Total Operating Expenditures per capita at Public Libraries, 1990-2002

Abstract

Total operating expenditures (TOTOPEXP) for each library were divided by that library's population served using two different measures of population. This ratio provides a much-watched indicator of support for a library and a number often used to compare different library systems. The two different methods of calculating the ratio result in numbers that are only very slightly different. While the ratios figured with unduplicated population as in the rank order tables are, as one might expect, slightly higher, the changes from 1991-2002 are less for these ratios than those figured on the raw population of the legal service area.

For each of the size groups, the current dollar values rose as did the estimated 1990 (and 1991) constant dollars calculated using the Consumer Price Index (CPI). Each group rose over the period but the percent increase for the quartiles are roughly inversely related to their size. The smallest libraries in the first quartile had the highest dollars expended per capita, greatest dollar increases, and highest percentage increases in this ratio by both methods of calculating the ratio. The ratios for all groups went up over the period and all beat inflation. There is evidence in the 1992-1993 data of three of these groups were affected by the recession at that time.

There is a Rank Order Table for this variable in 2002 and a summary table of the ranks by state from 1992-2002.

Charts and Tables

There are four charts and two tables reporting the results of an analysis of the total operating expenditures per capita. There are two measures of population in this dataset, the first is the purer POPU_UND, the "unduplicated" population of the legal service area, and POPU_LSA, the raw population of the legal service area. Given that some library populations overlap, the first variable attempts to correct for double counting and remove people who are counted in more than one library's service area. However, this variable begins in 1991 while POPU_LSA exists for 1990 so while both measures are used here to calculate per capita expenditures there will be no data for 1990 for the first calculation below using the unduplicated population. Given that all other similar charts start with 1990, these will, too. Per capita results using the raw population of the legal service area follows.

Each of the two calculations comes with two charts and one table. The first chart contains the values for the first three quartiles, that is, the smallest libraries, while the second contains the values for the fourth quartile and the 95% libraries. There are 10 lines in a single chart: one for each of the groups nominal and constant dollars (that is, discounted for inflation using the consumer price index) and it was a bit hard to see what was going on because the values are close for the various groups so they were separated into the two charts. The skew of raw values has been noted elsewhere but the value of the ratio of expenditures per capita for the groups are much closer than raw values. Why is this fact so?

This value is a ratio constructed between two skewed but highly correlated values. Correlation can be seen as a measure of the consistency of ratios across a series of values. What happens is that as the population served rises, the expenditures rise similarly and, as a result, the values all start closer together than the raw values (as is shown in the table below) and also rise closely so separating them makes what is going on clearer. These charts, then, use a linear vertical axis unlike the raw expenditure figures.

Per capita Expenditures Calculated Using UNDUPLICATED Population Served

First the smaller libraries:

The lines on the plot are both solid and dashed and come in a variety of colors. The solid lines indicate the nominal (that is, reported) dollars per capita and the dashed line indicates the calculated constant 1990 dollars. So, the first quartile libraries figures are presented in a solid green line and the constant 1990 dollar figures for these libraries in a dashed green line. Second quartile results are in red and third quartile are in blue. In this case, the smallest libraries fall between the second and third quartile libraries. A rough test of the ability of these libraries to stay ahead of inflation in expenditures per capita is to see if the dashed line for any group is above the 1990 values. In the case of unduplicated population per capitas, that test is very rough because the 1990 data cannot be calculated but the table below has the calculated values. Between 1991 and 2002 the value of the dollar fell about 24%.

And the larger ones:

The data in the tables below show that for the period analyzed these groups had an increase in total operating expenditures per capita calculated with unduplicated population that was real.

Data

The chart summarizes these values:

Note: The value for the Town of Ulster Public Library, Kinston, New York (newkey = NY0259) for POPU_UND for 1992 was 8,478, 1993 was 1, and 1994 was 11,507. In this analysis, the value for 1993 was changed to 9,993, the mean of the 1992 and 1994 figures.

Per capita Expenditures Calculated Using DUPLICATED Population Served

This section repeats the charts and First the smaller libraries:

And the larger ones:

Data

The chart summarizes these values:

Note: The value for the Town of Ulster Public Library, Kinston, New York (newkey = NY0259) for POPU_LSA for 1992 and 1994 was 12,329 while for 1993 it was 2. In this analysis, the value for 1993 was changed to 12,329.

Conclusions

Two methods are used here to calculate per capita expenditures. Using either leads to similar conclusions for these libraries: that expenditures per capita improved over the period and improved most for the smaller libraries. All groups beat the effects of inflation throughout the period except for a slight dip in the 1992-1993 period (which can be seen in the three of the five columns of constant dollars by both measures for those years). This dip shows the effect of recessions on these libraries similar to that seen in the 12 ARL libraries' expenditures.

The first quartile libraries did the best in keeping ahead of inflation while the larger the group, in general, the worse they did.

NCES estimates in 1990 that the total expenditures was $4,068,211,391 (N=8,754)¹ and $7,571,645,000 in 2001² (N=9,129), an increase of 86% in the total expenditures for all libraries in 2001 (with imputed data) and apparently without imputed data in 1990 so this percent increase is a high estimate. The comparable NCES figures for 2002 are not yet published.

Notes

¹ U.S. Department of Education, National Center for Education Statistics Public Libraries in the US.: 1990. E.D. Tabs, June 1992. p. 42. There is a disclaimer that this estimate may be low because of nonresponses.

² U.S. Department of Education, National Center for Education Statistics Public Libraries in the United States: Fiscal Year 2001. E.D. Tabs, June 2003. p. 66.

October 21, 2004