This week government 2.0 enthusiasts will meet in DC for the Gov 2.0 Summit organized by Tim O’Reilly with a quite interesting list of vendor sponsors and an impressive number of speakers from government and elsewhere. Most of the people who are driving and implementing change in the Obama administration will be either speaking or attending, and the #gov20 hashtag on twitter returns lots of tweets that witness great expectations for the event.
Tim O’Reilly, one the very first to spot the new Internet trends a few years ago and to introduce the term “web 2.0”, has been addressing government officials in DC and elsewhere with his vision of government as a platform. This has been circulating for a few months rather quietly, but now is vigorously emerging – in conjunction with the above-mentioned conference – as the new buzzword for “government 2.0”.
His views are summarized in a recent article on TechCrunch and is also available through an extended session hosted by the Federal Aviation Authority back in June. Many of the recent initiatives undertaken by the Obama administration as well as by governments around the world on web 2.0 are covered by Tim’s government as a platform: data.gov, recovery.gov, appsfordemocracy, different kinds of crowdsourcing, and so forth.
O’Reilly derives this concept from the technology industry, where the most successful companies have all been platform companies, enabling their partners’ and ecosystem’s success. Examples he often mentions are Microsoft, Google, Apple. Platform providers allow others to develop applications and move in directions that they would never think about. To some extent, this is true for the open government data movement too, where public data are made available for others (vendors, non-profit organizations, other government agencies or the public at large) to extract and create value.
He highlights some basic characteristics of platform organizations: they embrace open standards; they build simple system that can evolve; they develop for participation; they learn from their users, and especially those who do unexpected things; they lower barriers to experimentation; they build a culture of measurement; they celebrate developers; and they do not reinvent the wheel.
Undoubtedly most of the above characteristics do apply to government organizations that really want to catch the innovation wave triggered by web2.0.
However it seems to me that there are a few flaws in equating government to a platform. Let me examine some:
- Government operates in an inherently more regulated environment, while platform companies break new ground in domains that are either lightly regulated or not regulated at all. For instance, when looking at data, government struggles with regulatory distinctions between public data, data covered by freed of information, personal data, confidential data, and so forth.
- Government motivations are different from other platform providers. Technology platform providers ultimately want to make money: their partners’ and developers’ success determines how much money they will make. Government aims at creating “public value”, which is a combination of fulfilling its political mission, being efficient and providing services to its own constituents. Therefore the measures of success as well as the rules of engagement of those partners and developers are different, and more complicated.
- Government is both a platform provider and a platform consumer. Focusing on government as a platform provider misses the very important point of how government can use web 2.0 to leverage other platforms (e.g. data provided by citizens on a public social network) to discharge its own tasks (e.g. intercepting fraudulent behaviors, or protecting constituents in case of emergency). Reality is that government needs to run, source, and utilize a variety of different platforms in order to create public value.
- Government must remain a solution provider in domains where there is no business case for others (businesses or the public) to do so. Mashups and other cool web 2.0 examples abound in areas where mashers can create new services, new value, new business models. However most of what government does is stuff that nobody else is much interested in: child care, unemployment support, basic health care, public education, where often those who are the most frequent users are also those on the “wrong” side of the digital divide (least affluent, least connected, etc). While there is a lot that web 2.0 can do to better connect government with the non-profit sector, it is important to appreciate that being a platform may not be a priority in several domains for quite some time.
- Government is many different things at the same time to the same people. Government is an authority, a protector, a supplier, a democracy, and entertains with citizens all these relationships at the same time. This implies that it can be a platform provider, user, integrator, or just a unique solution provider, and all these roles overlap in such a way that it is difficult to determine what would be the sensible boundaries of a platform.
- Government remains accountable for anybody else’s mashups. Using the Apple or Google analogy, the liability for mistakes and glitches in applications built using their platform remain with application developers (unless they expose a problem in the platform). if something goes wrong with a mashup or “app-for-democracy” using government data that got a prize or some form or recognition by government, be assured that government will be criticized. So, will governments find themselves thoroughly testing and certifying third party applications? How does it work with O’Reilly’s suggestion to lower barriers to experimentation and permit failure?
