The Shift and the Denial: Scholarly Attitudes toward Family Change, 1977-2002
Norval Glenn, Alex Roberts and Thomas Sylvester
Institute for American Values, 2008 - 8 pages
Many Americans have been questioning the importance of marriage, and the rift among scholars has allowed both progressives and traditionalists to claim that the experts are on their side. Over the past 20 years, have leading scholars reached a consensus on marriage?
For the past four decades, family scholars have been engaged in an often intense debate about the changing place of marriage in American family life. One side — let's call it "concerned" or pro-marriage — has argued that the decline in marriage has been a troubling trend with demonstrably negative consequences for families, and children in particular. The other side — call it "sanguine" or pro-family diversity — has argued that families haven't necessarily been weakened by divorce and unwed childbearing and that the negative impact on children has been exaggerated. Divorce and unwed childbearing are the family structures through which most children experience fatherlessness and father absence.
This academic dispute has had serious implications for society. Many Americans have been questioning the importance of marriage, and the rift among scholars has allowed both progressives and traditionalists to claim that the experts are on their side. Over the past 20 years, have leading scholars reached a consensus on marriage?
To answer our question, we examined all articles published in the Journal of Marriage and Family (JMF) from 1977 through 2002 to identify those that deal with family structure effects on children. We selected the JMF because it was, and remains, the leading journal in family social science. Any trends in the field, we reasoned, would be reflected in its pages. We found 266 relevant articles — 12.2 percent of the total articles published.
To track scholarly conclusions over time, we assigned each article a "family structure effects rating" from 1 to 5. A "1" means that the author appears to believe that family structure, divorce, and unwed childbearing are generally unimportant to child well-being. A "5" means the author reports high concern about increases in divorce and unwed childbearing. A rating of "3" represents an intermediate stance. Our ratings are based on the authors' conclusions, not on the conclusions we would have drawn from the findings reported. The coding, done by one person (Thomas Sylvester), is necessarily subjective, though we took measures to minimize systematic bias (see www.familyscholarslibrary.org).
Part 1: The Shift
Overall, we found strong evidence that scholars have become more concerned about the effects of family change on children.
As shown in figure 1, scholars were initially somewhat sanguine about changes in family structure, with articles published between 1977 and 1982 carrying an average rating of 2.81. In the mid-1980s, the average article rating rose to 3.4 and remained there. This number is within the "concerned" range. The most common rating for the entire study period was "4," which we gave to 150 or 56.4 percent of the articles. Thus, although there hasn't been a radical shift of opinion, scholars have clearly become more attuned to the possible negative effects of divorce and unwed childbearing on children.
When we looked only at those articles focused narrowly on surveys and hard data (that is, the quantitative studies), we found the same trend with a clear shift toward the concerned view. This fact suggests that there was some empirical basis for the collective shift in opinion.
The strongest evidence that scholars have become more concerned about the effects of family change on children comes from our analysis of the 32 "synthetic" studies in our sample. (Synthetic studies summarize and integrate the results of other studies.) As shown in figure 2, there was a clear and consistent shift away from the sanguine view and toward concern during the period studied. This trend is particularly important because the synthetic articles reflect family studies literature in general, and not just in the JMF.
Another important finding of our study is that access to more and better data correlates with more concern about family breakdown. For example, the sanguine view receded and remained relatively unpopular as the number of studies on children and family structure increased. When an article took a clearly sanguine view, it was much more likely to be a theoretical or opinion piece than a data-based study.
We did not try to assess the methodological quality of the quantitative studies, but it is generally agreed that studies with large, representative national samples are better than those with small and local convenience samples. Therefore, it is important that the average rating for the 119 studies using national samples is 3.66 compared with 3.28 for the 82 other studies. Furthermore, the 87 longitudinal studies have an average rating of 3.68, compared with 3.37 for the 114 cross-sectional ones. The 60 longitudinal studies with nationally representative samples have an average rating of 3.70 while the 55 cross-sectional studies with non-representative samples have an average of 1.11. These differences suggest, although they do not prove, that the better the research design, the more likely the researchers were to conclude that there were important family structure effects on children.
