ONLINE Magazine
Table of Contents Previous Issues Subscribe Now!
Turbo-Powering the Flat Portal: The University of California's Labor Research Web
by Terence Huwe
Even the most cursory glance at recent literature reveals that enterprise information portal software has come of age. Portals continue to be hot. Prices are coming down, and many libraries and information centers have been trying them out. At the high end, packages like Livelink and IntraSmart offer a wide range of DHTML power and collaboration tools, though it's not hard to find critics. At the lower end, open source products like Zope have developed loyal followings and large communities. However, it's not always necessary to make a new technology investment to leverage a Web site into a content-based portal. The big surprise in today's Web portal sector is the continued effectiveness of flat Web design.

Flat Web sites: We've done that already, right? Yes, but in the era of the Invisible Web, flat is good. Flat Web sites are picked up by today's Web crawlers and can generate heavy repeat traffic. What's more, flat portals are ideal for libraries and information centers that don't want to get into the hardware business or get locked into a single product family. The challenge is to turbo-power them and link them to what searchers crave most: solid, reputable content.

This article profiles the University of California's new Labor Research Portal, which was designed by the Institute of Industrial Relations Library. The portal and its related Web sites all reside on central university campus Unix servers running various flavors of Apache Web Server software. With nothing fancy running in the background, this tale of portal production focuses more on those perennial "killer apps" that sprout in the soil of proactive libraries: a focus on community and relationships, with a bias for reference and outreach. The content strategies employed took their cue from the communities served—even when this ran against conventional wisdom.


Founded in 1945 at UCLA and UC Berkeley, the Institutes of Industrial Relations (IIRs) are "organized research units" that support faculty research activity across disciplinary boundaries. Within the IIR communities, sociologists, economists, business professors, anthropologists, and even city planners form a diverse community to study work and employment issues. In June 2000, the California legislature established a new program that would span the entire UC system and build upon the two IIRs. Called the Institute for Labor and Employment (ILE), the new institute brought a new, $6 million annual revenue stream to the University of California—serious money in the social sciences. Suddenly, it was possible to formulate a new statewide research community.

This was a golden opportunity for library-led content development. The library staff was ready with a vision that was simple but powerful: Create a Web presence that linked purchased digital resources with original research, and throw some reference and outreach into the mix.


Activist librarians are always looking for opportunities to channel effective information services to where these servers are needed. These activists share a common trait: They diagnose the patient before prescribing the treatment plan. This is a vital step and is often harder than it may seem. For example, effective knowledge management strategies that work in a law library may fail at a pharmaceutical firm. What's more, it's not just the industry sector that defines how organizations succeed or fail in effective use of information resources; more than anything, it's the people. In the case of the IIRs, a seven-fold increase in funding brought growing pains—and an enormous opportunity to create a content strategy that fit the patient.

The Institute of Industrial Relations Library had a proven record in Web administration. Since 1995, the library has managed Web services at Berkeley's IIR, with considerable success. Like many Web sites, it was a cottage effort, and by 1998 it housed several thousand documents on hundreds of directories. Web traffic averaged 2,000 hits daily, and the top-ranked domain for visitors was .com. Clearly, there was a market for quality content in labor research. That same year, the Library won the John Sessions Memorial Award from the AFL-CIO and American Library Association. Consequently, the library could tackle the huge new task of portal design with solid experience in Web and design—and organizational diagnosis.


For the IIRs and the ILE, the first challenge was to help create the organizational infrastructure for a statewide research program. This meant getting involved in local programs at UCLA, Berkeley, and other campuses. The Library wanted the opportunity to create the overall Web service and focus on key value points—original research, guides to the literature and Internet, and program materials. Three insights about the user population immediately emerged:

Support Faculty Independence—But Manage the Digital Content. Research faculty have an enormous workload, often teaching three courses, writing grants, supervising doctoral students, and writing articles and books, all at once. These hard-working professionals mostly want to be left alone to do their work—unless you can save them time. With that in mind, the library offered to oversee faculty working papers and port them to the Web. This formed a key alliance that defined the Library as the site for effective Web content management.

