Sunday, February 24, 2008

e-government 3.0

Wait! What? 3.0? How did that happen? I, along with many others, have long had a theory that government should - and often does - sit behind the private sector in terms of technology. A government department taking a leap into the unknown, what used to be (and probably still is) called the bleeding edge is a rare and brave one. Often, at least more than the laws of probability would allow, such efforts end in failure. Web 2.0 is probably more like e-Government 4.0 or maybe even 5.0. Thinking about this has been languishing in my blog drafts folder for many months - the article in the Economist last week prompted me to dig this out and refresh it a little. It's all about what do governments do next? They've had the "get everything online" phase - and many have done pretty well; most moved on to driving up usage of transactions (and, again, many have done well). But what next? What's in the third phase?

Electronic Government 21.06.2001.jpg DWP Conference Slides - 30.01.2003 - stages.jpg

The slide on the left was something I first put up in early 2001 to explain the stages I thought government would go through on the journey to a truly transformed operation. If I remember correctly, the words on the bottom (Publish, Transact, Interact, Transform) were from a report writtten the previous year called "e-Government @ IT's best." I no longer have that report, but I'm pretty sure that when I created the slide I reversed the second and third stages - I felt strongly that interaction would come after transactions, i.e. that transactions were going to be one way at first (taxpayer sends tax return) and would then become two way (government says, "thank you for your tax return. we need this bit of information. did you know that you can also do this. and here's your refund transaction" etc). The yellow octagon (bottom left) is where I thought we were then (i.e. heavily mail dependent, quite some way into the publish stage but with a big bump just ahead of us as we tried to orchestrate transactions).

Later, I put some dates on the slide suggesting that each stage might take at least a year to break through - but that online private sector companies were already moving ahead of government and making the move to some truly dramatic changes. I also used a much simpler background - I have no idea what I was doing with the whole sunset at sea thing. The opening slide from the conference where I showed this slide noted that the FT was saying that we'd already spent £3 billion on e-government in the UK alone, by January 2003. By mid-2003 I think I'd abandoned the slide - I can't find it in any later conference pitches. Evidently I'd realised it didn't say anything by then (and that I was, of course, hopelessly optimistic about the dates by which any of this could be achieved).

But in 2008, I think it has relevance again. Governments, generally, seem to have stalled at the "transact" stage. All of the big plans to move to real interaction, to transformation (dramatic or otherwise) appear - to me viewing from inside government but outside the e-government race (and it was a race) - to have stalled.

If e-government 1.0 was publishing - several thousand websites per country - and e-government 2.0 was about transactions (moderate take-up of a small number of transactions, high take-up of a couple and low take-up of the rest), then what is e-government 3.0?

For me e-government 3.0 is the set of actions that need to be completed to truly have a high level of online interaction between citizen, business and government. The creators of that interaction will be both private and public sector-based. Interaction means lots of things to lots of people. I mean: high levels of usage, bi-directional flows of information (that is, citizens send to government and government engages in dialogue, including offering access to services not yet taken up by that citizen), third parties engage in service provision wrapping their existing offerings alongside government services (so, for instance, third party pension providers will show you your state benefits alongside what your employer benefits could be and potential for additional benefits with separate schemes).

In e-Government 1.0 and 2.0, one of the big surprises for me - and this is a global, not local point - was the lack of engagement from non-government businesses in providing services - that is, few third parties took up the challenge of providing services alongside or in competition with government. From 2000-2003 I figured that we were only a step or two away from seeing that engagement grow strongly. Early on I was involved, loosely I think it's fair to say, with schema development for various financial transactions with government and lots of software vendors and perhaps that gave me a false sense of how the market might develop. In hindsight, there were a few clear reasons for the absence of third party involvement:

1) It's hard to compete with something that's free. If government does something reasonably well, why would you try and charge for something that appears to only offer modest advantages. In the UK an example here is HMRC's Self Assessment process. Tools that tried to compete were provided by Which (TaxCalc) and Microsoft (TaxSaver I think, although there was another one called DigiTax perhaps) - they were good, but they weren't good enough and they died out. There is also a famous story of an online fishing licence process that with its attendant publicity, in one go, pretty much killed off aspirations for many 3rd parties.

