There were some lively comments on my previous post, which was really only half-finished, so I’m going to write the other half here. But first, let’s get a few things clear.
I like transparency in government. I think it’s great that people campaign for a more open government, especially here in Canada. I am impressed with the kinds of things some people manage to achieve by scraping through government data.
Got that? OK. Now here is what’s wrong with Government 2.0.
- The rhetoric of citizen engagement too often masks a reality of commercialization (last time)
- Information is not always democratizing.
- Information is not always the problem.
- Transparency is an arms race.
- Privacy is the other side of the coin.
- Money flows to Silicon Valley.
So, one at a time.
Information is not always democratizing
In comments on the previous post, Kevin Donovan and Fazal Majid both pointed to this article by Michael Gurstein, who asks “who is in a position to make “effective use” of this newly available data?” and answers himself
‘open data’ empowers those with access to the basic infrastructure and the background knowledge and skills to make use of the data for specific ends. Given in fact, that these above mentioned resources are more likely to be found among those who already overall have access to and the resources for making effective use of digitally available information one could suggest that a primary impact of “open data” may be to further empower and enrich the already empowered and the well provided for rather than those most in need of the benefits of such new developments
It’s worth reading the whole article, and the example of land ownership record digitization in Bangalore, which “allowed the well to do to take the information provided and use that as the basis for instructions to land surveyors and lawyers and others to challenge titles, exploit gaps in title, take advantage of mistakes in documentation, identify opportunities and targets for bribery, among others. They were able to directly translate their enhanced access to the information along with their already available access to capital and professional skills into unequal contests around land titles, court actions, offers of purchase and so on for self-benefit and to further marginalize those already marginalized”
In a recent talk, Kentaro Toyama asks his audience (at the 8 minute mark), “You and a poor rural farmer are each given a single e-mail account and asked to raise as much money as you can for a charity of your choice. Who would be able to raise more money?” The answer, of course, is that “you” (like me probably urban, western), because we know wealthy people, we have lots of friends with e-mail accounts, we are literate and have lots of experience writing e-mails, and so on. The technology is identical for each, but the result is different because of the context in which it is being used. The Internet does not democratize. Instead, technology amplifies people’s ability to get things done, and the more ability you start with, the more you get.
How would this work with Open Government? How about the "Open311" services (http://open311.org) which let citizens report non-emergency issues to local government. The archetypal case is potholes that need fixing. If these services come into widespread use, they will be pushed to skew services towards neighbourhoods where people report most frequently (else why implement the service?), which will be the smartphone-owning home-owning better-off neighbourhoods. The better off get better services.
Information is not always the problem
The pothole reporting case is an example of this too. Round the corner from my house is a very rough patch of street that has been that way for a couple of years. The city knows it’s in bad shape. Why isn’t it being fixed? Because there are other, busier stretches of road that are higher priority. Information is not the problem; resources are the problem. There is no group of city workers sitting round waiting for a call about a pothole so they can go out and fix it.
In a different vein, it’s worth reading danah boyd’s talk about unintended consequences of transparency when not accompanied with interpretation. Raw data such as the registered sex offender list certainly counts as “open government” but she retells an expose from The Economist which began
with the story of Wendy Whitaker who was arrested in 1996 at the age of 17 for having oral sex with a classmate three weeks shy of his 16th birthday. She was convicted of sodomy against a minor, ended up in jail for a year and is now listed on the registry. "She sees people whispering, and parents pulling their children indoors when she walks by." Not only did she have to pay the price for her teenage indiscretions by going to jail, but she's forced to deal with them day-in and day-out for the rest of her life. Because of the registry. I wish that I could say that Wendy's story was rare, but I hear stories like this over and over again. People's lives ruined because of the registry.
Seeing the problem as one of information can lead us down the wrong path. As boyd says:
In focusing on the first step – transparency or access – it’s easy to forget the bigger picture. Internet access does not automagically create an informed citizenry. Likewise, transparent data doesn’t make an informed citizenry. Transparency is only the first step. And when we treat transparency as an end in itself, we can create all sorts of unintended consequences.
Transparency is an arms race
The Open Government initiatives are getting lots of information out into the public that wasn’t effectively public before. That’s good. But now that the information about, say, voting behaviour and donations is open there is a clear incentive for donors and donation recipients to muddy the waters. Intermediaries will appear. Money will be exchanged not as donation but in some other form. I’m sure there are endless ways of working the system, now it needs to be worked. Systems respond to changes, and in this case a move to transparency will be met by a move to obfuscation.