The shift in views is illustrated by generalizations made by authors in the JMF early and late in the period covered. In 1979, two authors concluded that there were "no consistent findings" on the effects of divorce on children. By 1991, things had changed. That year, Paul Amato and Bruce Keith published a widely cited meta-analysis of studies on the effects of parental divorce. After pointing out that a host of negative outcomes are associated with parental divorce, the authors stated, "The results lead to a pessimistic conclusion: the argument that parental divorce presents few problems for children's long-term development [which, the authors imply, had been the prevailing view] is simply inconsistent with the literature on the topic."
Some authors expressed differing views in articles published early and late in the study period. In 1986, Marilyn Coleman and Lawrence Ganong, experts on stepfamilies, wrote, "[In a] comprehensive review of empirical studies of stepchildren, we concluded that there were few differences between stepchildren and children from intact nuclear families." In 1990, however, these authors reported research suggesting a link between stepfamilies and low levels of child well-being. In 2000, they, with colleague Mark Fine, wrote, "Since at least the middle of the decade, it has been safe to conclude that stepchildren are at somewhat greater risk for educational difficulties and internalizing and for externalizing behavior problems than children living with two parents."
Finally, we think it is important to point out that only one of the 201 quantitative studies in our sample failed to find any evidence of family structure effects, although authors of another 28 articles (14 percent) reported only weak effects. Among quantitative family researchers, disagreements about family structure effects on children now seem to be almost entirely about their magnitude and importance rather than about whether or not such effects exist.
To say that a "new consensus" has emerged among family scholars about the effects of family structure on children is somewhat of an exaggeration. However, there now is widespread agreement that there have been negative effects from recent family changes that are strong enough and pervasive enough to be important.
Part 2: The Denial
As family scholars have become more aware of the importance of marriage, they have abandoned many arguments once used to support the sanguine view of divorce and fatherlessness. For example, in the earlier years of the divorce revolution, it was sometimes claimed that father absence was not problematic for children because children could find surrogates. According to one author, grandfathers, stepfathers, boyfriends, and other male relatives and neighbors "provide a wide variety of possible males who [can] serve as fathers to the child." This argument has not fared well over time. Although a few scholars continue to research "social fathering," no evidence exists that a substantial number of children in father-absent homes enjoy close, involved relationships with "social fathers." There is also no clear evidence that stepfathers or live-in boyfriends consistently improve the well-being of children. Thus, the "social father" argument has become rare.
Despite the overall shift in scholarly views, some family scholars maintain relatively sanguine views about the increase in single-parent families and its implications for child well-being. We found some recurring arguments used to support this perspective. Some arguments are valid, but others appear to be ideologically driven efforts to discount and minimize strong evidence that family structure has consequences for children. These kinds of arguments are, in our view, misleading and methodologically unsound. Let us look at four of the more common "sanguine" and misleading arguments:
1. The problem is parental conflict, not divorce. The typical cross-sectional study on divorce compares children in intact homes to those in divorced homes. In general, such studies find that children of divorce suffer from significantly higher rates of emotional and behavioral problems. A reasonable argument made by some sanguine scholars is that these problems may stem from pre-divorce parental conflict. Recent longitudinal studies have provided support for this argument, since they have shown that some apparent negative effects of divorce occurred before the divorce.
Scholars disagree, however, about the size of those effects. Some authors contend that preexisting factors account for "most of the negative effects" divorce presents to children. Others, such as D. R. Morrison and M. J. Coiro, say, "[F]actors associated with divorce itself, such as parental absence, changes in custody and relationships, and declines in parents' psychological well-being, explain increases in children's behavior problems over and above the effect of pre-disruption parental quarrels."
Much is yet to be learned about this issue, but existing research findings indicate that although pre-divorce influences do account for some of the effects once attributed to divorce, divorce and its aftermath typically have important additional negative effects on children. Furthermore, it is important to consider that the breakup of a marriage is a process with extension in time rather than something that occurs with the issuance of a divorce decree. Parental separation typically occurs well in advance of the decree, and the decision to end the marriage is often made by one or both parents well in advance of the separation. Therefore, using the date of the decree, or even of the separation, to discern between pre-divorce and post-divorce influences may not be very meaningful. Some "pre-divorce" influences may ensue from the early stages of the divorce process. For instance, parental conflict tends to increase when one parent decides to end the marriage.
2. Family structure does not matter "necessarily." Another common approach is to argue against a purported alarmism about the negative effects of divorce and unwed childbearing. "Certainly the consequences of divorce are not trivial," writes Judith Stacey, a vocal proponent of family diversity, "but divorce, in and of itself, does not harm the young nearly so much as [certain scholars] have claimed." David Demo says the negative consequences of divorce and single parenting "have been greatly exaggerated."
