Table 1.

Correlation coefficients between age at death and own fertility (mothers) and between age at death and a daughter’s fertility (grandmothers) for women in the Utah Population Database who were born in the nineteenth century. The mother sample includes all parous women born before 1900. Column 2 lists sample sizes. Column 3 lists correlations for samples pooled across the demographic transition (Figure 1), with significance levels in column 4. In column 5, the effect of pooling across the demographic transition is removed by controlling the woman’s birth year. Without this control, the shorter lifespans of the higher fertility pre-transition women combined with the longer lifespans of the lower fertility post-transition women have a negative effect on correlations between longevity and fertility. Line 2 removes women who died during the child-bearing years. Their inclusion has a strong effect on associations between death age and fertility because their early deaths impose low parities. The grandmother sample includes all women born before 1870 with complete information about the fertility of a randomly chosen daughter. We use the fertility of a daughter to parallel the Madrigal and Melendez-Obando (2008) measure of grandmother effects as the correlation between a woman’s longevity and the fertility of her linking daughter. Line four removes women who died before reaching their grandmothering years.

	1	2	3	4	5	6
	Sample	N	Parity by death age	p	Parity by death age controlling for birth year	p
1	Mothers	101,607	0.08163	<0.0001	0.12118	<0.0001
2	Mothers who lived >50	87,923	−0.06371	<0.0001	−0.01356	<0.0001
			Fertility of a Randomly selected daughter by death age		Fertility of a Randomly selected daughter by death age controlling for birth year
3	Grandmothers	51,012	−0.02033	<0.0001	0.00410	0.3547
4	Grandmothers who lived >50	43,698	−0.01777	0.0002	0.00825	0.0846