For over 30 years, Victor Calef, M.D., taught at the San Francisco Psychoanalytic Institute. He began giving seminars in 1952 and remained an active teacher and supervisor until several days before his death in 1983. Brilliant, pugnacious, loving, and infuriating, he taught by provoking students to prove him wrong with the same charismatic intensity that inspired us, most of the time, to want to prove him right. I designed the following research project after a morning's discussion with Vic left me sufficiently provoked that I set out to show him just how wrong he could be.
A candidate at the time, my practice was made up almost entirely of women. In the course of a conversation with Vic, I had mentioned how few men I saw and added that I was eager to see more. His advice was brief; analyze your prejudices against men and you'll get plenty of male referrals.
This didn't sit right. Not only was I unaware of such prejudices (admittedly, no defense in the analytic world), but I also had the impression that none of my female colleagues received any more male referrals than I did. However, continued discussion with Vic had the quality of writing on water, so I set out formally to investigate whether a problem did in fact exist for female analysts who sought male referrals.
I obtained data from 170 analysts from four psychoanalytic institutes across the U.S. I matched female analysts from those institutes with an equal number of male analysts from the same institutes. Of the total, 88 were candidates (44 women and 44 men); 21 were training analysts (11 women and 10 men); and the rest were graduated members of the institutes (29 women and 32 men).
Each analyst was asked to complete a monthly questionnaire for three months, describing all adult referrals received that month. The return rate on the 510 questionnaires was extremely high, resulting in very reliable findings (100 percent from three institutes and 98.3 percent from the fourth, thanks to the exceptional efforts of five analysts who helped collect the data: Drs. Barbara Deutsch, Muriel Laskin, K. Lynne Moritz, Barbara Rosenfeld, and Eleanor Schuker).
Among many findings, the most striking was this: out of 1,267 adult referrals received by the subject analysts, male analysts received roughly equal proportions of female and male referrals (52.1 versus 47.9 percent), whereas female analysts received a much higher proportion of female referrals (80.5 percent female versus 19.5 percent male). This difference in proportion was highly significant statistically. As the total number of adult referrals made to male and to female analysts was almost equal (631 versus 636), the effect of gender was related to the distribution of referrals between male and female analysts rather than to the absolute quantity. (Interestingly, this contrasted with the expectations of many colleagues who believed that female analysts receive more referrals overall than do their male peers.)
At this point, in looking at the data, I found a collaborator, Daphne de Marneffe, and together we examined the central finding in conjunction with the rest of the data. We discovered that it held up when we controlled for a host of variables: patient age, diagnosis, and ability to pay; analyst age, degree, years of experience, institute affiliation, status within the institute, number of referrals received, and hours spent in clinical practice; whether referrals were made by a man or by a woman, or by a candidate, an analyst, or a nonanalyst; and regardless of the initial purpose or final disposition of the referral. Of special interest was the fact that the finding held up strongly when we isolated only those referrals who turned out to be referrals for psychoanalysis. Finally, the results continued to be highly significant even when we removed from our statistical analysis those instances in which an analyst of a particular gender had been specifically requested by the referred patient.
As we examined these results, we were impressed by a number of things. First, while the study had initially been prompted by a concern about the paucity of male referrals to female analysts, we realized that a larger issue was at stake. Our data offered us strong evidence that the process by which referrals are made is not gender-blind but is instead heavily influenced by considerations of gender match between analyst and patient. This finding does not correspond to conventional analytic wisdom. With the exception of a few recent articles, the analytic literature has suggested that for most adult cases, the analyst's gender should not be a major consideration in making a referral. Our results suggested not only that considerations of gender in fact weigh heavily, but that, equally challenging to conventional wisdom, preference for one gender match far exceeds another (male patients were consistently referred to male analysts and were rarely referred to female analysts). What was equally interesting was that when we isolated those referrals made specifically by analysts (as opposed to non-analysts), our results showed exactly the same patterns. In other words, our study raises questions not only about the influence of gender on referral patterns, but also about whether analysts, in referring patients, do what they think they do as regards gender.