Of course, reasonable and legitimate arguments can be made about the magnitude of family structure effects, and no doubt some commentators have exaggerated that magnitude. However, no author of an article in the JMF during 1977-2002 argued that differences between children who grow up in different family structures are absolute and categorical. No one, to our knowledge, has ever claimed that all "children of divorce" become delinquent while no persons whose parents stay together ever do. There would seem to be little need, therefore, to rebut such extremist beliefs about family structure effects. But consider the following statements from JMF authors who took a relatively sanguine view of divorce and single parenting (all with emphasis added):
"In conclusion, the results of the current investigation support the belief that parental marital disruption does not necessarily culminate in adjustment problems in a college-student population."
"Marriage does not always enhance one's well-being . . ."
'these findings support the contention that family structure does not automatically adversely affect family members" well-being. . . ."
These statements are accurate and would serve to puncture extreme fatalism about the effects of family structure. The problem is that no serious family scholar holds such extreme views. The "not necessarily" arguments are essentially attacks on straw men. They do nothing to rebut the argument that an intact family tends to be the most supportive family structure for children. Instead, they sidestep this argument and try to undermine the pro-marriage view by painting it as absolutist.
Scholars rarely make "not necessarily" statements about other factors that influence child well-being. In our examination of the JMF articles, we found no examples of authors pointing out, for instance, that growing up in poverty does not inevitably result in poor child outcomes. Nor do scholars apparently feel obliged to emphasize that low levels of maternal education do not necessarily lead to educational difficulties for children.
3. Parent-child relationships and family income affect children far more than family structure. Abundant evidence shows strong and pervasive effects of family structure on a family's economic status and the nature and quality of parent-child relationships (called "family process" by researchers). Abundant evidence also shows that a family's economic status and parent-child relationships affect child outcomes. For instance, after divorce, the primary custodial parent tends to be stressed and can be caught up in establishing new romantic relationships, while the other parent typically has reduced contact with the child. One would expect, therefore, that divorce typically negatively affects parent-child relations and the quality of parenting, and considerable research has indicated that it often does.
Thus, to a large extent, family structure has its impact on child outcomes indirectly through its effects on family income and parent-child relationships. This means that if researchers statistically control for the effects of family income and parent-child relationships, it will appear as if family structure per se has relatively little influence on whatever outcome is being measured. What is important, however, is the estimated total effects — the combination of the estimated direct effects and the estimated indirect effects. It is rarely useful in this and in similar situations for researchers to compare the estimated direct effects of the more distal variable (in this case, family structure) with the estimated direct effects of the more proximal variable or variables (in this case, family income or parent-child relationships). Even when such a comparison is useful, it does not provide a valid measure of the relative causal power of the compared variables.
The logic detailed in the previous paragraph is generally understood among social science researchers, and thus it is odd that several of the authors of the JMF articles we reviewed inappropriately pit family structure against economic or process variables. For example, in an article titled "Does Family Structure Matter?" the authors champion "the perspective [that] stresses the importance of family processes in all family structures and posits that family structure per se is comparatively unimportant" — a view based apparently on a reduction in estimated family structure effects when "family process" variables were controlled. Similarly, two other authors write, "children's behavioral problems associated with [the inadequate supervision by single mothers] may be attributable to low income and the need for maternal employment rather than being the result of single-parent family structure per se."
Among scholars who take a sanguine view of family change, these kinds of arguments are common:
"[R]ecent research indicates that family processes better explain adolescent wellbeing than family composition."
"More consequential than family type for children's well-being is the quality of parent-child and other family relationships."
"[B]eing in a single-parent versus a traditional two-parent family may not be as critical a factor for adolescent outcomes as are parenting practices per se, particularly support and monitoring."
These arguments are weak, because family structure has its effects on children largely through its effects on family income and parent-child relationships. These statements above are almost analogous to saying, "It is not the microwave oven per se that heats the food; it is the radiation it emits." Such distinctions are technically correct, but they create competition between possible explanatory variables where it does not belong. This is not the way that social science is normally done. "Per se" arguments are almost never made about other broad variables that have their effects through specific mechanisms. For instance, no one argues, "It is not poverty per se that negatively affects children; rather it is poor schools, bad neighborhood influences, and other factors that correlate with poverty."
