Editorials

We’ve gotten a lot of mail in the past few weeks about the Berkeley Public Library’s plans for introduction of radio frequency identification devices (RFID) in a time frame close to proposed staff reductions. Library managers contend that there’s not a causal link; Library Director Jackie Griffin sent the Planet a copy of a letter to someone who raised the issue in which she said “the reduction in hours is a financial reality given our budget deficit as is the reduction in staff. Reducing the staff is not caused by RFID nor was the decision made to purchase RFID taken with reducing staff as a goal.” On the other hand, the initial technology buy for RFID is at least half a million dollars, probably more, and the money has to come from somewhere. Proponents suggest that implementing this kind of mechanized system, which enables readers to check out their own books, will save money by reducing workers’ compensation costs, but they haven’t supplied data which backs up this hypothesis. It’s not surprising that there are still some skeptics out there. 

We dropped in on a recent meeting of the Library Board of Trustees at which this topic was discussed. Our first observation was that the meeting room at the South Berkeley Branch Library was dramatically too small for the number of people who wanted to attend. People were lined up 10 and 20 deep out of every door. Since one well-advertised feature of the lavish new Central Library building is big meeting rooms, the choice of the little room at the South Branch indicates that someone in a planning capacity is out of touch with public opinion on how important the upcoming decisions are for Berkeley. This information, coupled with the fact that for the first time in recent memory a library tax measure lost in an election, suggests that Library Board members need to do a better job of exploring exactly which cost-cutting measures are likely to meet with civic approval. The minutes of that meeting and previous meetings where the staffing cuts were discussed have yet to be posted for the information of those who couldn’t squeeze into the meeting room, unfortunately.  

Last Saturday night we attended a sold-out gala at the main library where patrons paid handsomely for a filet mignon dinner and the privilege of sitting at table with an author with a Berkeley connection. The evening’s theme was “Love the Library,” and the enthusiastic participation of both authors and patrons suggested that a lot of people do indeed love the idea of having a fine local library. But libraries without librarians, or at least with fewer humans in the mix, are not apt to engender the same amount of affection in the long run.  

When we looked on the library’s website trying to figure out what time the dinner started, we discovered that their computer system had been on the blink for several days. That reminded us of another problem with RFID, a principle we learned from 20 years in the software business: It never pays to be an early technology adopter. In the few minutes we were able to squeeze into the back of the room at the South Campus Branch we heard a dispute over whether 100-plus users of the chosen system worldwide was a lot or a few. The answer, of course, is “it depends.” It’s certainly not a lot, but is it enough? Is there published evaluation literature on the system’s robustness and efficacy? A quick check on the web found one librarian’s blog complaint about persistent malfunctions in the RFID system she was using. 

Then there’s the whole complex question of whether RFID technology is a long-term threat to privacy. The Electronic Frontiers Foundation, a very respectable, technologically savvy watchdog organization, certainly thinks so, as does the ACLU. Both are adamantly opposed to any large-scale implementation of RFID technology in libraries given the current state of the art. In a recently published paper, UC computer scientists David Molnar and David Wagner say that “current conventional wisdom suggests that privacy risks are negligible unless an adversary has access to library databases. We show this is not the case.” 

Berkeley Public Library management touts its adoption of what they call “best practices” for preserving privacy while using state-of-the-art RFID. However Molnar and Wagner looked at BPL’s best practices and concluded that while they are indeed the best available given today’s technology, they are not good enough to overcome the inherent privacy risks which still exist: They do not in fact prevent the “unauthorized reading of tag data” which many fear. 

All in all, it seems that leaping ahead with RFID implementation presents some big problems. Each controversial topic (staffing model, RFID costs/benefits, privacy) needs its own intelligent discussion. At a minimum, the Library Board of Trustees needs to hold a well-advertised public meeting in a hall big enough to accommodate everyone who has information to share or an opinion to contribute. Otherwise they’ll hear the voice of the people of Berkeley only when it’s too late, at the ballot box.  

—Becky O’Malley