The future of government

March 7, 2008 1:35 AM

How should government change the way it provides services in response to the technological revolution happening around us?

I'm giving a speech to 500 civil servants on Monday. It's working title is Government 2.0 - yeah I know, it's a bit hackneyed.

Peculiarly, the event might just help shape future government policy in the UK. So I want get it right.

So far I'm touching on:

1. The hive mind as a force for good
2. Freeing up data to the open source community
3. New ways of consulting with information communities

This and more. Can you help me? All ideas welcome.

posted by baggymp to Society & Culture (15 answers total)

Increased transparency. There's no reason why all kinds of internal data from government agencies shouldn't be available to us. This goes double for budgetary items, because where we should know where our money is being spent. I don't mean the post office got X dollars, either. I want detailed breakdowns of where the money goes within government agencies.

Don't just publish dense spreadsheets of it either. Open up APIs, so that talented citizens can query this sorts of data and make it useful.

(this is all part of your #2, but it's one of my pet causes)
posted by chrisamiller at 1:51 AM on March 7, 2008 [2 favorites]

If the government took email as a genuine channel for communication it would be a win. I've lost count of the number of emails that just disappeared into a hole with no response when dealing with government (note this is also something that corporates also fail at).
posted by bystander at 2:43 AM on March 7, 2008

Sharing best practice - how can you use technology to get government employees to identify "what works" quicker? Two approaches to this, probably: raw data crunching, APIs for analysing stuff, and that sort of thing, but also human/community-driven tools. "Digg this lesson plan?"

360 degree analysis - could you have an epinions or metacritic of your local careers office / A&E ward / daycare centre? Can civillian information sharing and reporting be more effective than other sorts of (gameable) targets?

Communities - how can you use the web to help encourage local solutions and build partnerships between the public & third sectors? The current favourite example is, of course, geographically proximous Obama supporters finding each other online and collaborating way, way before the campaign actually started doing official stuff in their area. Could you use that to make streets safer, cleaner, better?

Crunching data - open up the right data (or all of it) and see what patterns people can spot. This is probably the least popular one because it's about accountability (and who really wants to be accountable?) and challenging the conventional wisdom. Emily Oster's talk on AIDS (it's not as bad as it looks, the best way to fight it is by reducing other diseases [because you have more motivation to protect yourself from AIDS if TB isn't going to kill you in your 40s], reducing poverty doesn't necessarily help) is a good example of the latter. Build a nation of Freakonomists!

Build stuff small - however the government uses 2.0 stuff keep it small, nimble, simple, don't try and do too much at once, test it out, see how it gets used, iterate iterate iterate. Sounds obvious, but when it comes to government & IT the status quo is very much in favour of spending billions of pounds on massive systems that don't actually work.
posted by so_necessary at 2:57 AM on March 7, 2008

You need to contact the guys over at The Open Rights Group and MySociety. But looking at your site, I'd be amazed if you don't already talk to them.

I've just written and cut a whole bunch of waffle about how I want government to be more like the internet (smart edge) and less like a telco (dumb edge), because to anyone else it probably just sounds like rambling. Ditto all the other ideas I'd like to get over (eg applying usability principles to organisations).

So I'll just say "get the raw data out there. the net will take care of the rest." And a couple of useful things that could be done to that end, IMO:

1) Replace Crown Copyright with a US-style public-domain-by-default system (a lot of people think freeing up that data will generate wealth).
2) Instead of waiting for a FofIA request, just throw the data out there anyway unless someone stamps "do not release" on it. Default to open, instead of default to close. It'll be cheaper that way.

(I'm not really concerned about open file formats, free software and all that traditional stuff... just get the data out there, in any format, and someone'll build tools to slice'n'dice it.)

On citizen interaction with government.. the problem is that it doesn't happen often enough to become habitual. Take this site. I come here every day, so I know how it works, I know where all the features are. It's become a habit. Maybe my local government has a really neat tool to report vandalism via my mobile phone, but I only interact with government once a year if I'm lucky, so the habit of using their tools isn't ingrained in me. I wouldn't even know where to start looking for such a thing.

