If it is assumed that the 1911 low-fertility occupations generally exhibited infant mortality rates (IMRs) of approximately 50 per thousand (see Appendix E) while the pre-1837 population exhibited a typical level of 150 per thousand (see Wrigley and Schofield, Population history, Table 7.19), the following simple calculation demonstrates the maximum likely scale of the effect of this difference in infant mortality on recorded fertility over the first 7.5 years of marriage.

The calculation is based on the following further assumptions.

1. (i) That both populations breast-fed for a full twelve months, giving the same exposure to the effect of lactational amenorrhoea.

2. (ii) That all the infant mortality occurred at the beginning of the first year of life. This is, of course, grossly unrealistic but it simplifies the calculation. As it will bias the outcome so as to overestimate the extent of the infant mortality effect, it does not call into question the conclusion reported in the text that the effect is exceedingly small; indeed, it substantially strengthens that conclusion.

3. (iii) According to Bongaarts's summary of the available evidence, twelve months' breast-feeding increases the period of postpartum infertility by an average of eight months (see above, ch. 8, n. 57). It has therefore been set in the calculation below that the birth interval is eight months shorter among those experiencing an infant death.

4. […]