ntouk.com - Jerry Fishenden's technology policy blog

I came across a paper I wrote back in April 2004 recently. It's called "Building the Agile Department: a Service Oriented Architecture for Government".

It was written for the Cabinet Office at the time whilst I was working on the development of the Government Gateway, although achieved a wider circulation including to the (then) Inland Revenue and Her Majesty's Customs and Excise. It was intended to provoke discussion, and practical action.

The paper looked 5-7 years out. So it was trying to see what could be delivered by around 2010. It set out its objective as being to:

"... provide a simplified background on the principles and benefits of a Service Oriented Architecture (SOA) -- an organisational commitment to build an enterprise application portfolio around a collection of well-designed, network-available, component services. It looks at the relevance of an SOA to government organisations internally as well as the wider joined-up government and e-government programmes. Its purpose is both to inform and provoke debate on the nature of government architectures, both now and in the future."

Some of its language remains eerily topical:

"Many departments are currently engaged in a major review of their use of Information Technology (IT). The drivers for this review include the e-government agenda and joined-up government programme, as well as recent developments such as the Gershon report and the Chancellor's 2004 Budget statement. Departments are keen to find ways to improve the efficiency and reduce the costs of their own IT services and operations. They also desire more agile and flexible IT operations that are better placed to respond to changing business needs."

It recognised that:

"[there is an] underlying business trend away from silod, functionally isolated business processes and systems to a model that is more agile, flexible, adaptable and cost-efficient. The previous focus on stovepiped internal business functions is increasingly being superseded by a realisation that the focus in fact needs to be the external consumers of public services and the overall end-to-end processes involved in the delivery of those services."

It foresaw a significant change in the way government approached IT, which even back then did not have a great reputation for cost-efficiency or success:

"This model is a fundamental improvement over that which prevails today. It foresees a time when the silod line of business/Head of Duty style systems are not needed. Instead, smart, re-usable common services are employed. For example ... all of the elements that go to make up the process of Self-Assessment are provided as a set of services -- spanning the Rules that perform the calculations, the authentication required, the help text etc.

This raises fundamental questions about how future systems should be conceived, architected and delivered. It will require a move away from the current expensive silod function-oriented project-by-project acquisition of infrastructure approach towards a wider architectural model that ensures all systems developments take place within the same framework."

So, where are we 6 years on?

It would certainly be timely in the current economic climate to review in detail what has changed in those interim years of substantial IT investment. To determine how far IT has moved towards an improved model focused on services rather than infrastructure, and on end-to-end process design rather than mono-functional silo.

A recent Centre for Policy Studies paper "It's ours: Why we, not government, must own our data" highlighted SOA as a key idea for government to adopt. So perhaps, these 6 years on, finally its time has come?

Or do you have better ideas?

Much has been learned about the practicalities of real world SOA in the meantime -- not least by the Government itself courtesy of its Government Gateway.

I've put a PDF copy of my 2004 paper here. Warts and all. I'm not sure, after all this time, if it was the final polished version or not (it seems a little rough around the edges) ... if you dig out another later (or better) copy, please let me know.

In the meantime, I'm considering what I would recommend now. If, for example, someone were to ask me to produce a series of actions to implement a "new model" of how to manage and provide IT services in a more cost-effective, agile way, based on real-world practical experience of designing, operating and managing large-scale systems ...

Technorati tags: future Britain IT technology policy open government public services

Is it just me, or are the wrong questions often asked about public sector IT? Most of the questions I keep being asked seem to be along the lines of "should we use more open source?", "should we use the cloud?" and "is product X better than product Y?".

Which seems to me to be starting in entirely the wrong place. These questions start with a potential solution in mind without defining what the public service problem or requirement is.

This technology-led approach contributes to the high levels of public sector IT spend that go on procuring and maintaining technical infrastructure, rather than on the efficient running of our public services. An average of 80% of the annual public sector IT spend (some 17bn to 21bn GDP) is estimated to go on watering, feeding and updating existing systems. But what does that infrastructure actually deliver? How does it play an active part in improving our public services, of "doing more with less"?

This obsession with the acquisition and ownership of infrastructure, rather than procuring services and capabilities to meet public service needs, betrays out-dated mainframe thinking in an internet world. We should be defining requirements and capabilities, not infrastructure. And then procuring services that meet those needs.

Yes, that might involve more open source, more use of commodity cloud offerings and swapping product X for product Y or vice versa. But who's to say? The best solutions will result only when we first understand properly what public service requirements we are trying to meet. Not by procuring infrastructure and technology as ends in their own right.

So I think we should start to rethink the questions that need to be asked. We need to start asking hard questions about the fundamental needs of our public services, not by assuming that the answer lies in specifying, procuring and building yet more IT infrastructure, or deploying product X rather than product Y.

