Expenditures for Library Materials at Twelve Association of Research Libraries Members, 1908-2006

Abstract

Mean expenditures for library materials at the 12 libraries are charted with an estimate of these expenditures in 1913 constant dollars plus binding. The variable measuring expenditures for materials and binding appears in the first year of the Gerould data. Subsequently, the binding was disaggregated leaving two variables: expenditures for materials and expenditures for binding. To create a continuous series, the two numbers were added together for the years after they were disaggregated.

According to figures from the US Bureau of Labor Statistics, the value of the dollar since 1913 "constant" (the first year available) had declined so that a 2006 dollar has an estimated purchasing power of a nickel in 1913. The data show that while the nominal annual or "current" dollars spent on materials and binding rose pretty consistently, the 1913 constant dollar expenditures also rose and the 2006 constant dollar figure is about 26 times that of the 1913 figure. That is, the expenditures at this group of libraries rose in real terms although the recent drop is striking.

Expenditures for Materials and Binding, 1908-2006

The green line traces the mean values for each year in the reported dollars and the red line shows the estimated 1913 constant dollar figures. The current dollar figure rises pretty uniformly although there is a dip in 1971 similar to that seen in the volumes added plots that will be dealt with again below. Given the magnitude of the current figures compared to the constant dollar figures, the increase in these constant dollar figures does not look as dramatic as they are. In fact, as mentioned, the 2006 constant 1913 dollar figure is about 26 times as great as the 1913 figure, meaning that the expenditures for materials and binding went up, roughly, 26 times in real terms over the period.

Two plots show the year to year changes in expenditures, one percentage change in the current dollar figures and percentage change in current dollar figures. Each shows similar but volatile number changes from year to year. With these two plots we have about reached the limit of charts as an analytical tool. True, we can see in both of them what appears to be the effects of the Great Depression and also the 1971 dip but given the volatile nature of these data, what we might ascribe to external causes may just be random fluctuations. So, the next stage will be to use more systematic tools on these data.

Given the volatility the two percentage change charts show, why do the volumes added figures seem more regular and understandable? One reason, I think, is that with volumes, we are looking at volumes "added"—that is processed by catalogers—rather than volumes purchased. I have speculated that cataloging backlogs are a means of regulating workflow when purchases can vary dramatically from year to year. Growth, then, is a variable that lags behind the inputs or changes in the economy. Expenditures, it seems can change rapidly from year to year and budgetted dollars usually cannot be held over from one year to the next.

This plot has both constant and current dollars and it obscures the pattern of constant dollar expenditures. What happens when just the mean current dollar figures for these institutions are plotted separately?

Note that the vertical axes of these two plots are different. The maximum value on this axis is $700,000 while it is $12,000,000 on the first one.

This plot finally makes some sense and allows us to speculate about what a mathematical model might look like. Note two things about this plot. First, the 1971 constant dollar expenditure figure is $352K and it falls after that in a pattern reminiscent of that seen with the first volumes added chart. Second, notice that expenditures in 1983 are close to those in 1971 and rise pretty continually since then. At the same time, as we see in the volumes added figures, growth in volumes was basically flat. Why do we have an increase in constant dollars and no corresponding increase in volumes? We are probably looking at the effect of increasing serials prices and format changes in purchased materials but it will take further work to be sure of this speculation.

September 12, 2007