Help Allies Achieve Their Goals. Unlike many academic libraries, the IIR Library is essentially a special library, and it serves a specialized segment of the university community. User groups include ladder-rank faculty, professional researchers, trade unionists, doctoral students, and not least, the general public. That's a diverse user population with a vast need for reference. The Library always emphasized reference, and this helped form a basis for the community to come, because the service brought people back, and reinforced the content strategy process.

Make No Enemies. Let's face it, the workplace is a dynamic arena where the staff and management follow many agendas and goals. How organizations handle strife can speak volumes about what is possible for librarians. In a business firm, power struggles are often sudden, intense, and have a final outcome. In academia, power struggles can last decades. The library cannot afford to make enemies or play favorites, and the Web portal needed to be balanced among extremely diverse groups.


Charting our own development strategy involved a certain degree of risk-taking, as the Library proposed to perform new work before it was certain of new funding. Building on proven relationships and strengths in the UC Berkeley sphere, the Library followed a five-point plan for developing its role. Because none of the transition planners had a deep personal interest in the technology, the plan was phrased in plain language.

Handle All Web Work. Before June 2000, library staff handled Web work for Berkeley's IIR, but did not employee a career Web developer. In order to solidify new staff positions in the planning process, the Library offered to handle Web services not only for Berkeley, but also for UCLA affiliates and the ILE, which would eventually span every UC campus through its affiliate structure. This made it easier for the Transition Team to support the Library's overall goal—to manage the digital publishing process.

Create a Coherent Mix of Original Research and Purchased Content. Deliberations among the Transition Team focused mainly on how to balance governance between campuses and with the many professional communities that are stakeholders in the ILE. The Library staff focused on the mix of digital content that would support researchers in a practical way. Three broad types of information were identified as crucial.

First, the ILE was primarily interested in promoting broad study of labor issues—and publishing the results not only in the usual, peer-reviewed avenues, but also as policy papers and reports for special audiences. This was the perfect format for a digital library of Web documents.

Second, many ILE faculty members were interested in building data archives with data sets from many different government agencies. To avoid competition with potential friends, the Library reached out to data archivists and faculty interested in data collection. The resulting partnerships reduced the risk of dispersing various functions of information management.

Third, although many of the transition planners were library users and knew a lot about the resources the University had already acquired, not all were expert searchers. Therefore, a vibrant collection of guides to databases and search tools was a high priority for the new portal.

Plan to Enliven Digital Conversations. As a university-community partnership, the ILE presented the University with a new opportunity for building community support. However, the faculty was of two minds on the matter of how to engage the non-academic labor research community. Some wanted the ILE to focus on research aimed at policy makers, while others wanted a practically oriented outreach program. Since both ideas made sense, and in fact could both be followed simultaneously, the Library proposed to manage digital conversations on discussion group lists. The staff also offered expertise in interactive form-making—both secure, CGI forms, and less secure (but quickly generated) PHP form-making.

Coordinate Information Acquisition and Development. Since researchers were interested in original work and wanted to manipulate raw data sets, a substantial opportunity existed for library staff to negotiate purchases from private publishers that repackage statistical data. By offering to do the negotiations, librarians could help create long-term business relationships that would support preservation strategies. This development also reinforced relationships with data archivists, who could house acquired data on their servers.

Emphasize Human Analysis and Counsel. The library had achieved significant levels of success in employing the same techniques of human analysis of the Internet that certain commercial portals, like Northern Light, have advocated. Therefore, the core strategy for creating Web reviews was to ask the staff to monitor the Internet—using the search tools like Google—and to track down the specific high-value sites that lie buried in academic and nonprofit Web sites.


