Given the controversy that coercive treatment has generated in psychiatry and law, it is surprising that there is not a wealth of data on the extent and outcomes of coercion. One would expect that the most basic data on the incidence of formal, legal involuntary hospitalization within a nation would be published by each government. Most nations, however, do not compile and/or publish data on involuntary hospitalization or on any other type of compulsory psychiatric treatment as they compile and publish demographic, economic, social and health data; accordingly, there is no such compilation by nation published by the United Nations or any of its agencies. Where national statistics are published, as in the United States, they are subject to variation by hospital type and region, and the individual studies we have make generalizations tentative. There is still much we do know. From legislation, court decisions, and legal and psychiatric writings, we know that the dominant philosophy in Western Europe and North America favors community treatment without coercion and favors a maximum of patient autonomy in all treatment modes (26;52;68;106). Congruent with this philosophy are data that show that the number of persons residing in psychiatric institutions and their lengths of stay have been declining. Some national data also indicate that formal legal involuntary admissions have been declining. When individual studies from within nations are taken together, they also indicate declines in legal coercive hospitalization, though they may not be representative of their respective nations at one point in time. Data on the extent of nonformal coercive mental hospitalization and other coercive psychiatric treatments are much less available and less informative of trends. Data on the admission process of voluntary hospitalization indicate that coercion is common, that other coercive psychiatric treatments are used with voluntary as well as involuntary patients, and that ECT and psychosurgery are seldom used with involuntary patients. Because of small sample sizes, lack of representativeness, and variation in definitions, we do not know whether these indications from individual studies can be generalized, that is, whether they actually hold throughout the nations with studies and whether they hold in nations where there are no studies. Most notable is the little we know about the extent of involuntary hospitalization and other coercive measures in less developed countries. We encourage more research to fill in the missing information. Most basically, efforts should begin to encourage governments and the World Health Organization to collect and publish periodically data on the incidence of formal, legal categories of mental hospital admissions. Until then, researchers might combine in international working groups to obtain such basic data from official governmental sources of nations representing political or geographic regions. We encourage researchers to coordinate their efforts in investigating outcomes of coercion so that studies may have comparable measures, a greater opportunity for natural field experiments, and if efforts are combined, larger sample sizes with the ability to control relevant variables statistically. Researchers in nations with in- and outpatient case registers should take advantage of the possibilities of record linkage, which eases the collection of follow-up data between hospital and community treatment; however, because of legal restrictions in some nations, reliability problems with secondary data, and lack of important variables in patient records, primary data collection following individual patients is necessary if we are to obtain answers to many of the aforementioned questions. We also need more qualitative and quantitative investigation at the front end of the coercive process to identify reasons for resorting to coercion and to identify mechanisms by which both formal and informal coercion can be avoided. In this research, attention should be paid to the definitions and criteria for coercion, placing our understanding in the wider context of caring and respect for patients, as the work of Hoge, Lidz, Westrin, and their colleagues are doing in the United States and Sweden (8;24;42;50;56;57;114). A major question in coercive psychiatric treatment is whether these treatments are abusive, that is, used for social control without beneficence. Numerous studies show that involuntary hospitalization and other coercive measures are used for both social control, i.e., to protect the individual patient and to protect others, and beneficence, i.e., to bring treatment to patients and reduce illness-induced suffering (31;32;93;94). However, a full assessment of coercive psychiatric technology would seek to answer the question of whether the social control and beneficence can be achieved without formal invocation of the state's coercive power. Because of methodological problems with existent outcome studies, we do not know whether we can avoid the coercive measures and yet achieve positive outcomes. It may be that by reaching out more than is currently done to meet the needs and address the concerns of mentally ill persons, particularly those who are severely and persistently ill, both social control and beneficence can be achieved (99). That is surely an important research task yet to be done.