These are just some of the reasons why viewing government as a platform is just too simplistic. For sure, some or most of the attributes that O’Reilly recognizes as key to “government as a platform” – such as openness, design principles, culture of measurement and so forth – remain very important attribute of a future government. What is required, though, is some more pragmatism about where such a model does make sense and can deliver value, and where it does not, and what other models can capture the complexity of government roles and responsibilities.
It is probably more accurate to say that “government has a platform” (and probably more than one), i,e. data, common processes and solution patterns it can share with others to create value.. But definitely it is not a platform.
The Gartner Blog Network provides an opportunity for Gartner analysts to test ideas and move research forward. Because the content posted by Gartner analysts on this site does not undergo our standard editorial review, all comments or opinions expressed hereunder are those of the individual contributors and do not represent the views of Gartner, Inc. or its management.
Comments are closed
16 Comments
Thougtful piece, detailing some very important and (to now) less explored constraints and incentives operating in the government realm. From what I can tell, though, there’s really no fundamental argument between your view and O’Reilly’s. He doesn’t flatly assert that “government is a platform,” rather he mostly advocates that government should act “as” a platform. As, not “is.” In fact, he appears to be limiting himself to be “encouraging more of this government-as-platform thinking,” which sounds far less sweeping and assertive, or definitional.
Of course government plays other roles… it’s just that in the technology area it hasn’t played this one well, but could and should, with positive results. I think you wind up agreeing with that at several points in your own piece. Anyway, good dialogue to be having, I believe, as it is in the very American tradition of hashing out the proper role(s) of government. I await Mr. Jefferson’s own blog….
@Lewis – Thanks a lot for your valuable comments. Indeed there is no fundamental disagreement and I have been pushing for the same thinking for the last two years, when addressing government clients about the benefits of openness and where the value of web 2.0 really is (and you can guess how tough it was before the last presidential elections :-).
This being said, it is very possible that by hyping the platform concept so much (indeed, government is also a platform, amongst many other things), people lose sight of how different its role is from that of a technology company. While the Apple or Microsoft or Google metaphors are attractive to exemplify how much value can be extracted from information that government sits on, that’s where similarities end. Besides what I’ve already written in the post, there is the fundamental un-coolness of government. Now, many say that with Obama and similar leaders, government will become cool again. I doubt it, beyond the few enthusiasts that are banging the drum of government 2.0. The Google or Apple or Microsoft ecosystems carry those brands as an essential component of their value: why do iPhone apps look cooler than – say – Nokia or Samsung apps although they are functionally equivalent? Government cannot win that game. On the contrary, I suspect that only when people forget about data.gov (and the likes) and whose data they are dealing with (the Weather service is a good example that O’Reilly has used a few times), value can be truly realized.
Further, public data provision is just a portion of what government does: for all the rest, such as dealing with personal or confidential data, or delivering essential services, the platform model is either very far away or very different (e.g. how about government using somebody else’s platform to deliver services or manage patient records?). But this would drive the discussion in a very different direction.
Much of this is spot on and a timely addition to the debate. Your final para is the critical one – some of government is and should be a platform. Other bits are not – such as administering and paying social security, adminstering policing etc.
But the conservative, security driven mindset from the bits of government that are not platforms discourages necessary experimentation with bits that should be more open. I appreciate why you have adopted the tone you have in your post, but your overall tenor might further encourage this mindset.
One debate that hasn’t emerged so far is the opacity and lack of democratic accountability of large multinational platform/cloud operators. Governments may be bad or incompetent but at least you can get rid of them or use other democratic routes when they fail. But even governments sometimes can’t make big multinationals behave.
Maybe the title of the article should be adjusted slightly to meld the 2 perspectives :
When government is not a platform.
At the gov2.0 expo/summit there are huge innovations being pitched. Don´t want to stifle the innovation,, just focus in the places where it is most valuable
Sorry but this post doesn’t resonate with me. It is a criticism of over generalizing by over generalizing. I feel that over criticizing amongst ourseleves pushes for needless separatism that devolved into argyements. Rather than mutual support, unity, eye on the prize that sets goals and achieves.
