This term is already out of favour (try hack)

“desktop as a service” (DaaS) – yes it can mean that, but is also known to mean “Database as a Service”. This is sowing confusion for ittle gain.

13 09 2010

While technically it is correct to use the term “SOA” as opposed to web services to express the basic concept of service building blocks, I suspect modern architects would consider the term a bit retro. At the same time, its just another TLA to lay people.

20 09 2010

Great stuff. ILC works for me.

The report encapsulates many of my thoughts, which will provide much fodder for my blog. I’ll do what I can to keep it on the agenda for improving information governance theory and practice.

Look on the Quarkside.

22 09 2010
Brian Lockwood

I like the localism.
Steve Hogan, “Pay Peanuts get Monkeys” is offensive to low paid workers. You may earn a big salary but I very much doubt you could (as one of many many examples) assist in open heart surgery (on ~20K or so a year) without fainting or vomiting. Justify your salary by all means but you don’t have to denigrate the low paid.

22 09 2010

I found the report to be over detailed and suggest that it is reduced in size to around 30 pages max and address the major issues only.

Secondly the report does not address how a post-bureaucratic IT system can be fitted into an existing bureaucratic system. This will require Project Managing the change and will require entirely new processes to make it happen.

While the bureaucratic system has control over its successor we can expect entrenched attitudes to slow its progress and increase its costs and to confuse its goals. To make the transition happen we require completely different methods of delivery. I suggest the IT ( I would prefer the use of IS instead) is delivered not to the bureaucracy but to the post-bureaucratic entity. The bureaucratic entity will wither away over time but it will not be without difficulties.

22 09 2010
Brian Lockwood

On the adoption of webmail for schools. This is a nice example of where you need someone in the job. Where students use school email, they often use it to transfer draft documents to and from school. They then need the document juts when everyone else on the RBC network also needs theirs and so they can’t access it. Never mind that Live Mail seems to frequently crash IE7/8 when dealing with attachments in 2003 TS edition.

There are some excellent free resources (e.g sogo, funambol) that schools can use with common authentication with AD. The attachment is then in school when needed, not stuck on a remote (busy) server, donloading attachments does not use up RBC bandwidth at busy times.

Local in education means ask the people in the classroom what works for them.

27 09 2010

A long train journey allowed me to concentrate on the details of Better For Less – at least insofar as it may affect leadership and policy. As usual there’s the Good, the Bad and the Ugly.

Good: Goal, Guiding Principle and Workstreams

Bad: Incomplete Lifecycle, Concepts and Governance

Ugly to follow

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: