yourlibrarian: SamatWork-no_apologies_86 (SPN-SamatWork-no_apologies_86)
[personal profile] yourlibrarian posting in [community profile] sim_studies
I attended the following presentation this week and found the discussion of needed research in the area of increasing the usability of online communities interesting. So I thought I'd post the write-up here in case others might find it of interest.

IT Research Challenges to Advance Open Government – Marti A. Hearst, UC Berkeley School of Information, iConference 2010

Hearst recently authored Search User Interfaces, and is currently on leave to work for the Obama administration. She began with some disclaimers about not representing the government and said she will be speaking about areas that are not her field of expertise. Government 2.0 is supposed to be a way of doing government with modern technology, and redoing parts of government that intersect with IT.

She wanted to speak about three main areas that are currently open for innovation. They are:

New technologies for public/government communication.

Open data and other technologies for government transparency.

Cost savings and efficiencies from new approaches to using technologies. This would include cloud computing and other current developments.

She began with an overview of government structure. In terms of employment, most of the workers are in the Exec Branch, only 1% are in the Legislative or Judicial Branches. So most of the IT spending occurs in the Exec branch as well.

Government agencies are like large public universities. They serve the public, are relatively autonomous, slow to change, and similar in structure and operation. There is lots of replicated functionality, among the agencies and sub agencies. And universities, like agencies, depend on government revenue for their funding.

The federal government is among the nation's largest employers. There are 15 executive cabinet agencies, with 1.9M employees. There are 70 independent agencies with about 180,000 employees. Additionally, there's the Exec Office of the Pres (EOP).

There is no Agency of IT. The Eisenhower Exec Office is where a lot of the action is. This building houses the OSTP which does Science and Technology Policy, there's the OMB that oversees eGovernment and IT, and the OIRA (Information and Regulatory Affairs). The government's new CIO is in the eGov office.

OMB used to be the Bureau of the Budget and assists the President in budget formation and review. It now has an expanded role with IT spending as a hook for linking reforms to management. It is in charge of procurement policy, accountability for management results, information security, and shaping management priorities. Each agency has its own technical people but this is where the high level stuff occurs.

She ran through a recent timeline on Egov or Open Gov starting in 1998 with the Govt Paperwork Elimination Act. By 2003 agencies had to provide the option of submitting required info electronically. In 2002, the Egov Act was instituted and in 2006 the Transparency Act, which required a single searchable website of federal spending awards information. In 2009, the Open Gov Directive moved on transparency.

The Egov Act of 2002 was designed to improve the management and promotion of electronic government services through the following acts:

Establish a Fed Chief Info Officer.
Establish a framework
Make the fed govt more accountable
Transform agency operations by collaborating and utilizing best practices from public and private sector orgs.

The Open Govt Initiative had three pillars: transparency, participation and collaboration. The approach uses cutting edge tools and practices. This includes online dialog tools, visualization and dashboards, social media, and challenges and prizes. It also changed the default stance to one of exposing info rather than hoarding it. It was also designed to require action.

The first requirement was for an online posting of three high value datasets from each agency. A senior official must take responsibility for the accuracy of info found in There was also the creation of an OMB working group. On it shows when the datasets were posted and by whom. The next deadline was that there must be an open gov web page for each agency, and a framework from OMB for describing spending data. The CIO and CTO would create an Open Gov Dashboard.

The EPA got their materials out early, but all agency pages can be found with the name of the agency / open . Still to come, the OMB will make a framework for incentive-backed strategies to find innovative or cost-effective solutions to improving open government. The idea is to put out challenges and prizes to get problems in gov solved.

Later still the OMB will issue a longer-term strategy for Federal spending transparency. Each agency must publish an Open Gov Plan that will describe how it will improve transparency and integrate public participation. OIRA will review existing OMB policies.

In addition, agencies with significant FOIA backlogs must reduce this by 10% each year. Each agency's Open Gov Plan shall be updated every 2 years to keep them involved in continual development.

The memorandum on FOIA instituted a presumption of disclosure for govt records. A new Exec Order reversing changes to PRA can only be done by an incumbent president who can assert constitutional privileges to withhold info and the president must get Attorney General and White House Counsel approval.

A recent example of transparency efforts is the visitor record log for the White House. They had also attempted to provide a visualization of how IT spending is distributed across the federal gov. is also a part of this effort to track spending, in this case across the country with stimulus money. The attempt is to make the sites appealing and easy to utilize. The UK has recently launched their own version of and there's hope that local governments in the U.S. will follow suit.

Viewing Citizens as Partners occurs through several different approaches:

Participation and collaboration: Try to find out citizens' needs and get citizen input for decision making. This includes public hearings, citizen juries, FACA committees, requests for comments, and idea generation tools. FACA committees stem from a 1972 act. They are small standing committees that advise the govt on all topics. Over 1000 of them exist with NSF panels as one type of FAC. It's a complex formal procedure though which includes high level officials, and take place by filing a charter with Congress, notification of meetings in the Federal Register, and complicated payment arrangements. It's one way for citizens to get involved but it's too cumbersome for quick advice. The in person gatherings are also not feasible for most people.

Online participation tools used in the last few months have asked for input on specific questions such as open publishing, asking gov employees how to save energy, critiquing implementation plans for, and asking what should be included in Open Gov.

Peer to Patent and Open Government Dialogue are examples of different ways of getting participation and discussion. They're experimenting to learn what works and what doesn't. Hearst also showed a MindMap of the Summary done for a Transparency Conversation. Some agencies have more experience in Egov. EPA makes heavy use of all forms of tools. It just released guidelines in the use of social media. Continuing issues include having to print out and manage official records and stating facts not opinions unless approved to do so. Hearst also showed a flow chart on "Should I Respond Online on the EPA's Behalf?" which employees can consult to decide how they should engage with public comments.

Hearst then provided a list of potential Online Discussion Research Questions which seemed highly pertinent to anyone dealing with or managing online communities, though of course more critical when issues of importance or aspects of "official" response are involved.

How to compile and summarize responses?
How to engage appropriate stakeholders?
How to moderate?
How to debate topics?
How to incorporate new entrants?
What are the types of conversation?
Are they valuable or not?
What role should they play in decision-making?
How to evaluate and assess?

Research questions for other areas were also posed:

Social Media – How to establish policies about who can post? How to balance providing official info with being conversational and responsive? What are the goals of this kind of communication? How to achieve them? How to handle archiving rules? Hearst cited the TSA blog as being particularly successful in some of these issues.

Information Visualization – Design guidelines for info dashboards? Automating tools for info mashups?

Open Data – How to search large data collections? How to devise shared metadata? What tools can aid in data normalization? How to link the content across datasets – what is the role of the semantic web? How to publish all posted content (such as on web sites) as data? How to discover who is using a dataset?

Proliferation of web sites – No one knows how many there are. Having many sites is not necessarily a bad thing but it can cause problems with duplication and confusion. The U.K. Gov did a massive consolidation of their web site recently. It deleted many websites and put them under a new umbrella. Apparently many smaller sites found their access went up dramatically. Should the U.S. do this as well, and if so, how?

She talked about how academic institutions tend to provide information. It's duplicated across departments, and things like courses and admissions tends to be replicated, but the publication of research is scattered and inconsistent and there still isn't consistency in submitting to institutional repositories, So should the emphasis be on archiving or curation? This is an example of unique content vs. administrative structure. There are wide ranging audiences for research content, especially for scientific agencies.

Why is it so difficult and expensive to do all this? Is there a turnkey solution for all but the content? How to automate some of the design process? Same with the usablity testing? Updating sites with stale designs?

How Can Someone Get Involved?:

Research collaborations – Find relevant agency contacts. This can be hard, but you can search for research papers that describe work with a particular agency. Contact the researchers for contacts. Talk to people at the NSF as well.

Govt Often Uses Info from outside groups – Non propfits, bloggers, university groups can do this work on their own initiative. Some examples: Technology developed by OMB Watch for was used in the re-launch of usaspending. The Sunlight Foundation compiled lists of strategies for govt to address openness in gov. UC Berkeley posted guidelines in a blog on how to improve the design of This had a big influence in its presentation. There are also Open Gov how-to workshops and websites.
There are Federal Register annotation tools that exist, making announcements and such more accessible.

Participate by Writing – OSTP request for comments was published in the Fed Register on May 21, 2009. It sought advice on Open Govt topics such as: What info should be more readily available? What are the limitations to transparency? What strategies might be employed to adopt greater use of Web 2.0 in agencies?

Participate with Classwork – Usability Clinic is one where professors teach usability courses and have students critique websites as a homework exercise. Commit to a particular time period so that orgs can sign up for it, and it can turn into longer-term projects.

Participate with Data Analysis: Expose inefficiencies or create new, useful functions. So for example you could analyze hiring latency on a per-agency basis. If data isn't there? Comment on the agency's opengov website. Ask for time-to-hire data for each agency. Be persistent.

Participate with Code – write code that makes it trivial to make beautiful, highly useful, 508 (ADA) compliant visualizations for wbe sites. Offer webinars to explain how to use it. Make the code compliant with govt specific constraints. More practical dispersion of research experience.

The federal government has numerous Impediments to Adoption of Off the Shelf Software – TOS, PRA, PII, OWACCS, COOP (disaster recovery requirements), 508 (access for disabilities), Cookies, and Gifts (restrictions on acceptance).