I hate being the lone voice calling us out of our bubble. Yes, Oreilly and SV are big and arrogant, but so are we. As a Californian I get this all the time personally too. This obsession with hype, fads, that many have is frustrating. What happened to getting the job done regardless of who is hyping, rather who is working on it.
“Government must remain a solution provider in domains where there is no business case for others”
please show me an area where an NGO, union, non-profit or other lobbying group is not involved. Does a group like Sunlight not count? This sounds like bubble talk to me. Don’t forget too that a lot of these groups have tendrils in both DC and SV, and they do great things.
“Government operates in an inherently more regulated environment”
Huh, do we really? I’ve worked in both and when it comes to data it’s nit all that different. Their are bigger diffs across industries than across the gov/priv divide.
A taste of irony insofar as political parties mount “platforms” to try to win elections. 🙂
While O’Reilly’s metaphor may not be necessary, it is possible to take up his perspective as another useful addition to the gov20 conversation.
Andrea’s critical remarks, though, are equally useful, providing a counterbalancing antithesis to O’Reilly’s thesis. I might suggest that instead of ‘simplistic’ we could try on the adjective, ‘overdetermined’. It’s usually predictable that one’s peculiar, highly localized position in the world will shape one’s views. Examples include the barber for whom everyone is a “head”, and the doctor who “diagnoses” world problems, medical or otherwise.
My angle on O’Reilly’s metaphor is that it might be innocently idealistic and perhaps out of step with a zeitgeist of hyper-dependencies on others to solve our problems. In the age of the “personal” fitness trainer wherein we even hand over our own bodies to someone else for domestication, the suggestion that people left to their own devices will take the data bull by the horns to envision, then materialize their own social visions seems naive.
For all that, some enterprising people will indeed mash the bull and it would be a shame to fail to provide the enabling tools. For me, the thrust and parry of the gov20 debate overstands any magesterial judgment of what may be wrong, may be right. Democracy invites pluralism as a resident value and this debate is an exemplar of that.
@william – You are right, there is always a risk that calling for caution gives more ammunitions to luddites and “security freaks”. But I guess there is already enough enthusiasm around for my own little call for caution not to cause any problem.
Your point about cloud providers is well taken, and it is something that GSA and others working on using cloud services will have to take into account when striking the balance between using internal resources (more or less virtualized) and moving more aggressively to cloud-based infrastructures and platforms owned and managed by vendors.
Let me also make a last point about vendors (hardware, software, IT services, IT consulting, management consulting, etc.). I have not seen many people commenting about what government 2.0 mean to them: it seems to me that this new experimentation-oriented, technology-savvy attitude shown by the federal administration is creating tremendous opportunities at a moment in time when the rest of the economy is not doing terribly well. So why should they caution their government prospects and clients?
@Steven – thanks for your reflections – I believe that controversy is beneficial to better understand pros and cons.
As far as your two points:
“please show me an area where an NGO, union, non-profit or other lobbying group is not involved. Does a group like Sunlight not count? This sounds like bubble talk to me. Don’t forget too that a lot of these groups have tendrils in both DC and SV, and they do great things.”
Of course there is a lot that NGOs, businesses, people organized in social networks as well as individual employees in government can do by using social software. But there are areas where replacing government is either too difficult or too dangerous or simply unprofitable.
Let me give a few examples. Unemployment support: who else than government should pay people while they do not have a job? Criminal justice: would you like your destiny to be crowdsourced, should you ever be on trial? Any kind of service to any disadvantaged community: sure, you can pull resources from the voluntary sector, but who do people go to when they are in need? Government is the service provider of last resort in several domains, but it also ensures – because of the democractic processes we have in place – that there is sufficient (although imperfect) equality and transparency in dealing with a variety of issues and services. Indeed a lot of other stakeholders are and will remain involved, but the balance between their contribution and influence and government accountability varies across service domains and policy areas.
