However, I don’t see ‘centralization’ the same way Brett fears it (ie. pre fall of the wall Eastern European model). I think it’s a question of building a platform so that Government is less in the way and citizen’s become empowered with tools to participate, to a certain degree, in the governing process.

I’ve just blogged about it arguing that a good model for such platform is Facebook with its consistent interface, perfected usability and the opportunities offered to developers to host their own applications:

http://www.architxt.net/blog/new-media/government-websites-like-facebook/

My view is that if we keep doing things the same old way we are going to keep getting the same incremental improvements and will only be able to expect progress at the current rates– which is not fast enough for many of us.

Maxine and Lewis make great points about content, but I think I would like to see more centralized guidance on interface standards and even technology choices to enforce findability of info and the many goals you state in your post like being “more unified, efficient and strategic.”

I also think Gwynne used exactly the right metaphor in saying “The pendulum is always swinging” because it is. And as for me, I would like for it to swing to more centralized for at least a while so we can optimize a few things. Then let it swing back to decentralized more later. Gwynne is also right, of course, about the focus being on freeing the data.

So thanks all for the great thread.

The organization costs, building costs, and sacrificing some degree of local control outweigh the advantages. What happens instead is that the market system creates a collection of innovators competing to be a dominant force in the 20,000+ websites at no cost to the government (beyond normal profit that is probably less than indirect employee overhead). This competition produces no upfront costs, leads to solutions at least as good as would be created under a centralized control system, and does it faster than any committee could decide which technology stack was correct (several religious wars here – OS/variant of Linux?, language?, db platform?) – when what matters is decent architecture and interoperability standards (messaging, microformats, data standards). The reason that some sites are outdated is because others pushed the envelope whilst others used the prevailing quality – which would probably equate to the standard in a non-competitive setting.

Anybody who has seen what happened in Eastern Europe during centralized control very quickly builds an appreciation for market forces. Centralization is hard and ultimately produces incorrect quantities of what is needed, late, with higher coordination cost.

The wastefulness of the redundancy is a good thing for both pricing mechanisms and innovation. Competition brings down the price and determines winning technologies – I don’t mean to imply the correct ones always win, just that the market can generally decide better than a committee. . So although it is done many times the cost of each occurrence is probably less than it would be otherwise, and the natural selection makes for a better outcome. Even in standards the market system decides what makes sense – standards live or fail based on adoption.

At the simplest level, the innovation and creative destruction are what the American economy is based on. So what if my agency website is based on old technology and looks pants. That was local work done by local people with local tax dollars – at least they didn’t give the money to one of the vast mega-corporations from the beltway – who would most likely be needed for the scale of coordination needed, and wait for the committee to decide what I wanted from my agency.

(I published this on GovLoop as well where I originally saw your post)

I agree with most of what you say, but I think you have the unit wrong. It’s not the website, but the information/data. Free the information from the presentation layer–fonts, colors and UX–and you can mix it wherever the end user is.

Think about how you use the web. How many sites you return to? I bet you usually use a search engine, links from a friend, links in another site, wikipedia, to get to what you want. And what you want it to accomplish a task.

I am with Lewis (and Tim O’Rielly) on the idea of government as a platform. Authentication of data/information is important, and requires gov-wide standards. And I am all about making the language government uses more friendly to citizens, but one big CMS is a big mix of apples and oranges, of flu prevention and SEC filings.

Let’s free the data and info and have government join the conversation.

Thanks for your “making me think” post.

Anyway, I like the idea of a consistent, open-source platform. It makes sense for government. I like decentralizing because that is the nature of the collaborative technology solutions. I am not in favor to offer the public any govt employee who authors a report or document as a spokesperson for that content especially if it relates to policy.

Online communication: Internally: yes, less silos, easier ways to share information, more dialog. Externally, we need –as I’ve said before– to address the distinction between informal (G-C/C-G) communication around projects, programs, events, rules, legislation etc. and formal public comment on the same, especially as it relates to policy making.

Most govt agencies have no strategy or solution here or even appear to have policies that distinguish between the two. In other cases, attribution required of citizens speaking in conventional public hearings is ignored when that dialog is electronic, or online. Lacks structure and uniformity.

So while unified designs, web platforms, editorial content, and updated design practices are have merit in this discussion, also keep in mind the need to build collaboration models that incorporate structure and standards from 230+ years of democratic processes between and among government and citizens. Thanks for allowing me to comment.

I’d like to see better centralization and consistency of the tools and basic infrastructure (CMS?) in order to allow the appropriate agency subject matter experts to focus on the content. I echo, based on first-hand experience, Jeffrey Levy’s excellent 5% comments.

I can also see the benefit of more inter-agency content development to avoid redundancy and make it easier for users to find information without having to know which agency does what. I think the current major portals (science.gov, etc.) have not succeeded in this challenging task.

The importance of agency-specific branding shouldn’t be underestimated, though. When policy makers and citizens don’t know that an agency provides fundamental data and info that is relied upon daily all across the country, those agencies have a harder time getting funding to continue providing those data in the future.