Volume 23 Issue 3 Article 5 11-2021 Long-term Effects of In Utero Exposure to “The Year without a Long-term Effects of In Utero Exposure to “The Year without a Summer” Summer” Hamid Noghanibehambari Texas Tech University, Department of Economics, Lubbock, USA, hamid.noghanibehambari@ttu.edu Farzaneh Noghani Texas Tech University, Rawls Business School, Lubbock, USA, farzaneh.noghani@ttu.edu Nahid Tavassoli Texas Tech University, Department of Economics, Lubbock, USA, nahid.tavassoli@ttu.edu Mostafa Toranji University of Tehran, Department of Economics, Tehran, Iran, mostafa.toranji@ut.ac.ir Follow this and additional works at: https://www.ebrjournal.net/home Part of the Environmental Public Health Commons, Health Economics Commons, and the Labor Economics Commons Recommended Citation Recommended Citation Noghanibehambari, H., Noghani, F., Tavassoli, N., & Toranji, M. (2021). Long-term Effects of In Utero Exposure to “The Year without a Summer”. Economic and Business Review, 23(3). https://doi.org/ 10.15458/2335-4216.1288 This Original Article is brought to you for free and open access by Economic and Business Review. It has been accepted for inclusion in Economic and Business Review by an authorized editor of Economic and Business Review. ORIGINAL ARTICLE Long-term Effects of In Utero Exposure to “The Year without a Summer” Hamid Noghanibehambari a, *, Farzaneh Noghani b , Nahid Tavassoli a , Mostafa Toranji c a Texas Tech University, Department of Economics, Lubbock, USA b Texas Tech University, Rawls Business School, Lubbock, USA c University of Tehran, Department of Economics, Tehran, Iran Abstract This paper uses the aftermath of the great Tambora eruption in 1815 as a natural experiment to explore the long-term effects of a nutritional shock during prenatal development. The volcanic explosion of Tambora formed substantial ash columnswhichhamperedsunlight,cooleddownthesurfacetemperature,reducedthelengthofthegrowingseason,and led to a severe harvest failure during summer and winter of 1816 in Europe and northeastern states of America. US decennial census 1850 provides evidence that cohorts in utero during the climate anomaly revealed lower literacy rates, lower labor force participation rates, a fewer number of own children, and a higher female-male ratio. The results are confirmed among the same cohorts in England, Canada, and Norway. The decennial census of each country indicates negative effects of exposure during prenatal development on labor market participation rates in adulthood. Keywords: Fetal origins hypothesis, Public health, Environment and health, Education, Labor market participation JEL classification: I15, Q51, N31, N51 Introduction I t is now well established that external stressors during prenatal development have adverse short term effects for newborns and negative long-term effects in adulthood. 1 Maternal stress, nutritional deficiency, and pollution during in utero are shown to be associated with infants’ health outcomes including low birth weight, preterm birth, birth defects, fetal death, infant mortality, complications in pregnancy, and congenital disorders (Almond, Mazumder, & Van Ewijk, 2014; Almond & Mazumder, 2005; Beach & Hanlon, 2016; Duncan, Mansour, & Rees, 2017; Hoynes, Schanzenbach, & Almond, 2016; Isen, Rossin-Slater, & Walker, 2017; Majid, 2015; NoghaniBehambari et al., 2020c; Nog- haniBehambari, Noghani&Tavassoli, 2020a,b; Page, Schaller, & Simon, 2017; Tavassoli, Noghanibe- hambari, Noghani, & Toranji, 2020; Van Ewijk, 2011). The lower health endowment of the affected individuals transforms the trajectory of their out- comes later in life. It lowers math and cognitive test scores at age 7, decreases the completed years of education, decreases the annual hours worked and labor force participation, diminishes the earnings at age 30, lowers life expectancy, increases the mor- talityrates,raisesthelikelihoodofold-agemortality due to specific causes of death such as cardiovas- cular diseases (Almond et al., 2014; den Berg, Lin- deboom, & Portrait, 2006; NoghaniBehambari, Noghani &Tavassoli, 2020a,b; Persson & Rossin- Slater, 2018; Yeung, den Berg, Lindeboom, & Portrait, 2014). There are some empirical challenges to assess the long-term effects of an in utero shock. First, the shock must be orthogonal to other individual and household level covariates including their genetic endowments. Second, the effect of initial health Received 20 August 2020; accepted 29 March 2021. Available online 8 November 2021. * Corresponding author. E-mail addresses: hamid.noghanibehambari@ttu.edu (H. Noghanibehambari), farzaneh.noghani@ttu.edu (F. Noghani), nahid.tavassoli@ttu.edu (N. Tavassoli), mostafa.toranji@ut.ac.ir (M. Toranji). 1 Almond and Currie (2011a,b) provide a review of the recent research which investigates the fetal origins of later-life outcomes. https://doi.org/10.15458/85451.1288 2335-4216/© 2021 School of Economics and Business University of Ljubljana. This is an open access article under the CC-BY-NC-ND license (http://creativecommons. org/licenses/by-nc-nd/4.0/). deficiency is not uniform across the life cycle. It could be revealed only in childhood, adulthood, or old age. The latency of these effects causes a detection problem. Third, families could reimburse orreinforcetheirchildrenwithalowerinitialhealth endowment. The reinforcement of families could push up the estimates of this long-term link since less healthy children are sort out for fewer family resources. These inherent aspects of the question make the non-experimental studies disputable. The agricultural catastrophe due to the cold sum- mer of 1816 provides a unique opportunity to examine the long-term effects of external shocks during antenatal development. It started with the unanticipated volcanic eruption of Mount Tam- bora in the Dutch East Indies (nowadays Indonesia) in1815.Thegreatestinrecordedhistory,thevolcanic ash disseminated over the globe over the months after the eruption. It cooled down the global tem- perature by 0.4e0.7 C and caused harvest failures across the world (Stothers, 1984). In the US, eastern and northeastern states experienced an unprece- dented cold summer in 1816, a year known as “the year without a summer”. The harvest failure was so severeintheseregionsthatfoodpricesincreasedby as much as three times during months afterward. However, western, southern, Midwestern, and northern states were not affected by the climate anomaly(Oppenheimer,2003). UsingCensusdatafortheyear1850,weshowthat cohorts who were exposed to the cold summer and the proceeding agricultural failure during prenatal development in affected regions had lower literacy ratesandlaborforceparticipationintheiradulthood. Beingexposedtotheprenatalshockisassociatedwith 2.6 and 1.4 percentage points decrease in the proba- bility of being literate and active in the labor force, respectively. Furthermore, the exposed individuals have,onaverage,0.11fewernumberofchildren. Investigating the long-term relation between health endowment in early life and adult outcomes has important policy implications. The significant long-term correlation could suggest public policies to weight more on the fetal period rather than contemporaneous adults’ outcomes directly if the target is to enhance education and labor force participation. Second, it could help health policy- makers to recognize more vulnerable groups in so- ciety.Furthermore,itpointstothefactthatawelfare program to provide pregnant mothers during times of hardship is not only a health policy but also an education policy or labor force protection policy for the next generations. This paper makes several contributions to the current literature. First, to the best of our knowledge, this is the first study to evaluate the economic and non-economic impacts of the infa- mous climate anomaly following the Tambora eruption. Second, it contributes to the environ- mental economics literature by providing the long- term labor market effects of a roughly 0.5 C reduc- tion in average annual temperature. Third, scant transportation system and lack of government pro- tective policies in the period of this study limit the scope of external deteriorating effects and so pro- vide more accurate estimations of the long-term correlations. The rest of the paper is organized as follows. In section 1, we provide a brief literature review. Sec- tion 2 describes the core dataset, sample selection strategy, and gives summary statistics of the final dataset. Identification strategy and main results are discussedinsection3.Insection4,wegooversome robustness checks and show the heterogeneity of the coefficients in different sub-samples. Possible issues and drawbacks of the analysis, some concluding remarks, and suggestions for future research are presented in section 5. 1 A brief literature review Prenatal shocks to embryos could have adverse impacts on health endowment at birth and through this lower initial health capital affect the outcomes later in life. Almond (2006) exploits the nationwide spread of the 1918 influenza pandemic as a natural experiment and explores its long-term effects dur- ingadulthood.Hefindsthatcohortswhoseprenatal development period intersected with the pandemic showed lower educational attainments, lower in- come, lower socioeconomic status, more rates of physical disability, and larger transfer payments in their adulthood ages. Following this paper, Almond and Mazumder (2005) find that cohorts in utero duringthespreadofthefluillustrate higherratesof impaired health outcomes in old ages compared to cohorts born a few months before or after. A positive or negative nutritional shock in utero has significant short term and long-term effects. Hoynesetal.(2016)investigatethelong-runimpacts of participation in the Food Stamp Program. Access to food stamps during early childhood is associated with lower odds of metabolic syndrome and raises the economic self-sufficiency of women in adult- hood. The important implication of their paper is the causal effect of a direct policy-driven channel, namely nutrition, during childhood, and health outcomes later in life. A small strand of literature uses fasting during the Islamic holy month of Ramadanasaplausiblyexogenousnutritionalshock ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 195 in utero and investigates its impacts later in life. Being exposed to Ramadan during prenatal devel- opment is associated with lower birth weight, a fewer number of male births, and a higher likeli- hood of disability during adulthood (Almond & Mazumder, 2011), reduction in thetest scores at age 7(Almondetal.,2014),anincreaseinthesymptoms indicative of coronary heart problems and type 2 diabetes during old ages (Van Ewijk, 2011), lower cognitive and math test scores at school and fewer working hours during adulthood (Majid, 2015). NoghaniBehambari et al., 2020c investigate the ef- fects of enforcements in child support policies dur- ing the 1970s and 1980s in the United States as a pathway to improve the welfare of single mothers and find that this improvement has led to a reduc- tion in infant and child mortality rates. Based on these studies, we expect that a regional harvest failure has adverse health effects on children in utero or their early years of life by reducing nutri- tional stock available to households. Famines are widely studied cases to evaluate the long-term effects of malnutrition in utero. Scholte, Van Den Berg, and Lindeboom (2015) explore the outcomes of individuals who were in utero during The Dutch Hunger Winter (1944/45) by sub-interval of gestation. They find negative and significant ef- fects of exposure during the first trimester of gestation on employment outcomes at ages 53 and above. Exposure in the second and third trimester increasesthehospitalizationratesbeforeretirement. In a similar work, Neelsen and Stratmann (2011) find that exposure to the Greek famine of 1941/42 leads to lower educational attainments for cohorts who experienced the famine in utero or during the first year in life. Part of this effect can be explained by sharp re- ductions in income. In various ways, income can have positive effects on birth outcomes. For instance, Noghanibehambari andSalari(2020) show that an increase in income due to changes in Un- employment Insurance (UI) benefits during preg- nancy can increase birth weight among UI eligible mothers. They show that a change in health insur- ance away from public insurance and more towards presumably better quality private health insurance could be some mechanism channels. A small strand of literature uses exposure to pollution during the antenatal development period and explores its adverse health impacts on new- borns and adult outcomes. Isen et al. (2017) use US administrativedatatoevaluatethelong-termeffects of the 1970 Clean Air Amendment Act. They exploit the variation in differential exposures of counties that were affected by the act and find that higher exposureintheyearofbirthisassociatedwithlower earnings and labor force participation at age 30. Chay and Greenstone (2003) use variation in pollu- tion exposure in US counties caused by the 1981e1982 recession to estimate the effects of pollution on infant mortality rates and finds that a 1% reduction in total suspended particulates (TSP) will cause a 0.35% decline in infant mortality. In a similar study, Tavassoli et al. (2020) show that in- dustrial pollution during the late nineteenth and early twentieth century was associated with higher infant mortality as measured by gender ratio. Theeffectoftheseshockscouldbedeterioratedby parental over or under investment on children. The parental investment could reduce the health endowmentgapbetweentheiroffspringorreinforce this gap if they decide to invest more in their healthier children (Currie, 2011; Frijters, Johnston, Shah,&Shields,2013;Yi,Heckman,Zhang,&Conti, 2015). Restrepo (2016) finds evidence that low educated parents allocate more resources to their offspring who had initially normal birth weight compared to their low birth weight children, while highlyeducatedmotherscompensateforthishealth gap among their children. Mortality rates during adulthood and old age could be partly explained by early life nutritional andeconomicconditions.Yeungetal.(2014)explore this path and find that an adverse income shock caused by a recession during pregnancy and the first year of life will increase the risks of old-age cause-specific mortality. It increases the probability ofdeath due tocanceramongmalesand females by about 8 and 6 percentage points, respectively. Also, it increases the female mortality due to cardiovas- cular diseases by roughly 5 percentage points. NoghaniBehambarietal.(2020b)investigatethelink between early life economic conditions and old-age cause-specific mortality using US data. They show that a 1 percent decrease in the aggregate business cycle in the year of birth is associated with 2.2, 2.3, 3.1, 3.7, 0.9, and 2.1 percent increase in the proba- bility of mortality in old ages due to malignant neoplasms, diabetes mellitus, cardiovascular dis- eases, influenza, chronic respiratory diseases, and allotherdiseases,respectively.denBergetal.