In view of the fact that these inappropriate and misleading comparisons are made by well-trained and generally competent researchers, we infer that as a whole they are motivated by a well-intentioned but misguided desire to downplay and divert attention away from the evidence for family structure effects.
4. Even raising questions about possible family structure effects can have undesirable consequences and thus should be discouraged. A few authors of the JMF articles we reviewed are overtly ideological, arguing not only that family structure effects are too small to be important but also that family structure effects research should cease. For instance, two authors write, "Scholars must stop asking about the price of deviance that faces individuals, families, and societies when couples divorce, when parents remarry, and when women, especially mothers, live without men." The tone of this statement indicates that the authors consider questions about family structure effects to be ideologically motivated and thus properly opposed on ideological grounds. Opponents of family structure effects research also are concerned that discussion of negative effects of single-parent families will stigmatize and cause feelings of guilt among persons in those families.
Concerns about stigma and guilt should not be lightly dismissed, but they are not good reasons for making some family research topics off limits. Doing so would be analogous to stopping research on the health consequences of obesity because the findings might make obese people feel uncomfortable. It would be blatantly anti-scientific and in the long run could even hurt its intended beneficiaries. Fortunately, family scholars in general have rejected the call to abandon family structure effects research, which has increased substantially over the past few decades (see figure 1).
Reviewing articles in the Journal of Marriage and Family, we find that an apparent majority of scholars have come to believe that family structure matters, and matters to an important extent, for children. This widespread agreement has emerged, in large part, because scholars have amassed a wealth of data on the subject, and the data support such concern.
On the other hand, some scholars have continued to insist that family structure is not very important. In view of the fact that there is always some ambiguity in social scientific evidence, there is room for honest and responsible differences of opinion on this issue. However, some of the arguments, rhetorical devices, and modes of data interpretation used by "sanguine" family scholars are so unconventional and contrary to accepted "best practices" that ideological bias is the only reasonable explanation for them.
This brief summarizes "Trends in Scholarly Writing on Family Structure Since 1977 in the Journal of Marriage and Family" by Norval Glenn and Thomas Sylvester (see http://www.familyscholarslibrary.org.) Norval Glenn, Ph.D., is the Ashbel Smith Professor and Stiles Professor in American Studies, University of Texas-Austin. Thomas Sylvester, J.D., and Alex Roberts are Affiliate Scholars, Institute for American Values.
This research brief comes from The Center for Marriage and Families, based at the Institute for American Values. It was published in February 2007.
This research brief was commissioned by the National Fatherhood Initiative and supported by Grant No. 2006-DD-BX-K003 awarded by the Bureau of Justice Assistance. The Bureau of Justice Assistance is a component of the Office of Justice Programs, which also includes the Bureau of Justice Statistics, the National Institute of Justice, the Office of Juvenile Justice and Delinquency Prevention, and the Office for Victims of Crime. Points of view or opinions in this document are those of the authors and do not represent the official position or policies of the United States Department of Justice.
About the National Fatherhood Initiative
National Fatherhood Initiative (NFI) was founded in 1994 to confront the most consequential social problem of our time: the widespread absence of fathers from children's lives. NFI's mission is to improve the well-being of children by increasing the proportion of children growing up with involved, responsible, and committed fathers in their lives. NFI accomplishes this mission through educating and inspiring all people, especially fathers; equipping and developing leaders; and engaging every sector of society.
About the Center for Marriage and Families at the Institute for American Values
Directed by Elizabeth Marquardt, the Center for Marriage and Families at the Institute for American Values issues research briefs, fact sheets, and other material related to marriage, families, and children. Its Scholarly Advisory Board includes William Doherty, University of Minnesota; Norval Glenn, University of Texas; Linda Waite, University of Chicago; W. Bradford Wilcox, University of Virginia; and James Q. Wilson, UCLA (Emeritus).
The Institute for American Values is a nonprofit, nonpartisan organization dedicated to strengthening families and civil society in the U.S. and the world. The Institute brings together approximately 100 leading scholars — from across the human sciences and across the political spectrum — for interdisciplinary deliberation, collaborative research, and joint public statements on the challenges facing families and civil society. In all of its work, the Institute seeks to bring fresh analyses and new research to the attention of policymakers in government, opinion makers in the media, and decision makers in the private sector.