Paper Hoarders vs. Byte Manipulators: Can We Reconcile Good Records Management With Good Information Management?


By David H. Rothman[1]


Whoops. At least 43,000 email messages vanished into the ether in June '99, and guess which bureaucracy they came from. The National Archives and Records Administration. Yes, the very same people to whom we've entrusted the original copy of the Constitution. The ironies go on. While the White House has rhapsodized endlessly on the beauties of automation, a certain Clinton appointee prints out all of his important email the day it arrives. His name: John Carlin. Job: Archivist of the United States.


The events at Archives dramatize the gap between documents management and information management in the federal government. Never mind the 43,000 AWOL messages. Some agencies have even fought for the right to delete electronic messages without much ado if paper versions exist. Email is the Rodney Dangerfield of government documents. Grouchy with good cause, one of Ralph Nader's organizations is suing Archives for the preservation of imperiled bits and bytes--in a case still ongoing as of early 2000: Public Citizen vs. Carlin. Point is, some powerful bureaucrats care less about computers and the Net than about paper and records, including the microfilm variety; yes, records please—not data. Meanwhile many younger executives worry less about preservation of paper and more about the ability to move bits and bytes around through spreadsheets or databases or other programs. They see this as the era of data mining, of information warehouses, not the concrete-and-steel variety.


Can the two sides be reconciled? David Bearman, head of Archives and Museum Informatics, a consulting firm specializing in preservation issues, points out the inherent clash between traditional records management and information management. One discipline focuses on stability and history. The other focuses on manipulation of data for the needs of the moment. What's more, formats for computer tapes, optical disks and the rest keep changing on us. Bearman correctly warns that if we want to preserve digital information, we

will have to move it continually to new formats. At the same time he is skeptical of bureaucracies massively warehousing information—the very process that some optimists think might ease the task of backups.


Still, certain information-oriented people have centralized information and records, tried to reconcile the two disciplines, and lived to tell about it. They are the information services crew at the World Bank, which, though international and only quasi-governmental, is still a massive agency whose needs overlap with many of those of our native bureaucracies.


To save money and speed up loan processing and other activities, the bank scans most every scrap of incoming correspondence; it digitizes the images and creates character-based files as well. According to consultant Richard Barry, ex-chief of information services at the World Bank, the Bank has used a blend of document management software from various vendors, notably Excalibur Technologies Retrievalware and an Oracle-based metadata database to provide the necessary contextual information about the records that is not necessarily evident within them.


"Why the devil hasn't Washington made a better marriage between its CIOs and records managers and gotten on with enterprise-wide electronic records management?" Barry asks. "A big reason is Y2K, which has gobbled up so much money and time in recent years. But now the Y2K work is behind us. Why not use that dividend and finally take on electronic records?"


But how can we blend records and information management—those two often-warring disciplines—without compromising each?


For chief information officers and wannabes, here are some ideas synthesized from conversations with Barry, Deputy Director for Army Records Edward Arnold, and others, especially Owen Ambur of the Information Resources Management Division at the Fish and Wildlife Service. Blame me, not them, if you disagree with anything in the ten rules below.


Rule #1: In blending your record and information management, focus on the needs of citizens and agency people alike. The needs may overlap more than you'd expect.


Even insiders may not be able to keep up with all those beautiful organizational hierarchies. So think, "Function." What do various offices do? And can that be turned into a search word that people anywhere in the agency can key in? Thus, a civilian Army employee in the future might find out anything and everything about rifle design if, say, he were writing a procurement regulation that would affect hardware in the field. He would not have to make phone call after phone call and then wait for various manuals or other literature to reach him. The Web could give him the basics and maybe even include the phone number of the leading Army and civilian experts on rifle design. Similarly, with a function-hip intranet, an information officer could better reply to queries from the New York Times or NBC, as opposed to having to wend her way through a nightmarish series of links.


With storage space so cheap, in fact, shouldn't as much releasable information as possible go on the Web for the press and the rest of public to view directly? In stocking your Web site--increasingly the way in which your records and facts will reach the citizenry--don't just play up the usual mix of budgetary details, administrative trivia and public relations hype. Instead think like a good magazine or TV network. What does John or Jane Q want and need? The Fish and Wildlife Service has Web-posted some captivating images of animals, complete with samples on the home page. What's more, fast links go to a very browsable library so that you can see agency photos of sea turtles or spotted owls. Serving the public, you may even want to go beyond your usual mission if you have valuable records of public interest. Ironically the secretive Central Intelligence Agency is a model in a limited way. The CIA took time from tracking dictators and terrorists to put online The World Factbook with non-sensitive facts about the world's countries.


Another customer-oriented example, at a grander level, is the government blue pages now under development. Citizens won't have to thread their way through an impossible hierarchy of links as often as they might otherwise. For example, an entrepreneur will just type "business" and see state-by-state lists that go far beyond the usual local outposts of Small Business Administration. She'll also see contracting offices that might be a source of opportunity for her at other agencies. The blue pages are very much a work in progress, full of glitches, but the idea is laudable.


Rule #2: Preserve records of the information and processes behind a decision, transaction or other action--not just the results. A classic example of a mistake from the pre-computer era is at the Bureau of Indian Affairs. BIA is in Legal Hell. The reason? Well, how much good does it do to say that the Indians got X million dollars over the years if you lack all the crucial information showing receipts and signoffs?


If anything, unless agencies take appropriate precautions, the problem could worsen with computers. How to cope with this? You'll want to use either images of documents or verifiable digital signatures to show proper authorization and procedure--not just authority at the document level, but also chains of authority encompassing the entire decision process. Too, you'll need to tie both information and documents to their sources. And if programs such as spreadsheets were used to create a set of facts, then you ideally will be able to reproduce calculations and do everything else to add context. The end result could be that most everything is a record (not impossible because of the compactness of electronic data). So be it. Most information won't have a continuing value, but you'll have disposed of it properly, typos and all.


As much as possible, try to automate the capture and preservation of documents.  Barry suggests that you consider an electronic document management system (EDMS) that is also a trustworthy recordkeeping system, aka a records management application (RMA) such as Tower Software's TRIM system

and Provenance's[2] FOREMOST system. (Barry himself has worked as a consultant for Tower Software on the future direction of software needs.) Unlike most EDMS's, they have the necessary functionality to do serious electronic recordkeeping. Also catch up with a standard that National Archives has blessed for all federal agencies—the Defense Department's 5015.2 RMA standard. The Joint Interoperability Test Command at Defense certifies software against that standard. Many popular office applications wouldn't come close to meeting 5015.2. Also, in choosing software and configuring systems, don't neglect the human element. To the maximum extent practical, preservation should be done by a trustworthy third party or by a system that allows only read-only access.


Rule #3: Get backup strategies and other safety schemes down cold. Within the limits of practicality, allow not just for error but also for the possibility for sabotage. The zapping of those 43,000 messages, which included the email of Carlin and many other top officials at Archives, happened during a weekend. Archives critics theorize that disgruntled employees may be to blame. Whatever happened, the backup at an Archives contractor failed to work. Real estate has an unofficial motto, "Location, location, location," and, with the Archives debacle in mind, perhaps information resources should have its own slogan: "Redundancy, redundancy, redundancy." Redundancy alone won't make a system trustworthy—other issues remain such as the integrity of the data—but it certainly is a must. Even tape backups may not be rigorous enough since, among other things, they may be cumbersome to use and not up to the minute.


Also, instead of storing data in one geographical location, think about a number of backups in government and private facilities. These days Exodus Communications and other vendors even offer protected warehouses for the safekeeping of servers and important Web files, which, yes, should be treated as records. Another safety strategy would be to require more than one person to push the button to destroy crucial records. Yet another would consider the risks of viruses and other delights. You might even think about using different operating systems and different programs in systems storing the very most important records.


Rule #4: On the migration front, take it for granted early on that your PCs, your mainframes, your everything, will be paperweights one day, and that your data formats will be the cyber equivalents of the Rosetta Stone—unless the right tools exist to read old disks, tapes and other media. Plan accordingly. Not to despair. Y2K was actually a blessing; whether they wanted to or not, many information managers learned how to move data over from ancient systems and cope with old formats. Just don't minimize the problems that will result if you don't plan ahead for easy migration.


Rule #5: Work out a systematized way of determining who can access a document and data, when this information will be available to them--be they insiders or outsiders. Try to build this access control into your systems as early as possible. Strive to comply with existing requirements such as the paperwork-reduction laws and the updated Freedom of Information Act. Because most everything can be a record--thanks to the new technology—these measures are all more important. Paradoxically, improved access controls can pave the way for a richer public record since agency employees will better know the rules of the game.


Rule #6: Strive for openness to truly improved management of records and information. You can't just throw technology at problems; your agency's people and policies must keep up with the times. Many bureaucrats have risen in an organization through their ability to play by structured rules and well-ingrained habits, such as the belief that secrecy will lead to more power by fostering the dependence of others on them. A good CIO will work to change these rules for the better. Greater sharing of information will help an agency carry out its mission, and ideally he or she can communicate this truth. Also, the CIO should recognize that one of the most attractive characteristics of the new technology, the ability to create greater accountability, would lead to the most resistance from old-style bureaucrats. 