So the backstory to that "really neat tool" goes something like this:

Someone (maybe one of the people who you're going to talk to on Monday) says "wouldn't it be neat if...", and they take the idea to a private contractor, who never ever says "that won't work", instead they say "great idea! that's be £0.5million please". (I know... I was that contractor. Never again.)

So they build the tool, run it for 3 years at £2500/month, and get 15 submissions. And everyone involved decides that government-backed IT projects are always failures.

How to solve that problem... well, first, get IT- and usability- experienced people on government's team so they can't be taken for a ride by a contractor. Give them the authority to say "no". Second, once you've set the data free, if there's a pressing need for a tool there's a good chance someone will come along and build it for you anyway. Third, either make interacting with government's web tools habitual so people learn them, or make them so easy you don't need to learn them. I don't have a clue how to accomplish that.

Damn it. Still rambling. But if you can find something useful in this, great.

Good luck! Wish my MP had a last.fm account. Mine was one of the 70 who signed up for the copyright extension EDM. Guess who I'll be voting against next time around...

(Oh, one other thing, that totally won't go down well. Stop worrying about fairness. It would be very useful if each class at my kid's school had a mailing list for parents. The teacher could email the list with what was going on that week, what homework had been given out, that kind of thing. The school can't do it "because it wouldn't be fair to the parents without email". There's similar logic happening when I email a teacher and get a written letter in reply.)
posted by Leon at 3:01 AM on March 7, 2008 [1 favorite]

While Government 2.0 sounds great, make sure that you aren't leapfrogging the basics that many government websites just don't provide yet. I'm an expat, so I rely on online services from my home governments (Orange County, California, the US) pretty heavily. Here's what I've found helpful:

- updated websites with accurate fee and mailing-address information (it's frustrating how often I have to call and confirm this!); to quote an article I'm linking below, bad information is worse than no information.

- quick response times to e-mailed questions, with reference numbers or codes so my questions can be traced and answered by real people if I choose to go into an office in person

- any kind of form that I need to sign officially EITHER being moved online with some form of online authentication (getting a new driver's license, changing my address for voter registration purposes) OR being put up in PDF format for me to print out, fax in, sign and scan and email in, or mail in

- equal prices for services regardless of whether I do them online, in person, or through the mail; along with this, support for me to use my (in my case) Latvian debit card to pay for a seal or certificate I need mailed out to me here

- Skype IDs for government offices so I can call and ask questions for free from wherever I am

And as great as online collaboration sounds, do make sure that participants take away the idea that they've got to make their web presences really easy to wade through. Here's a list of US websites for governments, rated with a letter grade, and an article about why one of the top scorers, the state of Michigan, did so well:

In Michigan, the IT department is working on mobile applications for tasks that people want to accomplish on short notice, Hogan said. For example, one future application will let people who book charter fishing tours use handheld devices to buy 24-hour fishing licenses. “These guys show up at the dock at 4:00 in the morning, and the skipper finds out they don’t have their fishing license,” he said.

posted by mdonley at 3:06 AM on March 7, 2008

(An aside: look at the language many of us are using in our answers. Clipped, fast, note-like. A lot of people seem to be very anxious to get their ideas over to you. I think that's interesting, from a "people are apathetic about politics" point of view)
posted by Leon at 3:10 AM on March 7, 2008 [1 favorite]

keep it small, nimble, simple, don't try and do too much at once, test it out, see how it gets used, iterate iterate iterate

This is worth repeating. I realize that beauracracy and red tape are what government does best, but building monolithic applications is an outdated paradigm (esp. on the web).

Solicit user feedback often, and respond to the community. Release beta versions, and invite the right kinds of users to test them. Find a way to get things done without 18 levels of approval, so that when you start to deliver, the software isn't already outdated.
posted by chrisamiller at 3:19 AM on March 7, 2008

keep it small, nimble, simple, don't try and do too much at once, test it out, see how it gets used, iterate iterate iterate

This is worth repeating. I realize that beauracracy and red tape are what government does best, but building monolithic applications is an outdated paradigm (esp. on the web).

