Dondena Interviews

Half of women who have four children have them with more than one partner. Perhaps that’s to be expected, because as Elizabeth Thomson, who is  Professor of Demography at Stockholm University and spoke to Dondena in February, says, ‘you don’t get ... to three or to four or to five very often with the same person.’ But surprisingly, in the United States at least, one quarter of women who have two or more children have had them with different partners. One quarter. So not everyone is doing it, but one-quarter is one-quarter.

Certainly its just plain interesting to see the make up of different families, but it’s also useful to know this given the aging population—will these children cooperate to help their elderly step-parents? Or must communities be aware that these complex family dynamics will contribute to a social problem? Maybe soon we’ll replace that age-old question of “How many kids do you have?” with “How many partners do you have?” Or, maybe we’ll just know.


Give us a short summary of the talk you’re giving this afternoon.

Increasing proportions of parents and children are living in families where at least some children don’t have the same parents. I’m calling this “childbearing across partnerships.” It’s known by others as multi-partner fertility. I don’t like that term because to me it connotes people running around having sex and having kids with all these numbers of people and very few have children with three or more partners. And also, fertility is supposed to be a term for the aggregate—I don’t have fertility, you don’t have fertility, we have children or we are child-bearing.

            When this first came on the scene, I was asked to give a keynote in the fall of 2006 to a conference at (University of Wisconsin) Madison at the Institute for Research and Policy, where Frank Furstenberg and Marcia Carlson, who is now my colleague at Wisconsin, were presenting some of this first information from a study called Fragile Families, which is mostly urban, young people and mostly in disadvantaged areas, so it’s not a representative sample of the United States.

            They were documenting this phenomenon and I looked at some of the papers, and I said, ‘Well this isn’t anything new, people have been having children with more than one partner for centuries.’ In the far past they used to do it after death, but more recently it’s after divorce, or in the case of cohabitation, separation. ‘So why aren’t we talking about step-family child bearing?’ I asked.  I’d actually done several studies of step-family childbearing, showing that the formation of a new partnership stimulates additional births – because they want to have a child together --that the woman or the man might not have otherwise have had. So (Furstenberg and Carlson) said, ‘Well, people sometimes have their first child outside of a union.’ I account for that, I look at them when they form unions, but they said, ‘Well, sometimes they have children with other people but not when they’re living with them or married to them, and there are different fathers,’ and I said, ‘Yeah but that’s not very common.’

            Nevertheless, if you look at sort of the family behaviors that we’ve been studying under the rubric of the second demographic transition, where the first transition is just the fall in fertility from high levels to about two kids, you see having children without a resident partner (non-union childbearing),  separation and divorce, re-partnering, and having children in step-families. And all the research that has looked at these four pieces of the puzzle is about one of them. So, childbearing across partnerships is kind of the cumulation.

            Most of the work has been descriptive—showing how common this is in the population. We decided to take a more demographic, or what I call a fertility-centered, look at it, so we start observing people after they have their first child, and we don’t worry about whether they’re in a partnership or not, we just look to see if they have a second child with the same partner or a second child with a new partner, or no second child at all. Then we take women who had two children with one partner and we look, do they have a third birth and if so is it with the same father? So we’re trying to look at the differentials in what I would call ‘fertility risk,’ or the chances of having children.

            There had been work on this in the United States, Trude Lappegård had done work on Norwegian men, and Ann Evans and Edith Gray had done some work in Australia. So we put together the U.S., the Australian, and the Norwegian data, and then I added the Swedish data to get more coverage.

            And basically what you can see in the descriptive part, if you look at women with four children, half of them have a child with more than one partner in these countries. So you don’t get, in a fertility regime where most people are having one or two children, to three or to four or to five very often with the same person. You get there with new people. But even if you look at just the second birth, in the U.S., of women who had two or more children, or even who have just two children, more than a quarter of them have had those children with different partners. So it’s not that everyone is doing it, but it’s a substantial phenomenon. In Sweden I think it’s only 10 percent of those births and in Norway a little more and Australia little more.

            Several collaborators and I wrote a paper that came out this year using French data, trying to see if these extra births in step-families were keeping fertility up around two children for women, and if we didn’t have them would fertility be even lower, which is true. But it doesn’t completely compensate for staying together and having kids.

            Now, all these studies we’ve done are in countries with high separation rates and high fertility, what we call “highest-low fertility,” which is somewhere around or just under two children. So it’s close to what you would need to replace the population. It’s the kind of fertility that doesn’t contribute to the aging population. And also, in countries with “lowest-low” fertility, under 1.5 children per woman, you might have delayed marriage but you don’t have separation or divorce.

            But in Italy, the latest data show that about 20 percent of births are outside of marriage, most of them are in cohabitation. We know that these relationships tend to be less stable, although initially there is a preponderance who marry after the birth, and Italy still has shotgun marriages, where pregnancy may occur outside marriage but marriage occurs before the birth. That is a little more common than in other settings.

            One of the things I end with in this talk is if we looked at all the second births in Italy, would it be a high proportion of them that occur because of a second partner? I don’t know the answer to that, but my colleagues here have the data, they could answer the question.


What are some of the interesting changes you’ve seen in family demography during your career?