(2006) use a longitudinal dataset covering the years 1812e2000 for individuals born in the period 1812e1912 in the Netherlands and document that household economic conditions early in life can explain adult mortality rates. They exploit booms and busts during childhood as an instrument for individual economic conditions. Applying a hazard analysis, they find that being exposed to a boom early in life is associated with a 9% reduction in 196 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 adult mortality rates. Using the same instrument, vandenBergetal.,2011showthatbeingbornunder a recession increases the probability of cardiovas- cular mortality rates later in life. den Berg, Gupta, van den Berg, and Gupta (2015) introduce a causal pathway for this effect. They show that marital sta- tus, as a determinant of adult mortality, is itself affectedbyeconomicconditionsearlyinlife.Among women, longevity is reduced upon marriage in case they are born under adverse economic conditions. However, marriages have protective effect for men. Marriedmenenjoylongevity,andthemaritalstatus does not depend on economic conditions in early life. On the other hand, male mortality rates commove, negatively, with business fluctuations in their early childhood. Other relevant papers confirm the influence of economic and non-economic conditions during prenatal development and early life on later out- comes (Almond & Currie, 2011a; Banerjee, Duflo, Postel-Vinay, & Watts, 2010; Beach & Hanlon, 2016; Bhalotra, Karlsson, & Nilsson, 2017; Carlson, 2015; Case, Fertig, & Paxson, 2005; Cutler, Miller, & Nor- ton, 2007; den Berg et al., 2006; Duncan et al., 2017; Frijters, Hatton, Martin,& Shields, 2010,2013; Lin& Liu, 2014; Lindeboom, Portrait, & den Berg, 2010; Maccini&Yang,2009;Maruyama&Heinesen,2020; Myrskyl€ a, 2010; Myrskyl€ a, Mehta, & Chang, 2013; NoghaniBehambari, Noghani &Tavassoli, 2020a,b; Olafsson, 2016; Parman, 2015; Strand& Kunst, 2006; Tavassoli et al., 2020; Torche, 2011; Torche & Kleinhaus, 2011; von Hinke Kessler Scholder, Wehby, Lewis, & Zuccolo, 2014). 2 Backround, data, and varibale definitions 2.1 The great Tambora eruption The explosion of Mount Tambora in 1815 was the largest volcanic eruption in recorded history. It claimed the lives of approximately 88,000 people in the islands close to Sumbawa, Indonesia, the place of occurrence. The following ash eruption was the earth's largest in magnitude since the most recent Ice Age period (Stothers, 1984). In the year 1816, an unflagging and dry fog shadowed the northeastern parts of the United States. It continued during the spring and summer over a long period on a diurnal basis. A New York reportreads “thedryfogreddenedanddimmedthe sun to such an extent that sunspots became visible tothenakedeye”(Stothers,1984).Thehazereduced the sunshine and cooled down the surface temper- ature.Thus,thelengthofthegrowingseasonalmost halved in most regions including Southern Maine, Southern New Hampshire, and Eastern Massachu- setts. Snow began to fall in June in Albany, New York, andDansville, Main. Harsh frosts spreadover a wide area from Connecticut and New England to as south as Trenton, New Jersey (Oppenheimer, 2003). The inadequate feeding ground killed much livestock in New England during the winter of 1816e1817 (Baron, 1992). Briffa and Jones (1992) estimate that the summer temperature across western and central Europe was 1e2 Ccoolerin1816thantheaveragefortheperiod 1810e1819. Post (1977) provides evidence of large increases in indices of wholesale grain prices in Europe and North America which peaked in the year 1816 and subsisted relatively during the following years. The price of bread was so high that even customary wage earners could not afford it. In Europe, the consequences of the Napoleonic Wars aggravated the situation. The harvest failure and subsequent famine triggered social riots in most European cities with the proclamation of “Bread or Blood”. In Canada, however, such social distur- banceswerenotrevealedmostlyduetoanembargo on grain exports between July and September 1816e17 (Post, 1977). Several features of this phenomenon offer an appropriate context to test the Fetal Origin Hy- pothesis. First, the weather anomaly arrived completelyunprecedented,unforeseen,andwithout warning. Second, the location of the eruption was too far away to have any direct effect on US mothers. The main channel was the indirect effects which caused a persistent decrease in the sunshine, afallinthelengthofthegrowingseason,andfinally harvest failure. Third, preliminary trade roots hampered the import of food from neighboring states. The inadequate transportation network iso- lated the affected regions and aggravated the agri- cultural damage. Fourth, there are no government interventions or welfare policies in this period to neutralize the shock. Therefore, we expect that the indirect effect coerced only a nutritional effect on pregnant mothers following the crop damage of the 1816 summer. Fifth, the anomaly occurred only in the eastern regions of the US. It leaves the in- dividuals born in other states unaffected and so provides a control group. Comparing individual outcomes during adulthood who were born in eastern and northeastern states versus western, Midwestern, southern, and northwestern states (first difference), and those born in 1817 versus in the proceeding and following years (second differ- ence) can reveal the long-term effect of the in-utero exposure. The only channel for omitted variables to bias the estimated coefficients is that individuals’ ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 197 idiosyncratic shocks commove positively with year- regionsthatwereaffectedand negativelywith year- regions that were not. 2.2 Sample construction Inordertoassessitssocioeconomiceffects,weuse individual-level data from Census 1850 (100%). The data are extracted from Ruggles, Genadek, Goeken, Grover, and Sobek (2017). In the main analysis, we restrictthecohortstothose whowerebornbetween theyears1810e1820.However,theresultsarerobust to changes in the rolling window. Due to data limi- tations, we refer to historical reports and historical climate and temperature reconstruction studies in order to identify the exposed and non-exposed re- gions.Themainclimateanomalyhadbeenobserved in eastern and northeastern regions while western, southern,mid-western,andnorthwesternstateshad not been affected (Keith R Briffa, Jones, Schweing- ruber, & Osborn, 1998; Lough, 1992; Oppenheimer, 2003). Therefore, we divide the sample into two re- gions: states that were mentioned in historical re- cordsanddefinitelyexposedtotheweatheranomaly, andimmunestateslocatesinotherareas. 2 Since the main increase in grain prices commenced circa summer, fall, and winter of 1816e1817, cohorts who were born in 1817 should have experienced the nutritional shock during pre- natal development. Therefore, 1817-born cohorts in affected regions constitute the treatment groups while other cohortsserve asthe control groups.Ina robustnesscheck,weexchangethecontrolgroupby cohorts born during 1816e1818, combined 1816 through 1818, and combined 1817e1818, separately, and show that the most affected groups are actually the 1817-born cohorts. A summary of the characteristics of the final sample is reported in Table 1 for two illustrative cohorts born in 1817 and 1819. The average labor force participation rate for 1817-born cohorts in affected regions are, qualitatively, lower than all other cohorts. Their average literacy rates during adulthood are lower compared to the 1819-born persons in the same states. The ratio of females is the most compared to all other cohorts. Their number of children, however, does not follow this pattern. 