The Most Dangerous Branch of Science? Reining in Rogue Research and Reckless Experimentation in Social Services.

AuthorDwyer, James G.

TABLE OF CONTENTS ABSTRACT 1 TABLE OF CONTENTS 3 I. INTRODUCTION 5 II. AN ILLUSTRATION OF RECKLESS SOCIAL-SERVICE INNOVATION 12 A. Federal Law Background of Child Protection Practice 13 B. Advocacy Research Driving Policy Experiments 14 III. VIEWING THE PROBLEM THROUGH A LENS OF EXPERIMENTATION AND RESEARCH ETHICS 30 A. Inadequate Pre-Implementation Review 30 B. Entities With Self Interest in Outcomes Conduct the Research 38 1. Financial Self-Interest 39 2. Personal Ideology or Policy Preference 40 3. Studying Oneself 41 C. Impacted Persons Not Treated as Subjects 46 D. Even if Impacted Persons Are Treated as Subjects, Protections are Inadequate 48 E. Some Autonomous Persons Are Incapable of Voluntary Informed Consent 51 IV. EXISTING RULES 53 A. Preventing Unpromising Experimentation 53 B. Counter-Acting Researcher Bias 57 C. Protecting Subjects 61 D. Making Proxy Consent Meaningful 65 E. Preventing Coercion 68 V. BUILDING A MORE ROBUST ETHICAL FRAMEWORK 69 A. Treating Innovation as Experimentation subject to Pre- Approval and Study 69 1. A New Screening Mechanism 70 i. Impact Studies 70 ii. Ombudsperson 71 iii. Addressing the Agency Problem 73 iv. Specifying Standards for Approving Experimental Programs/Policies 74 2. Treat Social Services Experimentation as Research 75 i. Pre-Approval 80 ii. Piloting 82 iii. Legitimizing Participation 82 iv. Concerns 84 B. Minimizing the Prevalence and Influence of Bad Research 87 1. Treat all Substantially-Impacted Persons as Subjects 87 2. Mandate Consideration of Impact on Non-autonomous Persons 88 3. Rigorous Review of Study Design Before Approval 89 4. Expand the Concept of Conflict of Interests 92 5. Reduce the Influence of Large Funding Sources 92 6. Rationalizing Consent to Participation in Research 93 VI. CONCLUSION 94 I. INTRODUCTION

Research regulation captures little public attention, even when a crisis like a pandemic creates urgent need for potentially harmful medical experiments. But the approval process for human-subject research is a central aspect of life for most university departments and their faculties and for other institutions conducting research--not just medical research, but also psychological, sociological, anthropological, historical, etc. (1) And social science research--the focus of this Article--is immensely important to all areas of social policy. (2)

Social scientists in the United States. have complained, throughout the half century in which the federal government has imposed ethics rules on human-subject research, that requiring advance approval of their work by Institutional Review Boards ("IRBs"), as federal regulations have been interpreted to require, is generally unnecessary and--when institutions take it seriously--so burdensome and poorly-executed that it dooms many research plans that could yield important knowledge or serve as valuable learning experiences for students. (3) Their work, they say, is rather harmless, not posing physical dangers to subjects the way biomedical research does and rarely raising serious concerns about non-physical impact. (4) Most social scientists have accepted IRB review as a fact of life, and some even ascribe value to the process. (5) But others have issued strident calls for exempting all or nearly all social science studies from that process, charging that it violates free speech rights and causes public harm, far outweighing any benefits, by inhibiting acquisition of needed knowledge and generating great expense. (6)

This Article counters the claims that social science poses little danger and, further, that social scientists can be relied on to self-regulate. It urges greater scrutiny of research proposals in social science, or scrutiny of a different kind. The greatest dangers from social science research have yet to be widely acknowledged and addressed. The threat is not so much harm to research subjects, which has been the focus of IRB-review defenders and critics, (7) but rather that bad research is produced and generates bad policy harmful to a much greater number of people. (8) Intuitively, this is especially likely with policies impacting people less able to participate in policy making and to challenge research findings, or as to whom there is widespread prejudice or devaluation--for example, welfare recipients, gender-questioning youth, prisoners, domestic violence victims, juvenile delinquents, and persons with a mental disability or mentally illness. (9)

