Ellis, Reid, and Kramer’s article advances a comprehensive model of the mortality-fertility nexus. The authors’ tour de force review traverses disciplinary boundaries to collate empirical support for a ‘two-tiered’ model of the impact of different sources of extrinsic mortality on life history strategies. To support their argument that a second tier exists – that ambient clues to higher mortality in one’s surroundings can prop up fertility – Ellis, Reid, and Kramer summarize extensive evidence that mortality exposures in the mundanity of everyday life, as well as exposure to mass casualty events, correspond with higher fertility behavior, including having children earlier in the life course and ultimately having more children.
Like much of the social science literature linking aggregate conditions to individual behavior, Ellis, Reid, and Kramer largely leave untouched questions of how extrinsic mortality affects individual life history strategies through the proposed second tier. That is, whereas prior research has conceptualized the first tier of mortality exposures as operating through physiological and biological mechanisms to affect individuals’ energetic constraints, Ellis, Reid, and Kramer take a decidedly insouciant approach to outlining the potential mechanisms linking the presumed second tier to individuals’ fertility-related behaviors.
Even so, Ellis, Reid, and Kramer imply that observed linkages between extrinsic mortality conditions and subsequent fertility are the result of conscious, volitional processes whereby individuals are intentionally and strategically responding to ‘ambient cues’ with higher fertility behavior – a sentiment shared by prominent demographic analyses of the mortality-fertility nexus (Heuveline & Poch, Reference Heuveline and Poch2007; Nobles, Frankenberg, & Thomas, Reference Nobles, Frankenberg and Thomas2015). Ellis, Reid, and Kramer conclude that regardless of the mortality signal, the ‘outcome is largely the same: increased allocation of resources toward a faster pace of reproduction and a higher offspring number…’ (52). Indeed, the adopted terminology reveals a tacit assumption that individual fertility behavior amounts to a coherent life history strategy rather than simply life histories.
Yet, empirical evidence from a study that I led in Malawi complicates this narrative (Smith-Greenaway, Yeatman, & Chilungo, Reference Smith-Greenaway, Yeatman and Chilungo2022). Recognizing that unintended fertility constitutes a sizable share of total fertility globally, with nearly one half of pregnancies being unintended, either unwanted or occurring sooner than desired (Bearak et al., Reference Bearak, Popinchalk, Alkema and Sedgh2018), we sought to examine whether mortality corresponds with fertility only via volitional processes, or if non-volitional ones are also of relevance. To answer this question, my colleagues and I collected longitudinal data on a representative sample of women living in Balaka, Malawi – a resource-constrained context where rates of child and adult mortality remain at global highs. Over an initial two-year period between 2009 and 2011, women completed interviews at four-month intervals, allowing us to carefully track short-term fluctuations in extrinsic mortality conditions in their day-to-day life to understand how it affects their fertility. Specifically, we asked women how many funerals they attended in the prior one month, allowing us to approximate women’s mortality exposures in their familial and social networks. We also asked women about their fertility intentions, including whether they intended to have a child in the coming months, and pregnancy-tested them at each interview, spaced four months apart. These unusually rich, closely spaced panel data allowed us to observe whether short-term fluctuations in women’s mortality exposures corresponded with their subsequent odds of becoming pregnant. Further, with prospective fertility intention data, we disaggregated pregnancies according to whether women intended versus did not intend them.
The results directly challenge the notion that all sources of extrinsic mortality led women to volitionally allocate resources and time to reproduction. Instead, while our study showed that having attended more funerals in the prior month corresponded with a substantial increase in the odds that women become pregnant in the short term, the association was driven by pregnancies that women did not intend to conceive. Further, we found no evidence that funeral attendance elevated the desire to become pregnant over any time horizon.
These results imply that mortality exposures do not (always) work through volitional pathways but instead may signal the burden that high mortality imposes on women’s lives – a burden that acts as a barrier to their ability to allocate resources to avoid pregnancy. Across all contexts, dealing with death is time-consuming – it upends routines, it is emotionally taxing, and it can fray social support systems as it sears unfillable holes in individuals’ family or social network (Smith-Greenaway, Verdery, & Carr, Reference Smith-Greenaway, Verdery and Carr2025).
Of course, this does not negate evidence that extrinsic mortality corresponds with individuals’ fertility and – in select cases – does elevate women’s fertility desires and preferences (see, for example, Smith-Greenaway & Lin, Reference Smith-Greenaway and Lin2023); however, it does demand that we think more carefully about whether individuals accurately detect and encode the mortality environment around them (Ciacio et al, Reference Ciancio, Delavande, Kohler and Kohler2024; Delavande & Kohler, Reference Delavande and Kohler2009) and how, specifically, it comes to shape their life courses. Mortality exposures are not only cues. They are not only signals humans then robotically process to assess their own risks or the risks to their offspring. Humans are social creatures with linked lives (Elder, Reference Elder1994) – across diverse cultures and across centuries, evidence demonstrates the great care we take in burial rituals, memorialization practices, and grieving – attesting to the personal and social significance of deaths. Approaching the study of the mortality-fertility nexus with a bereavement lens would better acknowledge that mortality events do not only signal the threat of death but also represent an actual death that requires individuals to mourn and grieve the dead and to eventually reassemble their lives. Of course, deaths too serve as morbid reminders of our own mortality; yet, the higher fertility observed following mortality exposures does not always signal humans pursuit of reproduction but can instead reflect their inability to prevent the default potential of reproduction that arguably defines women’s lives. Thus, engaging such questions of how extrinsic mortality comes to shape individuals’ behaviors and outcomes has the potential to crack open lines of inquiry currently sitting at the margins of life history literature.