That's a bit idealistic. Open-ended iteration doesn't sit well with the contract-spec-build-pay way external IT projects are traditionally set up. If you want to iterate you're better off doing it in-house, but if you do outsource you still need someone experienced on your end to guide the process, else you're going to be treated as a meal-ticket by your contractor.

I wish I could talk about the project I was involved in. It was small, nimble, simple, and... nobody used it. The concept was broken because it didn't meet the users' needs (it was intranet stuff, so I'd say there's an 80% chance baggymp was exposed to it, but didn't bother to use it more than once or twice), but nobody on gov's side had the knowledge, or cared enough, to put the brakes on. Extreme programming is useful but it doesn't guarantee project success. You have to service a need, and you need to be plumbed into your users' lives.
posted by Leon at 3:53 AM on March 7, 2008

Perhaps you should examine whether all this technology has positively affected productivity in terms of output per worker or, conversely, number of workers employed to produce a certain output. If it hasn't, ask "Why not?"

Could make for a lively discussion.
posted by ZenMasterThis at 5:59 AM on March 7, 2008

Two resources that might provide additional helpful food for thought: the Hansard Society's Parliament for the Future report, and the Feb. 16, '08 Economist issue that focused on government and technology (with a strong European/UK bent, imo, which the Economist is wont to do).

Here's an article called "Government offline: Why business succeeds on the web and government mostly fails". And if you look at this main index page for the issue, you should see on the bottom right under the heading "A Special Report on Technology and Government" a whole pile of articles with analysis on your very topic. (Not that you won't get good stuff from here as well, but if you need to cite a source, I don't know that AskMetaFilter has quite the cachet that you would hope for. Yet.)

To me the big problem with the way government provides services goes something like the following:

1. Spend six months forming a committee of high-level internal bureaucrat stakeholders to assess The Problem, Need and Audience. Decide that members of the technology community and private sector can't be trusted to be involved. Dismiss the participation of the low- or mid-level public employees who actually work either with the org's IT or with the public users that would receive the modernized services.
2. Spend six months publishing a report.
3. From that report, develop a new committee of internal bureaucrat stakeholders. Delay when there is political fallout from agendas of members of the original committee who did or did not get asked to be on the new committee. Spend six months resolving.
4. The new committee realizes that the report has already become partially obsolete. Additional delay ensues when new committee decides it needs to go on lengthy fact-finding junkets to other governments that have implemented some technology advances.
5. Issue a new report. Result of that report is "Build a big bloated enterprise thing. Find a vendor to provide."
6. Vendors worldwide begin to salivate at the €/£10M tender that is coming down the pike. They know what happens next:
     a. Submit a tender and win.
     b. Get big fat government checks
     c. Spend twelve months developing use cases.
     d. Spend six months realizing the existing technology in the government org is so bad, and the users so behind, that the original proposal can never succeed -- all thanks to a lack of communication between the original committee of stakeholders, and the jobsworths not wanting to openly admit that they are using obsolete systems and have an obsolete level of knowledge about the Problem.
     e. Try to salvage something from the project... but in the meantime, know that the real technology, out in the real world, continues to zoom past this project.
     f. Ask for more money when the scope has changed fundamentally and the funds are gone.
7. ???
8. E-Government!!!!!!

I don't believe that government can improve the level of technology it uses to provide services, as long as it brings a same-old, same-old government approach to technology. The public sector needs to start with the same approach to technology that the technology community uses.

But it's delicate -- people don't like to admit that they don't know about something. It's hard to get an agency to confess, "Yeah, we've been doing this the wrong way for ten years, and we know that the public hates it but we can't hire good tech people to join us and fix it because our compensation isn't competitive with the private sector." Or, "Yeah, we know that everyone in this office is gray-haired and still uses typewriters, but we're so worried about being made redundant before we reach pension that we're going to close ranks and claim that the domain knowledge and processes simply can't be modernized in that way."