3 Empirical model and main results In order to capture the average outcomes of adult cohorts who were born in 1817 in states exposed to the harvest failure of 1816, we run different formu- lations of the following Difference-in-Difference model: y isz ¼b 0 þb 1 YOB1817 i þb 2 SOBexp i þb 3 YOB1817 i SOBexp i þb 4 X i þGðYOB i Þþz s þx z þe isz ð1Þ where i indexes the individual, s the current state of residence, and z the state of birth. y refers to indi- vidual outcomes including activity status in the laborforce,literacy,gender,andthenumberofown children. The indicator YOB1817 equals one if the individual is born in the year 1817 and zero other- wise. SOBexpcontains a set of dummies to identify the exposed regions as explained in section 2.2. Some individuals' characteristics are captured in X including a fourth polynomial function of age, Duncan Socioeconomic Index (SEI), 3 and race. Gð:Þ is a third-degree polynomial function to capture the secular trends in the birth year. Besides, for each outcome, we also include other related outcomes as Table 1. Summary statistics. Birth Place Eastern States Birth Place West-Northern States (YOB: 1817) (YOB: 1819) (YOB: 1817) (YOB: 1819) % Active in Labor Force 49.74 (50.00) 50.38 (50.00) 53.02 (49.92) 53.29 (49.90) % Literate 92.67 (26.07) 93.45 (24.75) 72.19 (44.81) 74.39 (43.66) % Female 48.13 (49.97) 47.44 (49.93) 45.04 (49.76) 44.34 (49.69) No. of Children 2.019 (1.972) 1.864 (1.826) 2.408 (2.313) 2.267 (2.142) % Whites 95.99 (19.62) 96.59 (18.15) 87.93 (32.58) 90.02 (29.98) Number of Cases 166,859 140,632 3,546 3,207 Note: Standard deviations are reported in parentheses. 2 Identifiedexposedstatesare:Maine,Vermont,NewHampshire,Massachusetts,RhodeIsland,Connecticut,NewJersey,Delaware,Maryland,Virginia, New York, Pennsylvania, West Virginia, North Carolina; while the following states are recognized as the unaffected regions: Minnesota, Texas, New Mexico, Arizona, Nevada, California, Colorado, Kansas, Oklahoma, Missouri, Arkansas, North Dakota, South Dakota, Wyoming, Idaho, Montana, Washington, Utah, Nebraska, Iowa, Louisiana. 3 Duncan Socioeconomic Index or equivalently SEI is a constructed score for each occupation based on the income and education level for that occu- pation. The higher the index, the higher the presumed socioeconomic class of the person. The score varies from 1 to 96 (top coded). 198 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 the control variables. Specifically, for the outcome literacy, we control for the person's gender. For the outcome labor force participation, we control for gender, literacy, and number of own children. Finally, for the outcome number of own children, we control for gender and literacy. In z and x are included a set of dummies for the current state of residence and another set for the state of birth. Finally, e is a disturbance term. All standard errors are clustered on the place of birth. If individuals’ idiosyncratic error terms, e.g. the family genetic inheritance, varies systematically for the1817-borncohortinaffectedregionscomparedto othercohorts,thenequation(1)suffersfromomitted variable bias. However, the eruption happened without warning and to an unprecedented degree. Theanomalyhasnotbeenexperiencedoverthepast hundreds of years. 4 Therefore, its outbreak and regional variation were perfectly orthogonal to in- dividualsandhouseholdcharacteristics. Male embryos are more susceptible to external shocks during the antenatal period. Therefore, harsh circumstances are associated with a higher ratio of females in live births due to higher fetal death and infant mortality among male newborns. Hence, the gender ratio (female-male) has been used as a proxy for infant mortality (Sanders & Stoecker, 2011). 5,6 The main results of regressions introduced in equation (1) are presented in Table 2. Shown in the full specification in column 2 of each panel, expo- sure to the climate anomaly during prenatal devel- opment is associated with 1.2 percentage points lowerprobabilitytoparticipateinthelaborforce,2.7 percentagepointslowerprobabilityofbeingliterate, 0.10 fewer children, and 1.1 basis points higher likelihood of being female observed during adult- hood. These numbers are equivalent to 2.4, 2.3, 3.6, and4.4percentchangefromthemeanoflaborforce participation, gender ratio, literacy,and the number of children, respectively. Whether or not including individual characteristics and fixed effects, the co- efficients are significant at the conventional levels. Since the gender ratio is used to a proxy infant mortality and fetal death, a higher likelihood of being female (i.e. a higher female-male ratio) im- plieshigherratesofmortalityamongnewbornswho were in utero during the cold summer. Lower labor force participation and education (proxied by literacy rates) have been linked to initial health en- dowments byseveral studies that explored the Fetal Origin Hypothesis (Almond & Currie, 2011a,b). These results suggest that there are long-term con- sequences for those cohorts in utero during the harvest failure of summer and winter of 1816e1817. 4 Robustness checks In this section, we go over some alternative specifications to check the robustness of the main results.AsshowninPost(1977),thewholesalegrain prices peaked in 1816 and subsided in the following years. However, they did not reach the initial levels until 1819. Therefore, cohorts born even two years after the harvest failure could still be affected. We replace the treated time in equation (1) with a different combination of cohorts born between 1816 and1818,separately.TheresultsareshowninTable 3andTable4.Theimpactsaremorepronouncedfor birth cohorts 1817 who experienced the 1816 sum- mer and winter in utero. There is no evidence of an impact on the 1816-birth cohorts. As shown in the first panel of Table 3, the 1818-born cohorts reveal lower labor force participation and the coefficient is significant. Still, its magnitude is slightly lower than that of the 1817-born cohorts. Aggregating both 1817- and 1818-birth cohorts boost the marginal ef- fects. The exposed cohorts who were born between the years 1817 and 1818 have 1.4 percentage points lower likelihood of participation in the labor force during adulthood. Thesecond panelillustratestheresultsofliteracy. Althoughthecoefficientsofthe1816-and1818-born cohorts are negative, they are not significant. The reductioninliteracyoccurredonlyforthoseinutero during the harvest failure, i.e. 1817-cohorts, and for the 1817e18 aggregated cohorts. Similar results are obtained for the number of children and gender. Depicted in the left panel of Table 4, the effects on the number of children are negative but insignifi- cant. The increase in the probability of being a fe- male, as shown in the right panel, is the same for 1817-born cohorts and 1818-born cohorts. Using 1817e1818 year of birth as the treated time reveals quiteasimilarmarginaleffect(0.9percentagepoints versus 0.8 percentage points). Next, we split the sample by race and gender and run the full specification model for each group individually. The results are illustrated in Table 5 4 As stated in Oppenheimer (2003) and Briffa and Jones (1992), the cold summer of 1816 was the coldest of the past six centuries. 5 Someevidenceandapplicationofthisfactisprovidedintheliterature(Almond&Edlund,2007;Bruckner&Catalano,2007;Torche&Kleinhaus,2011). 6 Almond and Mazumder (2011) use in utero exposure to Ramadan, the holy month of Muslims in which they fast from sunrise to sunset, as a natural experimentandfindthatthenutritionalshockhasnegativehealtheffectsonnewborns.However,inalongerhorizonanalysis,theyalsoshowthatexposed individuals have higher female to male ratio. The sex composition in the adulthood could point to the initial gender ratio bias at birth. ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 199 Table 3. Long-term effects of in utero exposure to the cold summer of 1816 for different threshold of time treatment. Labor Force Participation Literacy YOB ¼ 1816 YOB ¼ 1817 YOB ¼ 1818 YOB ¼ 1816-18 YOB ¼ 1817-18 YOB ¼ 1816 YOB ¼ 1817 YOB ¼ 1818 YOB ¼ 1816-18 YOB ¼ 1817-18 Treated States Birth year 1816 0.004 (0.004) 0.005 (0.007) Treated States Birth year 1817 0.012** (0.005) 0.028*** (0.006) Treated States Birth year 1818 0.011** (0.005) 0.003 (0.007) Treated States Birth year 1816e18 0.006 (0.008) 0.013 (0.010) Treated States Birth year 1817e18 0.014** (0.007) 0.023* (0.012) Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes State of Birth FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes State of Residence FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes F(Birth Year) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes G (Age) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 902,225 902,225 902,225 902,225 902,225 Note:Standarderrors,clusteredonthestateofbirth,arereportedinparentheses.Fð:Þisathird-degreepolynomial.Gð:Þisaquadraticpolynomial.Controlsforlaborforceparticipation include: race dummies, Duncan Socioeconomic index, number of children, literacy, and gender. Controls for Literacy include: race dummies, Duncan Socioeconomic index, and gender. Controls for Number of Children include: race dummies, Duncan Socioeconomic index, literacy, and gender. Controls for Gender include: race dummies, and Duncan Socioeconomic index. While only the interaction coefficients are shown, the main effects are included in all regressions. Table 2. Long-term effects of in utero exposure to the cold summer-winter of 1816. Labor Force Participation Gender (Female ¼ 1) Literacy Number of Children (1) (2) (1) (2) (1) (2) (1) (2) Treated States Birth Year 1817 0.021*** (0.006) 0.012** (0.005) 0.020*** (0.006) 0.011** (0.004) 0.060** (0.025) 0.027*** (0.006) 0.226 (0.282) 0.102** (0.043) Treated States 0.011* (0.006) 0.110 (0.256) 0.013* (0.006) 0.182 (0.129) 0.267*** (0.025) 0.022 (0.025) 0.151 (0.282) 3.159** (1.467) Birth Year 1817 0.024*** (0.006) 0.014** (0.005) 0.020*** (0.006) 0.009** (0.004) 0.069*** (0.024) 0.030*** (0.006) 0.148 (0.188) 0.121*** (0.038) Duncan Socioeconomic index 0.018*** (0.000) 0.017*** (0.000) 0.001*** (0.000) 0.009*** (0.001) Race: White 0.133*** (0.011) 0.085*** (0.007) 0.352*** (0.017) 0.550*** (0.065) F (Birth Year) No Yes No Yes No Yes No Yes Year FE No Yes No Yes No Yes No Yes State of Birth FE No Yes No Yes No Yes No Yes State of Residence FE No Yes No Yes No Yes No Yes G (Age) No Yes No Yes No Yes No Yes Number of Cases 1,242,262 1,242,262 1,242,262 1,242,262 902,225 902,225 1,242,262 1,242,262 Note:Standarderrors,clusteredonthestateofbirth,arereportedinparentheses.Fð:Þisathird-degreepolynomial.Gð:Þisaquadraticpolynomial.Controlsforlaborforceparticipation include: race dummies, Duncan Socioeconomic index, number of children, literacy, and gender. Controls for Literacy include: race dummies, Duncan Socioeconomic index, and gender. Controls for Number of Children include: race dummies, Duncan Socioeconomic index, literacy, and gender. Controls for Gender include: race dummies, and Duncan Socioeconomic index. 200 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 Table 5. Long-term effects of in utero exposure to the cold summer of 1816 for different race-gender groups, outcome: labor force participation. DV: Labor Force Participation White Non-White Males Females (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se Treated States Birth Year 1817 0.007 (0.006) 0.007 (0.005) 0.004 (0.013) 0.005 (0.010) 0.008* (0.004) 0.008* (0.004) 0.010* (0.005) 0.010* (0.005) Birth Year 1817 0.005 (0.005) 0.010** (0.004) 0.016 (0.012) 0.020* (0.009) 0.008* (0.004) 0.012*** (0.004) 0.011* (0.005) 0.011* (0.006) Treated States 0.015 (0.314) 0.258 (0.261) 0.279 (0.251) 0.330 (0.376) 0.549* (0.311) 0.566* (0.323) 0.300 (0.254) 0.200 (0.276) Controls No Yes No Yes No Yes No Yes F (Birth Year) No Yes No Yes No Yes No Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes State of Birth FE No Yes No Yes No Yes No Yes State of Residence FE Yes Yes Yes Yes Yes Yes Yes Yes G (Age) No Yes No Yes No Yes No Yes Observations 1,105,262 1,105,262 42,820 42,820 595,352 595,352 552,730 552,730 Note:Standarderrors,clusteredonthestateofbirth,arereportedinparentheses.Fð:Þisathird-degreepolynomial.Gð:Þisaquadraticpolynomial.Controlsforlaborforceparticipation include: race dummies, Duncan Socioeconomic index, number of children, literacy, and gender. Table 4. Long-term effects of in utero exposure to the cold summer of 1816 for different threshold of time treatment. Number of Children Gender (Female ¼ 1) YOB ¼ 1816 YOB ¼ 1817 YOB ¼ 1818 YOB ¼ 1816-18 YOB ¼ 1817-18 YOB ¼ 1816 YOB ¼ 1817 YOB ¼ 1818 YOB ¼ 1816-18 YOB ¼ 1817-18 Treated States Birth year 1816 0.032 (0.036) 0.005 (0.004) Treated States Birth year 1817 0.029 (0.040) 0.008** (0.004) Treated States Birth year 1818 0.042 (0.045) 0.008** (0.004) Treated States Birth year 1816e18 0.071 (0.056) 0.003 (0.006) Treated States Birth year 1817e18 0.046 (0.057) 0.009* (0.005) Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes State of Birth FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes State of Residence FE Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes F(Birth Year) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes G (Age) Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 1,242,262 Note:Standarderrors,clusteredonthestateofbirth,arereportedinparentheses.Fð:Þisathird-degreepolynomial.Gð:Þisaquadraticpolynomial.Controlsforlaborforceparticipation include: race dummies, Duncan Socioeconomic index, number of children, literacy, and gender. Controls for Literacy include: race dummies, Duncan Socioeconomic index, and gender. Controls for Number of Children include: race dummies, Duncan Socioeconomic index, literacy, and gender. Controls for Gender include: race dummies, and Duncan Socioeconomic index. While only the interaction coefficients are shown, the main effects are included in all regressions. ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 201 Table 6. Long-term effects of in utero exposure to the cold summer of 1816 for different race-gender groups, outcome: literacy. DV: Literacy White Non-White Males Females (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se Treated States Birth Year 1817 0.015** (0.007) 0.016** (0.006) 0.031 (0.028) 0.032 (0.024) 0.008 (0.011) 0.007 (0.008) 0.023* (0.011) 0.022** (0.007) Birth Year 1817 0.016** (0.007) 0.014** (0.006) 0.047* (0.027) 0.034 (0.023) 0.014 (0.010) 0.012 (0.007) 0.029** (0.010) 0.025*** (0.007) Treated States 0.159*** (0.014) 0.039* (0.020) 0.220*** (0.063) 0.327 (0.375) 0.312 (0.371) 0.371 (0.292) 0.156*** (0.015) 0.059*** (0.014) F (Birth Year) No Yes No Yes No Yes No Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes State of Birth FE No Yes No Yes No Yes No Yes State of Residence FE Yes Yes Yes Yes Yes Yes Yes Yes G (Age) No Yes No Yes No Yes No Yes Number of Cases 871,807 871,807 30,999 30,999 465,831 465,831 436,975 436,975 Note: Standard errors, clusteredonthe state ofbirth, arereportedin parentheses. Fð:Þ is athird-degreepolynomial. Gð:Þ is aquadraticpolynomial.Controls for Literacy include: race dummies, Duncan Socioeconomic index, and gender. Table 7. Long-term effects of in utero exposure to the cold summer of 1816 for different race-gender groups, outcome: number of children. DV: Number of Children White Non-White Males Females (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se (1) b/se (2) b/se Treated States Birth Year 1817 0.068 (0.148) 0.051 (0.044) 0.195 (0.146) 0.224*** (0.060) 0.029 (0.153) 0.014 (0.055) 0.142 (0.128) 0.110** (0.049) Birth Year 1817 0.109 (0.099) 0.093** (0.040) 0.049 (0.106) 0.084* (0.047) 0.155 (0.097) 0.054 (0.053) 0.004 (0.086) 0.153*** (0.043) Treated States 1.572*** (0.306) 0.298 (0.333) 1.905 (2.266) 1.365 (1.280) 0.781 (2.019) 2.604 (1.895) 0.355 (1.058) 0.713 (0.777) F (Birth Year) No Yes No Yes No Yes No Yes Year FE Yes Yes Yes Yes Yes Yes Yes Yes State of Birth FE No Yes No Yes No Yes No Yes State of Residence FE Yes Yes Yes Yes Yes Yes Yes Yes G (Age) No Yes No Yes No Yes No Yes Number of Cases 1,181,861 1,181,861 60,453 60,453 644,019 644,019 598,295 598,295 Note: Standard errors, clustered on the state of birth, are reported in parentheses. Fð:Þ is a third-degree polynomial. Gð:Þ is a quadratic polynomial. Controls for Number of Children include: race dummies, Duncan Socioeconomic index, literacy, and gender. 