2) Governments are not necessarily able to be open about the complex methods of interaction. The UK still, I believe, leads the field with its govtalk certification (see but even that still contains many draft schema dating as far back as 2002 and lacks important schema in, say, the health records domain. The last major update to the documentation was in March 2005 and that only really to note that the OeE was no the eGU. Interacting with government is complicated and takes a lot of documentation to figure out. When Sun partnered with Software AG to design and build a local box (known as a DIS box) that would connect a department to the Government Gateway, reams of documentation were necessary to handle the various messages that could travel back and forth.

3) Allowing third parties in makes things complicated for government. At present, if the main provider of a service is government and they realise that there's a problem with their online service, a quick change and everything can be fixed. What if you've actually certified 5 external providers to provide that same service and furnished them with protocols, schema and instructions? And then find there's a problem. The quick change is no longer possible and overall service can suffer quite dramatically.

4) Finally, at the beginning government figures it needs to seed the market. It doesn't know how many people are going to use a service, and it can't make promises to one third party over another about the business opportunity. So figuring out how much market share is there for the taking by 3rd parties is a challenge. Government doesn't know. To stay with a UK example, HMRC's Self Assessment saw 3.7 million returns last tax year (plus a few thousand more in the couple of days after the filing deadline) - that's a pretty good number, probably close to 50% of the total returns but it's taken 7 years to build to that number and it's probably a good 2 million short of the total that could be delivered today with minimal effort from those sending in tax returns. I don't think anyone - certainly not me - starting out on the e-government trail in 2000, would have thought that it would take this long to get 50% of the returns in electronically.

I think e-government 3.0 has to take these, and other, issues head-on in an effort to really drive up usage by the population - and use the space that creates amongst the internal workforce to drive real change to processes, dramatically simplifying them, joining them up where it makes sense to do so (and recognising that "sense" is sometimes a hard word to define at the get go) - bringing it all together under a new overall mission, allied to the somewhat waning transformational government agenda. I believe people need a new flag to rally under - the momentum has gone from the original drive to get things online, progress is only incremental and the buzz has gone. E-government should, by now, be business as usual of course; there is, perhaps, no need even to continue to use the word but, at the same time, the original aims of governments that embarked on the course have not yet been entirely realised and so we need something to rally the troops.

Here are the 6 things I would propose a government do to channel their efforts towards busting the next barrier to progress in e-government:

1. A clear statement on the direction of travel for the government adopting this plan. For instance, they will no longer develop new e-government services where there is potential for commercial involvement and competition, where there are existing services offered by such third parties or in a specific list of defined transactions that are as yet undeveloped, difficult to develop, or have the potential to be widely used in specific markets. Essentially, a government would publish a list of 50 or 100 transactions that it wanted the private sector to adopt, embrace, drive usage on and join up. There I've said it: join up. Some transactions that already exist do so only as stove-pipe transactions. Government would allow third parties to join those up and not compete with them.

2. An OpenAPI will be developed, documented and published detailing how to get to and from government. I noted above that govtalk had gone a long way down this path. but there is plainly more to do. I was intrigued to see an announcement today regarding health clinics in the USA developing connections to Google for storing health information - with the idea being that it can be retrieved as the patients move from state to state or doctor to doctor. This strikes as a perfect example of an API as I imagine it. Plainly there are, in this case, privacy issues to be concerned about - but Google reads your email and may know as much about your health if you mail updates to friends and family as anyone. This would be put on a fast path, using the hit list of transactions in (0) above as the base. Government would do what the software industry does and always support at least n-1 versions, perhaps more.

