Blog post

US Congressional Report Challenges Open Government: It Was About Time

By Andrea Di Maio | February 17, 2011 | 10 Comments

open government data

Those who have been reading some of my earlier posts about the Open Government initiative in the US Federal Government (see here and here for a start) know that I have been critical about the way open government has been implemented. The Open Government Directive had made it a compliance-oriented rather than a value-oriented exercise and has put too much emphasis on publishing data sets for the sake of transparency, or creating opportunities for external engagement, without asking fundamental question about the purpose this would serve and the target audience.

At the end of January a report by the Congressional Research Service “The Obama Administration’s Open Government Initiative: Issues for Congress” restated the exact same objections I raised in previous posts, some of which didn’t really make me many friends.

Although recognizing the potentially positive impact of open government, it stresses some of its downsides:

Conversely, Congress may find that increased transparency and public attention make the federal government more susceptible to information leaks of sensitive materials. Additionally, increased collaboration and participation may make the sometimes slow process of democratic deliberation even slower. Congress may also choose to evaluate the monetary costs associated with implementation of the open government policies

When it comes to data sets

Although the datasets released to the public may be useful in many ways, it is unclear how some  of them will increase the transparency of the operations and actions of the federal government. […] this type of transparency does not give Congress or the public  much insight into how the federal government itself operates or executes policies. […] Crowdsourcing may improve data and operations, but only for agencies that make useful data accessible.
Releasing these datasets to the public also assumes that the public will have the knowledge, capacity, and resources to evaluate the data, offer valid insights, and reach replicable results and verifiable conclusions.

On the latter, I said that open government is for as captive audience of usual suspects (lobbyists, corporations, NGOs, political parties) having a vested interest in data.


Irresponsible manipulation of the datasets may allow certain groups or  individuals to present unclear or skewed interpretations of government datasets, or come to  questionable conclusions. Moreover, users of this data will have to know exactly what datasets  they seek, especially in agencies that release hundreds of thousands of datasets. Counter  intuitively, agencies that release vast amounts of datasets may become even less transparent  because the public will be unable to decipher which data are important to their needs. […] Making the data public, in this way, does not necessarily make the data more  accessible or usable. Without the ability of the public to access and use the datasets that are  released, the government may not be more transparent

I did make many of these points since the very beginning of the open government initiative, and I have been warning other jurisdictions not to enter a competition based on how many data sets, or how many Facebook pages or idea contests they would run vs. other jurisdictions.

What very few have done, and should have been hardwired in the directive, is to link openness to value and mission objectives. There is still time, although the clock is ticking as the Congress – which is no longer as favorable as it was at the beginning of the Obama administration – starts looking more closely into this matter.

I hope that some of the people who judged my comment about the Deputy CTO for Open Government Beth Noveck leaving as disrespectful and ungrateful will see with greater clarity and less emotion why a change of course is sorely needed.

Comments are closed


  • Bowen says:


    I’m wondering if the problem is that Open Data and Open Government are being conflated. Open Data without explanation is useless. Without open access to government officials and without open process many of the issues that you’ve raised will continue to bedevil Gov 2.0.

    I do, however, disagree with your concerns about leaks of sensitive materials. Part of the culture change that must happen is to gain a true understanding of what is “sensitive” and what isn’t. Simply put, unless it’s part of an ongoing legal concern, contains personal, identifiable information or is rigourously classified, it should be available if government produces it.

  • Data for data’s sake isn’t a solution to any of our problems with government.

    Frankly, we’ve had so much data cheerleading in the gov 2.0 game that we’re missing how to play the game better.

    Saving Taxpayers Money
    -Governments are getting multiple requests a week for new data, data that companies want to resell and republish. Fine, but that’s taking resources and time from an already cash-strapped government.

    Consistency and Predictability
    -Data has been released in non-standardized formats, creating different formats for the same type of information across different governments, making it difficult to scale products.

    Greater Transparency and Citizen Access
    -Data, not documents, has been emphasized by the data cheer squad, so the public may be able to report a pothole or view crime stats on a map, but they can’t engage with the road repair budget, or see crime task force reports.

    Creating a greater open government community
    -The US has done very little to coordinate across governments, and rarely looks to international work to learn from. I was shocked to find that I was only one of two Americans at an international training on legislative data standards; I’m now coordinating work across governments to train them on Akoma Ntoso and find ways to share legislative processing software. –I feel pretty alone in this work.

    All of this has created a lot of data being under-used, a number of companies engaged in a data gold rush (which I predict will lead to a gov 2.0 bubble burst), and a lot of important government information (documents) being sidelined.

    I may not agree with the entire report, but it’s about time we think about open government as not predicated on data, but on access to the larger information environment of government.

  • Steve Ardire says:

    Hi Andrea – this is your best post ever on Open Government initiative

    My 2cents: unless you create meaningful, interoperable information you’re not going to get transparency or accountability on operating or executing policies nor have dynamic, agile applications that really engage citizens in issues with deeper insights.

  • Susan Hess says:

    Well done Andrea! I can’t add anything! Sarah, Steve well stated as always. Bowen as you stated there is a huge disparity between different agencies of what is construed as “leaks of sensitive data”. If commonly understood datasets, definitions coupled with federated security policies were adopted across agencies I highly doubt it would be as large of an issue.

    Just my humble opinion.

  • @everyone. Excellent comments.

    Open Government is alive and well, and has been chugging along improving how government functions long before it was hijacked by the Open Data movement, claiming Open Government’s benefits as its own. Now Open Data has been reviewed an found wanting. Let’s set it aside and get back to the task of Open Government.

  • John Sheridan says:

    Andrea continues to promote self-examination among those of us with responsibilities in the Gov 2.0 arena. This post about open data is no different. I wonder if I might further explore this issue by way of analogy.

    In various parts of Australia, there are semi-regular kerbside junk days. On these days, householders can leave certain items outside their house and register their involvement online. Interested people can then drive to these sites and pick out anything in which they are interested and take it away for free. Obviously, not everyone is interested and some stuff is never picked up.

    Open government data is a bit like this. Government puts it out there but only the interested pick it up. That’s not a bad thing. But as in my analogy, success should be measured by that which is picked up, not that which is left out. Leaving it out in the first place is the necessary precursor though.

    You never know what people will do with the stuff you leave out. Should we assume the worst? Will my old bike be used in a bag snatching, resold at an enormous profit by a junk dealer quick off the mark, or help some kid who doesn’t have one? Should I be prevented from leaving out anything that isn’t in every way safe? Should I put a label on the bike “Do not ride down hill at high speed while showing off to your sons so they think you are tough”? Should I have to leave a helmet out too? Or can I assume that the sensible majority know how to use a bike safely?

    One of the compelling arguments for releasing such data is that it has already been collected, collated and paid for. It has, in fact, been ‘used’ by government. The thing about data is that it isn’t diminished by that use (unlike the suffering suspension on my old bike). It’s being released because it may have additional, yet to be explored, value – value that its collators couldn’t fathom or didn’t have the resources to explore. To prevent that value being discovered is worse than leaving my old bike under the house because, while only one person can ride that bike at a time (ok, maybe two), many can use and reuse the data.

    Of course, leaving out the bike doesn’t tell people what I used it for, or for how long, or how fast I rode, or how far. Analysing its condition might help though. Knowing that I had a bike and stopped using it may make people who know me (or who are just curious) wonder why. Did I spend all my money on a new bike? Did I get lazy and stop riding? Did I have some awful accident and am trying to hide it? For those who need or want to know those things, there are other ways of finding out. And, if I am trying to hide the reason, then the bike is a clue. I’m not pretending that leaving out the bike is a substitute for answering these questions. Sometimes a used bike is just a bike.

    Opening up government data isn’t a substitute for opening up government. Recycling my bike won’t save the planet. But it’s a whole lot better than keeping it shut up under the house where no one can use it.

    (This is my private opinion. The views expressed are mine alone.)

  • @John – Thanks for your very thoughtful comment.I love your analogy and let me use it to explain why I am skeptical.
    A bike or any other object that people normally use and recycle through garage sales, kerbside junk days or – as they called them when I was leaving in Brussels – “braderies”, has a well defined purpose. A bike, a hammer, an old lamp all have a clear purpose and use case. Of course, as government is accountable, a few disclaimers here and ther e(such as “do not ride too fast or you may hurt yourself” pretty much like those on cigarettes may be in order). But my problem is that open data is not bikes or lamps. It is pedals, chains, bolts, spokes: it is raw material to build something that might be a bike or a lamp or something else. Pretty much like one of those complex Lego sets (like the Taj Mahal or the Tower Bridge) with special pieces: if those pieces are sold alone in a garage sale, they may seem purposeless, unless people get all the others needed to build the complete set.

  • John Sheridan says:

    Ah @Andrea, there you go making us think again.

    I still have all my childhood Lego. Unlike that of my sons, most of it wasn’t tied to specific models. It was more generic. After I had used it for its original purpose, it could be reused in many different ways, ways that Mr Lego hadn’t suggested or maybe even thought of. (Maybe it was open standards Lego – but let’s not go there, the bruises are still showing!)

    Even the pieces from those newer themed sets can be used out of their original context. It’s true that, without the instructions, that original context might not be able to be reconstructed but surely that reinforces the analogy – new uses for old components.



    (These are my own, personal views.)

  • @John – True, Lego pieces have a standard API that allows to connect them in unplanned ways. But their only purpose is to be connectible pieces and there is an entire value chain to get them to your kids’ playroom: packaging, instructional manuals, retailers and so forth. Open government data is like taking pieces that have not been build for the purpose of being connected, and throwing them to the public. The bike analogy is probably better: most people wouldn’t know what to do with spare parts, unless they have a clear purpose (manufacturing, selling, restoring, repairing bikes). And most people would not buy spare Lego pieces on a kerbside junk day, unless to complete a construction for which they are missing pieces; on the other they would probably buy a full set, if the price were better and the quality good enough.
    My point is that packaging open data into something that actually makes sense to somebody is key to their use. What most open gov plans seem to miss is the recognition that the only ones who have an interest in using data are intermediaries: they will buy spare bike parts or Lego pieces, and then assemble or package in that meet potential clients’ interest and purpose.

  • President Obama’s Open Government Directive tied together two conceptual processes, open data and open government, that have widely divergent methods and goals. Open Data is not Open Government (one being concerned with the Legos; the other with the supply chain). And Gov 2.0 is not Open Government (one being concerned primarily with tech-enabled collaboration and Web 2.0, the other with transparency). By conflating these three ideas, we have confounded analysis. And the Obama Administration (not “we”) has created a framework of data sets and plans that don’t necessarily move the ball for either open government or open data.