Another problem with government using a governmental approach to technology is that then vendors, contractors and consultants aren't held to the same performance standards that they would be in the private sector. Out in the real world, if you don't deliver on a £5M project, you'd get sacked, then sued for damages. But in the public sector, there's rarely any action to be had; the failure and loss just get written off. As so_necessary said upthread: "Sounds obvious, but when it comes to government & IT the status quo is very much in favour of spending billions of pounds on massive systems that don't actually work." Leon addressed this too.

I don't think that you'll ever really get open source working for government until you address the force of Microsoft's lobby. Remember, this is a company that has basically thumbed its nose at $2.5B (yes, B) in fines from the EU over the anti-trust case. And the only way to address the Microsoft lobby is to change the minds of legislators and leaders, to get them to understand ways that open source is better for the public.
posted by pineapple at 6:08 AM on March 7, 2008

Short list:

1. Focus on reduction of paper and latency. Electronic transactions are a huge cost and time saver for everybody, if done correctly. This requires adoption of cryptography standards so that citizens can communicate electronically with the government in a way that cannot be forged or intercepted (email alone is not secure). Open cryptography standards are best (for a zillion reasons -- c.f., Schneier's blog)

2. Individual privacy needs to be taken seriously and regulated. This is going to be a huge problem.

3. Transparency. This has already been covered.
posted by qxntpqbbbqxl at 7:31 AM on March 7, 2008

Without the people, you have nothing.

Thus what is required more than anything else a reform in the selection, remuneration and advancement of civil servants such that citizen-facing personnel are likely to be tech-savvy and tech-friendly (not the same thing, perversely) and their supervisors are going to be those who have proven themselves tech-savvy and tech-friendly.

In the US (and it may well be the case in the UK), the failure to build up a strong technology competence in the civil service has led to a near-total reliance upon private contractors for high level technology. Rather than driving pay-raises and promotions for those who face citizens and supervise those who do, tech savvy becomes rewarded only for those in the ghetto of tech support or the niche of consultant and vendor selection.
posted by MattD at 8:43 AM on March 7, 2008

Also, you ought to emphasize that there's more to leveraging the wisdom of crowds than just bunging up a website and inviting people to comment willy-nilly. For instance, Healthcare For London did something rather like this for their 2008 public consultation. They put a MORI survey online (very sensible) but they also had a thing (that I can't find now) where you could just submit ideas and quasi-Digg other people's suggestions.

#1 was "i like beer."

As always with UGC/human-flavoured online stuff, you have to put a lot of effort into creating useful long term communities, otherwise the discourse will inevitably be dominated by the sub-clinically deranged.
posted by so_necessary at 9:35 AM on March 7, 2008

Response by poster: Thank you all. I've redrafted what I'm going to say having considered a number of your comments. The words "Build stuff small, test it out then iterate, iterate, iterate" are going to appear. I'm also going to talk about:

Transparency, data mash-up and release, building confidence around data sharing,increased customer focus,wikinomics, open source solutions, My Society's "Power of Information" report and the need for in-house technology competence.

This and the usual stuff they make you speak about.

If I get the chance, I'll stick it on my blog tomorrow morning.

Once again, thanks for your help.

One last thing - someone emailed me to say we need an AskMeFi for public sector technology projects as a way of spreading best practice. What do you think?
posted by baggymp at 8:18 AM on March 9, 2008 [1 favorite]

Who do you envision participating? Would it be the implementers? The agency? Who would be supplying the answers? Who would be encouraged to ask?

I think the reason AskMe works so well is that it was borne of a community that self-selects and rewards quality posting and quality knowledge sharing. I would worry that in a more public venue where the barriers to participation were low, that the result would be more "Yahoo! Answers," less "AskMetaFilter."
posted by pineapple at 9:00 AM on March 10, 2008