Econ paper

8.4 Nonli 门ear Effects on Test Scores of the Student-Teacher Ratio 287

PPm时呻阶向of.优伽仇f岛缸阳e臼ω⋯s岱s research in t白he町i让r areas of sp 巳cia剖Ii泣za矶tiωiωO∞n. Most

research in 巳c∞onomicωs first a叩pp 巳臼ars In巳臼c∞onoml 比cs j户ou町rnals ，so 巳c∞onomiβst岱s一or t由h巳让 libraries 一subscrib 巳 to economics journals.

How elastic is th 巳 demand by libraries for eco-

nomics journals? To find out ，we analyzed the rela-

tionship between the number of subscriptions to a

journal at 巳S. libraries (耳)and the journal's library

subscription price using data for the year 2000 for 180

economics journals. Because the product of a journal

is not the paper on which it is printed but rather the

ideas it contains ，its pric 巳 is logically measured not in

dollars p巳r year or dollars per page but instead in

dollars per idea. Although we cannot m巳asur 巳 "ideas"

directly ，a good indirect measure is the number of

times that articles in a journal ar 巳 subsequently cited

by oth 巳r researchers. Accordingly ，we measure price

�=【剖'川事:‘�..Library Subscriptions and Prices of Economics Journals

Subscriptions 1200

1000

800

200 . 105 15 20 25

Price per citation

(a) Subscriptions and Price per citation

In(Subscriptions) 8

DemandwhenAge = 5

A H V

1 坷 ， &

句

JA吁 F

、 J

r b

飞 d q L 唱

l n u - .

123 4 In(Price per citation)

(c) In(Subscriptions) and In(Price per citation)

In(Subscriptions) 8

5 4 3 2

� �

O -6 -5 -4 -3 -2 -1 0 1 2 3 4

In(Price per citation)

(b) In(Subscriptions) and In(Price per citation)

There is a nonlinear inverse relation between the number of U.s.library subscriptions (quantity) and the library price per citation (price)，as shown in Fig- ure 8.9a for 180 economics journals in 2000. But as seen in Figure 8.9b ，the relation between log quan tity and log price appears to be approximately Ii门- ear. Figure 8.9c shows that demand is more elastic for young journals (Age = 5) than for old journals <Age = 80l.

continued