The USDA Economics and Statistics System, a joint partnership between the Albert R. Mann Library and three economic agencies of the U.S. Department of Agriculture (USDA), began three and a half years ago. The partnership with these agencies, the Economic Research Service, the National Agricultural Statistics Service, and the World Agricultural Outlook Board, supports a network site for up-to-date and historic agricultural economic information.

Currently, the site maintains 79 full-text serials and nearly 200 historic data series. A significant portion of the data received is time sensitive, which requires the library to post the information to the Internet within minutes. We disseminate this information via WWW, gopher, telnet, ftp and e-mail, and this work is supported by 11 librarians and 7 library assistants.

You may ask why would a library seek out a partnership which requires a considerable commitment of staff time and resources?

The answer lies in the mission of the institution. Mann Library supports research and education for a land grant agricultural college. The land grant concept, as originally defined in the Morrill Acts in the 19th century, sets aside public land to fund institutions of higher education which would provide instruction in the agricultural and mechanical arts. For over a century, the Department of Agriculture has promoted agricultural research and practice at these institutions, and the land grant libraries have actively acquired USDA publications about this research.

Mann Library is also a depository library which collects agriculture and related information published by the Federal Government, and makes this information available to the general public.

The library collects USDA information in whatever format it is published: paper, microfiche, digital; and has a particularly strong collection in the area of agricultural economics. Therefore, we were especially excited when an opportunity arose for the library to collaborate with the Economic Research Service and acquire complete sets of their digital information.

In one agreement, the library was able to acquire hard-to-get information, and experiment with new ways to deliver this information to citizens who might never enter an agricultural library. When we consider the fact that we distribute over a half million full-text documents and binary data files annually from this network service, we are convinced we made the right choice.

Mann Library’s experience with the USDA partnership has been both a challenge and a change for us. Information in a digital environment is far more dynamic than print. In our arrangement with USDA, data is arriving at all times during the working day. We must frequently juggle our staff schedule in order to have staff ready to receive, review and process the information. An added complication is that the University holiday calendar is significantly different from the Federal one, and on some of our days off we require staff to be able to process the data from their home.

Also, much of this data is time sensitive, and we have to hustle to get it posted to the network server. Almost daily there are minor obstacles to navigate: editors miss deadlines and the information arrives late. Or Internet performance is uneven, and, not so rarely, unavailable. And sometimes we discover small errors which need to be fixed before the data goes out.

Finally, we have to be creative in solving access problems experienced by remote users. For example, last week I had an exchange with user X. This person wanted a text file, but could not retrieve it with his WWW browser for an unknown reason. We couldn’t reproduce the problem with our software, but we couldn’t assume this patron’s problem wasn’t real. After an exchange of information with this user, I sent the file, a full 450 Kb, via e-mail. It arrived complete, he was happy, and mentioned in his thanks that he had a daughter who graduated from Cornell.

The partnership has also revolutionized our work. In the past, we developed a top-notch system to process depository documents. With the USDA information, we are developing a top-notch system to process digital government information. To bridge the two we rewrote our job descriptions.

Four years ago, when we loaded the first USDA data onto our server, our Information Technology (IT) personnel were afraid to give our acquisitions staff an account, much less unrestricted privileges, on the UNIX-based network systems in the library. To solve the problem, we provided sustained, in-depth training, purchased up-to-date equipment and software, and re-engineered our work processes.

Today, the acquisitions staff routinely modify both processing software and the data archive contents on those same UNIX systems, and the IT staff are looking to transfer other library computer functions to the division. We didn’t remove or replace the staff; instead, by showing we could master digital skills, we replaced inappropriate IT staff preconceptions.

We also rethought library functions. You might think acquisition library assistants rarely answer reference questions. Not so with USDA information. The acquisitions staff routinely answer access and content questions from remote users.

Also, we have integrated the USDA titles into our other electronic titles on the Mann Library Gateway. The classic HTML Web page is staggering under the growing load of titles. Devising better home pages is no longer thought of as a "reference only" or "cataloging only" prerogative. Instead, it is an effort distributed throughout the library.

Finally, we are breaking down the division between librarian and programmer. We are not looking to make librarians into programmers, but we are insisting new computing projects be designed, and the appropriate staff be trained, so that the project can reside, be upgraded and maintained at the division level of the library. This practice was initiated with system enhancements to the USDA system.

We are also having to reconsider collaboration with other libraries. Mann Library is mostly self-sufficient, but we’re aware there is more USDA information to be managed than any one library can manage. Ultimately, much of what we do will have to complement other networked academic/USDA partnerships. We are aware of the need to plan ahead to prevent overlap of effort, and to guarantee that essential services are not overlooked. We hope to facilitate the process through meetings and personal contacts.

We believe agency/academic partnerships benefit both the agency and the library, but the arrangement requires periodic re-examination. Three years ago, both the library and the Economic Research Service agreed that there was a critical need to create network access to digital information. Today, that need has been met; indeed, the partnership has developed excess file delivery capacity. Re-evaluation establishes new priorities for both partners, and while it is not clear which priorities will now rise to the top, the process must take into account changes in both technology and the political climate in the University and Washington.