The existing UK model seems to provide the worst of both worlds. In theory almost everything is outsourced, but the client is still often making the call on much of the infrastructure and the technology it embeds. Why? The result is both high cost and low impact in terms of improving our public services, whilst all the time risk is retained within the public sector (and ultimately underwritten by the taxpayer).

From commodity services such as word processing and email to more complex service requirements, such as realtime taxation and welfare, we need to move to a model that defines very precisely the needs of the UK public sector. We also need to update our assumptions around governance, architecture and procurement to enable the public sector to become much more agile, relevant and cost-effective in the acquisition of services and capabilities.

The coalition partners both made clear during their time in opposition that they had smart ideas to ensure IT was better managed and integrated into the cost-effective operation of the UK's public services. The recent freeze on public sector IT is a good first step, providing time to plan a new direction and much more rigorous approach to governance, architecture and procurement.

So now would seem to me the perfect time to start asking the hard, fundamental questions about our public services that we need to answer first if we are to ensure the public sector gets the IT capabilities it requires. Only when that is done can we help ensure that an improved, more cost-effective and impactful approach for public sector IT will be successfully delivered.

Technorati tags: future Britain IT technology policy open government public services

With thanks to Edgar Whitley for spotting it, the new coalition government document sets out some key principles for public sector IT (see page 21), including:

• We will take steps to open up government procurement and reduce costs; and we will publish government ICT contracts online.
• We will create a level playing field for opensource software and will enable large ICT projects to be split into smaller components.
• We will require full, online disclosure of all central government spending and contracts over 25,000.
• We will create a new right to data so that government-held datasets can be requested and used by the public, and then published on a regular basis.
• We will require all councils to publish meeting minutes and local service and performance data.
• We will require all councils to publish items of spending above 500, and to publish contracts and tender documents in full.
• We will ensure that all data published by public bodies is published in an open and standardised format, so that it can be used easily and with minimal cost by third parties

And also aims to reign in some of the excess salaries in the public sector, which have out-stripped the private sector.

Technorati tags: future Britain IT technology policy open government public services

I've been asked more than a few times lately what I would like to see happen with public sector IT after the election. That's a big topic of course ... on which I could dive into a whole series of urgent changes that need to happen in parallel.

The recent crowd-sourced work on the ideal government IT strategy, which impacted several election manifestos, made clear just how much needs to be done.

But I'll put aside the more structural issues around implementing effective governance, architecture and procurement for the moment, as there's plenty of detailed work happening on that already. Instead, here is an interim shortlist for a few short, sharp pragmatic projects to get public sector IT motoring.

1. Implement an integrated single tax and welfare system. This has been a gleam in the eye since at least the late 1990s, if not before. But the difference now is that the means to deliver it, cheaply, efficiently, quickly and in carefully managed chunks has been around for at least 4 years. Given the economic situation, this has to be priority 1 -- and will in turn help drive a whole host of significant public service reforms once implemented. The only thing that baffles me is why it was not done before.

It will prove a significant catalyst for progress.

2. Complete the modernisation of DLVA tax disc renewal. Get rid of the paper and make the process a rolling one, rather than an annual exercise. Whether a vehicle has road tax or not is easily checkable electronically, the piece of paper in the front window is a waste of time and resources. Let people move to monthly direct debit so it's a continual renewal rather than an annual big bang. And reduce the price for online or rolling renewal to reflect the fact it's cheaper and more efficient than continuing with manual processes.

3. Learn the lessons of taking paperwork, admin and bureaucracy out of the system from the changes to the DVLA tax disc system (above) and apply them systematically and rigorously to government's other large and inefficient transactional systems. Drive the operational efficiencies, reduced bureaucracy, hard cash savings and improved services that properly managed change will bring.

4. To deliver effective online public services, we need to have a good underlying identity system and a way of letting third parties, rather than solely government employees, directly interact with those services. So it's long overdue time to review, update and implement the existing policy documents on the use of intermediaries, trust frameworks and online identities. Drive them rapidly across public, private and voluntary sectors. Ensure the Government Gateway delivers its original vision of enabling online public services to support a whole range of trusted identities and identifiers, not just those issued by the Gateway itself. Deliver that within a year. We were once ahead of the US and other big players on this, but now we're behind. So it's time to play catch-up and recover from the lost, wasted years of ruderless direction.

5. Immediately open up procurement to small and medium suppliers, particularly around low hanging fruit such as website design and innovative, modular services. There's a wealth of talent and innovation out there, it just needs to be unleashed. Demand aggregation has stifled and distorted the market, concentrating power in the wrong places. Transparency of procurement and contracts will help drive and support the change. Breaking apart the existing monolithic, market distorting model will drive major cost savings, enable innovation in the design of public services and for rapid, tangible progress to be made.