While the ILE leadership addressed organizational concerns like the impact of its research on state government, the role of UCLA and UC Berkeley within the new organization, and related issues, the library staff performed an audit of existing content and formed a strategy for new content creation. Server logs presented a clear guide as to what users wanted to see: new and timely reports on "hot" issues like the need for a living wage in San Francisco, or the impact of wages on airport safety, which researchers had already been studying prior to the September 11, 2001 disaster. In fact, server logs showed big upticks every time the library uploaded new faculty working papers and announced them. Professor Michael Reich's paper on a living wage in San Francisco generated several thousand downloads during its first week online; a follow-up paper studying a living wage for workers at the Port of Oakland generated similar interest. The key challenge for this portal, then, was clearly how to make access hassle-free.

Other known users said they wanted to find a place on the Web where they could rely on detailed directions to rich data sets—and not be forced to slog through big, sophisticated Web sites to find the answers. Chris Erickson, a professor of business administration at UCLA, said, "I need Web guides that will give me specific tables and cross-tabs of labor data about Southeast Asian countries. I know how to search sites like the World Bank to find this, but it can take up to 15 minutes to find exactly what I need."


Since 1995, the library had been publishing two series that met the proper conditions for turbo-powering: Internet Research Guides and Berkeley Labor Guides. The Library's Internet Research Guides generated the second-highest traffic on the UC Berkeley IIR Web site and were essentially exhaustive directories of URLs having to do with labor issues.

These were reconfigured in two ways. First, the really big directories were locational tools, such as lists of all the labor unions sites in the U.S. and around the globe. These were edited and expanded. As a new initiative, the library staff started publishing more selective guides to the "best" URLs—generating shorter, but more purposeful lists of URLs. Second, the Berkeley Labor Guides series was vastly improved and given top priority as a source of a smart searching solution. The Berkeley Labor Guides had an unusual value point, encompassing much more than just the Internet, including bibliographies of books and articles, relevant Library of Congress Subject Headings, and other finding aids.


The third area of content strength lay in the Library's long experience with faculty working papers. Since the late 1990s, working papers had migrated to the Web in a big way. Computer scientists rely heavily on them to stay current with archives, and the Los Alamos model for "E-Print Repositories" has received wide attention. The Library had migrated the IIR-Berkeley's full series of faculty working papers to the Web, handling "calls" for papers and working closely with the faculty editor of the series, Professor David I. Levine. Although no one on the ILE Transition Team had started talking about it yet, the Library staff knew that e-print repositories would open the door to many productive partnerships and to help turbo-power the portal. With that in mind, the Library reached out to the local leaders in the field—the California Digital Library's eScholarship project. The resulting partnership with eScholarship offered an innovative solution for providing "persistent copies" of research papers—with the imprimatur of the University's premier online aggregator.


In October 2001, the library rolled out its first version of the Labor Research Portal. Its main access points are the ILE home page at the Office of the President, and UC Berkeley's IIR Web. Even before it was fully edited and revised, it was snagged in the net of the University of Wisconsin's Internet Scout Project and given "Scout Selection" status. This was an early and welcome vote of confidence. The rollout coincided with two interesting publications that were also immediate hits: new Berkeley Labor Guides on employee democracy, labor culture, and globalization; and a virtual photo gallery of labor art.

The "Labor Culture" Berkeley Labor Guide incorporated lots of public domain graphics and focused on arts and music at a new level of visual richness. It also defined a new format strategy for the series: continuous improvement and revision. Video and sound clips will appear in succeeding editions of the Labor Culture guide.

The Labor Art Exhibit was a joint effort of Berkeley-based librarians and labor specialists, appearing both online and in physical display space. The IIR Directors Lounge became the site for permanent and guest exhibits, with future hopes to circulate exhibits to UCLA. The first guest exhibit was a photo essay of a transportation strike that took place in Tracy, California, in the fall of 2000. The IIR-Berkeley Web site hosted the virtual version on its "Labor Art Exhibits" site, which was linked to the portal.

These two new visual initiatives sparked interest in the California labor community, helping to jump-start dialogue about long-term acquisitions strategies for the library. For example, Fred Glass, Communications director of the California Federation of Teachers, expressed an interest in incorporating his film archives into the library's collections—specifically so the archives could be ported to the Web. Glass produced an award-winning documentary series called Golden Lands, Working Hands, which was shown on several PBS stations.