Privacy is the other side of the coin
There is one essay in the Open Government collection (by Jeff Jonas and Jim Harper) that addresses privacy concerns. Once data is targeted to be made public, it becomesimportant that improper or inappropriate data not be bandied about. So they recommend such strategies as limiting backups (!), destroying old data, “minimal disclosure in transfer between projects” (which goes counter to the Gov 2.0 direction) and more. It’s almost like the arms race, except instead of mischievously hiding data that could harm yourself, IT professionals are dutifully hiding and destroying data that could harm others.
Data anonymization and aggregation are often cited as ways of handling these issues, but recent developments in “re-identification” have shown that such efforts may be doomed. The identification of individuals within the Netflix Prize data set (paper) led to the cancellation of the second Netflix contest, and has cast a chill on crowdsourcing contests. Legal scholar Paul Ohm’s recent long but highly readable paper on Broken Promises of Privacy suggests that we need to re-assess the basis for many of our data privacy laws because information previously thought to be anonymous has now got to be treated as potentially privacy-damaging.
Bill Schrier, CTO for the City of Seattle, addresses privacy in an essay in the Open Government collection. As just one example, many elected officials maintain lists of email addresses for use in contacting constituents, and these lists are a common target for public disclosure requests. It is difficult to prevent misuse of these lists for spam emails and other inappropriate uses. Open Government is good, but receiving penis or breast-enlargement emails as a result of emailing your councillor? Not so much. Other examples he has come across include a chilling effect on grievances because complaint investigations are public records, and the city being forced to provide a complete list of full legal names and dates of birth to a local radio station, “two of the three pieces of information needed to steal employee identities”. Schrier is an open government advocate, but making this data public is “not a trivial or inexpensive task” if it is to be done with care. As he asks, “most of [the problems] can be overcome. But do we really want to make it that easy for everyone to obtain and use that data?”
Money flows to Silicon Valley
I’m not quite sure how to articulate this final point, but as I haven’t seen it elsewhere I’ll give it a go. Let’s take an example. Google and the City of San Francisco developed a format the General Transit Feed Specification, which “defines a common format for public transportation schedules and associated geographic information.” Using this, cities can push their transit schedules to Google for publication on Google Maps.
Useful, for sure. I love to be able to plan trips and Google Maps is a convenient way to do it. But there is a downside, which is that a purely local transaction, of me planning a bus trip in my home city, now involves sending money (via advertising) to California. The winner-take-all nature of the Web means that an increasing number of apparently local exchanges are done via California, such as personal ads via Craigslist. Is that a good thing for the local economy? How would we feel if it weren’t Google but a Chinese company that got the information and the money along with it?
So am I opposed to Open Government? No, but I’m far from convinced that the digital approach to the problem will actually lead to an increase in openness.
Good points, Tom.
You're a little more pessimistic than I would be. But I'd like to make one point about your first assertion: "Information is not always democratizing."
Yes, true. But it strikes me that the same thing can be said of the deployment of any technological framework; its use can only be made by those who have the skills and the cultural capital to access it.
There's a case to be made for paternalism here. So yes, the most underprivileged will probably not be able to use Government 2.0. But their well-wishers could and by that I mean NGOs, watchdog groups and so on. So there's a possibility here that Gov 2.0 will make it easier for these groups to do their job, which in turn may end up helping the underprivileged.
Posted by: scritic | November 17, 2010 at 09:03 AM
I'm glad you're continuing to write about Open Government -- I also find myself frequently turned off by the uncritical cheerleading that can occur in the OG space.
I think you're overstating a few points, though. Not to nitpick, but GTFS is a poor example: it really is an open standard, and Google did us all a favor by herding the world's transit agencies toward a common format. Here in Washington, DC, the local transit agency releases its data in GTFS, but (last I checked) hasn't yet reached a deal with Google Transit. But as a developer, I can use their GTFS data for analyses or to build my own applications quite easily. In fact the local transit policy community is already benefiting from this data.
This is not to say that public/private partnerships deserve more careful consideration than they're currently getting -- people don't seem to understand that we've been down this road before -- just that GTFS isn't a great example.
The point about democratization is an excellent one, but again, I would quibble with your choice of example. Telephone penetration really is pretty good in this country. I don't think that adding new interfaces to service systems necessarily disenfranchises those who use legacy systems. In some cases, they can help make sure that that doesn't occur: DC geocodes all service requests, and can track how well they're dealt with on a ward-by-ward basis. The stats show whether the poorer parts of the city are being treated fairly.
But this isn't to deny that the effect can exist. I've never been very happy to hear that legislators are engaging directly with their constituents through Twitter, for instance -- you can be sure that back at the office a large staff is robotically dealing with a huge backlog of communications from constituents who are on average older and poorer than Twitter's userbase.
Posted by: Tom Lee | November 18, 2010 at 10:22 AM
scritic - I agree with you about technology, and in general I guess I'm a social constructivist rather than a technological determinist. At least, most of the time. And to the extend Gov 2.0 does let watchdog groups etc do their job it can help - and more power to these groups. My concern is that commercial interests may be more able than watchdog groups to use the information, but we'll see.
