Info

(.29)

(.25)

(.10)

(.33)

(.32)

(.34)

J Fractions may not sum to ! .00 due to rounding error.

J Fractions may not sum to ! .00 due to rounding error.

their family history, yet they have tested positive). The small numbers of respondents in this group (N=10) may be partially responsible for the absence of any detectable effect, however. As more K2082 subjects are followed through the four-month (and one-year) interviews, we will want to take a closer look at this group.

The presence or absence of life insurance is a qualitative measure of consumer demand. The number of life insurance policies a woman has is another, more detailed indication of demand. In our sample, the number of life insurance policies held range from 0 to 9, but only 22 of the 286 women (7.7%) reported having more than two policies. Ordered probit estimates of models 1—3 using number of policies as the dependent variable did not converge probably because so few women had more than two policies. As an alternative, we present the simple crosstabulation of number of life insurance policies a woman has (collapsing all of those who have two or more policies together) by the testing/family history groups in Table 3. The associated chi-square test reveals that one cannot reject the hypothesis that the number of insurance policies held is equal across the testing/family history groups. That is, the figures in Table 3 do not support the adverse selection hypothesis at the bivariate level.

Was this article helpful?

0 0

Post a comment