OMB Watch's second letter, addressed to Sally Katzen of the Office of Management and Budget and David Barram of General Services Administration, responds to the new PowerPoint presentation on the firstgov.gov site. While the presentation is much clearer than the first version, it also raises many questions and leaves many issues unanswered. August 10, 2000 Ms. Sally Katzen Deputy Director for Management Office of Management and Budget 260 Old Executive Office Building Washington, D.C. 20503 Mr. David Barram Administrator General Services Administration 1800 F Street, N.W. Washington, D.C. 20405-0002 Dear Sally and Dave: Thank you for taking the time to meet with us on August 2 to discuss our concerns regarding the development of FirstGov. As we mentioned during that meeting, OMB Watch, along with many in the public interest community, feels FirstGov can be an important contribution to improving public access to government information. For that reason, we strongly support the objectives of FirstGov. On August 9, we discovered your revised PowerPoint presentation on the FirstGov web site <www.firstgov.gov>. Thank you for considering some of our points. The presentation is much clearer than the first version, and demonstrates your conviction to strengthen public access. At the same time, your presentation raises many questions and leaves many issues unanswered. Accordingly, we would like to raise several questions with the hope of getting clear answers. 1. How will the FirstGov approach to "topics" help to break out of the silos of government information currently available? It appears from your presentation that FirstGov will have topics. We are pleased to see this since it was a significant criticism we had of the earlier presentation of FirstGov. However, your presentation of topics raises many issues. On Page 9 of your presentation you state that "specific pages of ‘topic lists' would reside at an agency, either for that agency's information or as an interagency topic list, organized by subject (e.g. trade, toxic waste)." Does this mean that FirstGov will develop a Topics Page that links to agency sites? If so, then what happens when you click on "toxic waste"? Will it take you to EPA, even though many, many agencies have information about toxic waste? If so, then doesn't this perpetuate the existing problem where information is accessible only by searching each agency individually? On Page 19, you say regarding topics that "[p]ages would be specific to an agency or be a combined page where various agencies had banded together to provide easy navigation to government information for specific subject areas." This seems to suggest that your intent is to deal with the silos problem by fostering interagency topic pages. If so, many questions arise. Will FirstGov provide the resources for these interagency dialogues? Some agencies have extremely useful web sites with helpful topics; others are less useful.If the intent is to shift responsibility for FirstGov to the agency web sites and encourage collaborations between and among agencies, the public will experience great unevenness in accessing information – a key problem and reason for creation of the portal – unless resources are provided to help the agency pursue these activities. Will these interagency dialogues happen without structure or will there be a planned framework? Where will these interagency topics reside, on the FirstGov topics page, agency web pages, or both? From our perspective, providing structure and leadership is absolutely necessary. (More on structure and leadership below.) The shift in responsibility to agency web sites creates other questions. As you click-through the FirstGov web site to an agency web site, you have lost the added value of the FirstGov site. For example, most portals, such as Google or Yahoo, keep the search engine available to the user, even as they explore lower level topic pages. Under the FirstGov model, it would appear that once you click a topic you are no longer in the FirstGov framework and no longer would have direct access to the search engine. Finally, we are uncertain about the difference between "topics" and the "Concierge Bulletin Board." The example in the PowerPoint presentation of using the Bulletin Board for accessing all agency press releases is a good one to discuss. Shouldn't that simply be a topic under the topics page? If something qualifies as a "subject of special interest," the criteria you establish for the Bulletin Board, then it should qualify as a Topic also. Whether it is on the Bulletin Board or Topic page, how will there be a site for all agency press releases unless FirstGov takes action to build something across agencies. Yet your description under the "topics" section seems to suggest otherwise. We would encourage you to move Bulletin Board topics to the "topics page" and take the time to develop the cross-agency searches that are necessary. An alternative to the Bulletin Board might be a scrolling news bulletin section, like the "Breaking News" section of www.zdnet.com. In this way, the government can post hot news information. 2. What is planned for the search engine to help reduce user burden and protect privacy? You point out that a commercial search engine can produce 79,000 web pages when searching for the term "superfund." You point out that the advantage of the Fed-Search search engine is that it will focus only on government web sites. You say in the presentation that typing superfund "will take you to the portion of the EPA's web site devoted to that effort." However, it is not that simple. There are multiple web pages within EPA that reference Superfund. There are web pages at the Agency for Toxic Substances and Disease Registry, Office of Environmental Management in Department of Energy, NASA's Marshall Space Flight Center, at national laboratories such as Brookhaven, HUD's Empowerment Zones and Enterprise Communities, National Institutes of Health, U.S. Army Corps of Engineers, and National Oceanic & Atmospheric Administration to name just a few that discuss Superfund. This, of course, does not include NTIS, which sells publications regarding Superfund, the various congressional committees, judicial actions, or state and local governments. It also does not capture related issues, such as Brownfields, which also has many web hits. So a search of "superfund" is likely to yield a number of hits similar to a search on Google or other commercial services, but all the sites will be within the government. A successful search engine not only searches broadly, but it finds a way to get the relevant web sites listed high on the search retrieval. It would be helpful to describe how search results will be ranked so that they display what the user is looking for. In the Superfund example, many of the above web sites deal with descriptions of a Superfund site and the remedial programs that are underway. Others deal with a description of the Superfund program, the statute and regulations. One, we found, deals with art related to Superfund. It would be helpful for the user to understand these distinctions without having to go through hundreds of possible hits. This is not intended to diminish the value of the search engine being developed by Fed-Search. To the contrary, the question is what are the plans for improving the search engine so that it is user context sensitive. If there is no plan, is there a process that can be described for developing such a plan? We are also concerned about how the database/index will protect personally identifiable information. Is it the agency responsibility to stop the Fed-Search spider from grabbing such information? How often will the agency sites be spidered? Is it Fed-Search's responsibility to insure that the public cannot search on people? What is the policy regarding privacy of the information found? 