Tom Lee - It is very helpful to get a view from someone more involved in doing the actual work than I am. I'll definitely read up about GTFS, and readily admit I may have been hasty.
Posted by: tomslee | November 18, 2010 at 09:29 PM
Another good reference on this is Lessig's 'Against Transparency' which I think was unnecessarily derided when he published it: http://www.tnr.com/article/books-and-arts/against-transparency
As to money flowing to California: Steve Song has made similar points with telecoms in Africa, but I'm not sold it is as strong of a case against open government because this data was not available previously, so it isn't as if Silicon Valley is replacing previous companies: http://manypossibilities.net/2009/07/mobile-operators-and-blue-gum-trees/
Furthermore, Silicon Valley's failure to expand to China shows that local context matters. I'm not sure how much, but it can definitely play a role in limiting the imperialism of Californian firms.
Posted by: Kevin Donovan | November 19, 2010 at 05:42 AM
I agree about Lessig's article. I suspect the reaction was because it was not what his regular audience wanted to hear.
I guess Silicon Valley is not strictly speaking replacing local companies in that example - but I do think it's a general phenomenon of the Web: Amazon replacing local bookstores, Craigslist replacing local advertising outlets etc. And while context does matter, particularly at the borders between languages, there are fewer language barriers than national barriers, which gives web companies a bigger part of the world to operate in than most brick-and-mortar outlets. But those are very general trends, so I'm sure there are lots of caveats and qualifications not explored.
Posted by: tomslee | November 21, 2010 at 09:14 AM
There's a lot to say here, so I'll just go section by section.
Information is not always democratizing
Democracy does not entail economic equity, and nobody is claiming that Gov 2.0 will fix corrupt or problematic governments. The fact that this information can serve to highlight these issues puts the bell on the cat - exactly a good thing.
The logic of the 311 argument is simply invalid. Reports depend on the the absolute numbers of people with 'net access, not the relative number, wealthier areas have higher penetration rates for tech, yes, but there are far fewer of them to spot and report problems.
While resource scarcity is the primary bottleneck to service delivery, it is unlikely to hinder service delivery to have extra information. You could argue the benefit is small, but so is the cost. The benefit is probably greater than the cost and so it is worth doing.
Information is not always the problem
Not only did she have to pay the price for her teenage indiscretions by going to jail, but she's forced to deal with them day-in and day-out for the rest of her life. Because of the registry.
This case subject to the same critique as it's trying to support. The information wasn't the problem, the faulty legal system that allows such an offence is the problem. If legal system was not at fault, then she is rightfully justly listed.
Access to that information allows us to know about that kind of injustice and take action to prevent it, if it was indeed an injustice. Had that list not been available, whatever system-fault that put Whitaker in jail would remain unnoticed.
Transparency is an arms race
This is wrongheaded in several ways. If corruption is an inevitability then increased access can't do any harm - you'd have to tell a pretty fancy tale about how enhanced information would make corruption worse.
Also, it requires a bizarre mental gymnastics (which is unmotivated by anything other than vaguely cynical conjecture here) to assume that increased transparency will ipso facto cause increased obfuscation.
Privacy is the other side of the coin.
When discussing privacy, we have to be explicit about our standard. If you're saying technology is only acceptable iff it improves privacy relative to the state of affairs without that technology, that's an unprecedented standard.
We have phone books, credit records, etc. which are readily available sources of identifiable information. Claiming that the Netflix case put the users at an unreasonable threat to their privacy by dint of insufficient care is unconvincing- if it takes a team of computer science specialists to configure an algorithm with pathetically low accuracy to de-anonymize the data, your Netflix data is safer than your credit card number, SIN etc. The existence of the Stopwatch Gang does not prove that Canadian banks are insecure.
Identity theft etc. is and was primarily an offline problem.
Money flows to Silicon Valley
This is pretty unadulterated anti-globalization rhetoric, I don't think much else need be said here. If you don't like globalization and think its effects are bad - opposition to Gov 2.0 shouldn't even register as a concern. This problem, if it is a problem, does not begin or end with making government data available to citizens.
Posted by: Jeremy Vernon | November 26, 2010 at 11:03 PM
Jeremy - Thanks for taking the time to contribute.
If there is one aspect in which I'm "opposed to" Gov 2.0 it is what I wrote in the previous post - that the initiative and effort of civil liberties activists (like yourself, a bit of googling suggests) is being used by some to promote this thing called "Gov 2.0", while the reality is that a number of large commercial outfits are ready to make a bundle off it. It is phrased (by Tim O'Reilly and others) as a social movement, but O'Reilly and others have a commercial agenda, and I think that should be a concern for activists whose efforts are being invoked in the rhetoric.