Who might be potential allies at your agency? Auditors, perhaps. And it can't hurt to have sympathetic lawyers on your side. Just be careful. Lawyers often care more about posterior covering than about client service and may often push for the needless destruction of documents. Try to balance this threat somewhat by enlisting the support of records managers and archivists, who are more concerned with the actual value of the material to your organization and society. In fact, if you're CIO, perhaps you can even try to bring these functions under your wing if they are not already—one way to bridge the gap between paper and electrons. At the World Bank, Richard Barry began this integration as far back as the late '80s, and he believes it greatly smoothed the transition to electronic records. On his Web site check out relevant papers on information management and records management distinctions and the integration of IM and RM.



Rule #7: Use education to make the nuts and bolts as easy as possible. Arrange for a Web page to include links to vendors with possible solutions--as well as links to organizations in business and government that have successfully executed. Don't just rely on vendors alone for information. Put up Web links to trade publications with product reviews--taking into consideration that in at least a few cases advertisers may influence the contents of the reviews. Too, encourage your eligible people to benefit from specialized organizations such as the Federal Information and Records Management Council ,the Association for Information and Image Management International, ARMA International and the Society of American Archivists. Also, remember an obvious resource for the eligible--the Chief Information Officers Council. In fact, anyone can benefit from the Council's Web-posted publications, such as a guide to Best Practices In The Federal Government; it explores many document-related issues.


Among other handy Web resources, incidentally, would be ARMA member Alan Zaben's remarkably comprehensive list of links, Owen Ambur's page with links to some major laws affecting access to public information, Public Citizen's area on the Carlin case and other topics related to electronic record-keeping, OMB Watch's information policy page, and a records-management page from the U.S. National Archives and Records Administration itself. However badly Archives bungled its e-mail storage, the agency has come through with flying colors here with a well-organized collection of links to various laws, regulations and initiatives.


Rule #8: Remember that technology is just a means to an end. Many CIOs have been technologists rather than information managers. It isn't enough, however, to put up a Web site with the latest interactive features and the most advanced Java.

The site must truly ease access to the information resources that your agency has developed over the years. Substance over technoglitz!


Rule #9: At the same time don't be shy about using new technologies that do

advance your agency's mission. Rich Kellett, director of the Emerging IT Policies Division at the General Services Administration, wrote recently in Federal Computer Week: "Real-time government is about responding to the increase in the number of electronic devices the public uses to receive or to provide information in real time." Rather than just filing away an agency director's important speech, you might put it on the Web in a timely way--and perhaps even do a live RealAudio or RealVideo broadcast if the interest is there.


Rule #10: Understand that bureaucracy is inseparable from paper for the moment. Some paper will remain, and not just for ease of reading. Many documents might be just plain too expensive to convert for the moment; warehouse space might be cheaper, until time has passed for the records to be safely burned. Similarly, microfilm may cost too much to convert. What's more, you may lessen resistance to your bits-and-bytes-oriented plans if they are not of the all-or-nothing variety. Be flexible within reason. Leave the binary thoughts, the 1s vs. the 0s, to the machines.




David Rothman, author of NetWorld, The Silicon Jungle and other tech-related books, has been writing about technology since the 1980s.






Why We Need a Well-Stocked National Digital Library


Every student in Ted Nellen's 11th-grade English classes passed the New York State Regents' examination in his subject in spring 1999--despite a miniscule budget at Murry Bergtraum High School for the purchase of reading material. Most of the kids came from low-income families there in Manhattan. And a big reason for this success was that Nellen helped his students pull down many megabytes of literature from the Internet. Understandably he wishes that we had a well-stocked national digital library online with thousands of books from which to choose--and not just valuable classics in the public domain.


Despite years of talk along those lines, including some rhetoric from Vice President Al Gore, Washington has yet to give us a national digital library rich in contemporary books. In fact, Congress has extended copyright terms and thus will make it harder for students in the future to download books for free.


The real answer, if we're to be fair to both copyright holders and the public, is a national digital library fund that would pay writers and publishers by the number of downloads. The money could come from a mix of public and private sources, including, yes, Bill and Melinda Gates, who, to their credit, have already committed many billions to charities of various kinds. The same program could work with local public libraries and encourage the spread of book-friendly, tablet-shaped machines that could be used for other purposes, too, ranging from Web-shopping and email to simplified income tax preparation--thus cost-justifying the national digital library. The program could also help teachers catch up with Nellen and present the Net to children in the right context. For a detailed proposal called TeleRead, which I have been developing since the early 1990s, see my US NEWS & WORLD REPORT commentary--David Rothman


ã 2000         David Rothman

[1] David Rothman, author of NetWorld, The Silicon Jungle  and other tech-related books, has been writing about technology since the 1980s. This is a pre-publication version of an article that was published in The Year in Computing 2000 magazine, Vol. 4 No. 1, p.156. It has been subsequently updated with current URL references on July 7, 2001.

[2] Now TrueArc.