A recent article in The Crimson reports a largely favorable assessment of the major restructuring of the Harvard library system in 2012 into a single, centralized administration. According to the article, the library has saved $25 million since 2009, most of this attributable to policy changes enacted between 2010 and 2012, in which the University "reshuffled and streamlined administrative positions in the library system to reduce inefficiencies and reallocate resources to better balance the needs of a centralized library."

A central aspect of this "streamlining" was the creation of a new "shared services" bureaucracy that was supposed to allow a smaller organization to perform its tasks more efficiently by pooling resources and minimizing duplication throughout the library system. For staff such as myself, the reality of "shared services" on the ground has been the creation of a highly inefficient system of bureaucratic fragmentation, with staff at each library divided both between local and shared contingents, and with shared services staff now accountable to separate offsite managers.

In technical services, our experience suggests that budgetary savings since 2012 have been due not to increased efficiency, but rather to severe cuts in quality and to the outsourcing of library work. These trends have allowed the library to continue processing books with smaller staffing, but this has come at the sacrifice both of the quality of its collections and of its contractual commitments to its employees.

In February of 2014, I participated in a workflow initiative on copy-cataloging aimed at harnessing the supposed new advantages of the integrated library system. Although we were led to believe that the workflow initiative would allow staff to develop new policies in a participatory way, in reality, we were handed a set of policies prepared in advance by managers and expected to approve them with little or no revision. Several of us expressed strong concerns that the new policy guidelines would severely reduce the quality of cataloging at Harvard, and thus reduce the ability of patrons to find materials in our collections. Our concerns were strong enough that all five of the copy-catalogers working on the project resigned. The new standards were nevertheless adopted by central administration for processing books in the main library collection.

It seemed clear to the copy catalogers involved in the project that the new cataloging standards were designed to facilitate outsourcing by allowing substandard vendor supplied cataloging to be accepted without meaningful revision by staff. These new, reduced cataloging standards are now being used to overlook the errors already found in a recently introduced workflow called "direct to destination" where outside, for-profit companies catalog for Harvard and send titles directly to shelf without review by local expert staff.

This is now the overall trend in the Harvard Library: In collection development and technical processing, there is now a strong impetus from central administration to outsource work to vendors wherever possible. More and more books are purchased on blanket orders with cataloging and shelf-ready processing supplied by vendors. This means that our collections are less and less often being conscientiously built by trained bibliographers who know both the collections and the needs of the disciplines they serve, and more and more by vendors with a narrow interest in selling books.

Librarians and library staff at all levels have a basic accountability to the university and the intellectual mission it serves. By outsourcing our work to vendors, Harvard is ceding the independent function of the library as a curator of information in the service of knowledge to a handful of private corporations whose main interest lies in maximizing profits.

In the short term, such policies may look good to people concerned about budgetary savings, but the long term cost to the quality of Harvard's collections may be inestimable.

Noah D. Cohen is a library assistant at the Harvard Law School Library.