Institutions versus Foster Homes for Parentless Children: Rich and Healthy, or Poor and Sick?

When I was a teenager, we used to think it very funny to ask people their opinions: “Is it better to be rich and healthy, or poor and sick?” This “joke” comes back to me nowadays when I think about discussion of the advantages and disadvantages for abandoned children’s development of being cared for in an institutional setting or in high quality foster homes.

Most discussion of this issue is based on the findings of the Bucharest Early Intervention Project (BEIP), a randomized controlled trial of the two care settings carried out almost ten years ago by Charles Zeanah and colleagues. Further publications and recommendations based on the BEIP finding that foster care had beneficial effects are appearing every day, and proposed legislation in the U.S., the Children in Families First Act (CHIFF) is based primarily on the BEIP reports.

A few months ago, the publication Zero to Three published an article, “A Matter of Timing” (Fox, N.A., Zeanah, C.H., & Nelson, C.A., Vol. 34 (3), pp. 4-9). The article repeated many of the conclusions of earlier reports of the BEIP, as well as focusing on alterations in brain development possibly associated with institutional care in the early years.

I responded to the article with a letter which was kindly published by the ZTT editor, Stefanie Powers, along with a response from Fox, Zeanah, and Nelson. In my letter, I expressed concern with possible biases from lack of blinding of evaluators--  an almost insurmountable problem of this type of research, but one that requires cautious interpretation of findings. I also queried the comparison between the “poor and sick” institution and the “rich and healthy” foster homes, and particularly the apparent blanket conclusion that all foster homes were superior to all institutions, a conclusion that appears to evade the fact that this randomized controlled trial did not in fact isolate or hold constant some relevant variables. In closing, I expressed concern about the emphasis on the BEIP in the proposed CHIFF legislation, and the fact that the practical outcomes of inappropriate interpretation of these research outcomes could be very serious indeed.

Fox, Nelson, and Zeanah  kindly replied to my comments in the  March 2014 issue of Zero to Three (pp. 4-5).They summarized some of the statements of the earlier article, and very reasonably emphasized the findings of the brain measurements as unbiased by expectations of researchers. They went on to note that every nonrandomized earlier study of institutional versus home care had the same findings—that home care was superior—which is hardly surprising because in those nonrandomized  studies selection biases would certainly have been at work. (And surely the BEIP, with all its difficulties and expense, would never have taken place if the nonrandomized studies had had different outcomes.)

In responding to my comment that comparisons of high-quality institutions to foster homes had not been done, Fox, Nelson, and Zeanah referred particularly to a 1975 publication by Tizard and Rees, comparing children in  high-quality institutions to a contrast group of which the children had been full-term, healthy babies, both parents were white, the father was employed and working-class, the mother was not employed outside the home, and there was no slightly older child in the home. The institutions studied by Tizard and Rees were certainly of very high quality compared to the Romanian orphanage of the BEIP, but there were some serious problems by today’s highest standards of care for infants and toddlers. Tizard and Rees noted that “Although staffing was generous, many different nurses handled the children. The average number of staff who had worked for at least a week with each child in the past 2 yr was 25.8, range 4-45. The major turnover was in untrained assistants and student nurses, but in all except two institutions the trained nurse in charge of the child’s group had changes at least once in the past 2 yr.” (p. 64). In addition, a number of the children had parents who visited from time to time. These institutions, although of high quality in various ways, did not meet the criterion of having assigned caregivers that I stated in my letter, nor the criterion of having low staff turnover that is usually considered developmentally important. The Tizard and Rees study thus did not do much to supply data comparing an excellent institution to foster care, as the institutions did not meet important criteria, and the contrast group were not in foster homes.

In their response to my letter, Fox, Nelson, and Zeanah concluded, “Our research provides the most definitive evidence to date that foster care is a more effective alternative than institutional care for abandoned infants and young children”. And, yes, I must agree with this, that it is “the most definitive evidence”. However, this does not mean that it is actually definitive, and it is regrettable that the CHIFF legislation confuses “most definitive… to date” with “definitive”.

Relevantly, the March-April 2014 issue of Infant Mental Health Journal contains several articles that contribute to this discussion. One of these “Psychopathology in young children in two types of foster care following institutional rearing” (Tibu, F., Humphreys, K.L., Fox, N.A., Nelson, C.A., & Zeanah, C., IMHJ, 35, 123-131), compared children from the BEIP who had been assigned to high-quality foster homes to a group that remained in institutional care for a time, then were placed in government-sponsored foster care (some of this group had already been adopted or returned to biological parents; the comparison group here were those who had not had those placements). The groups had thus spent different amounts of time in foster care. Information about the two groups also differed in  sources of bias, as there could have been biological or behavioral reasons why some were adopted or returned, and others not. Children in the high-quality foster homes had significantly fewer reports of ADHD. Girls in the high-quality group had significantly fewer internalizing problems--  an unusual finding, as girls are usually seen to be less sensitive to environmental factors than boys are. The authors pointed out the need for high-quality foster care to support good development.  (This comparison may have been one of “rich and healthy” care to “poor and healthy” care, rather than to the “poor and sick” care of the  low-quality institution.)

A second useful article from this issue of IMHJ  is by Victor Groza and Kelly Bunkers (“Adoption policy and evidence-based adoption practice: A comparison of Romania, Ukraine, India, Guatemala, and Ethiopia”, INHJ, 35, 160-171). This is a complex article reviewing a number of issues about adoption practices and needed policy, and I am going to choose a few comments from it that may be of particular interest to readers. (I don’t think I am “cherry-picking”, but no doubt someone will call me on it if I am.) Groza and Bunkers note that there is already a comprehensive legal framework on these issues, derived from the United Nations Convention on the Rights of the Child and emphasizing that decisions about adoption or other placements are always to be made with the goal of achieving the best interests of the child rather than furthering adult concerns of other kinds. They point out the need for careful record-keeping about domestic adoptions (a point well applied to intercountry adoption as well), and the difficulty of having a good evidence base without this—relevantly, they refer to traditions of secrecy about adoption that can interfere with the collection of information.

Groza and Bunkers looked at efforts to encourage domestic adoption in countries that were once much involved in intercountry adoption. These have generally been successful, but in the case of Ukraine a public relations campaign that emphasized how difficult it was for children to grow up in institutions may have stigmatized the children and alienated institution staff. Groza and Bunkers noted that it was not clear how these efforts had progressed after a change in government in 2010, and of course the current unrest in Ukraine in March 2014 makes it hard to predict what will happen next--  reminding us that there are many factors in any nation’s life that influence the number of abandoned children and the ways available to care for them.

An interesting point made by Groza and Bunkers compared situations in countries with many resources to those in the low-resource countries examined in this article. They commented, “Regardless of the [low-resource] country reviewed in this article, the main reason children entered the child protective system was abandonment.  The second reason was that a parent, usually the birth mother, relinquished the child, most often immediately or soon after birth. In contrast, most children in high-resource countries enter the child protection system because of a combination of neglect and abuse within their biological family. Often, parental rights are terminated involuntarily through the judicial system” (p. 167).  

Obviously, systematic investigation is revealing more and more about characteristics of abandoned children and their developmental outcomes in different settings. What is “most definitive… to date” may not be so at a later date. In any case, extreme care must be taken about generalization from events in low-resource countries to those in high-resource countries, and vice-versa. Research on the needs of parentless children is not yet ready to be depended on by legislation. If such legislation is baldly stated to be based on values or a priori assumptions, well and good; if it is claimed to be  based on the “scientifically proven”, that is a mistake.