Dondena Interviews

Another renegade sociologist turning to biology for clues about society, Ryan Masters of the University of Colorado Boulder, spoke to the Dondena Centre in November about recent work on the effect of economic recessions on mortality. Looking into the wombs of cohorts born some 85 years ago, he can make some fascinating claims about the in utero effects of societal events.


Tell us about your talk today.

My first motivation with this is try to find an actual cohort mechanism that we can directly test in U.S. data. And the reason why I think this is important is that most of the research that tries to take a cohort-based perspective of trends in mortality don’t actually have mechanisms. We allude to all these things and sometimes bring in aggregate-level data to try to encapsulate what we think the changes are, but this is one of the first times that we are actually able to pinpoint an in utero process that does have a pretty substantive effect on adult mortality risk. And then show that it actually wanes across cohorts.


Ok. I want to make sure that I’m understanding this correctly. When you say in utero exposure, you mean a woman who is pregnant during an economic recession, how does this affect the mortality of the fetus?

Exactly. Yes.


You can see my skepticism.

Um, yes.


Because when I read this I thought this can’t possibly be what he’s studying.

That is what I’m trying to study. I’m trying to pinpoint not just in utero processes, but actually trimester-specific. So where in gestation might one’s susceptibility to later-life mortality risks actually fall? There’s been two lines of investigation. Often times people are interested in first-trimester exposures, because that is often times the gestational period where developmental processes are sort of set. So there’s these trajectories of physiological growth. And others are interested in the third trimester because that’s the greatest susceptibility for those set trajectories to be interrupted by either nutritional deficiency or massive exposure to stress or something else.


And so what did you find?

What I find is no evidence of the first trimester in these data, but I find very strong, significant effects in the third trimester to economic recessions.


This is not such good news for people who are expanding their families right now.

Well, it’s hard to say. One of the key take-aways, as I mentioned before, is it’s a cohort-specific effect and so the data I’m actually looking at are men and women in the United States who are aged 50 to 85 between 1986 and 2006. So these are men and women who belong to cohorts who were born in 1900 to 1956. And what I find is that these economic recessions that transpired back across the 20th century had significant effects. But I can’t speak obviously to the long-term detrimental effects of being exposed to recessions today.

But what I do find, as I mentioned before, is that this waning cohort process means that these in utero effects were there for cohorts born at the early part of the 20th century, but they’re essentially nonexistent by the 1950s. And so I’m trying to explain why we would presume to expect significant effects of in utero exposures to recessions for cohorts born at the early part of the century but not extant effects for those born at the mid part of the century.


So how does this fit in with the research that you do broadly speaking?

Broadly speaking I’m interested in what are the sort of the underlying mechanisms that have driven these massive reductions in mortality across the United States over the last 50 years. On the one hand, a lot of people point to period-based mechanisms. What are the things we’re doing in the current present day to intervene, to mitigate the deleterious effects of cardiovascular disease? So, new surgical technologies, pharmacological technologies, other medical advances in knowledge and access. But my perspective tries to make room for these cohort-based processes as well: So what happened? What transpired across the 20th century that has changed the conditions of early life? What has changed? Intergenerational transfers of health as well—so the ability for one generation to, sort of, bestow health capital on the subsequent generation. And what are the driving mechanisms behind those things? Reductions in infectious disease, implementation of work projects, nutritional transformations. These things that have long latency, lasting effects on the U.S. population health.


Tell me about the data that you worked with to do this study? Was it great? Was it a nightmare?

It’s not without limitations.

The United States is unlike a lot of western European countries that enjoy these long-standing, full-cohort histories of mortality records: Sweden, France, even the UK, for example, you can actually chart the lived experiences of cohorts that were born back as early as the 1890s and beyond. In the United States, we don’t have those records. So unfortunately when we try to build in these cohort-based processes of longevity, we have to rely on survey-based data.

I used the National Health Interview survey 1986-2004. This is an annual survey of the non-institutionalized U.S. population, and it asks a number of great questions about health-related lifestyle: behaviors, diagnoses, functional limitations, etc. It also has great records on socioeconomic data as well. And what the National Center for Health Statistics did, which was a great public service, was to actually link respondents of the National Health Interview Survey with the National Death Index at the Centers for Disease Control. So we can actually use these National Health Interview Survey Linked Mortality Files, as they’re called, to follow upwards of 21 years of subsequent survival for these respondents. We’ve got waves now 1986-2004, linked to mortality records up to December 31, 2006, and I’m able to then look at these cohort members that were born from 1900 to 1956 and chart up to 21 years of survival during their adulthood.

In order to then link it to whether there was an economic recession taking place while they were in utero, I then use the National Bureau of Economic Research’s public data, in which they have quarter-year business cycle and also the official records of recessions. Based off the quarter-year of birth that is provided in the National Health Interview Survey, I’m able to link that to the National Bureau of Research quarter-year economic activity to see what was transpiring during their trimester-specific in utero environment.


Tell me about in your field right now, what are you seeing that you like quite a lot, and what would you like to see more of when it comes to research or process or money, funding, things like this?

Well, what I like is that there’s a resurgence in almost formal demography, to put it crudely. That there is, I think, a generation of demographers being trained who are really attentive to quality data, because, as I mentioned before, the United States isn’t as blessed with rich, available data as some other countries are, and so some people have gotten quite creative in trying to find ways to implicate either cohort-based or period-based mortality processes. And I think that takes some ingenuity, it takes training, it takes some grit, I guess, to stay at it. It’s no longer the case that you can just rely on secondary data sources to do our work, and I think people have highlighted some unique and clever ways of doing that.

What would I like to see more of? I would like to see, especially sociologists, because this is my home field, possibly moving past what Jeremy Freese and his colleagues called our bio-phobia—our reluctance to actually study physiological processes, or even physiological outcomes. We sort of have a fear of biomarkers, a fear of genetics, as providing alternative explanations that possibly limit our contributions to understanding social structure and human population health. I think there’s been some inroads, and people are making some fantastic strides trying to incorporate physiological processes and actually highlight how social structure comes into play. But there seems to be quite a bit of tension there.


You guys do not have an easy job. It seems like a lot of people are very much against this. But I have to say as a lay person—I’m a journalist, not an academic at all—that seems to me to be the most important thing that sociology could possibly do, is link these physiological processes and social structures and things. I thought, in fact, that was the whole point of the field.

Right, we certainly shouldn’t self-censor and we shouldn’t cede certain areas of knowledge and understanding to other disciplines. I think there is a massive push for multi-disciplinary, trans-disciplinary work, I certainly think there’s more room for that. And we as sociologists could be doing a better part to try to engage that.


Do you feel like young scholars are more interested in this?



You guys are training a good number of people, who can then go forth and multiply?

I think so! I was a member of Robert Wood Johnson Foundation’s Health and Society Scholars Program, a post-doc role trainingship that Mark Hayward was on the national advisory committee for. And their whole mission was to try to train, or possibly even retrain, social scientists and health researchers to rub shoulders, collaborate more, delve into multidisciplinary avenues of population health research. It was partly their mission to establish a new disciplinary field of population health that draws from all sorts of disciplines and understandings. And so those types of programs have really reshaped the kind of training new scholars are getting. Even if it’s indirectly, they know that it’s out there, they know they should keep their ears up and listen to justifications for more multidisciplinary avenues of research. I certainly think there’s a changing understanding of what it is to be just a social science researcher, as opposed to a researcher in general who draws from all sorts of fields.


So you will leave Bocconi and then what’s the next great academic thing you have on your calendar?

I will be going to the University of Nottingham to give a talk at their Methods and Data Institute on age-period-cohort methodologies.


It has to be professionally satisfying to be in on sort of the ground level of something like this, and helping build what this area of study is going to become.

I wouldn’t toot my horn to that extent. This debate has been persisting since the 1970s, and there’s been some recent advances across the 2000s with new techniques and such. I would applaud those folks’ efforts, and also the critics of these folks. I’m just interested in the stuff, and I see some areas in which I think they can be applied nicely and others where we have no business trying to apply them. I’m coming in late in the game really, just to add my two cents.


Learn more about Ryan Masters here. Learn more about the Dondena Seminar Series here.
















Last updated 07 May 2015 - 12:09:18