You have
Access
Ellis, Reid, and Kramer’s article advances a comprehensive model of the mortality-fertility nexus. The authors’ tour de force review traverses disciplinary boundaries to collate empirical support for a ‘two-tiered’ model of the impact of different sources of extrinsic mortality on life history strategies. To support their argument that a second tier exists – that ambient clues to higher mortality in one’s surroundings can prop up fertility – Ellis, Reid, and Kramer summarize extensive evidence that mortality exposures in the mundanity of everyday life, as well as exposure to mass casualty events, correspond with higher fertility behavior, including having children earlier in the life course and ultimately having more children.
Like much of the social science literature linking aggregate conditions to individual behavior, Ellis, Reid, and Kramer largely leave untouched questions of how extrinsic mortality affects individual life history strategies through the proposed second tier. That is, whereas prior research has conceptualized the first tier of mortality exposures as operating through physiological and biological mechanisms to affect individuals’ energetic constraints, Ellis, Reid, and Kramer take a decidedly insouciant approach to outlining the potential mechanisms linking the presumed second tier to individuals’ fertility-related behaviors.
Even so, Ellis, Reid, and Kramer imply that observed linkages between extrinsic mortality conditions and subsequent fertility are the result of conscious, volitional processes whereby individuals are intentionally and strategically responding to ‘ambient cues’ with higher fertility behavior – a sentiment shared by prominent demographic analyses of the mortality-fertility nexus (Heuveline & Poch, Reference Heuveline and Poch2007; Nobles, Frankenberg, & Thomas, Reference Nobles, Frankenberg and Thomas2015). Ellis, Reid, and Kramer conclude that regardless of the mortality signal, the ‘outcome is largely the same: increased allocation of resources toward a faster pace of reproduction and a higher offspring number…’ (52). Indeed, the adopted terminology reveals a tacit assumption that individual fertility behavior amounts to a coherent life history strategy rather than simply life histories.
Yet, empirical evidence from a study that I led in Malawi complicates this narrative (Smith-Greenaway, Yeatman, & Chilungo, Reference Smith-Greenaway, Yeatman and Chilungo2022). Recognizing that unintended fertility constitutes a sizable share of total fertility globally, with nearly one half of pregnancies being unintended, either unwanted or occurring sooner than desired (Bearak et al., Reference Bearak, Popinchalk, Alkema and Sedgh2018), we sought to examine whether mortality corresponds with fertility only via volitional processes, or if non-volitional ones are also of relevance. To answer this question, my colleagues and I collected longitudinal data on a representative sample of women living in Balaka, Malawi – a resource-constrained context where rates of child and adult mortality remain at global highs. Over an initial two-year period between 2009 and 2011, women completed interviews at four-month intervals, allowing us to carefully track short-term fluctuations in extrinsic mortality conditions in their day-to-day life to understand how it affects their fertility. Specifically, we asked women how many funerals they attended in the prior one month, allowing us to approximate women’s mortality exposures in their familial and social networks. We also asked women about their fertility intentions, including whether they intended to have a child in the coming months, and pregnancy-tested them at each interview, spaced four months apart. These unusually rich, closely spaced panel data allowed us to observe whether short-term fluctuations in women’s mortality exposures corresponded with their subsequent odds of becoming pregnant. Further, with prospective fertility intention data, we disaggregated pregnancies according to whether women intended versus did not intend them.
The results directly challenge the notion that all sources of extrinsic mortality led women to volitionally allocate resources and time to reproduction. Instead, while our study showed that having attended more funerals in the prior month corresponded with a substantial increase in the odds that women become pregnant in the short term, the association was driven by pregnancies that women did not intend to conceive. Further, we found no evidence that funeral attendance elevated the desire to become pregnant over any time horizon.
These results imply that mortality exposures do not (always) work through volitional pathways but instead may signal the burden that high mortality imposes on women’s lives – a burden that acts as a barrier to their ability to allocate resources to avoid pregnancy. Across all contexts, dealing with death is time-consuming – it upends routines, it is emotionally taxing, and it can fray social support systems as it sears unfillable holes in individuals’ family or social network (Smith-Greenaway, Verdery, & Carr, Reference Smith-Greenaway, Verdery and Carr2025).
Of course, this does not negate evidence that extrinsic mortality corresponds with individuals’ fertility and – in select cases – does elevate women’s fertility desires and preferences (see, for example, Smith-Greenaway & Lin, Reference Smith-Greenaway and Lin2023); however, it does demand that we think more carefully about whether individuals accurately detect and encode the mortality environment around them (Ciacio et al, Reference Ciancio, Delavande, Kohler and Kohler2024; Delavande & Kohler, Reference Delavande and Kohler2009) and how, specifically, it comes to shape their life courses. Mortality exposures are not only cues. They are not only signals humans then robotically process to assess their own risks or the risks to their offspring. Humans are social creatures with linked lives (Elder, Reference Elder1994) – across diverse cultures and across centuries, evidence demonstrates the great care we take in burial rituals, memorialization practices, and grieving – attesting to the personal and social significance of deaths. Approaching the study of the mortality-fertility nexus with a bereavement lens would better acknowledge that mortality events do not only signal the threat of death but also represent an actual death that requires individuals to mourn and grieve the dead and to eventually reassemble their lives. Of course, deaths too serve as morbid reminders of our own mortality; yet, the higher fertility observed following mortality exposures does not always signal humans pursuit of reproduction but can instead reflect their inability to prevent the default potential of reproduction that arguably defines women’s lives. Thus, engaging such questions of how extrinsic mortality comes to shape individuals’ behaviors and outcomes has the potential to crack open lines of inquiry currently sitting at the margins of life history literature.
Acknowledgements
I thank Ellis, Reid, and Kramer for their thought-provoking article, and the editor and editorial staff at Behavioral and Brain Sciences for the opportunity to provide this commentary.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
None.