3. Who "owns" the Fed-Search database/index? It appears from your presentation that Fed-Search, a nonprofit organization, will be given special authority to spider every site in the government to develop the database/index. The creation of the index does add value to the government information, potentially making it a commercial product protected under copyright law. Is there clear legal agreement between Fed-Search and the government that the database will be public domain, open source information? Your description of the co-branding process (i.e., Certified Partners) implies that non-governmental entities will be permitted to access the database/index only if they agree to certain requirements. This would imply that the Fed-Search database is not public domain. If correct, then is the government willing to permit any entity to spider all its web sites to establish its own database/index? If there is a special relationship with Fed-Search or anyone else, on what legal basis can that be justified? If the Fed-Search database is public domain, then why is there a need for the co-branding process since anyone interested in using the database/index would be able to? If the database/index is "owned" by Fed-Search, what authority does the government have to link it to other government responsibilities, such as those required under the Government Information Locator Service or EFOIA? It would be helpful if the Fed-Search database creates a GILS record so that the agency has a workable index at least of web documents. Similarly, the database could prove useful to helping agencies develop their indices of agency information systems as required under EFOIA. 4. What is the relationship between Fed-Search and FirstGov? It appears that GSA has operational responsibility for FirstGov and that the President's Management Council is providing overall direction. Yet the heart of FirstGov is the search engine, which is being developed and maintained by Fed-Search, a nonprofit organization, with its own Board of Directors. What is the relationship between the two initiatives? Are there formal agreements, such as licensure arrangements? Will Fed-Search have any role in developing FirstGov? If so, what? Will Fed-Search respond to public comment about the search engine? 5. What are contractor responsibilities? It appears that GSA is entering into one or more contractual relationships to develop FirstGov, although the PowerPoint presentation is not very clear on this point. If so, what are the responsibilities of the contractor(s)? Will they design and house the web site? Will they develop the topic list? How will contractor responsibilities dovetail with Fed-Search's activities? How much money is GSA spending on this initiative? 6. Will FirstGov be a vehicle for strengthening public access to government information? FirstGov has enormous potential for providing a floor plan for accessing government information. For example, if FirstGov is effective it will provide access to information on the web. But much of agency information is not on the web. Why not require agencies to do an inventory of agency information holdings and post the catalogue on the web? In this way, FirstGov could at least provide the public with information about what they are not getting. Why shouldn't FirstGov explain how to get access to non-web based information? Couldn't there be a requirement to require agencies to respond to public requests for such information? And why not provide an explanation of the Freedom of Information Act for those who may need to resort to that tool to obtain information? Why shouldn't FirstGov describe certain standards that are consistent throughout government web sites, such as privacy principles? And if FirstGov is a vehicle for making government more transparent, why not a locator of government employees? This could be a topic, if FirstGov is developing topics. Why shouldn't FirstGov explore ways of developing common corporate identifiers (e.g., corporate IDs, facility Ids) since none now exist? The public will look to FirstGov to provide cross-agency information. Yet how can you find out about a company's compliance with regulatory requirements if there is no mechanism to search across agencies. In fact, most agencies do not have such identifiers to search across agency programs. What plans are there for future access to databases that exist within agencies? Will FirstGov experiment with database access through ODBC calls or through other means? What ideas are there for linking with state and local government information? OMB Watch has started a survey of state legislative sites and will have information that it can share about locations and quality of sites like Thomas in the fall. These are just some ideas that represent the potential of FirstGov. No one expects the government to achieve these outcomes by September 30. But there is an expectation that the government will seize the initiative and provide a process for addressing and implementing new visions for meeting the public's right-to-know. It is vitally important that the President's Management Council build on the opportunities created by FirstGov to establish government-wide standards and practices for public access. It has been said that one reason FirstGov has taken so long to build is that agencies did not uniformly meta-tag web pages. Regardless of the merits of this argument, it demonstrates the need for enforceable leadership. 7. How will GSA and the President's Management Council insure that the process for input on FirstGov is an open one? We understand the expedited timetable you are working under to make FirstGov operational. We know that you will make mistakes and that you will continue to revise the direction of FirstGov – even as you receive comments such as these. However, this should not be an excuse for disregarding important process considerations. As we mentioned in our last letter, the government should solicit comments from the public. We are pleased with the emphasis in the PowerPoint presentation on soliciting public feedback. Would it be possible to also seek feedback in face-to-face and print fashion? Are there plans for focus groups to discuss FirstGov design and content? Will these be handled by a contractor, GSA, or another entity? When will comments to the CBD notice on co-branding be made public through the developing FirstGov site? Why is a nonprofit organization, the Center for Excellence in Government, convening a meeting on August 10 on behalf of GSA to discuss FirstGov? Shouldn't this meeting be convened by GSA with public notice, and shouldn't there be a public record of the meeting? How was the nonprofit organization selected? Who is funding this meeting? Who was invited to this private meeting and what was the criteria for selection? Why is press not permitted? While this may prove to be a useful meeting, the appearance of secrecy runs counter to the very purpose of FirstGov and openness in government. As a final point, would it be possible to keep different versions of the FirstGov PowerPoint presentations on the web site? It will likely prove to be useful historical material. Yours truly, Gary D. Bass Executive Directorr