Turbo-charging the flat portal depends heavily on links to other organizations that employed systems administrators and hosted reputable content. The Labor Research Portal did not have to do everything; instead, the Library staff linked the portal to where the action was.

Within the UC system, the key partner is the California Digital Library (CDL). The CDL is committed to linking vast arrays of library e-content throughout the University of California system, without stepping on the toes of local campuses and departments. It has built a strong advisory structure to keep relations within the UC system smooth. CDL receives a great deal of attention for all it does, but three particular CDL programs had special value for the Institute for Labor Employment and the Institutes of Industrial Relations.


Counting California offers access to all California statistics available online. It's another example of a really great idea with a lot of power that essentially is about as low-tech as you can get while still being on the Web. Counting California brought together data archivists, librarians, and statisticians, with the goal of scanning and porting PDF files of statistical publications to the Web. There's nothing fancy about the current edition of Counting California—except that it works quite effectively as a reference tool. The Labor Research Portal links to Counting California at the top level.

The vast majority of labor-related reference questions—whether originating in the library or via the Internet—require statistical information about the U.S. and California work force, or work force trends. But even though many questions follow similar pathways (such as union membership statistics, strikes and lockouts, etc), no two questions are the same. People don't think in the same categories that the Bureau of Labor Statistics or Bureau of the Census provides: They want statistics by industry, by job type, by region, and by characteristics of the population—all in one handy table. While packaged data exists, essentially the labor reference librarian's task is to teach the researcher how to compile custom tables—over and over again. Counting California helps the researcher at the Ur-level of their quest—the very pages of statistics that form the raw material of statistical profiling can be downloaded and reused. Because of this, the Labor Research Portal emphasizes its own ready-reference tables, which can have a summary format.


One of the biggest challenges for a research university is to gain control of its intellectual capital, particularly in digital format. eScholarship's series of working paper repositories were launched in early 2002, with the goal of throwing a net around all of the interesting and valuable research in progress, which often appears all over the Internet on personal or departmental sites. Taken together, the IIRs and ILE families of research units presented a very large sector of the social sciences faculty, statewide. Every campus received faculty research support from the ILE, and if an organizational partnership could be achieved, CDL and eScholarship would benefit greatly. At the same time, eScholarship had acquired the internal rights to use e-journal software that was developed by Bepress []—another time-saving feature for the ILE.


Pioneered at UC San Diego, CDL Searchlight got its start as a straightforward CGI search program and was quickly adopted as a standard for searching multiple databases in a single search interface. CDL Searchlight shows users how many hits they can find on a search string in all of the databases they identify (with a default to "all" databases). Librarians and super searchers can respond to such an obvious solution by saying, "So what else is new?" but surprisingly few students and faculty have actually discovered and used the program. Linking Searchlight to the Labor Research Portal had two key benefits: It "pushed" CDL's end-user services into the Web world of a large group of social sciences users, and it created new opportunities for the Library staff to emphasize tutorials and bibliographic instruction.


So far, all of the value points illustrated draw on two very simple ideas—focusing the flat Web on high-quality content and forging partnerships with allies who are creating the Web's next generation of finding tools and services. By partnering with others who can bring value (and programming know-how), the Library gained two very powerful benefits. It was freed of the labor of building large libraries of CGI programs and services, and it was empowered to lead the new organization's dialogue about digital publishing.

The turbo-powered flat portal enables librarians to emphasize what they do best—organize, evaluate, and provide access to high-quality resources—while building a new role as publishing partners. As the Institute for Labor and Employment begins to produce a rich corpus of knowledge about work in California, the digital archive is already positioned for inclusion in long-term digital preservation strategies like eScholarship, with the full panoply of power searching tools available. It is much more common for original research to ebb and flow as faculty members shift their research goals or move to other institutions; in this respect, library involvement in content management is a vital organizational talent.


