Abstract
In newspaper articles and the publishing industry, there is an opinion that
a decrease in the sale of books affects an increase in the lending of books in
public libraries. However, there are few proof cases of the correlation and causality
between these variables. We sought to empirically study whether the two
variables truly demonstrate correlation or causality. We used data from 1970 to
2015.
The regression coefficient for the two variables of level data showed very
high significance. However, this was a case of“ spurious regression.” The“ spurious
regression” theory was proved by Granger and Newbold（ 1974）. We then
performed regression analysis using stationary time series data, and the data
after removing the time trend. We conducted a unit root test to derive the stationary
time series data. Additionally, we used the Hodrick-Prescott filter to
remove the time trend. Analysis of these data did not return a significant
regression coefficient. We conclude that the lending of books at the public
library has no effect on the sales of books.