The biggest change is the amount of data we have now. Part of the reason we have so many family demographers is that we have a lot of family sociologists who started doing family demography. And then we have demographers who move into family sociology, so we’re drawing from two well-established disciplines to deal with these kinds of issues.

            I think the other thing is that these kinds of family problems, like divorce, non-marital child bearing, and complexity of families, I think it has been very visible because of the economic problems of such families in the United States. There’s not a lot of research in Europe, except in the UK, but there is evidence now that even in the highly supportive social welfare states that when parents separate the income level goes down, the children have some problems in school, that sort of thing. So I think it’s a social problem that draws people in.


So it’s not because we’re nosy? And we want to make sure we’re normal?

No! I think it’s because it has social policy and social welfare consequences in a way that other areas of sociology don’t.

            But the demography of aging is also very hot. In the demographic community that’s become a major concern. People are looking at, how long can people really live? Do they live healthy or do they end up being very, very dependent?

            The reason I think this family complexity is really important is that we already know that when separations occur, relationships between fathers and children tend to suffer more than relationships between mothers and children, into the father’s old age. And then, we don’t know, maybe these multiple kids from multiple partners will give people a wider safety net. But we don’t know, because there’s conflicts associated with separations, complexities associated with half-sibling relationships, and we don’t know whether in old age these parents will have problems because the children they produced don’t get along.


What’s not going on today in family demography that ought to be?

I said that the reason I thought it had gotten so big, and what had really changed, was data. So we have tremendous amounts of data that we sometimes can hardly analyze. But lots of times you get into a question and you don’t have the data that you need. One thing that I really wish we could do collectively is to build into the national statistical system the collection and preservation of life histories.

            Demographers can look at the links between our educational progress, our careers, our work, our unemployment, and our family biography. And those data you can’t get in vital statistics because you just look at these events and they’re often not linked longitudinally. In the Nordic countries they’re linked longitudinally because they’re tied to your person number and researchers with very careful constraints can go in and link people across time. But they can’t observe cohabitation. They can estimate it in some of the registers that are based on apartment-level identification numbers, where you live, but even that’s not exact. So I would like to see demographers collectively focus a lot more on trying to build a sustainable statistical system that doesn’t focus on snapshots, but focuses on the accumulation of lives. That’s a huge undertaking.

            In the 1990s we had the Fertility and Family Surveys. We have the Demographic and Health Surveys that cover Eastern Europe and most of the poor countries of the world, but they’re snapshots. In the snapshot you can go back and ask about histories, but it’s very time-consuming. But it would be very nice if you could even just get samples and follow them. Or you even if routinely, let’s say every five years, you collected information about people’s histories. You could follow life courses over time, over places, and learn a lot more about the economic crisis or a shift in policy. You could really see not just the broad, rough outcomes, but you could see, if you are in a certain place in your life span, under a certain set of economic conditions, partner conditions, with children or whatever, how these events and these circumstances influence the next steps that you take. Because change doesn’t affect everybody the same way. If you’ve already started your relationship you’re not going to be prevented from carrying it forward, but you might be prevented from having a child, or helped to have a child.

            In this area I’m working in, I would like it if demographers could come together more, and not try to reinvent the wheel, but build on what we already know about step-families and what the implications are for children. There’s a lot of research on that, but there’s very little on half-siblings. Having children with new partners is increasing because there’s been an increase in separation, an increasing pool at risk. But it still was significant, even in the 1950s, and I’m sure that if we went back to the ‘30s and the ‘20s we would find that it was happening as well. But we sort of treat it as an anomaly. The other really worrisome thing is that the difference by economic situation seems to be growing, at least in these countries with high levels of stability. My colleague Sara McLanahan talks about “diverging destinies.” What she’s talking about is that people at the lower end of the economic ladder have not only that to contend with, but they have these family disruptions and complexities to contend with, so they have a kind of double life burden, and their children especially can have difficulties from there.


You leave Milano today and what’s the next exciting academic thing on your calendar?

I will go back home tonight to my Stockholm apartment and there are two things that are going on. We’re in the process of collecting the Swedish Generations and Gender Survey data, that will be the last of this set of surveys, and the data will come out in the fall. So that’s something I’m working on that I hope will have a successful conclusion. But the next big academic event is a kick off meeting in March for a big European Commission grant that my colleague Livia Oláh got, under the FP7. It’s called Families and Societies and she’s got 28 partners in 15 countries. I’m going to be on the board so I kind of get to sit back and watch it happen. We have a great new post-doc working with us, and we’re looking at issues of how people fall through the policy cracks if they’re not in the family organization that the policy is thinking about when it’s developed.


For somebody who is just a few years from retirement you have an awfully full plate.



Have you slowed down, or not yet?

Not yet. Part of the reason I took the job at Stockholm University is that I’d finished a term as director of the demography center at Madison, and I love Stockholm, and my alternative was becoming department chair or becoming associate dean or going full time into the classroom. It’s a very high level; the Madison campus gets the best students in the state. But I’m not a master teacher. So I said, ‘Let’s see, my grants have kind of curled down, I’m not going to get money right away because I haven’t been publishing a lot while I’ve been director, so what am I going to do next to make the last 10 or 15 years of my career interesting?’ And this kind of plopped in my lap. And it has been interesting.


Learn more about Elizabeth Thomson's research here. Learn more about Dondena here.

















Last updated 10 December 2016 - 05:39:13