Much of the power in enterprise information portal software lies in its interactive capabilities. Portals are widely accepted, and most universities have developed such services already. Portals are most often found at the campus, "college of..." or "school of..." levels and can be customized to meet departmental needs. It really doesn't make a lot of sense to create another substratum of enterprise information portals at a lower organizational level, in such a richly supported environment. However, it does make sense to maximize the unique, subject-oriented materials that reside within the many domains of research universities. With that in mind, the Library staff has the following recommendations for information professionals who want to turbo-power their flat Webs.

Point at What's Unique. How many library Web sites focus on what the physical library does and describe what the digital library does as a sideline? To distinguish your offerings from the vast array available, focus on the unique content you have. In many cases, this will entail original research and program materials, but in some cases, the unique offering is your own perspective on the field. For example, The Labor Research Portal offers an exhaustive list of trade unions on the Web. One of our visitors, a labor historian, said, "I never knew there was a Web site for the union representing exotic dancers!" The value point was the library's organization of Web links into a useful format—turbo-powering the obvious.

Emphasize Human Analysis. A corollary to the above, but in the era of the Invisible Web, human analysis has proven to be enormously important and is seen by many as absolutely vital. Flat Web sites are discovered by Web crawlers, but many database-driven sites are not. Reviewing and annotating the Web—even if restricted to your own organization's key industries or research areas—is a core competency for information professionals and can boost the value of your Web services dramatically.

Focus on Partnerships. Maybe you employ a digital librarian or programmer who can program Perl, configure Internet Scout's Linux-driven portal prototype, or run a Zope portal—or maybe you don't. If you don't, that's not a reason to wait for a bigger budget. Opportunities for partnerships abound, but most people never ask for help. Perform a "skill audit" for likely partners, and ask them if they want to join forces.

Think Long Term. The IIR and ILE communities reside within an academic community that expects to be in business for hundreds of years. However, many of the leaders of the hidebound university forget to consider issues like legacy system access, the continued usefulness of print, and the risk of losing digital resources by mishap or by a failure to remember that they exist. Turbo-powered portals shine a spotlight on unique materials, and remind top administrators of the many challenges that are attached to digital publishing and preservation—provided these portals focus on the subject area in an engaging manner.

Don't Lock Yourself into the Wrong Web Authoring Platform. The ILE and IIR Web services run on central campus academic servers that cleave to the non-extended HTML and the open Web in a classic Unix environment. It may or may not make sense to acquire a particular vendor's turnkey product, depending on how much support you can expect in your environment. Even in corporate environments, it's important to weigh whether to join forces with an IT department or create a library IT plan; both strategies have worked, depending on the firm. As long you operate within an open Web environment, your options also remain open, so choose wisely when you acquire portal software.


With no new technology outlays, the ILE Labor Research Portal achieved three goals. First, it gained the attention and enthusiasm of its user community, because it provided a universe of information and navigation tools that no other West Coast academic labor site had yet offered. Second, it set the stage for a fully integrated content strategy that linked Web publishing with academic research. Fortuitous timing brought powerful allies in the shape of the California Digital Library, which had similar goals. Third and most important, the portal symbolized the mainstreaming of the Library into the research agenda of the Institute for Labor and Employment, and formalized a "voice" for information professionals in the organization.

Rather than investing substantial effort in acquiring and mastering the hot new software of the day, the Library staff focused on its core competencies to bring order to the dynamic but chaotic world of Web resources. With all its options still open for moving to enterprise information portal software, the staff were left to wonder whether the Unix-driven open Web environment, so common in research universities, was in fact a more powerful alternative. Regardless, the "turbo power" that was employed had to do mostly with human analysis, good writing and research, and a bias for action. Whenever information professionals focus on these factors, the power of the Web itself becomes an ally to the library, its roles, and its relevance.

Terence K. Huwe [] is Director of Library and Information Resources, Institute for Labor and Employment, University of California.

Comments? Email letters to the editor to

[Contents] [ONLINE Home] [Subscribe] [Top] [Information Today, Inc.]