202 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 through Table 7. Females are the most affected groups regarding literacy, labor outcomes, and the number of children. Literacy outcome of whites shows significant reductions due to the exposure, while the effects on their labor market status are insignificant (see Table 6). The only anomaly in the sensitivity results ap- pears in nonwhites for the outcome of number of children in Table 7. This coefficient is positive and statisticallysignificantata5percentlevel.Sincethis effect is not consistent with other findings and ap- pears only in one subsample and only for one outcome, we do not believe that it negatively affects the overall findings of this paper. 4.1 Exposure and labor outcomes in England, Canada, and Norway The cold summer of 1816 affected most parts of thenorthernhemisphere.Thefamineduetoharvest failure combined with social disturbances was re- ported in most of west, north, and central Europe. The typhus epidemic that accompanied the famine was reported in almost every village and town in England (Post, 1977). In Montreal, Canada, snowfall started at the beginning of June. In Quebec, 30 cm snow was accumulated in the middle of 1816 sum- mer. Based on this evidence, as a robustness check to the main findings in the US, we use census data of other countries to check the exposure effects on later life labor force participation. However, the choice of the outcome and the countries in this section are limited by data availability. We use 1881 census data for Canada, 1865 census data for Nor- way, and 1851 census data for England. 7 All these censuses are based on a 100% sample. To examine the effect of climate anomaly and its aftermath during 1816e1817, we use a cohort anal- ysis explained by the following equation: y idz ¼b 0 þb 1 IðYOB¼1817Þþb 2 IðYOB¼1818Þ þb 4 X i þGðYOB i Þþz d þx z þn idz ð2Þ Inthisformulation, Ið:Þisanindicator function, z indexes the country of birth, and d indexes the sub-districtresidenceofindividuali. 8 Establishinga causal relation in equation (2) is hampered for a couple of reasons. The first is the lack of region- specific variation. If the cooler summers and more ruin of the harvest occurred in regions in which lower socioeconomic households live, then the co- efficients b 1 and b 2 onlyshowaspuriouscorrelation since such families would have had children with lower initial health endowment even if the cold summer had not taken place. Second, there are universal changes in cohort quality and nationwide factors,e.g.macroeconomicconditionsthataffectall individuals uniformly. This equation fails to distin- guish between cohort-specific universal changes fromthepureeffectofharvestfailure.Third,fertility decisions could have responded to the climate anomaly.Thiswillcauseaselectionissueifthereare correlations between their decision to childbearing during hard times and their characteristics. If households with higher socioeconomic status choose to have children even at the time of agri- cultural calamities and lower socioeconomic status households postpone the fertility timing, then the coefficients will underestimate the true effects. The best way to analyze the effects is to have longitu- dinal data that cover households’ information before and after the harvest failure. However, such datasets are scarce specifically during thevery early years of the nineteenth century. The correlational link between the climate crisis exposure in utero and labor force outcomes in adulthoodarepresentedinTable8.Thefirstcolumn of each panel reports the coefficients for a regres- sion including polynomial functions for age and birth year. This is quite similar to the baseline cohort analysis in the literature (Almond, 2006). Exposed individuals show, on average, 0.9 percent- agepoints,0.7percentagepoints,and0.2percentage points lower participation rates in Canada, Norway, and England, 9 if born in 1817. The negative long- termeffectsarestatisticallysignificantandrelatively large. Being born in 1818 has quite similar effects in Norway and England, but insignificant effects for Canadianindividuals.Theseresultssuggestthatthe climate anomaly and subsequent harvest failure in Europe and North America had considerable long- term effects. Recall that the average temperature in 1816summerwas1e2 Ccoolerthan1810s averages (Oppenheimer, 2003). This summer temperature reductionaffectedthehealthendowmentofinutero cohorts due to the negative nutritional shock to pregnant mothers. As a result, the 1817-born gen- eration revealed adverse labor force outcomes 7 OnthecontrarytotheUScensus,otheroutcomesorcontrolvariablesarenotavailableinthesecensuses.Intheregressions,weonlycontrolforgender that is the only available variable. We choose these years based on data availability for the year with 100% sample counts, as well as availability of labor force participation status. 8 Theregionofresidenceisspecifictoeachcountryandrestrictedtodataavailabilityincountry-specificcensus.Weusethemostdisaggregatedregions: county for England and Wales, Province for Norway, and district for Canada. 9 The census data covers England and Wales together. ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 203 compared to non-exposed cohorts born just before and after the shock. This evidence can support the negative effects found among individual outcomes in the United States. 5 Discussion and conclusion Initial health endowment at birth has significant predictive power for a broad array of individuals' later life outcomes. For instance, a nutritional shock during prenatal development can leave newborns with lower health capital at birth. Through this channel, it taints individuals' outcomes during adulthood. Chief among these outcomes are edu- cation and labor force participation. The main challengeinthisstudyistofindashocktopregnant mothers strong enough that affects the health endowment of newborns and meanwhile orthog- onal to mothers' characteristics. This paper explores this question using a plausibly exogenous shock during the early 19th century: the great Tambora eruption in 1815, the largest eruption in recorded history. Although occurred in Indonesia, the after- math ash columns were so large that they spread over the globe. The formed haze located above the troposphere hampered the sunshine and led to a severe temperature reduction in the summer and winter of 1816e17. The cooling summer and diminished length of growing season resulted in dreadful harvest failure. The climate anomaly affected most eastern and northeastern US states, while it did not reach the southern, western, mid- western, and northwestern states. The exogenous nature of the haze that was uncorrelated with in- dividuals’ characteristics at exposed region-years, the lack of governmental protective policy, high latitude location of the haze which rules out the possibility of direct effects through pollution, and inadequate trade roots provide an isolated and exogenous nutritional shock to pregnant mothers. Usingcensusdatacoveringtheyears1850e1880,we show that in utero exposure to agricultural failure due to this climate anomaly had statistically and economically significant effects on later life outcomes. Exposed cohorts revealed, on average, 1.2 per- centage points lower likelihood of being active in the labor force, 2.6 percentage points lower proba- bility of being literate, and 0.8 percentage points higher likelihood of being a female. The effects are more pronounced for 1817-born cohorts compared to a case in which 1816- and 1818-born individuals are considered the treated cohorts. A gender decompositionshowsthatfemalesaremoreaffected by the climate shock. A race decomposition shows Table 8. Long-term effects of in utero exposure to the cold summer of 1816 in Norway, England, and Canada. Labor Force Participation Canada Norway England (1) b/se (2) b/se (3) b/se (1) b/se (2) b/se (3) b/se (1) b/se (2) b/se (3) b/se Birth Year 1817 0.011** (0.005) 0.011** (0.005) 0.009** (0.005) 0.008* (0.004) 0.007* (0.004) 0.007* (0.004) 0.002** (0.001) 0.002* (0.001) 0.002* (0.001) Birth Year 1818 0.005 (0.005) 0.005 (0.005) 0.006 (0.005) 0.007 (0.004) 0.007* (0.004) 0.007* (0.004) 0.005*** (0.001) 0.001 (0.001) 0.005*** (0.001) F (Birth Year) Yes Yes Yes Yes Yes Yes Yes Yes Yes Year FE No Yes Yes No Yes Yes No Yes Yes Region FE No No Yes No No Yes No No Yes Country of Birth FE No No Yes No No Yes No No Yes G (Age) No Yes Yes No Yes Yes No Yes Yes Number of Cases 130,719 130,719 130,719 160,528 160,528 160,528 2,396,671 2,396,671 2,396,671 Note: Standard errors, clustered on the state of birth, are reported in parentheses. Fð:Þ is a third-degree polynomial. Gð:Þ is a quadratic polynomial. The regressions also control for gender. 204 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 thatliteracyoutcomesofwhiteswereinfluencedtoa considerably larger degree by the shock compared to nonwhites. As an alternative check, we used census data of three other affected countries, namely England, Norway, and Canada. Applying a simple cohort analysis, we found that 1817-born cohorts who were exposed in utero to the harvest failure of summer and winter of 1816e17 illustrate lower labor force participation rates during their adulthood years. All in all, there are some drawbacks in this study that hinder a causal link. First, the estimated co- efficients show only the average treatment effect on the treated cohorts. No heterogeneity is identified across individuals. In order to build up individual- specific shocks, it requires information on in- dividuals and households before and after the weather irregularity. Moreover, it should track in- dividuals into their adulthood. Such a dataset sim- ply does not exist in the early years of the 19th century. More importantly, the main channel through which the nutritional shock could affect individuals' outcomes is the initial health endow- ment. A good approach is to provide evidence of this first stage effect. For example, the effects on fetal death, maternal mortality, birth outcomes, and infant mortality could be proxied for initial health endowments. Again, lack of data impedes such es- timates. Second, households might respond to shock in two ways. They could postpone fertility decisions in order to have healthier and more pro- ductive children. If this fertility timing is correlated with households' characteristics, like socioeconomic status, then the long-term estimates suffer from selection issues. Another channel is reimbursement or reinforcement behavior of families as a response to the initial health status of children. If households over-invest in their healthier children and allocate fewer resources to their children with lower initial health endowment, then the long-term coefficients overstate the true effects. More noticeably, if such reinforcement behavior is correlated with families' features, like socioeconomic status, or mothers’ characteristics, like education, the estimations will be biased. Thethirdissuecomesfromthenatureofthelong- term analysis. Those individuals with low health endowment show, on average, higher rates of fetal death, infant mortality, and toddler mortality. The mortality rates could potentially sort out stronger individuals to reach adulthood, the period in which their outcomes are observed. This selection issue will bias the coefficients and understate the true effects. References Almond, D. (2006). Is the 1918 influenza pandemic over? Long- termeffectsofinuteroinfluenzaexposureinthepost-1940US population. Journal of Political Economy, 114(4), 672e712. Available at: https://doi.org/10.1086/507154. Almond, D., & Currie, J. (2011). Human capital development before age five. Issue PART B. In , Vol. 4. Handbook of labor economics. Elsevier. Available at: https://doi.org/10.1016/ S0169-7218(11)02413-0. Almond,D.,&Currie,J.(2011).Killingmesoftly:Thefetalorigins hypothesis. The Journal of Economic Perspectives, 25(3),153e172. Available at: https://doi.org/10.1257/jep.25.3.153. Almond,D.,&Edlund,L.(2007).Trivers-Willardatbirthandone year:EvidencefromUSnatalitydata1983e2001. Proceedings of the Royal Society B: Biological Sciences, 274(1624), 2491e2496. Almond, D., & Mazumder, B. (2005). The 1918 influenza pandemic and subsequent health outcomes: An analysis of SIPP data. The American Economic Review, 95(2), 258e262. Almond, D., & Mazumder, B. (2011). Health capital and the pre- natal environment: The effect of Ramadan observance during pregnancy. American Economic Journal: Applied Economics, 3(4), 56e85. Available at: https://doi.org/10.1257/app.3.4.56. Almond, D., Mazumder, B., & Van Ewijk, R. (2014). In utero Ramadanexposure andchildren'sacademicperformance. The Economic Journal, 125(589), 1501e1533. Banerjee,A.,Duflo,E.,Postel-Vinay,G.,&Watts,T.(2010).Long- run health impacts of income shocks: Wine and phylloxera in nineteenth-century France. The Review of Economics and Sta- tistics, 92(4), 714e728. Baron, W. R. (1992). 1816 in perspective: The view from the North- eastern United States. The Year without a summer? World Climate in 1816 (pp. 124e144). Beach, B., & Hanlon, W. W. (2016). Coal smoke and mortality in an early industrial economy. The Economic Journal, 128(615), 2652e2657. Bhalotra, S., Karlsson, M., & Nilsson, T. (2017). Infant health and longevity: Evidence from a historical intervention in Sweden. Journal of the European Economic Association, 15(5), 1101e1157. Briffa, K. R., & Jones, P. D. (1992). The climate of Europe during the 1810s with special reference to 1816. Briffa, K. R., Jones, P. D., Schweingruber, F. H., & Osborn, T. J. (1998). Influence of volcanic eruptions on Northern Hemi- sphere summer temperature over the past 600 years. Nature, 393(6684), 450. Bruckner,T.,&Catalano,R.(2007).Thesexratioandage-specific male mortality: Evidence for culling in utero. American Journal of Human Biology: The Official Journal of the Human Biology As- sociation, 19(6), 763e773. Carlson, K. (2015). Fear itself: The effects of distressing economic news on birth outcomes. Journal of Health Economics, 41, 117e132. Available at: https://doi.org/10.1016/j.jhealeco.2015. 02.003. Case, A., Fertig, A., & Paxson, C. (2005). The lasting impact of childhood health and circumstance. Journal of Health Eco- nomics, 24(2), 365e389. Chay, K. Y., & Greenstone, M. (2003). The impact of air pollution on infant mortality: Evidence from geographic variation in pollution shocks induced by a recession. Quarterly Journal of Economics, 118(3), 1121e1167. Available at: https://doi.org/10. 1162/00335530360698513. Currie, J. (2011). Inequality at birth: Some causes and conse- quences.TheAmericanEconomicReview,101(3),1e22.Available at: https://doi.org/10.1257/aer.101.3.1. Cutler, D. M., Miller, G., & Norton, D. M. (2007). Evidence on early-life income and late-life health from America's Dust Bowl era. Proceedings of the National Academy of Sciences, 104(33), 13244e13249. den Berg, G. J., Gupta, S., van den Berg, G. J., & Gupta, S. (2015). The role of marriage in the causal pathway from economic ECONOMIC AND BUSINESS REVIEW 2021;23:194e206 205 conditionsearlyinlifetomortality.Journalof HealthEconomics, 40, 141e158. Available at: https://doi.org/10.1016/j.jhealeco. 