An interview with Christopher Wildeman

Christopher Wildeman, Ph.D., is an Associate Professor of Sociology at Yale University. He is a leading researcher on the effects of parental incarceration and infant mortality. This is an unedited version of an interview with Dr. Wildeman conducted over email:

1. The United States has the highest rate of infant mortality in developed countries. Can you speculate as to why this discrepancy might exist?

Although there is a huge, contentious literature on this issue and there is no consensus on the myriad causes of the high American infant mortality rate, my best sense is that the social conditions of the impoverished play a crucial role in explaining the high American infant mortality rate. Other causes are also important, to be sure, but my sense is that the primary drivers of the high American infant mortality rate are the very high infant mortality rates of marginalized populations – especially African Americans, individuals with low levels of educational attainment, and individuals living in areas of concentrated disadvantage. This likely suggests something about the greater levels of inequality in America – especially in terms of access to adequate health care, living conditions, and the whatnot – but it isn’t clear to me based on the research I am familiar with exactly what the primary driver of this is.

2. Does your research indicate that if the U.S. reduced the amount to which it incarcerated individuals that it would in turn reduce the infant mortality rate?

Assessing whether incarceration is a cause of anything at the individual level or the population level is incredibly difficult, so it is hard to be certain that changes in the American incarceration rate over the last 40 years or so are responsible for any of the other changes that have happened in that time span. Nonetheless, my analyses, which do as much as the data allow to rule out alternative explanations do suggest that decreasing incarceration might be one way to improve the health and wellbeing of American children – possibly even decreasing the infant mortality rate. My analyses suggest that had the American imprisonment rate not increased from the 1990 level, the total American infant mortality rate might have been nearly 4 percent lower in 2003, and the absolute Black-White disparity in the infant mortality rate might have been about 7 percent lower. Of course, this doesn’t necessarily mean that decreasing the imprisonment rate to the 1990 level now would have the same result, as some of the damage may have already been done. But it does suggest that some of the distinctively high American infant mortality rate may be attributable to our distinctively high imprisonment rate. This conclusion is only tentative, of course, and we still need a great deal more research before we know for sure that incarceration does anything to damage population health or the health of infants.

3. In your research on mass incarceration and infant mortality you found reasons why women were at an increased risk for infant mortality while their partners were in prison. Can you explain those reasons in detail?

Unfortunately, there was no good way to test these mechanisms using the data that I utilized. But my hypothesis is that the increased risk of infant mortality is due to some combination of the stress of having a partner incarcerated, the diminished financial resources available to the household as a result of his incarceration, and the changes in childcare resulting from his absence that led to this increased. Future research really must do more to figure out whether this relationship is actually causal, however, and isolate which of these mechanisms is most important. The mechanisms part of the equation is especially vital since understanding the mechanisms through which paternal incarceration increases the risk of infant mortality means that we can reduce the effects of paternal incarceration on the risk of infant death even without reducing the incarceration rate. Since the American imprisonment rate is unlikely to decrease dramatically in the near future, this is an important point. I should also note in closing that at least in the dataset I utilized, the increased risk of infant mortality attributable to paternal incarceration was only found in families where the father was not abusive to the mother. So in the 60 percent of the families where the mother did not report abuse, paternal incarceration dramatically increased the risk of infant mortality; but in the 40 percent of families where the mother did report abuse, paternal incarceration had no discernible effect on the risk of infant mortality. (To clarify, this 60-40 split is not representative of the general population, where rates of domestic violence are much lower. Sixty percent of the fathers who were incarcerated were reported to have never been abusive to the mother, while 40 percent of the fathers who were incarcerated were reported to have ever been abusive to the mother.)

4. Could you speculate as to whether this relationship between public health indicators and mass incarceration is distinct of the United States, or if it is an international issue?

This is an incredibly hard question to answer, in large part because the crimes American prisoners have committed differ so dramatically from the crimes prisoners elsewhere in the developed world have committed. My hypothesis, however, would be that the consequences of incarceration for infant mortality (and child health more broadly) might be smaller elsewhere in the developed world than in the United States because of (1) the less serious offences committed on average by American prisoners than other prisoners, suggesting that might contribute more to family life, on average, than prisoners elsewhere in the developed world and (2) the stronger social safety net present in many other developed countries. Of course, these are just hypotheses, but I have conducted some preliminary analyses that seem to suggest that the consequences of imprisonment for population health may be most severe in the United States. (I should say that I am not quite sure whether I want you to print this last part, however, since the results from that paper have not yet been released into the public domain in any way.)