“Government operates in an inherently more regulated environment – Huh, do we really? I’ve worked in both and when it comes to data it’s nit all that different. Their are bigger diffs across industries than across the gov/priv divide.”
I agree to some extent, and I used data as an example. Processes and behaviors are more regulated, because all government does is to implement policies and regulations. Incidentally, I have also worked in and outside government, and I definitely see a difference. It is true that incidents happen (how many CDs or USB stickes with personal data were lost by government officials around the world?), but this has to do with how policies are implemented and not with how many there are.
@bob – I fully agree, let’s not use discussions about what government is or is not to create barriers to experimentation. But, equally, let’s make sure that there is sufficient understanding of where early experimental successes can be scaled up,fully deployed and replicated. I am sure you remember the early days o e-government, where a “best practice” (btw, I hate the term) was to buld a web site for an agency or a department. People were saying “let’s build them and they’ll come”. As we all know, many of those web sites were underutilized, or even totally ignored by people. Unless we identify both early successes and early failures, unless we understand the ingredients that make a gov 2.0 seed trigger sustainable innovation and create sustainable value, we are still all playing in a sandbox.
Totally agree that characterizing government as a platform, or even suggesting government act as a platform, is too limiting.
Maybe better to imagine government as an engine/enabler/contributor to an ecosystem for the provision of services and information to the public (and segments of it)?
Great job, Andrea, I´ve just linked this post into “Open Government Europe” group in LinkedIn.
Andrea, you make some very good points. And of course the issues are too complex to fit into a single Techcrunch or Forbes article, or a 45-minute speech, or even three days of conference. A metaphor is just that: a way of framing the issues such that people can see something they might otherwise miss.
Clearly, there are many functions of government for which the “vending machine” is just what we want. We want reliable, repeatable services, and we’re happy to pay for them.
But there are many other functions of government for which the generative power of platforms is the right metaphor, and for which we can greatly profit by understanding the difference between good and bad platform design.
And it’s not just about data. For example, the interstate highway is a platform, one that has fostered commerce and shaped our society. Public transportation is a platform. So is urban planning. Cities like Portland that have done a good job with both have reaped large benefits.
Police, fire, and other public safety functions — not to mention the rule of law — are platform functions. You have only to go to lawless countries to see how the private sector fails to function when the public sector doesn’t do its job, as a platform, of enforcing clear rules.
The key point is that if government realizes that it is a platform provider, it may make different choices. Its goal is not to crowd out the private sector by providing competing “applications” of its own (though it may launch applications that are core to the functioning of government or that raise the bar), but to set and enforce clear rules for how “applications” interoperate, to provide foundation capabilities on which both the private sector and other agencies of government can operate, and to make sure that bad actors don’t crash the system.
Metaphor, not reality. Government isn’t a computer platform. But it can learn a lot from the best practices of computer platforms. The most powerful platforms lead the way by showing new capabilities, but they don’t compete with their application ecosystem. (Again and again, when they do, we see them become less vital.) The most powerful platforms encourage what Clay Shirky called “good surprises.” (See http://gov2summit.blip.tv/file/2591940/)
One of my key goals with introducing this metaphor was to start the search for policy decisions with large downstream consequences. The example I have used was the Air Force decision to allow civilian use of GPS data. The Air Force built the platform for their own purposes. They could have kept it for military use only; but because they released it for the private sector to build on, we’ve seen huge benefits for citizens, and multibillion dollar industries built on the back of that government platform.
(I won’t do more than mention the all-too-familiar example of government investment in the Internet, originally a research network, or the Human Genome Project, which helped keep genomics from being the patented preserve of a few companies.)
Thinking government is primarily a service provider has gotten us to the point we are now, with nearly 40% of GDP devoted to government; thinking how government can be a platform provider might just help us figure out how to keep that from going to 50% or more, while actually providing more services.
Being a platform doesn’t mean abdicating responsibility for the functions of government. It means thinking differently about how best to perform those functions.
“Government operates in an inherently more regulated environment.” That’s not immutable. Improving our view of what government does for us should inform our decisions about what regulations we should create. Not the other way around.
