The link between climate change, food security and fertility: The case of Bangladesh

likely to stay home and work on the farm rather than migrate [62]. In Bangladesh, dry spells were associated with increased migration of male household heads, while farmers were 40% more likely to migrate after a dry spell compared to their peers engaged in other livelihoods [31]. Sellers and Gray revealed that the intention to have another child increased and the use of family planning decreased in response to delays in the monsoon in the previous year, espe- cially among wealthier and more educated women. As a result of abnormally high tempera- tures in the past five years, women with farms were less likely to give birth because they made greater use of family planning, especially among poorer and less educated women [61]. The short- and long-term effects of climate variability on fertility can be different. A recent study conducted in 18 sub-Saharan African countries shed light on the short- and long-term effects of climate variability on fertility. In the study, Eissler et al. found that temperature increase, both short- and long-term, has negative effects on fertility, while precipitation, both short- and long-term, has differential effects on fertility. For example, a one standard deviation increase in temperature for 12 months (short term) and 60 months (long term) leads to a 0.042 and 0.101 reduction in ideal family size, respectively. In contrast, a one standard deviation increase in precipitation is associated with a 0.052 decrease in children in the short term, while a one standard deviation increase in precipitation is associated with a 0.059 increase in wom-

en’s preference for ideal family size in the long term [65]. The effects of climate variability may have indirect effects on fertility through food produc- tion. Food insecurity leads to a range of demographic impacts, including but are not limited to changes in fertility. Reducing human fertility–a means of controlling population growth–may be one way to cope with inadequate food supply [66]. Changes in fertility due to food insecu- rity are usually explained by the effects of drought and associated famine. However, a study in sub-Saharan Africa found that fertility rates were lower among populations enjoying food security than among populations not enjoying food security [67]. Fertility rates were found to decline after drought and associated famine. For example, there was a significant decline in conception rates in Ethiopia during the years of drought and famine between 1970 and 1980 [68]. Similarly, fewer conceptions were reported during the months of January through May, when there were extreme food shortages in Finland due to the 1967–68 famine [69]. The his- toric famine in China from 1958 to 1961 had a similar effect on fertility [70, 71]. A study of the effects of the 1974 famine in Bangladesh showed that the impact on fertility varied from lower to higher socioeconomic groups. The separation of couples explains the pronounced negative effect on lower socioeconomic groups as spouses migrated to other regions of the country searching for sources of income, which affected the frequency of sexual intercourse [72].

Drawing on insurance mechanisms and replacement theory, Finlay argues that fertility responds to both natural disasters and infant mortality [73]. In a study in India, Pakistan, and Turkey, Finlay concludes that people have a “positive response” to child mortality [73]; people who have lost children in disasters prefer replacement [74, 75] and view more children as insurance against expected risk [74, 76]. In Indonesia, women who had lost one or more chil- dren during the disaster were more likely to plan to have more children after the tsunami. Before the tsunami, women who did not have children “were more likely to start families in communities where tsunami-related mortality rates were higher” [74]. Lindstrom and Kiros examined short- and long-term fertility responses to infant and child mortality in Ethiopia. They found that fertility responses tend to be high among people who have experienced infant mortality, which is referred to as replacement behaviour or replacement effect [77]. Owoo et al. conducted a study in Ghana and found that women living in a place with higher mortality shocks tended to have higher fertility preference [78]. The literature section shows that Bangladesh’s economy is largely dependent on agriculture, which is extremely vulnerable to climate change. The literature also shows that the global PLOS ONE Climate change, food security and fertility PLOS ONE | https://doi.org/10.1371/journal.pone.0258196 October 21, 2021 4 / 18