In addition to exploring the role of gender in how referrals are made, we were eager to explain our specific finding regarding the dearth of male referrals to female analysts. We considered a number of possible explanations but found only two that fit the data. (Several of the hypotheses most commonly advanced by colleagues to explain our findings did not hold up. Specifically, our data could not be accounted for by explanations based on the fact that more women request a same-gender therapist than do men, or the fact that there are more female than male referrals overall, or the fact that there are more male than female analysts in practice.)
Of the two explanations that did fit the data, one was based on the idea that there exists a general preference to match both male and female patients with same-gender therapists, a preference which in the study appeared more dramatic for women than for men due simply to the peculiar demographics of our sample (i.e., there were more female than male patients but an equal number of male and female analysts). When we checked this explanation against our data, we discovered that it could indeed account for a portion of our findings—but only a portion. We then examined the final possibility that our findings reflect a reluctance to refer male patients to female analysts which is not matched by an equivalent reluctance to refer female patients to male analysts.
We realized that these two possible explanations
had different implications, making it especially important to sort out the relative contributions of each. A preference to refer both men and women to same-gender analysts may challenge our theories about how referrals are or should be made, but it is not inconceivable that such a preference could reflect clinical wisdom, i.e., could serve the best interests of patients. On the other hand, we find ourselves hardpressed to come up with models of clinical wisdom that might account for the one-sidedness of a reluctance to make cross-gender referrals to female analysts but not to male.
The first explanation—a general preference for a same-gender match—raises questions about whether analysts rely in practice on “real” factors regarding patient-analyst match to an extent not reflected in formal theory. Such questions certainly merit further investigation, though the limitations of our data place them beyond the scope of our study. However, in relation to our second explanation—the specific reluctance to refer men to women—our data did reveal trends suggesting underlying attitudes that seem to be in the interests of neither patients nor analysts, but instead reflect familiar prejudices in the culture at large.
For example, we noted that the male patients who were referred to women were consistently younger than those referred to men, the analysts themselves being matched on age. To the extent that an analyst is perceived as an authority occupying a position of power, we wondered if there may exist some reluctance to place men in a situation where they occupy the less powerful role—a reluctance that is mitigated by age, so that those men who are referred to women are younger and less established than those referred to male analysts. Similarly, since the trend we observed subsided somewhat as female analysts moved into their sixties, we wondered if female analysts are perceived to gain the authority deemed necessary to treat men at a significantly more advanced age than are male analysts. Or, do anxieties about women as erotic transference objects interfere with referring men to women, especially younger women? Perhaps as women age they are perceived as less sexual and therefore safer. In general, we wondered whether familiar derivatives of phallic castration anxiety might fuel feelings that female analysts, themselves viewed as “castrated,” could be dangerous to men. Or, again a derivative of phallic castration anxiety, perhaps women are viewed as simply lacking what it takes to help men. Perhaps female analysts are seen as frighteningly reminiscent of an all-powerful mother imago that exerts excessive regressive pull and thereby threatens male identity.
While the nature of our data does not permit us to do more than speculate along these lines, we hope our speculations will stimulate further investigation of the dramatic trends we have documented.
Finally, we note a serendipitous finding that may bear on the question of whether anything can be done about the problem we have identified. Some months after we presented our findings to members of the San Francisco Institute (the first we had studied), two female analysts in San Francisco reported that they had started receiving, for the first time in their analytic careers, a significant number of male referrals from male colleagues. These women inferred that their colleagues, having been made aware through our presentation that they had not typically referred male patients to female analysts, had changed their behavior to correspond to what they believed they had been doing all along, i.e., referring male as well as female patients to female colleagues. It may be that, at least in the case of analysts referring patients, simply increasing the awareness of referral behaviors, which (we believe) few analysts in principal espouse, may help alter the behaviors themselves. The question, of course, is whether the biases contributing to the behaviors will prove amenable to straightforward educational correction, or whether those biases concern more deep-seated matters requiring more complex solutions. Perhaps further research will shed light on this question.
Elizabeth Lloyd Mayer, Ph.D., is a Training and Supervising Analyst at the San Francisco Psychoanalytic Institute. This research was conducted with the collaboration of Daphne de Marneffe, M.A., and in cooperation with the COPE Workshop on Issues in the Psychoanalytic Training of Women; it was funded in part by the Fund for Psychoanalytic Research.