“Technology platform providers ultimately want to make money: their partners’ and developers’ success determines how much money they will make.” False. Value created for consumers will determine the success of a platform provider. Same for government-as-platform.
“Government is both a platform provider and a platform consumer.”
This is only a criticism if you continue to think of government as One Thing. It’s many things, just as a jungle is the plants, animals and landscape that comprise it.
“However most of what government does is stuff that nobody else is much interested in:” Nobody is much interested in competing with government, it doesn’t play fair. Government-as-platform would change that equation greatly. Solutions might even be found in certain domains that don’t require the government platform at all.
“it is difficult to determine what would be the sensible boundaries of a platform.” Government wouldn’t provide a single platform. Government today isn’t an individual animal, it is a verdant jungle. Government-as-platform would be no less complex. Each “business unit” would provide services that make sense for them. _Bionomics_ was a while ago, that metaphor should be foundational by now.
“Government remains accountable for anybody else’s mashups.” False. Government is pretty good at disclaiming responsibility in a wide variety of areas.
@Tim O’Reilly –
Tim, thanks for your response. I do not think our viewpoints are so different. What I am most concerned with is to make sure that both gov 2.0 enthusiasts and skeptics get a balanced view of how new levels of openness and trasparency made possible by technology help government better perform their function, but do not necessarily imply that governments give up those functions.
You correctly point out that governments have always developed infrastructure that has worked as an enabler for others (society, businesses) to thrive: roads, public transportation, the Internet are all good examples. The fundamental difference with the platfom we are taklking about today (i.e. data), is that in all those exampels government was a first mover, filled a gap that nobody else was in a position to fill.
Data is different. There is already an abundance of data, from government, private sector, NGOs and – most importantly – people. Governments sit on a wealth of geographical information and map, and yet people use Google Maps or Virtual Earth, which are not from government already. Public museums publish pictires of their masterpieces, and yet pictures taken by people are already in flickr. Governments publish ratings on hospitals and clinics and yet there are already plenty of those on social networks.
One problem with your analogy is that in all previous cases where governments have developed platforms, the barriers to entry and the capital investment required were huge. The Internet has changed that equation, so each of us becomes part of one or many platforms.
This is why I said that government has a platform (and actually more than one) rather than being a platform: the challenge is to find the right role among many other platforms that already exist and – most importantly – will emerge, and how to be perceived as a trusted, authoritative platform, in a conundrum of platforms and data that will represent multiple versions of the truth.
I do also agree with your point that one should not look at government as a service provider only, and this is the point of my later post about “What government is and what it is not” (http://tinyurl.com/m8ocu7). Government is many different things at the same time to the same people: the role of data and platforms (both government’s and others) need to be understood in the context of those different and overlapping relationships between government and its constituents.
The bottom line is that talking about “government as a platform” has been a useful contribution to provide alternative perspectives, but its expiry date may be approaching. My contention is that – after doing pretty much the same things over the last two years (conferences, unconferences, mashup contests, open data deployments) – even those governments that are leading the “gov 2.0” charge need to reflect on the fundamentals of governing and ask themselves the hard questions about real, sustainable benefits of government 2.0.
One example of where the discussion should be going is about looking at basic government services and understand how they can be a platform (or use other platforms) to create constituent value and increase operational efficiency. This will also call for tackling the thorny issues of how to identify people, how to deal with personal data, how to enable citizens to control their own privacy. It is time to move on from open data and look at open services. Also in this case, though, the real issue for government will be to strike a balance between being a provider, a consumer and an integrator.
@Scott – thanks for your comment.
YOU SAY:
“Government operates in an inherently more regulated environment.” That’s not immutable. Improving our view of what government does for us should inform our decisions about what regulations we should create. Not the other way around.
MY REPLY:
Do not forget that government is people. We want it to be regulated to make sure the same rules apply to all. Further, one should get into the difference between the legislative and the executive branch. I do agree that people can be better involved in legislation, but I would argue that we need an independent and well-regulated body to execute those laws.
