Babies of the War: The Effect of War Exposure Early in Life on Mortality Throughout Life

Abstract

There is increasing evidence that circumstances very early in our lives, and particularly during pregnancy, can affect our health for the remainder of life. Studies that have looked at this relationship have often used extreme situations, such as famines that occurred during wartime. Here we investigate whether less extreme situations during World War II also affected later-life mortality for cohorts born in Belgium, France, The Netherlands, and Norway. We argue that these occupied countries experienced a considerable deterioration in daily life situations and show that this resulted in strongly increased mortality rates and lower probabilities of survival until age 55 among civilian populations who had been prenatally exposed to wartime circumstances. However, this mortality effect among the prenatally exposed is entirely concentrated in the first years of life, particularly infanthood. Once we condition on having survived the first years of life, those who had been prenatally exposed do not have higher mortality rates. This suggests that “culling” is important and that effects found in earlier studies may have been biased downward substantially.

Acknowledgments

We thank participants at the conference on “Long-Run Impacts of Early Life Events IV: Evidence from Developed and Developing Economics” held at the University of Michigan and the editor, Eileen Crimmins, for helpful comments.

Notes

¹ One of the very few other World War II studies that has not focused on famines is Van Os and Selten (1998). They focused on increased stress levels during the May 1940 German invasion in The Netherlands and showed that prenatal exposure increases the chances that the offspring will develop schizophrenia. There is, to our knowledge, no literature that studies similar types of effects by focusing on other wars.

² Denmark is excluded, since circumstances were less adverse during much of the war. Denmark remained a functioning democracy, and conditions (including rations [Egle 1943]) remained relatively favorable until mid-1943, when the government was dissolved and martial law instituted by the occupiers.

³ Note that the drops in female life expectancy are not mainly due to the Holocaust. This is clearest when comparing France and The Netherlands: calculated as a share of the total population, around six times more Jews were murdered in The Netherlands than in France, yet France experienced a sharper drop in life expectancy than The Netherlands.

Figure 1. Female life expectancy at birth by year.

⁴ Only for Norwegian males do these effects not reach significance, which fits with the smaller deterioration in civilian life circumstances that is evidenced in Figure 1.

Figure 2. Probability of surviving to age 55: females.

⁵ Note that the dummies for 1940 and 1945 also cover people who were born before and after the war.