3. Government would develop rules engine capability "in the cloud" - i.e. processing would no longer need to be behind the full set of government firewalls, but could be handled a step away with only the final results of transactions entering government. Government would commit to make these engines highly available, scaleable and up to date. It would likely partner with suppliers who have proven that they can do this - inevitably, that is a small list but it plainly includes the big Internet companies such as Amazon, Google, eBay and Microsoft; there would also need to be local (national) companies in this list to ensure that the local economy developed along the same path as the titans, where possible.

4. A way to encourage small businesses to get into the provision of such services would be set up. I imagine a Dragon's Den style competitions with businesses receiving funding to deliver services, based on the hit list with bigger incentives for the harder ones or where government already offers reasonable services. To date, I think governments around the world have fought shy of such an action, worried that it somehow breaks procurement rules or gives some players an unfair advantage. As I see it, anyone who wants to play can get a meeting, but they have to show up with a business case and plan that the angels (or is it dragons?) say is viable and that government is prepared to put its money into. To ensure the government lives up to its end of the deal, it will commit to invest in X deals for £Y a year - and whatever X and Y are, they'd be public and reported on monthly on the national portal - that would be in the UK, etc

5. To facilitate longer term development and operation, a model for paying transaction fees for businesses meeting government certification and providing accurate, timely transactions would be developed. The fees would start low and grow as volumes grew, incenting providers to seek out bigger markets, giving players that were growing fastest the opportunity to achieve real economies of scale. At some point, fees would level off and perhaps even reduce year on year, incenting the providers to find ways to deliver more efficiently. Government would pay these transaction fees by abandoning development of services where competition was being introduced, not entering new markets, no longer needing to upgrade servers and so on for its own websites. Government would have to provide some degree of technical and business support, and this would be factored into the costs.

6. Third parties would be granted the right to white label government content, with acknowledgement, using simple tools that made necessary policies, rules and instructions available to them. Ideally, content would come from the national portal which is the most likely place for the investment to be made to allow this but also may become, if it isn't already, the only source of up to date, joined up content.

Making these 6 points happen would be no easy task. It would require a step up in co-ordination inside most governments, a new reaching out to the private sector (and a large number of cynicism bypasses there), a rapid set of decisions on where the priority areas were and some fast delivery to prove that it was viable. It's very likely that there are some folks out there waiting in the wings, who have services that could be offered as pilots for this or have tools that could be used to develop services quickly, but they don't know who to talk to, how to get to them, how to fund it or how they will construct a business model.

I'd envisage some small businesses getting off the ground this time round - there were attempts at this during the early dotcom boom days; some survived, most did not - and growing into strong national transaction players who would look to do interesting stuff like take paper processing off the hands of government as a vehicle to make it electronically more quickly. There could be a mixed economy of big and small players emerging over time. Just as likely though is that existing industries - banks, accountants and so on - would step into providing their clients additional services and value.

In the background, government would work on its bigger picture for transformation - once assured that increasing numbers of transactions were coming in electronically, the smart folks on the inside of government would look for ways to redeploy staff to more customer focused areas, to develop new business models that could be based on the assumption that all transactions for, say, a start up business would be electronic and so on.

This would take time - 5 years or more. That might mean 10 for some governments. But it seems to me that governments have a better sense of time than most. At the moment, though, few have set out what it is that they want from their online services over the next 2-3 years let alone 5-10. I think a model like this could just see some real breakthroughs that would significantly benefit national economies - both on consumer and producer sides.

Dan Jellinek Transform slides - 13.05.2004 - inandout.jpg DWP Conference Slides - 30.01.2003-opening up.jpg Mark Forth Slides - Year 3 - 17.11.04 - storm.jpg


  1. Anonymous4:20 pm

    I would just like to point out TaxCalc is still around. Acorah now publish the software and TaxCalc 2008 will be lanuched later this year.

  2. you're right. they're at

    thanks for pointing that out.

  3. Anonymous8:45 pm

    This is an interesting blog on eGovernment: