Name authority control is one of the theoretical foundations of the organization of information and relies on the intellectual work of catalogers to link variations of an author’s name under one heading so that a searcher can be reasonably certain that, if the name they seek exists in the catalog, they can find it and that all works by a given author are traceable under a single heading.² Both Jeng and Tillett made the point that this practice enhances the recall and precision of name searching in the library catalog.³ The justification for name authority control is that patrons, who may be less skilled in finding names, will not find the correct entry if information professionals do not supply the links between different forms of names. In 1985, Burger defined authority work in computerized catalogs as creating authority records, gathering these records into a file, linking authority records to bibliographic records, maintaining authority files and systems, and evaluating the results.⁴ He stated that the role of authority control is to facilitate access and that it aids in both the finding function and the gathering function of the library catalog.

Bangalore and Prabha found that in practice, libraries may opt for full authority control of all potential access points, moderate control of only headings with cross-references, or minimal control of only frequently used cross-references, and that deciding on the level of authority control is based on several factors, including collection size, patron body, and funding.⁵ Some libraries centralize authority control in one unit or with an individual while others decentralize the process. Bangalore and Prabha found that “library approaches to authority control vary widely.”⁶ Zhang noted that while authority work is routine for large research libraries, it has been more challenging for small and medium-size academic libraries.⁷ She went on to state that even in large research libraries the levels of control vary.

In the mid-1970s, changes in the ways that libraries performed name authority began to accelerate. The Library of Congress Name Authority File became the National Name Authority File (NAF) with the creation of Name Authority Cooperative Program (NACO) of the Program for Cooperative Cataloging (PCC) in 1976. In 1977, the Library of Congress (LC) started keeping a computerized LC/NAF, which it made available to other libraries in quarterly microfiche beginning in 1979. The LC/NAF then became available in machine-readable format through OCLC in December 1980.⁸ This was an important development for library online cooperative union catalogs, which existed in part as venues for libraries to share cataloging records. As libraries made the transition from card catalogs to computerized catalogs, and as those catalogs became linked to the online union catalogs, consistency in heading formation became essential for maximized search and retrieval.⁹ According to Taylor, vendors became involved and began offering various name authority control services around 1983.¹⁰

The new cataloging rules, RDA, soon will usher in changes to cataloging. The last major cataloging rule change had sweeping effects on cataloging policies and practices, including name authority. Notably, after the implementation of AACR2, catalogers were concerned that their name authority files that existed prior to AACR2 would no longer be in compliance with national rules, and they struggled with solutions. Some libraries chose to avoid the conflict and closed their card catalogs in favor of switching to online catalogs.¹¹ Tracing family names is a new direction in which RDA is expected to move.¹² Families as a form of collective author have been used in certain information institutions such as archives, but not in libraries under the AACR2 rules. Although AACR2 did not cover family names, MARC21 has had an option to accommodate family names with an indicator of “3” in the name fields.¹³

Several potential future directions for developments in name authority control are being explored. The Virtual International Authority File, for example, is a joint project between the national libraries of Germany and France, the LC, and OCLC to “match and link the authority records for personal names in the retrospective personal name authority files of the Deutsche Nationalbibliothek (dnb), the Library of Congress (LC), and the Bibliothèque nationale de France (BnF).”¹⁴ Another potential direction for name authority concerns international name authority control. In recent years, catalogers have been addressing the question of international authority and discussing the ramifications of different means of authority control. The International Federation of Library Associations and Institutions (IFLA) proposal of an ISBN–type number for authors, the International Standard Authority Data Number (ISADN), to replace name authority control was initially introduced in 1980, and recently the idea has been gaining momentum. IFLA has a working group charged with defining the functional requirements of authority records, studying the feasibility of the ISADN, and serving as a liaison to other groups concerned with authority files.¹⁵ Recently, Functional Requirements for Authority Data: A Conceptional Model, edited by Patton, was published after being approved by the Standing Committees of the IFLA Cataloguing Section and the IFLA Classification Section.¹⁶

Concern over the cost of authority control, reflected in hours of cataloger time, is frequently discussed but is not a new issue. The following study question from a 1936 cataloging course syllabus from Columbia University School of Library Service illustrates this: “Catalogers are sometimes criticized because of the time and energy they spend in ferreting out the names of authors who wish to conceal their identity. What can you say in answer to this criticism?”¹⁷ In a 1953 reference to name authority, Tauber criticized the practice: “The cost of this research has increased considerably the cost of original cataloging, often needlessly and with no definitely beneficial results. The value of the information obtained, and its use by the public, has long been questioned in relation to the cost.”¹⁸ Bangalore and Prabha recorded the amount of time spent on each aspect of cataloging for seventy-five monographs. Name and series authority accounted for, on average, one-quarter of total cataloging time. In the most extreme example among the observed cases, the name/series authority work, took 87 percent of the total cataloging time for a Ukrainian language book, and in several examples the authority work took approximately 50 percent of the cataloging time.¹⁹ Sullivan estimated that subject, name, and series authority work at Yale Law Library took one-third of the time to catalog a book.²⁰

Authority work entails more than checking authorities when cataloging an item. Even more time-consuming is the database maintenance task of continually comparing the in-house authority file to the long list of updated headings issued weekly by the LC.²¹ Calhoun and Oskins stated, “It is safe to assert that if it is not kept up to date in some manner, inaccurate and conflicting headings will accumulate rather rapidly in a local system file, even one which has been cleaned up and brought into conformance with the LCNAF and SAF before being loaded.”²² Database maintenance is at the heart of the correct functioning of authority control. To control internal costs, in the late 1980s many libraries turned to vendors for a one-time cleaning of the automated authority file or ongoing authority file maintenance, or both. Such vendor services are expensive, and many studies have explored both the accuracy and the cost-effectiveness of going with vendor services.²³

To examine whether the costs involved in name authority control are justified, many researchers have analyzed whether name authority control actually improves the search results received by patrons in a computerized library catalog. An early study by Taylor in 1984 and a 1995 study by Bangalore both examined user transaction logs concerning name searches in OPACs to determine the usefulness of name authority control.²⁴ Both authors found most failed name searches in OPACs would not be helped by name authority control. Bangalore stated that she believed “that online linkage of authority records to bibliographic files is expensive, but justifiable.”²⁵ She added, “Many authority records have no cross references, and many of the records with cross references are not the forms sought by users.”²⁶ Both she and Taylor suggested that computer programming changes could boost name search matches such as inverting name searches automatically, imbedded spell-check, “order, spacing and punctuation inaccuracies, and common keyboarding errors.”²⁷

In another study criticizing authority control, Jeng argued that authority control does not serve the needs of end-users as well as better search interfaces could and concluded that it is not a good use of time and resources.²⁸ Using another approach to evaluating authority work, Pappas examined error rates and found in an analysis of eight RLIN members’ authority-controlled access points that 16 percent of personal name fields in MARC contained authority errors.²⁹ This represented one-quarter of all errors found in the 1,710 authority controlled headings in four hundred records analyzed. More recently the trend of advocating against authority control may be reversing. In 2004, Byrum stated that there was a “generally skeptical attitude about the cost-effectiveness of name authority work that prevailed in the 1970s and into the 1980s,” but that there is increased recognition by administrators of the value of authority control.³⁰

The final issue examined here is not one of name authority control per se, but of cataloging practice in general. Are libraries losing cataloging and other technical services positions? If so, how are they accommodating these changes? Wells collected data from seventy public graduate university libraries in the Southeast and found that 63 percent reported having lost technical services positions since 1990.³¹ Of those that had lost positions, three-quarters reported that the lost positions were librarian positions. Cataloging backlogs, cataloging new materials, and authority control were identified as the job tasks most strongly affected by the loss of positions.

Within cataloging, what types of employees are doing name authority work? Wolverton found that 94 percent of the doctoral institutions he studied stated authority work was done by Masters of Library and Information Science (MLIS) librarians, and 78 percent also had paraprofessional staff involvement.³² Wells reported that respondents, half of whom were from public libraries, in her study indicated that 80 percent of the authority work was done by MLIS librarians and 42 percent by paraprofessionals.³³ Bordeianu and Seiser surveyed libraries of the Association of Research Libraries in 1999 and found that 84.5 percent used paraprofessionals for copy cataloging and 67 percent used paraprofessionals for original cataloging.³⁴ They state that “several studies performed during the past 2 decades have indicated that paraprofessional participation in cataloging has been increasing steadily since 1977.”³⁵

In theory, authority control is central to the ideals of information organization. Several authors have observed that, in practice, it is expensive and time-consuming in an era of stagnant library budgets and the loss of technical services personnel, and it is largely invisible to patrons.³⁶ In addition, studies such as those by Taylor and Bangalore found that authority control would only have a minimal effect on improving hits from patron searches.³⁷ Despite these issues, are libraries still pursuing name authority control for their collections? Are rates of participation in name authority control affected by institution type or collection size? The central hypothesis for this study was that graduate institutions and the largest libraries would engage in most aspects of name authority control and smaller libraries and institutions offering lower degree levels would do fewer aspects name authority control. This hypothesis was based on the assumption that libraries with larger and more complex collections would derive more benefit from good name authority control, that they would be more likely to be authorized to contribute original name authority records to the NAF (i.e., have NACO–trained librarians), and—since they presumably have more financial and personnel resources—that they would be more likely to do their cataloging and name authority work in-house and less likely to outsource to vendors. The authors asked the following research questions:

1. What effect do institution type and collection size have on name authority work?
2. What name authority work is being done?
3. How are name authority processes conducted at different libraries?
4. Who does the authority work?

The authors used stratified random sampling to select one hundred United States institutions from each of the following categories: two-year colleges (community colleges); four-year colleges (baccalaureate institutions), and graduate institutions. The samples were taken from the 2008 editions of Peterson’s Graduate Schools in the U.S., Peterson’s 4 Year Colleges, and Peterson’s 2 Year Colleges.³⁸ Contact information for someone in the library, preferably the library head of cataloging, head of technical services, or similar position was obtained from reference sources. During late November and early December 2008, the authors sent e-mail messages to the identified person inviting them to participate in a survey posted on the Web tool Survey Monkey or to forward the survey to an appropriate spokesperson for their library. After two weeks, a follow-up e-mail message was sent to stimulate additional responses. Graduate institutions had a response rate of 40 percent, four-year schools had 35 percent, and two-year schools had 24 percent, for a total of ninety-nine responding libraries out of the three hundred initially contacted. Although a probability sample was taken, results of the study should be generalized with caution to academic libraries as a whole because the response rates were fairly low and the final samples, especially for community colleges, were small.

Many of the survey questions were based on previous studies by Wells and Wolverton and were meant to expand upon the results that they found.³⁹ Wells compared public libraries with community colleges, four-year colleges, and university libraries in Mississippi. While she received a reasonable number of responses from public libraries (thirty-two), she only had data from sixteen community colleges, eight four-year schools, and seven universities. Wolverton had an excellent sample of 193 respondents, but he only surveyed doctoral institutions, so his results cannot be generalized to smaller libraries or to libraries serving other types of academic institutions. Both of these researchers did an excellent job of addressing an information need in the cataloging authority control literature, and this study is meant to expand upon their contributions by comparing different institution types (as Wells did) with larger sample sizes. Other questions in this study were drawn from ideas from the literature and from the work experience of the study authors. See the appendix for the survey questions used.

The authors compared data with percentages. When the percentage differences were sizable, chi-square for independent groups was used to test for statistical significance of the difference. While percentages give an indication of differences between groups, the chi-square test indicates whether the differences between groups on that question were large enough that they were unlikely to have occurred by chance. The chi-square was calculated using frequencies in 2 x 2 contingency tables and tested for significance at the .05 level, which requires a chi-square result of greater than 3.841.

Because name authority work is part of cataloging in general, the first survey question was a background question that asked how much cataloging was done by the library (in-house) compared to cataloging provided by a vendor (outsourced). Two-year institutions were most likely to report doing “almost all” cataloging in-house at 88 percent compared to 58 percent of graduate institutions, a statistically significant difference. The differences between graduate and four-year institutions and between two-year and four-year institutions were not significant. Graduate institutions were most likely to state that “most cataloging is done in-house, but some is outsourced.” When the answer categories “almost all” and “most” were combined, graduate institutions and two-year schools were equivalent at 88 percent of both types, along with 80 percent of four-year schools.

Concerning cataloging practices by collection size, 90 percent of mid-size libraries and 80 percent of small libraries stated that “almost all” cataloging was done in-house, both significantly more likely to do so than large libraries at 43 percent. When “almost all” and “most” were combined, all collection sizes agreed at 83 to 90 percent (see table 2).

The Name Authority Cooperative Program (NACO) of the Program for Cooperative Cataloging allows authorized, trained participants to contribute original name authority records to the NAF regulated by LC. In this study, large libraries were significantly more likely than small libraries to be NACO contributors, and graduate institutions were significantly more likely to be NACO-authorized than two-year schools (see table 2).

Bangalore and Prabha discussed variations in the application of authority control depending on collection size, patron body, and funding, noting that there was a great variety of authority practice among libraries.⁴² This study tested their assertion by examining institution type and library size in relationship to name authority control in cataloging. While both of these variables were related to differences in practice, analyses showed more variation by collection size than by institution type.

Initially, the authors asked respondents how much cataloging was done in-house at their library as opposed to outsourced. For all institution types and library sizes, 80 to 90 percent stated that “almost all” or “most” cataloging was done in-house. Surprisingly, large libraries and libraries at graduate institutions were significantly more likely than other libraries to outsource some of their cataloging, while other libraries were more likely to do almost all of it in-house.

One of the central hypotheses in this study was that large libraries and graduate institutions would be most likely to do various kinds of authority work and most likely to do the work in-house. The first part of this hypothesis was partially supported—large and medium libraries were more likely to report that they “always” or “usually” did name authority, although there was little difference by institution type. This study and Wolverton’s both found that almost all graduate institutions do at least some authority work, with only 5 percent of Wolverton’s and 8 percent of this study’s graduate institution respondents stating that they never did authority work.⁴³

The twenty-five libraries that did not do name authority work reported several reasons for not doing so. Slightly less than half (48 percent) stated that they would like to do authority control but did not have the staff to do it. Another 40 percent stated they would like to do it but did not have the budget to support it. One-fifth (20 percent) stated that name authority control was not that useful because of their collection type or size. Few libraries (16 percent) chose the option that name authority control is of less value in an online catalog. A two-year and a four-year school said that name authority control costs more than the value it adds to records. These findings echoed those of Wells, whose respondents cited predominately lack of funds or lack of staff time as reasons not to do authority control.⁴⁴ In Wolverton’s study, of those libraries that did not do authority work, a similar 41 percent claimed insufficient funds, but half as many as the current study (24 percent) stated insufficient staff.⁴⁵ Wolverton’s study also found that 47 percent claimed they did not have time to do the work, a variable that could be related to insufficient staff.

An aspect of doing name authority work is establishing new headings, a task for which not all libraries are authorized to contribute outside of their local authority file. Wolverton’s study only collected data from institutions with the Carnegie classification as Doctoral/Research Extensive or Intensive, and 41 percent of his respondents stated they were NACO- authorized, compared to 35 percent of the graduate institutions in this study, a similar result.⁴⁶ Comparing within institution types in this study, graduate institutions were significantly more likely to be NACO-authorized than two-year institutions, but not significantly more likely than four-year schools. By collection size, large libraries were significantly more likely to be NACO-authorized than small libraries, but not significantly more likely than medium libraries. Being NACO-authorized was related to being more likely to do name authority: 82 percent of NACO-authorized compared to 63 percent of non NACO-authorized libraries said they “always” or “usually” did name authority, though this difference was not statistically significant.

Personal names were by far the most likely to be controlled by libraries in this study, and uniform titles least likely. Large libraries were significantly more likely to do name authority control than small libraries on names other than personal names. Medium libraries also were significantly more likely to control some types of names—corporate, geographic, and conference—than small libraries. Wolverton found a slightly lower percentage of graduate institutions controlling personal names, 88 percent compared to 100 percent in this study, but his respondents reported a slightly higher percent controlling conference names, series, and uniform titles (79 to 88 percent compared to 69 to 78 percent in this study).⁴⁷ It is difficult to compare Wells’ results across institution types because her sample sizes were too small to calculate percentages by type.⁴⁸

Almost all library types reported controlling name authorities for monographs. Serials and electronic materials were significantly more likely to be controlled by graduate institutions than by two-year schools. By collection size, while name authority was controlled for monographs by almost all libraries, large libraries were significantly more likely than small libraries to control serials, electronic materials, and other nonbook materials. Medium libraries also were significantly more likely to control electronic materials than were small libraries.

The timing of when authority work is done can vary between institutions. While institution type was only moderately linked to when name authority work was done, collection size was strongly related to the timing of the authority work. Small and medium libraries were significantly more likely to do their name authority during the cataloging process and large libraries significantly more likely to do it postcataloging, such as during database maintenance. Very few libraries reported doing name authority during a precataloging process.

Two-year schools were significantly more likely to verify name authority on received vendor records than were graduate institutions. The only significant difference by collection size was that medium libraries were significantly more likely than large libraries to do name authority work when copy cataloging. Libraries were most likely to use OCLC for verifying name authority, followed by using the LC authorities website and libraries’ local authority files. Libraries were unlikely to use vendor files, regional cataloging systems, or other libraries’ databases. Small libraries were significantly less likely to use verification sources that cost money, but there were no meaningful differences between library types and sizes on the use of no-cost verification sources. Wells’ respondents were most likely to access name authority through their local authority file (53 percent, a bit more likely than respondents in the current study) and OCLC (45 percent, much less likely than respondents in this study).⁴⁹ Only 13 percent of Wells’ respondents used the LC internal system, as the no-cost LC authorities website was not yet available at the time of her study.

When the authors asked respondents about their catalog’s authority control module, they found few differences between libraries. Depending upon institution type or collection size, 47 to 60 percent of respondents reported that their authority control module was integrated with their catalog. Small libraries and libraries at two-year colleges were more likely than other libraries not to have an authority module, and small libraries were particularly less likely to have an authority module that was integrated with the bibliographic file. In 2000, Wells’ study found a higher percentage of libraries without authority modules—36 percent compared to the 0 to 27 percent in this study.⁵⁰ This difference could be a result of eight years of continuing technological development or perhaps because half of her respondents were public libraries while this study only examined academic libraries. Wells also had several respondents who reported not having an online catalog at all, whereas no libraries in this study lacked online catalogs.

Wolverton found that authority work was done by MLIS librarians in 94 percent of doctoral institutions, very similar to the 92 percent of graduate institutions and 93 percent of four-year schools found in this study.⁵¹ His study reported a higher rate of participation by paraprofessional staff—78 percent compared to 57 percent of graduate institutions and 37 percent of four-year schools found in this study. Wells noted a slightly lower rate of participation by MLIS librarians (80 percent) and a 42 percent rate of involvement by paraprofessionals.⁵²

Bangalore and Prabha suggested that authority work is centralized in some libraries to one or a few individuals while in other libraries it is decentralized.⁵³ In this study, two-thirds of those respondents whose cataloging departments were large enough for this question to be applicable reported that authority work was decentralized to each cataloger while one-third said it was centralized to one or a few people.

Librarians have expressed concern about the deprofessionalization of cataloging as cataloging procedures have changed over the past thirty years. As quoted earlier, Bordeianu and Seiser stated a decade ago that “paraprofessional participation in cataloging has been increasing steadily since 1977.”⁵⁴ Wells and Wolverton asked their respondents about librarians versus paraprofessionals in name authority control, and this study asked this question again.⁵⁵ All three studies found that between 75 and 90 percent of institutions reported that name authority control was done by MLIS librarians. Paraprofessional involvement in authority control was reported by 42 to 78 percent of libraries. Graduate institutions were much more likely to involve paraprofessionals in this activity than two-year or four-year schools. Other than graduate institutions, fewer than half of libraries reported paraprofessional involvement in authority control, so one might conclude that authority control is still predominately treated as a professional activity in library practice.

There was a substantial difference between library types and sizes concerning outsourcing. Contrary to the study hypotheses, large libraries and graduate institutions were the least likely to do “almost all” of their cataloging and name authority work in house, but instead they were more likely to report some outsourcing for cataloging and authority work. They also were more likely than small libraries to contract for ongoing vendor services, such as an initial cleansing with ongoing reviews, vendor notifications of name authority changes for existing name headings, and receiving cataloging records with purchased materials. Large libraries were much more likely to report doing their name authority control during database maintenance rather than during cataloging itself; this also may be a reflection of the use of vendor services. Small and medium libraries were more likely to purchase one-time or occasional vendor services, such as file cleanups.

This study had a slightly higher rate of graduate institutions reporting using vendors for authority work—53 percent compared to 45 percent in Wolverton’s research. However, the two groups used vendors differently. Vendors performed a one-time run of the authority files of 35 percent of respondents in this study, compared to 23 percent of Wolverton’s respondents. Additionally, 40 percent (compared to 60 percent of Wolverton’s respondents) employed an initial run with ongoing reviews, and 45 percent (compared to Wolverton’s 65 percent) had vendors supply notification of changes in existing authority headings. This study, then, showed graduate institutions with less of an ongoing relationship with vendors than the previous study found.⁵⁶

Wells found that two-thirds of technical services departments at graduate institutions had lost positions through restructuring or downsizing, and this study also found two-thirds of graduate institutions had lost cataloging positions.⁵⁷ This trend was less pronounced in two-year schools and small libraries, where approximately 40 percent reported lost positions. A possible reason for this may be that smaller libraries have fewer positions to lose. An administrator might hesitate to cut the only staff or librarian position in a small cataloging department, but reducing staff might seem more manageable in a larger department with several employees. The loss of staff positions only affected the name authority control practice in some libraries, and this was most likely to happen in large libraries. These data suggest that different institution types and sizes lost positions at different rates, and the results of the losses were experienced differently by different kinds of libraries.

This study had two main aims: to build upon previous research on name authority control in order to enhance knowledge of how name authority work is being conducted by different types of academic libraries, and to capture a snapshot of how name authority work is currently being done before the new cataloging rules lead to changes in practice. The results will allow future research to study the effect of the new rules on name authority practice. For example, the concept and use of uniform titles will be somewhat different in the RDA cataloging rules, therefore a useful future research question might consider whether the control of uniform titles will increase under the new rules. This study may provide information that will help explore the effect of international developments in name author control on name authority practice in the United States. The results of this study also could be useful to administrators who are evaluating cataloging practice in their libraries. Knowledge of how similar institutions are participating in name authority could be helpful in decision-making processes for allocating budgets and staff positions or for considering various outsourcing services.

Much of the literature concerned whether name authority control added enough value to records to justify the cost. While this study did not examine the efficacy of name authority for the end user, the results show that name authority work is very entrenched in academic library practice, almost universally so for the control of personal names and for monographs. This seems to indicate that its value is accepted in the profession.

The foundational hypothesis of this study was partially incorrect and partially supported. Contrary to expectations, libraries at graduate institutions and the largest libraries were the most likely to outsource some of their cataloging and some of their authority control, while smaller libraries and libraries at nongraduate institutions were more likely to do this work in-house. The remaining part of the central hypothesis was supported—large libraries and those at graduate institutions were more likely to participate in more aspects of name authority control than other libraries, controlling a wider variety of names and controlling names for more types of materials. While previous studies analyzed work processes by institution type, this study found that the larger differences in practice were more associated with collection size than with institution type. Researchers should keep this in mind when designing future studies.

The authors identified several areas of statistical significance when libraries were compared by collection size. Large libraries were significantly more likely than small libraries to be NACO contributors. They were significantly more likely to control every type of listed name other than personal names. Large libraries were more likely than small or medium libraries to do their authority work postcataloging, such as during database maintenance, and more likely than small libraries to use vendor notices for heading changes and to contract with vendors for file maintenance. They were more likely to use purchased sources (such as OCLC) for verifying name authority. Additionally, they were more likely than small libraries to have paraprofessional staff doing authority work. Medium libraries were significantly more likely than large libraries to verify name headings when doing copy cataloging and significantly more likely than small libraries to control corporate, geographic, and conference names. They also were more likely than small libraries to use purchased sources for heading verification. Small and medium libraries were significantly more likely than large libraries to state that they did “almost all” cataloging in-house. An analysis of which items were significantly different by collection size suggests that the majority of practices in which large libraries were significantly more likely to engage than small libraries concerned the use of purchased vendor services and therefore were budget-related.

Fewer areas revealed a statistically significant difference by type of institution. Graduate institutions were significantly more likely than two-year schools to be NACO-authorized and were significantly more likely than two-year schools to control name authority for serials and electronic materials. They also were significantly more likely than four-year schools to control uniform titles. Two-year schools were significantly more likely than graduate institutions to state that “almost all” of their cataloging was done in-house and to state that they verified name authority on the records that they receive from vendors. For libraries by type of institution, areas with statistically significant differences generally concerned types of name authority work being done.

This study did not collect data about budgets and funding, but the findings highlighted many questions about the effects of funding on name authority practice. A variety of patterns in the data suggest that name authority procedures followed by different-size libraries could be related to limits imposed by the smaller budgets of small libraries compared to the budgets of larger libraries. Small libraries were less likely to use vendor services for cataloging and name authority work. They also were less likely to purchase services such as OCLC for verifying name authority. When they did use vendors, they were more likely to use a one-time service or an occasional cleanup rather than ongoing reviews, ongoing vendor notification of name authority changes, or other ongoing vendor work. They also were less likely to have an authority control module as part of their library catalog. The most common reason given by small libraries for why they did not control name authorities was because they did not have the staff to do so. Future studies could benefit from asking more explicit questions concerning the effect of the budget on name authority decisions.

Replicating this study with public libraries and perhaps school library media centers might be valuable. Collecting this additional information would allow a much more in-depth comparison of name authority practice across library types. A larger sample size would be helpful so that the results have more weight and because larger samples lend themselves to additional statistical techniques.