Reply to Kelly, Cutright, and Hittle

Publisher Website
Google Scholar
Abstract
The major thrust of KelIy et al.'s "Comment" is that the relationship found by me between level of social security and fertility is not at all causal but entirely spurious, and that it is eliminated when level of economic development is properly controlled. They demonstrated this by de­ veloping a "modernization index" and then controlling for this index. The result­ ing partial correlation coefficients are both smalI and statistically insignificant. In contrast, I controlled for infant mortal­ ity, per capita gross domestic product, and per capita newspaper circulation and found sizeable betas that were sometimes statistically significant. As a general rule, if the relationship between two variables becomes significant when nonintervening variables are controlled, the relationship between the original two variables is believed to be spurious. Accordingly, the work of Kelly et al. appears quite compelling.

This publication has 1 reference indexed in Scilit: