Cross-country relationships between life expectancy, intertemporal choice and age at first birth

Abstract

Humans, like other animals, typically discount the value of delayed rewards relative to those available in the present. From an evolutionary perspective, prioritising immediate rewards is a predictable response to high local mortality rates, as is an acceleration of reproductive scheduling. In a sample of 46 countries, we explored the cross-country relationships between average life expectancy, intertemporal choice, and women's age at first birth. We find that, across countries, lower life expectancy is associated with both a smaller percentage of people willing to wait for a larger but delayed reward, as well as a younger age at first birth. These results, which hold when controlling for region and economic pressure (GDP-per capita), dovetail with findings at the individual level to suggest that life expectancy is an important ecological predictor of both intertemporal and reproductive decision-making.

Introduction

Humans, like other animals, typically discount the subjective value of delayed rewards relative to those available in the present (Berns, Laibson, & Loewenstein, 2007). Explanations for this delay discounting phenomenon tend to emphasize that the uncertainty of future rewards makes capitalising on immediate opportunities a beneficial strategy in many circumstances (Andreoni and Sprenger, 2012, Daly and Wilson, 2005, Stevens and Stephens, 2010). Indeed, ‘intertemporal choices’ between immediate and delayed rewards are highly sensitive to context in humans (Lempert & Phelps, 2015). One common prediction about the role of ecological context, grounded in evolutionary theorising, is that intertemporal decision-making should on average shift towards immediate rewards when local mortality rates are high (e.g. Daly and Wilson, 2005, Frankenhuis et al., 2016, Hill et al., 2008, Kruger and Zimmerman, 2008). This is because a higher mortality risk equates to a lower likelihood of capitalising on delayed rewards due to the possibility of death. This is expected to take place both for an individual who may come to change their decision-making based on exposure to relevant information in their environment, but also at the group level whereby shared ecological factors like higher local mortality rates should produce on average steeper delay discounting.

Various lines of evidence support this proposition at the individual psychological level, including findings that exposure to natural disasters, violence or mortality cues is associated with a preference for immediate rewards over delayed ones (Lahav et al., 2011, Li et al., 2012, Pepper and Nettle, 2013, Ramos et al., 2013). Thus far, the evidence on this front comes from between- or within-participant analyses within the same country (e.g. Ramos et al., 2013), and analyses have tended to focus on specific cues of mortality risk, such as exposure to violence, rather than local mortality rates more generally. Here we therefore extend this work by asking whether variation in life expectancy across countries acts as an ecological predictor of the average intertemporal decision-making in those countries.

A similar logic applies in the domain of reproductive scheduling (Chisholm, 1993, Nettle, 2011, Wilson and Daly, 1997). As local mortality risk increases, people are expected to reproduce earlier, and to produce more offspring throughout their reproductive careers (Charnov, 1991, Ellis et al., 2009, Stearns, 1992). The benefits of accelerated reproductive scheduling when mortality risk is high are thought to arise from both an increased chance of reproducing, and increased time available to care for offspring, before death. There is evidence that both within and between countries, women's average age at first birth is younger when mortality rates are higher (Low, Hazel, Parker, & Welch, 2008). Local mortality risk indicators also predict total fertility, such that people in higher mortality-risk conditions tend to have more children on average throughout the lifespan (Guégan et al., 2001, Zhang and Zhang, 2005). We therefore sought to also replicate these previously reported relationships between life expectancy and age at first birth. Steeper temporal discounting has also been associated with having more sexual partners, an earlier age of first sexual activity, more relationship infidelity, greater odds of having a past or current pregnancy, and lower contraceptive use (Chesson et al., 2006, McCoul and Haslam, 2001, Reimers et al., 2009). However it has thus far gone unexamined how average intertemporal decision-making patterns relate to reproductive scheduling patterns across different ecologies.

The current study therefore had two main aims. Firstly, we aimed to explore the relationship between life expectancy, and both intertemporal choice and age at first birth. Secondly, we aimed to explore the association between intertemporal choice and age at first birth. We hypothesized that, across countries, (i) lower average life expectancy would be associated with a lower percentage of people willing to wait for a larger later reward, (ii) lower average life expectancy would be associated with younger age at first birth, as found in prior studies, and (iii) a lower percentage of people willing to wait for a larger later reward would be associated with younger average age at first birth.