YOU SAY:
“Technology platform providers ultimately want to make money: their partners’ and developers’ success determines how much money they will make.” False. Value created for consumers will determine the success of a platform provider. Same for government-as-platform.
MY REPLY:
I do agree about the value. However for technology platform providers (name any) this turns into revenues. I do think this is true (and the is nothing wrong with that). I would also argue that public value is a combination of value for citizens as well as political return for those in charge. I have not seen anybody touching upon how government-as-a-platform would change depending on which coalition is in power (and I do live in a country which is almost a case study for the relationships between government and information).
YOU SAY:
“Government is both a platform provider and a platform consumer.”
This is only a criticism if you continue to think of government as One Thing. It’s many things, just as a jungle is the plants, animals and landscape that comprise it.
MY REPLY:
This is not meant to be a criticism at all. The fact that government is many different things to the same people is very true, and one I’ve made that point even more explicitly in my next post.
YOU SAY:
“However most of what government does is stuff that nobody else is much interested in:” Nobody is much interested in competing with government, it doesn’t play fair. Government-as-platform would change that equation greatly. Solutions might even be found in certain domains that don’t require the government platform at all.
MY REPLY:
Din’t we say that with e-government (1.0) alerady? Web sites and portals should have increased people’s interest in and connection with government. Do you think that worked? And, if it didn’t, why not?
YOU SAY:
“it is difficult to determine what would be the sensible boundaries of a platform.” Government wouldn’t provide a single platform. Government today isn’t an individual animal, it is a verdant jungle. Government-as-platform would be no less complex. Each “business unit” would provide services that make sense for them. _Bionomics_ was a while ago, that metaphor should be foundational by now.
MY REPLY:
I wholeheartedly agree on this. Which is why I do not think a single platform for data )such as data.gov) is the right way to go.
YOU SAY:
“Government remains accountable for anybody else’s mashups.” False. Government is pretty good at disclaiming responsibility in a wide variety of areas.
MY REPLY:
Maybe people would not sue government for incorrect mashups, but reputation and brand would be seriously damaged, which turns into loss of political capital. Do not forget that governments are the lenders of last resort (as the current crisis has abundantly shown), and when things go wrong people will always turn to them.
Great discussion, if for no other reason that the additional insights from Tim O’Reilly. I found his response compelling.
Whenever any set of ideas gets a label or a brand – ‘eGovernment’,
“Government 2.0′ – you can be sure those ideas are bound to lose some of their value. What happens is that people trade in the brand and forget to keep in mind the complexity of the ideas and concepts. Perhaps it’s an understandable shorthand, but risks degenerating into the kind of simplistic either/or discussions which Andrea’s initial post was trying to resist I guess. So in that sense, a useful addition to the discussion which all too easily can get distracted by slogans and rhetoric (this is all about politics, after all!).
There is a very big issue underlying this conversation which i think needs much more scrutiny. And it is this – that we are witnessing some profound adjustments in the definition of what it means to be “public” or to be engaged in “public” work. It is worth reflecting that there is an answer to Andrea’s earlier question about unemployment benefits. He noted that as a provider of last resort, government is the only one who can pay people when they are out of work. It simply isn’t true. There are those who will argue strenuously that the welfare state and the advent of large ‘industrial’ government replaced a thriving and growing mutual movement. Now, the discussion will inevitably get around to questions of scale and scope, that’s true. But the point remains true – there are other ways to be ‘public’ than to rely on the the state.
So one of the intriguing ‘back to the future’ possibilties inherent in the Govt2 discussion, I think, is the opportunity to reinvent models of ‘public’ action which are no owned and controlled by the state but which harness the ‘small pieces’ of civil society action which can be ‘loosely joined’ with others to create a perfectly good ‘public’ system without being run by a government agency.
There may be other examples, but the key point is to avoid (much like the discussion about regulation) the risk of assuming that our current and relatively recent notions of ‘public’, which we tend to conflate with ideas of ‘government’, are themselves not fixed. It may be that the platform-enabling potential of the new social technologies are acting as a solvent on some of the assumptions about whta government and what it should do which we sometimes asume are rusted on.