As for the points on this post, I'd respond to some of your objections by saying "look at my 2nd para". But here goes on some of the others.
"nobody is claiming that Gov 2.0 will fix corrupt or problematic governments" - Well, Don Tapscott and Anthony D. Williams are,for two.
"The logic of the 311 argument is simply invalid" - it seemed a natural extension of the real-world example of land records in Bangalore. If my own extension is not very convincing, well OK - but I think the mechanisms Gurstein sets out are ones that Gov 2.0 advocates need to think about, and in the writings I have seen they are just not even considered.
"Had that list not been available, whatever system-fault that put Whitaker in jail would remain unnoticed." - So you are saying that because Whitaker's suffering has been magnified (as a result of making the list public and widely available), there is increased pressure on fixing the problem? Sounds a bit like what you are accusing me of - the worse it gets (for Whitaker), the better it is (for chances of fixing the problem). I am not convinced.
Talking of which "increased transparency will ipso facto cause increased obfuscation" - what I meant was "will cause increased efforts to obfuscate". What the outcome of the arms race will be, who knows? But I think the early examples of success may not be reproduced in the future as the other side accommodates itself to the new reality. It doesn't mean the efforts are worthless, but they are a long slog, not a magic bullet. This goes along with much of the thread that Evgeny Morozov is most outspoken about, concerning the possibilities for digital activism in repressive regimes.
On privacy "we have to be explicit about our standard". I agree, and I'm dismayed by how little attention the issue of privacy gets within the rhetoric of Gov 2.0. So maybe we are on the same side there.
As for the final point - it is the most tentative, and I think I need to crystallize it a bit more. It was a bit premature to include it here.
Posted by: tomslee | November 27, 2010 at 09:03 AM
Much to consider in these last two posts, Tom. Critique is an important part of civil discourse, particularly online, and I'm glad you've taken the time here to consider open government and Gov 2.0. That said, I would take issue with the record, as you've presented it:
1) Privacy is not overlooked, at least where I'm reading. I've written a fair bit about the topic recently and was a major focus of the Gov 2.0 Summit this fall. The talk by danah boyd you cite and link was, in fact, delivered at the Gov 2.0 Expo. I see and hear a lot of thoughtful commentary on privacy and the broader consequences of open data, not least because of crossover between IT security, health and government audiences in audiences.
2) Similarly, I thought the "commercial aspects of Gov 2.0" were clear long before I began publishing at O'Reilly. Several of my recent stories have focused on that aspect of Gov 2.0, in fact. White House open government officials like Kundra, who has framed open data as a lever for accountability, civic utility and economic value creation. As that seems to be a major thread in your critique, perhaps you might reconsider your frame here. Do you think the government releasing weather or GPS data for use in the private sector is "privatization?" Or health data? If so, you are naturally welcome to that position but it is not the definition that I understand from the debate around privatization of aspects of the military, or of proposals for Social Security.
3) Your other premises, expressed in that list, deserve the consideration that several comments here already offer. As to whether it's a good thing for companies to "make a bundle off of something" or money to flow to the Valley, if that activity creates new jobs or reduces the cost of services, I think many people would answer yes. Whether open government data flows in that direction will continue to be an open question, as are so many issues in this area, though there are a few early case studies to throw in the mix.
4) As to whether Wikileaks is open government, I wrote that post on Sunday to work out the question for myself and in hopes of providing a lever for discussion. Some people say it is. Others do not. Not acknowledging that would be intellectually dishonest. It's certainly not the White House version, but then honest debate will reflect the views of Jarvis, Sifry or activists elsewhere that expand that definition. If I didn't put forward a strong position then, know that it's because I wanted more time to reflect. Since you seem to favor slower reactions here, I hope you'll endorse that process.
Posted by: Alex | December 06, 2010 at 06:13 PM
Alex - thanks for the thoughtful response. Looking back, my post yesterday does read more harshly than I intended concerning your reaction. I did not intend to criticize a pause for reflection.
Rather than debate (because I'd like to go and think about your response and others') maybe I can say a few things that prompted my post. My reaction concerning privacy came mainly from reading the collection of essays in the "Open Government" collection, in which privacy concerns featured in perhaps three of 33 essays, and from some other reading I have been doing on the issue of privacy, particularly Paul Ohm's "Broken Promises of Privacy" paper (http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1450006). Concerns about privatization came partly from praise (that I do link to) of the "Big Society" program in the UK, and presentation of that initiative as in the spirit of open government. Your colleague Nat Torkington has concerns about that too, at http://radar.oreilly.com/2010/08/gov-20-as-means-not-end.html, and I think those concerns need to be worked out.
Posted by: tomslee | December 06, 2010 at 10:34 PM