Research that is poorly designed, executed, and/or analyzed because of bias or incompetence is likely to produce false findings or invalid conclusions. (10) This might or might not affect the individuals who populate the studies, but the potential for harm to members of the broader groups they represent is itself sufficiently great to warrant ex ante quality constraints on much social science research. (11) Yet the commission principally responsible for developing research regulation in the 1970s explicitly directed that IRB's "should not consider as risks the possible consequences of application of the knowledge gained in the research (e.g., the possible effects of the research on public policy.)" (12) In doing so, however, that commission had in mind use of valid research. (13) It never addressed the problems bad research might cause or the possibility that research in some domains should undergo greater scrutiny because of potential policy use. This Article's principal concern is with those problems, and it calls for consideration of that possibility.

Existing mechanisms relied on for quality control are review of research proposals by funding agencies, post-hoc peer review solicited by publishers, and public deliberation of legal reforms. (14) These are manifestly inadequate. Despite them, much poor-quality research is produced, published, and used successfully for policy and legal advocacy. (15) Some funding agencies, as shown in Part II, have long been part of the problem rather than the solution, promoting research designed to advance political agendas rather than scientific knowledge. Even when funded by unbiased sources, researchers might inject their own policy biases, cut corners, misunderstand the phenomenon, or lack competencies, yet their reports can find a publication outlet. (16) And, advocates and policy makers routinely rely on studies without regard for their quality or the reputation of the publisher. (17) Even if someone challenges the quality of research relied on or the soundness of conclusions drawn from it, advocacy groups, legislators, agency officials, and judges typically have little desire, patience, or capacity to respond to questions about research quality. (18)

This Article proposes, as an additional ex ante check against flawed research, enhancing a different step in the process that extends from research idea to policy implementation--namely, IRB assessment of human-subject research proposals. Currently, that review need not (and typically does not) include assessment of research design; it focuses on the safety of research subjects and their informed consent to participation. (19) Existing ethical standards and regulations were developed principally for medical experiments, as a reaction to notorious studies that physically harmed participants. (20) Moreover, though existing regulations address the possibility of distorting motivations, they target only profit motive, as when drug manufacturers seek positive reports on their new products. (21) That typically is not the sort of conflict of aims that infects social science.

This Article therefore urges that IRBs be required rigorously to examine research design, and it suggests strategies for motivating their members to do so. It also recommends that IRBs be charged with guarding against a broader set of interest conflicts--in particular, when researchers' funding comes from an entity that also lobbies governments to promote a policy agenda. Just as campaign-finance law reflects recognition that wealthy advocacy organizations could buy legislative votes if permitted, research-ethics law should reflect recognition that such organizations can buy "scientific findings" if permitted. (22) Many social scientists are highly dependent on grants not just for pursuing their own research ideas but also for retention and promotion in their jobs. (23) They also, to a far greater degree than in medicine, are likely to be driven by their own ideological views to design and analyze studies in such a way as to ensure felicitous results. (24) Once results are published, no matter how flawed, advocates for causes run with them to policy makers, and the effects are difficult to undo. (25) The need for more stringent ex ante scrutiny is pressing.

More dramatically, this Article suggests viewing new social service programs and practices as a form of human-subject experimentation, just as new medical interventions are. It recommends scientific--not just political--gatekeeping. Social welfare policy innovation impacting vulnerable populations is often fueled by "the latest research," which might be fatally flawed, but revealed as such only after the idea it promotes is implemented. The history of "treatments" for gay and trans persons illustrates that clearly. (26) But whereas the Food and Drug Administration approval process for new vaccinations, medications, and medical devices is robust, familiar, and the focus of most scholarship on human-subject experimentation, little scholarly or government attention has been paid to ethical standards and procedures for experimentation with behavioral and other non-medical interventions. Yet innovation in social services can also have life-altering effects, and it likewise entails danger of ethical lapses and undercounting adverse impacts on vulnerable populations. (27) Driven by ideology, it can be downright reckless.

Viewing social-service innovation as human-subject experimentation invites the question whether it...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT