According to a recent analysis published on Oxford University Press’s OUP Blog, the population of librarians in the United States has fallen by 1/3 since 1990.

I tweeted a brief reference to this study when it was first brought to my attention, which generated some questions from onlookers: How is “librarian” defined? Are school librarians included in the population studied? Have some of the jobs remained the same but with different job titles?

Since large trends in the world of librarianship are likely to be of some interest to readers of the Scholarly Kitchen, I thought I’d see if I could find answers to some of these questions and also share other key findings from the study.

One fundamental thing to understand about the study is that it’s based on US census data (as organized by the Minnesota Population Center), which apparently means that the professional designation “librarian” is defined by the respondent. Given that the designation is self-defined, and given that the study clearly assumes that some “librarians” do not have undergraduate degrees (see the “Gender and Education Wage Difference Chart” for clear evidence of that assumption), it appears that wherever the study refers to librarians we might more accurately substitute the term “library workers.”

The “Number of Librarians” chart provided by the study’s authors shows a slow but steady increase in the number of U.S. library workers between 1880 and 1950, at which point the number began to climb sharply:

This makes intuitive sense, since the postwar period was marked by significant increases in government spending on education (and especially higher education) generally.

What is perhaps more interesting is the fact that this accelerated rate of growth continues right through the recessions of the 1970s and 1980s without any pause. The 1990s mark a sudden and sharp reversal, one which itself has continued without pause despite the fact that the 1990s were a boom period for the US economy (interestingly, the decline seems to have eased up a bit since 2000). Why the number of self-identified librarians continued to grow strongly through two recessions, only to begin falling sharply during a period of relative prosperity, is something of a puzzle.

Other interesting findings include:

• Contrary to the conventional wisdom, librarianship (or at least self-identified “librarianship”) has historically been a relatively well-paid profession, and continues to be: in 1940, median librarian wages were 153% of the national average, and in 2009 they were 137% of the average — though at various times between those two decades they have been much closer to the median wage of all workers. This is especially interesting given that the study clearly includes non-MLS library workers among the “librarian” population.
• The unemployment rate among library workers is relatively low – one-fifth the national unemployment rate in 2009.
• Library work remains a female-dominated profession, though the female-to-male ratio used to be very different and has fluctuated over the years. In the 1880s a majority (52%) of library workers were men, whereas in 2009, 83% were women.
• The study suggests that men and women in professional librarian positions have an almost identical median wage; however, among those without college degrees there is an even greater income gap between the genders than exists in other professions.

There are also some interesting data points about marriage rates, age, race, and the distribution of public and private settings for library workers over time. But it’s the general population trend that I find particularly striking. I’m trying to decide how worrisome I should believe it to be. Does the depopulation of our workforce signal an encroaching senescence of the profession itself? Does it suggest that the boundaries that separate library work from other kinds of information work are breaking down?

Given the counterintuitive patterns of growth during recession and decline during boom periods, it seems unlikely that the trend can be explained in terms of recent fiscal hardship.

One thing is certain: it will be very interesting to see what the trend line looks like after the next ten-year census.