Were the outcomes of professionally arranged adoptions better than the outcomes of independent adoptions? This was one of the most widely cited comparative studies claiming to show that they were. It was conducted by two physicians, Catherine Amatruda (a Yale colleague of Arnold Gesell’s) and Joseph Baldwin, after the passage in 1943 of a Connecticut law requiring that all adoptions be investigated by a social agency. Mandatory inquiries of this kind were examples of the minimum standards that reformers had been advocating for decades, although this particular statute was considered inadequate because it did not require that investigations occur before children were placed.

The study compared 100 independently-placed babies with 100 babies whose adoptions had been arranged by agencies. Only data gathered at the time of placement was used, which meant that this research did not technically qualify as outcome research at all. There were both practical and ideological reasons for this. The researchers may have believed that it was too soon to follow up on outcomes since only a few years had passed. But they also assumed that adoptions judged “good” at the time of placement would necessary prove “good” later on. This ignored two key issues that later studies explored: reliability and validity. Research on reliability asked whether a meaningful consensus existed about the characteristics of “good” (or “bad”) placements. Research on validity tested whether initial placement decisions, reliable or not, accurately predicted outcomes measured later on.

Amatruda and Baldwin found that each group of 100 adoptions contained roughly the same proportion of “good” and “bad” babies and families. Good children were identified on the basis of normal developmental and mental testing, and so were babies categorized as “poor adoption risks,” a division that exemplified the hold eugenics still had on thinking about adoption qualifications at midcentury. Good homes met “modest standards” that parents offer “a reasonable modicum of security and stability, a happy home life and a decent up-bringing for the child.” Homes that did not do so—because of divorce, alcoholism, criminal records, drug addiction, or domestic violence—were categorized as “unsuitable.” The ratio in each group under study was three (good) to one (bad).

Agency adoptions were not distinguished by access to better human material. What made them different—and better—was their record of matching like with like. Agency placements made mistakes in matching only eight percent of the time, whereas independent adoptions mismatched twenty-eight percent of the children and parents. Good children were much likelier to end up in bad homes and vice-versa when professionals were absent from the adoption process. These findings, according to Amatruda and Baldwin, were empirical proof that “social agencies do better adoption placements than does the well-intentioned or expedient laity.” The naturalness of matching was so self-evident that Amatruda and Baldwin never wondered whether “low quality” children might need “high quality” homes. Matching itself was their measure of success.

Like other research substantiating the superiority of professionally managed adoptions, the larger goal was to decrease the avoidable risks that desperate birth mothers and foolish adopters frequently took by making it harder for them to arrange unregulated, non-agency adoptions. In 1957, six years after this research was published, Connecticut became the second state in the country (after Delaware) to ban independent adoptions altogether.