In the background, I'd also like to see a freeze on all new major "IT projects" (or, more accurately, ambitious change projects that in fact change little, generally automating old and inefficient processes instead). We should empower CIOs to aggressively sweat the assets of existing systems and drive down costs of existing contracts. In parallel, we need to have joint teams of empowered policy and technology specialists tasked with delivering real, urgent change on the ground -- with design of public services remodelled around the citizen not just talked about.

I've not seen much evidence that the annual spend of around 21bn GBP on public sector IT is delivering anything like the return on investment in public services that such a high figure should demand. In fact, public services are often not even mentioned in government IT strategy documents, which remain focused on technology for its own sake and confuse commodity and bespoke procurement requirements.

In the private sector, I had to show how any investment I proposed would deliver a ten to one return within a year. On that basis, public sector IT would be expected to show how it is delivering some 210bn GBP worth, annually, in terms of better public services. The existing evidence is to the contrary, showing public sector productivity has declined during a period of unprecedented expenditure and an average annualised growth of some 15% or so in public sector IT expenditure over each of the past 4 years.

Whatever the outcome of today's election, the current model will have to change. And there are encouraging signs that such change is coming. We should see this as the perfect opportunity to finally deliver well-designed, cost-efficient, public-service-led IT not see it as a threat, however much it may rock the boat of some existing vested interests.

So here's to tomorrow, with the hope that when all the election dust has settled, we'll have a clearer and more effective approach to public sector IT. One that finally delivers on the elusive promises of the past two decades.

Technorati tags: future Britain IT technology policy public services

An early start for a breakfast in central London with a mix of some of the biggest UK IT players and some of the smallest. A large round table, coffee, tea and a mix of bacon butties, pastries and fruit.

The topic: how do we fix government IT?

As it was held under the Chatham House Rule, I can't reveal who precisely said what ... but here are some of the more interesting soundbites ...

"Public sector IT has failed. The current model is a mess, with no clear leadership or ownership of the overall IT strategy"

"It's not industry's fault what is happening. Every time we try to show leadership or inject innovation into thinking, the government IT side push us away."

"The current CIOs just don't listen. They're remote from the reality of the changes that need to be made."

"We should stop talking about IT projects. There should be no such thing. The trouble is the CIOs aren't involved in public services, so IT is all they think about. Until this is fixed, IT is never going to work properly in the public sector."

"Bringing in CIOs from the private sector has failed. They have no understanding of the public sector ethos. They're not CIOs. They're IT managers."

"Cranfield needs to play a leading role in fixing IT skills in the public sector and bringing leadership and management in-house."

"The shared services agenda is ridiculous. Everyone wants to share their services with everyone else, but no-one wants to use a service someone else has built. It's a solution looking for a problem."

"Government demands completely bespoke solutions at high cost and high risk just to fix 1% of an obscure problem, rather than accepting something off the shelf which is an 80% fit and costs a fraction of the price."

"There needs to be accountability for in-house government IT consultants. They propose ideas that look good on paper, but which fail massively when implemented. Yet they are never held to account."

"There need to be incentives to persuade government departments to join-up."

"The biggest blocker to small players getting involved in government IT is the liability insurance. It's designed to keep the market in the hands of only the biggest players. And even some of them won't sign up to government's absurd and pointless terms."

"Government can never truly outsource risk anyway, so why does it both with such onerous contractual terms?"

"We need proper benchmarking of costs and benefits. We need to be able to show who is doing a good job, who a crap job. And we should fire CIOs and suppliers who are in the latter camp."

"Accounting officers need to be accountable for IT too, otherwise it'll never be taken seriously enough in Whitehall."

"We spend three times as much bidding for a public sector contract as we do a private sector one. It's a huge waste of everyone's time and money."

"Government procurement rules aim to minimise risk rather than maximise gain."

"Government management Boards are focused on policy not delivery."

"Permanent Secretaries don't want CIOs on their Boards. That would implicate them in the 70% failure rate of government IT."

"It's time government focused on outcomes, not inputs and processes."

"The process needs to be simplified to let more small businesses get onto frameworks and hence able to supply the public sector."

"Every time we propose savings and improved services through the use of IT, we get blanked."

"Don't be afraid to mandate things."

"Don't miss the chance to drive massive savings whilst still not kicking-off new projects."

"Don't underestimate the complexity of getting real IT leadership back into departments."

"Don't break what isn't broken. SI's are up for reform, able to take risk and responsibility. We didn't choose the current system or appoint the CIOs we currently have to deal with. We're happy to see change."

"Be careful SIs don't make cuts just by sacking their smaller subcontractors. Ensure transparency of how many SMEs they employ on public sector contracts.

"There are no real CIOs in Whitehall. The current lot are IT Managers. Government needs to bring in a new set of IT leaders who sit on Boards and take real ownership of IT. Only then will you start to fix the problems."

Technorati tags: future Britain IT technology policy public services