2014.02.004. den Berg, G. J., Lindeboom, M., & Portrait, F. (2006). Economic conditions early in life and individual mortality. The American Economic Review, 96(1), 290e302. Duncan,B.,Mansour,H.,&Rees,D.I.(2017).It'sjustagame:The Super Bowl and low birth weight. Journal of Human Resources, 52(4), 946e978. Available at: https://doi.org/10.3368/jhr.52.4. 0615-7213R. Frijters, P., Hatton, T. J., Martin, R. M., & Shields, M. A. (2010). Childhood economic conditions and length of life: Evidence from the UK Boyd Orr cohort, 1937e2005. Journal of Health Economics, 29(1), 39e47. Frijters, P., Johnston, D. W., Shah, M., & Shields, M. A. (2013). Intrahousehold resource allocation: Do parents reduce or reinforce child ability gaps? Demography, 50(6), 2187e2208. von Hinke Kessler Scholder, S., Wehby, G. L., Lewis, S., & Zuccolo, L. (2014). Alcohol exposure in utero and child aca- demic achievement. The Economic Journal, 124(576), 634e667. Hoynes,H.,Schanzenbach,D.W.,&Almond,D.(2016).Long-run impacts of childhood access to the safety net. The American Economic Review, 106(4),903e934.Availableat: https://doi.org/ 10.1257/aer.20130375. Isen, A., Rossin-Slater, M., & Walker, W. R. (2017). Every breath you take, every dollar you'll make: The long-term conse- quencesofthecleanairactof1970.JournalofPoliticalEconomy, 125(3), 848e902. Lindeboom, M., Portrait, F., & den Berg, G. J. (2010). Long-run effects on longevity of a nutritional shock early in life: The Dutchpotatofamineof1846e1847. Journal of Health Economics, 29(5), 617e629. Lin, M.-J., & Liu, E. M. (2014). Does in utero exposure to illness matter? The 1918 influenza epidemic in Taiwan as a natural experiment. Journal of Health Economics, 37, 152e163. Lough,J.M.(1992).Climatein1816and1811e20asreconstructed from western North American tree-ring chronologies'. The Year Without a Summer,97e114. Maccini, S., & Yang, D. (2009). Under the weather: Health, schooling, and economic consequences of early-life rainfall. The American Economic Review, 99(3), 1006e1026. Majid, M. F. (2015). The persistent effects of in utero nutrition shocks over the life cycle: Evidence from Ramadan fasting. Journal of Development Economics, 117,4 8 e57. Available at: https://doi.org/10.1016/j.jdeveco.2015.06.006. Maruyama, S., & Heinesen, E. (2020). Another look at returns to birthweight. Journal of Health Economics, 70, 102269. Available at: https://doi.org/10.1016/j.jhealeco.2019.102269. Myrskyl€ a, M. (2010). The relative effects of shocks in early- and later-life conditions on mortality. Population and Development Review, 36(4), 803e829. Available at: https://doi.org/10.1111/j. 1728-4457.2010.00358.x. Myrskyl€ a, M., Mehta, N. K., & Chang, V. W. (2013). Early life exposure to the 1918 influenza pandemic and old-age mor- tality by cause of death. American Journal of Public Health, 103(7), e83ee90. Neelsen, S., & Stratmann, T. (2011). Effects of prenatal and early life malnutrition: Evidence from the Greek famine. Journal of Health Economics, 30(3), 479e488. Noghanibehambari, H., Noghani, F., & Tavassoli, N. (2020). Child support enforcement and child mortality. Applied economics letters. Available at: https://doi.org/10.1080/13504851.2020.1869157. NoghaniBehambari,H., Noghani,F., & Tavassoli, N. (2020). Early life income shocks and old-age cause-specific mortality. Eco- nomic Analysis, 53(2), 1e19. NoghaniBehambari, H., Noghani, F., Tavassoli, N., & Toranji, M. (2020). Long term effects of in utero exposure to “the year without a summer”. SSRN Electronic Journal. Available at: https://doi.org/10.2139/ssrn.3668739. Noghanibehambari, H., & Salari, M. (2020). Health benefits of social insurance. Health Economics, 29(12), 1813e1822. Avail- able at: https://doi.org/10.1002/hec.4170. Olafsson, A. (2016). Household financial distress and initial en- dowments: Evidence from the 2008 financial crisis. Health Economics, 25,43e56. Available at: https://doi.org/10.1002/hec. 3426. Oppenheimer, C. (2003). Climatic, environmental and human consequencesofthelargestknownhistoriceruption:Tambora volcano (Indonesia) 1815. Progress in Physical Geography, 27(2), 230e259. Page,M.,Schaller,J.,&Simon,D.(2017).Theeffectsofaggregate and gender-specific labor demand shocks on child health. Journal of Human Resources, 54(1), 0716e8045R. Available at: https://doi.org/10.3368/JHR.54.1.0716.8045R. Parman, J. (2015). Childhood health and sibling outcomes: Nurture Reinforcing nature during the 1918 influenza pandemic. Explorations in Economic History, 58,22e43. Persson, P., & Rossin-Slater, M. (2018). Family ruptures, stress, and the mental health of the next generation. The American Economic Review, 108(4e5), 1214e1252. Post, J. D. (1977). The last great subsistence crisis in the Western World. Johns Hopkins Univ Pr. Restrepo,B.J.(2016).Parentalinvestmentresponsestoalowbirth weight outcome: Who compensates and who reinforces? Journal of Population Economics, 29(4), 969e989. Ruggles, S., Genadek, K., Goeken, R., Grover, J., & Sobek, M. (2017). Integrated public use microdata series: Version 7.0 [dataset] (p. 10). Minneapolis: University of Minnesota. Available at: https://doi.org/10.18128/D010.V7.0. Sanders, N. J., & Stoecker, C. F. (2011). Where have all the young men gone? Using gender ratios to measure fetal death rates. Scholte,R.S.,VanDenBerg,G.J.,&Lindeboom,M.(2015).Long- run effects of gestation during the Dutch Hunger Winter famine on labor market and hospitalization outcomes. Journal of Health Economics, 39,17e30. Stothers,R. B. (1984). The great Tamboraeruption in 1815 and its aftermath. Science, 224(4654), 1191e1198. Strand, B. H., & Kunst, A. (2006). Childhood socioeconomic po- sition and cause-specific mortality in early adulthood. Amer- ican Journal of Epidemiology, 165(1), 85e93. Tavassoli, N., Noghanibehambari, H., Noghani, F., & Toranji, M. (2020). Upswing in industrial activity and infant mortality duringlate19thcenturyUS.JournalofEnvironments,6(1),1e13. Available at: https://doi.org/10.20448/journal.505.2020.61.1.13. Torche,F.(2011).Theeffect of maternal stressonbirth outcomes: Exploiting a natural experiment. Demography, 48(4), 1473e1491. Available at: https://doi.org/10.1007/s13524-011- 0054-z. Torche,F., & Kleinhaus, K. (2011).Prenatal stress,gestational age andsecondarysexratio:Thesex-specificeffectsofexposureto a natural disaster in early pregnancy. Human Reproduction, 27(2), 558e567. Van Ewijk, R. (2011). Long-term health effects on the next gen- erationofRamadanfastingduringpregnancy.JournalofHealth Economics, 30(6), 1246e1260. Available at: https://doi.org/10. 1016/j.jhealeco.2011.07.014. van den Berg,G.J., Doblhammer-Reiter, G.,Christensen,K., den Berg, G. J., Doblhammer-Reiter, G., & Christensen, K. (2011). Being born under adverse economic conditions leads to a higher cardiovascular mortality rate later in life: Evidence based on individuals born at different stages of the business cycle. Demography, 48(2), 507e530. Available at: https://doi. org/10.1007/s13524-011-0021-8. Yeung, G. Y. C., den Berg, G. J., Lindeboom, M., & Portrait, F. R. M. (2014). The impact of early-life economic conditions on cause-specific mortality during adulthood. Journal of Population Economics, 27(3), 895e919. Yi, J., Heckman, J. J., Zhang, J., & Conti, G. (2015). Early health shocks, intra-household resource allocation and child out- comes. The Economic Journal, 125(588), F347eF371. 206 ECONOMIC AND BUSINESS REVIEW 2021;23:194e206