There's a new way to comply with those constraints which is to use This was a way to get people to submit free apps that are compliant with those constraints. has a goal to provide a way to let outside experts help improve government use of IT. It's an independent non-profit but will work closely with Fed govt agencies. Anil Dash is the director (early blogger) and it is sponsored by the MacArthur Foundation and AAAS. The idea is to have people work on code together to create solutions. Philosophy is open source and collaborative. Ideas are generated with open conversations, then document and publish what is learned, code is versionable, and work will be seen as experimental, hence more open to risk. This means not just academics but also online leaders will tahe part with people having short-term stints leading projects. There will be applications for fellowships; being active in the conversation and the shared code repository will likely make you eligible.

For Students – Presidential Mgmt Fellows, 2 year paid fellowship, few people in IT apply, and more are needed. Often leads to paid positions. Students must apply in the fall before they graduate.

There was some Q&A afterwards:

Comment: Where is the discussion of policy and ethics such as privacy protection and balancing access and security? Wanted to point out that it was the 1980 act signed by Carter that set up all of this and placed IT in the OMB.

Hearst replied that her history only went back so far for time purposes and that while privacy issues were important, this was an area with a good deal of research and activism. This was less true of other areas she had discussed.

Comment: What criteria guided the release of data to the web?

Hearst said there was a very short time frame for this release and it was supposed to be high quality, useful data. The idea is that once data starts going up it can jump start the process by getting public input regarding what is and isn't there yet. This also only applies to agencies under the Exec Branch, there's a hope that it will change expectations about what is possible and pressure will spread to other branches and levels of govt. The idea is to have a perpetual beta with this concept.

Comment: There's a conflict between the idea of permanent beta that runs up against the culture of slow, manageable and right. How will expectations be handled?

Hearst said that [eople have been pretty open to releasing information but soliciting it from the public and then acting on it has been a challenge. What about guidelines regarding how managers respond to emails? Some insist on having all conversation go through one person, others allow for open response by any employee. There seems to be no consistency. That is something that needs to be researched and developed.

Comment: The presentation focused a lot of ways to get the ordinary public involved, but there is a lot less focus on getting the academic community involved. Traditionally this had been done through govt grants, but little has been done about this recently, and they seem to be asking the academic community to lower themselves to the level of citizen response. He suggested that only the very dedicated or desperate individuals will participate in this way, but academics need more incentive to take part through grants and the older, more formal processes.

Hearst mentioned the challenges and prizes but said that they did not want to exclude others to focus on academics. She said that traditionally academics have not thought much about working with government other than through requesting grants.

Someone else asked where the money's going to come from for small government bodies. She suggested that there be workshops offered to these groups by iSchool participants on how to set up and manage these efforts. Hearst discussed the 311 movement (citizen reporting of need for services), and the software that's available for any community to use.

Comment: There's an unfunded mandate problem, with low compliance by some agencies whereas some others saw it as part of their mission. What is the combination of carrot and stick from the OMB?

Hearst says she hears this a lot from academics, but the requests right now are fairly low cost because they want people who are already doing things in the needed areas to shift it into this effort. She said the challenge is to turn this into standard operating procedure.

Comment: Someone talked about requests such as why people need to fill out tax returns, can't the IRS prefill them and send them for review and signature? He said that Intuit is opposing this due to their investment in tax prep software. What personal stories can she share about obstacles she's faced?

Hearst says that people focus a lot on obstacles and what the Open Gov movement wants to focus on solutions. She referenced how states have been spoken of as labs for the federal government, trying out ideas that later get adopted on a larger level. But they're now trying to have the Fed do experiments on its own.

My own impression is that the comment regarding grant work was the most telling of the presentation. While Hearst talked about prizes and challenges, such an approach is poorly targeted to academics. It is, however, well targeted to corporations or entrepreneurs. These days a lot of academics have to raise half their salaries through grants. They can't put in several years of work on the off chance that they might win a prize. This is particularly true given that all grants must include overhead expenses which go to the academic institution, so their employers would also lose out on such a process. In addition, the sort of work Hearst describes -- collaborative online work with no specific outcome -- goes against the publishing need of academics who are rewarded much more lightly for anything but traditional (and frequent) publications.

At the same time, his comment about who would participate in such projects strikes me as being on-the-nose. No online communities, regardless of how passionate the user base, tend to have a great many people doing all the work. It's generally a small core group, and even these people rotate in and out of intense activity. So who would have the time and motivation to take part in these various endeavors? Corporate reps -- either openly or undercover. Businesses are already hiring many PR people to work through social networking sites to promote their products and spin their image. They're the ones who have a stake in what gets decided by public policy, and to make sure that things go their way. If this effort at public input starts making inroads into public policy and expenditure, I'm pretty sure that "public participation reps" will be an area of future employment.


Social Informatics and Media Studies

April 2010

111213 14151617

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags