Information regarding the extent of the nation's investment in scholarly research
in the humanities is limited. Reliable data on expenditures for academic research
in the humanities became available only in the early 2000s, when the
National Science Foundation
(NSF) first began including the humanities field on the annual academic research
and development (R&D) survey that it has been using to collect information about
science and engineering R&D since 1972. On the other hand, information about
research libraries, the laboratories of the humanities, is plentiful. Several decades'
worth of data from the
Association of Research Libraries
(ARL), as well as the U.S. Department of Education's
National Center for Education Statistics
(NCES), make possible an analysis of trends in the size of collections, in budgets,
and in library staffing, as well as the extent to which these are keeping pace with
the size of student populations. These data also permit an examination of the relative
costs of scholarly serials and monographs. However, these data do not deal specifically
with humanities collections or the librarians who oversee them. Both the ARL and
NCES describe the scholarly and human resources of research libraries in their entirety.
Regrettably, data that can be used to compare the national investment in humanities
libraries with that in other types of libraries do not currently exist.
However, data concerning numbers and costs of new scholarly books in the humanities
are available from
Blackwell’s Book Services, one of the major suppliers
of books to academic libraries. The number of new humanities titles processed each
year by this distributor provides a rough measure of the nation’s return on investment
in scholarly research in the humanities. This information, along with data on the
average list price of new titles in the humanities relative to those in other fields,
also reveals something of the market conditions with which humanities scholars must
contend as they seek an audience for their work.
|
Indicator IV-10
|
Research and Development Expenditures at Colleges & Universities
|
|
|
The NSF has been surveying the nation’s public and private colleges and universities
concerning their expenditures on science and engineering R&D since 1972. In
2003, the foundation began seeking information on R&D in fields other than the
sciences and engineering, including the humanities. The expenditures considered
in the NSF survey are for both “sponsored research,” which is subsidized by federal
and nonfederal agencies and organizations, and “university research,” which is separately
budgeted under an internal application of institutional funds (see the
NSF survey questionnaire).
While they are the best available, these data underestimate the size of the national
investment in college- and university-based humanities research. They do not capture
two key forms of financial support for humanities faculty wishing to pursue research.
These are: (1) leave from teaching and (2) fellowship monies used by faculty to
cover both living expenses (when leave from teaching is without pay) and research-related
costs (e.g., source materials and travel). Additionally, the NSF surveys only institutions
that perform science and engineering R&D. Thus, the activities of institutions
that do not perform science and engineering R&D but that may conduct substantial
amounts of research in humanistic fields are not included in the NSF sample. Finally,
some universities responding to the academic R&D survey reported only their
science and engineering expenditures. For these reasons, the figures supplied here
should be treated as “lower-bound” estimates of total investment in academic humanities
research.
Because of relatively low response rates to the survey’s new humanities-related
items in 2003 and 2004, the data reported here are only for years 2005 and beyond.
As Figure IV-10a shows, between 2005 and 2008, college and university spending
on humanities research steadily rose from approximately $218 to $254 million, an
increase of 16%. But as the figure also reveals, monies dedicated to humanities
research were dwarfed by those spent on research in engineering and the medical
and biological sciences. When expenditures in all the scientific and engineering
disciplines are considered (including agricultural science and others not depicted
in Figure IV-10a), 2008 spending on humanities research amounted to 0.49% of the
amount dedicated to science and engineering R&D.
Figure IV-10b depicts the proportion of reported research expenditures subsidized
by the federal government over the 2005–2008 time period. Whereas the federal share
of all humanities R&D dollars was roughly one-third in 2005, by 2006 this share
had dropped to approximately one-quarter. Although a decline in the federal share
of research funding was felt by most fields in this year, the drop experienced by
the humanities was the largest (both in absolute and percentage terms) recorded
for that year. This reduction is also notable in view of the fact that the federal
share of humanities research dollars was already low compared to that of all other
fields except business and management. The federal share of humanities funding remained
near this reduced level through 2008.
|
Indicator IV-11
|
Research Libraries
|
|
|
This indicator relies primarily on data collected by the ARL. The association’s
current membership includes 123 of North America’s largest research libraries, 107
of which are located in the United States and 16 in Canada. Robert Molyneux, formerly
of the U.S. National Commission on Libraries and Information, has analyzed these
data, focusing on a sample of 12 ARL research libraries1 located
on the campuses of large public universities in the United States and looking back
to the first years of the twentieth century. Molyneux has limited his analysis to
these institutions to avoid distortion due to year-to-year differences in the numbers
and types of institutions that comply with the ARL’s request for data.
His analysis reveals that after an initial decline, libraries’ expenditures on their
collections grew steadily, if slowly, between the early 1920s and the mid-1950s
(Figure IV-11a). Then, starting in the late 1950s, the pace of growth accelerated
rapidly. This spurt in spending lasted until 1971, when it ceased abruptly. Average
library expenditures on materials and binding first dropped by approximately 5%
and then remained near this reduced level for a decade. The early 1980s saw the
beginning of another surge in expenditures, with spending increasing in small increments
nearly every year for two decades. This trajectory was reversed in 2005, when a
drop of 6% in mean library spending on materials occurred.
Figure IV-11b reveals that until the early 1980s the trend in volumes added
by these libraries closely tracked the trend in spending: modest growth followed
by an extended period of accelerated growth, a sharp drop, and then a plateau. But
from there the trends diverge. While spending had begun to pick back up by 1982
and expenditure growth continued apace into the early 2000s, the average number
of volumes added was relatively constant from the mid-1970s onward.
Other statistics available directly from the ARL for its entire membership depict
trends in the costs of serials versus monographs, as well as changes in levels of
purchases and expenditures for these two types of materials from the mid-1980s through
the mid-2000s. Between 1986 and 2000, the median unit cost of serials rose a dramatic
120% (Figure IV-11c).
However, with the proliferation of electronic serials that began in the late 1990s—and
the duplication of content in different access formats (print and electronic)—the
ARL unit cost per serial subscription data have grown less reflective of actual
serial prices. Unit cost per serial subscription is calculated by dividing total
expenditures for serials by the total number of serials, but that total has grown
tremendously due to counting of multiple formats of the same serial title. The decrease
in serial unit cost reported in the figure between 2000 and 2005 are thus the product
not of actual price declines but of such double counting.
In an effort to avoid such duplication and to produce data that more accurately
reflect serial price trends, ARL has moved to tracking serial titles as of
2006-07. (For more about the Association’s efforts to address the challenges involved
in measuring library serial costs in an age of electronic access, see the ARL publications
“Reshaping ARL Statistics to Capture the New Environment” and “The Impact of Electronic Publishing on Tracking Research Library Investments in Serials”.)
Trends in monograph prices and the relationship between prices and expenditures
are rather different than those observed for serials. Monograph costs were more
stable between 1986 and 2005. Thus, while the median unit price increased somewhat
during the late 1980s, it had begun to decline in the early 1990s and by 2001 monograph
prices had returned to their 1986 levels. Prices remained in this vicinity through
2005. During the period of gradual price decline beginning in the early 1990s, libraries
reduced expenditures, which kept monograph purchases approximately 20% below 1986
levels for several years. As prices began declining more sharply in 1999, an increase
in library spending resulted in purchases returning to their baseline levels by
2002. But while serial spending and serial purchases increased steadily thereafter,
monograph expenditures and purchases began to drop once more. By 2005, expenditures
were 10.8% lower than in 1986, with purchases at a level that was 6.8% lower.
Other data gathered annually by the ARL on its university libraries indicate the
relationship between levels of crucial university library resources—collections
and staff—and the size of the student population between 1986 and 2005. (The ARL
does not collect data relating library resources to the number of faculty, the other
major users of library services.) These data show that while the median number of
serials purchased per student held steady from the mid-1980s through the mid-1990s,
this number began to rise in 1999 and by 2005 purchases represented a 30% increase
over the 1986 level (Figure IV-11d). In contrast, the median number of monographs
purchased per student dropped sharply during the late 1980s and thereafter remained
well below 1986 levels. In 2005, libraries purchased approximately 37% fewer monographs
per student than in 1986. The number of library staff, excluding student assistants,
per student also dropped (Figure IV-11e). After an initial increase, the
figure declined in every year but three. The cumulative effect was a 25% decline
in the median number of staff per student.
The NCES’s
Academic Library Survey
(ALS) is a census of the nation’s academic libraries. The term academic libraries,
as used by NCES, refers to libraries that are located on the nation’s college and
universities campuses and that serve as informational resources for those institutions.
Many of these have a research orientation and supply a variety of supports—large
collections, access to the growing number of electronic indices, and the expertise
of librarians—to humanities scholars. ALS data on the proportion of college and
university resources flowing to academic libraries provide another measure of the
extent of support for scholarly research in the humanities. These data reveal that
between 1975 and 2000, academic libraries commanded a decreasing share of college
and university budgets (Figure IV-11f). Although the decline was gradual,
it was steady. Thus, over the period, the median proportion of the nation’s postsecondary
schools’ budgets spent on libraries decreased by over a third, from just under 4%
of total institutional spending to 2.4%.
Any subsequent edition of the Humanities Indicators will include additional statistics
generated from an analysis of the ALS public-use data files currently available
from NCES. Unlike the published data on which this indicator relies, the raw ALS
data can be disaggregated by type of institution; this will allow for the exclusion
of academic libraries, such as those on the campuses of associate’s degree–granting
institutions that are not designed to support original scholarly research.
Note
1 The 12 member libraries included in Molyneux’s sample are those at
the Universities of California (Berkeley), Illinois (Urbana), Indiana, Iowa, Kansas,
Michigan, Minnesota, Missouri, Nebraska, Ohio State, Washington, and Wisconsin.
In every academic year since 1907/1908, all 12 of these institutions have provided
responses to key items on ARL’s annual survey of its membership.
|
Indicator IV-12
|
Academic Publishing
|
|
|
Blackwell’s Book Services, a major supplier of books and bibliographic support products
to colleges and universities, collects extensive data on the content and price of
titles the company recommends for acquisition by academic libraries. These data,
available for humanities titles generally as well as for various individual categories
of humanities titles, provide rough measures of both research output in the humanities
and the publication opportunities available to humanities scholars. The Humanities
Indicators includes data on book publications rather than publications in scholarly
journals because the book remains the primary means by which humanities scholarship
is disseminated. Moreover, while the natural and social scientific communities place
considerable emphasis on journal publications, academic employers in the humanities
continue to focus on book publications in gauging humanities scholars’ productivity.
Figure IV-12a depicts the number of new academic humanities titles that were
published between 2000 and 2005. (The Blackwell’s humanities category encompasses
the disciplines of philosophy, history, religion, folklore/mythology, and language.
The company also has separate categories for literary texts, literary criticism,
fine arts, and performing arts; these categories, along with Blackwell’s humanities
category, are accounted for in the figure.) Over the course of those five years,
the overall trend was upward, for a net increase of over 1,500 titles. The greatest
increase occurred in 2001, and although the number of new titles dropped substantially
the following year, there were still approximately 500 more titles published in
2002 than in 2000. The number of titles spiked again in 2003. Then, from 2003 to
2005 the number of titles continued to increase, but less dramatically.
Figure IV-12b depicts the trends in the number of new titles in the humanities,
using Blackwell’s categories. Among these categories, literary texts saw the greatest
absolute increase in the number of new titles between 2000 and 2005. But in terms
of percentage growth, the fine arts category was the standout, with 64% more new
titles released in 2005 than five years earlier. The numbers of religion, history,
and performing arts titles also increased by substantial percentages.
As Figure IV-12c shows, when price figures are adjusted for inflation, the
average list price of new humanities publications increased only modestly over the
early 2000s. In 2005, humanities titles, on average, cost $52.05, up from $50.98
in 2000, a smaller percentage increase than for most other fields. This figure also
reveals that humanities volumes were consistently the least expensive category of
new book title—and by a considerable margin. In 2005, the average price of physical
science books, the most expensive category, was 140% higher than that of humanities
titles.
Among humanities titles, language and literary texts were the extremes in terms
of average list price (Figure IV-12d), with language titles being consistently
more expensive and literary texts less expensive than other types of humanities
books. While the price of other categories of humanities titles ranged from approximately
$55 to $65 in 2005, new language titles were priced at $90.65, on average. In contrast,
literary texts, on average, were priced at $25.96. Language texts also experienced
the greatest percentage price increase (15%) over the 2000–2005 period. The prices
of literary criticism and religion titles increased by smaller percentages (9% and
5%,). Other types of humanities titles saw price declines over the period, with
fine arts titles dropping by the greatest proportion (12%).
|