Social Science Accountability Beyond the State

As always, Kenneth Prewitt raises an issue about which we should all be concerned and he does so with his customary reflection and insight. In this intervention, I would merely like to “zoom out” a little, in both time and space, in order to consider what a broader, if less fine-grained, perspective might reveal about where we are today as social scientists who hope that we and our work will continue to be of value.

As we all know, social science arose with the modern state (hence the derivation of the word “statistics”) and developed as a reflection of the conviction that, as James Madison put it in 1788 “a good government implies two things: first, fidelity to the object of government, which is the happiness of the people; secondly, a knowledge of the means by which that object can be best attained” (Federalist Paper No. 62).

For nearly two centuries, American—and then many other—governments strove to be “good” in this sense. They assumed more and more responsibility for the “happiness” of the people for, and increasing to whom, they were responsible. This happiness has been understood in different ways, of course (an issue to which I will return), but in general it seems to have meant “welfare” or “well-being.” As an object of government, “well-being” might seem difficult to measure, much less promote, but the social sciences blossomed in the effort. Levels of education, assessments of health, measures of income, degrees of mobility—thousands of indicators of the welfare of the people sprang up, and soon even measures of popular satisfaction with the government’s fidelity to its purposes appeared in polling and surveys.

Policy developed out of that knowledge: policies to assimilate immigrants, manage urban transportation, promote social mobility, mitigate poverty, sustain agriculture, and foster innovation were all based on social science research that was itself becoming increasingly sophisticated and powerful. Indeed, as it developed it created for itself an autonomy and status, particularly in universities, that permitted not only utilitarian but “curiosity-driven” research not unlike that of its natural science counterparts. This too served useful public purposes, if inadvertently so; as Prewitt observes, “unintended social benefits appreciated retroactively” have been as important in economics as in physics. Intended or not, all of this social science was “the knowledge of the means” by which governments fulfill their purpose, and as such, it was astonishingly successful: if good government is knowledgeable government, social science has been extremely useful.

But that utility rests on a common understanding of the “object of government” and of the “happiness of the people” and these, far more than social science itself, have been caught up in controversy in recent decades. Since the 1970s, in the United States, and indeed around the world, the conviction that, in their efforts to promote happiness or welfare, governments had become too large, too expensive, too demanding, and too intrusive became an article of faith. From Ronald Reagan and Margaret Thatcher, Mikhail Gorbachev and the “Washington Consensus” of the International Monetary Fund and the World Bank, the contraction of the state in favor of the market was heralded as the new solution to the perennial problem of promoting happiness.

And that move to the market was accompanied by a change in the conception of happiness itself, now defined not as social welfare but as individual freedom. As Reagan put it: “our national purpose in this country—and we have lost sight of it too much in the last three decades—is to be free: to the limit possible with law and order, every man to be what God intended him to be.”

In this reading, the “object of government” has become a libertarian minimalism—law and order—and “the happiness of the people” is no longer social, but individual; no longer a claim on government service, but a release from government intrusion. The “means by which [the happiness] can best be attained” require relatively little knowledge on the part of our newly modest governments; in principle, how citizens fare in this neoliberal world is no longer really a matter of government concern. The increasing importance of non-governmental organizations and of non-state actors is but an indication of the contraction of the state and the limits on government in the twenty-first century (and of the limits of our imagination in describing the world in which we live).

The social sciences, having become a pursuit in their own right, continue on, however, increasingly looking to replace the shrinking government with beneficiaries who must be satisfied and by whom they will be held accountable. Thus the producers of useful knowledge about society supply such knowledge, including measures of “welfare” or “happiness,” to and for the anyone in the market who will pay for it—from the consumer products industry to the media, political campaigns and advocacy communities—as well as to and for anyone in the academy who will read it—that is, each other.

Accountability has become an issue for the social sciences because, unlike the governments that they were originally designed to serve (who were, in Abraham Lincoln’s words, governments “of the people, by the people, for the people”) all the non-state actors and non-governmental organizations that secure their services and use the knowledge they produce are themselves “accountability-challenged.”

The business sector in the United States genuflects to the shareholders of publicly-traded firms, but privately held firms account for well over 98 percent of all businesses in the US and over 85 percent of US firms with five hundred or more employees—and that overwhelming preponderance of privately held business is true globally as well. So too, the not-for-profit sector: universities, foundations, and advocacy groups accumulate large endowments managed by families or self-reproducing boards with little or no accountability.

Sometimes, thanks to the Brundtland Commission’s acknowledgment of their interest in sustainable development, the not-for-profit world claims to be accountable to “future generations” although the institutions by which such accountability is to be secured are, at best, mere glimmers on the horizon. For consumer products and services enterprises, including universities, “customer satisfaction” may sometimes serve as a proxy for accountability but it is not, in fact, the same. Government regulation is similarly a feeble surrogate and is, in any case, tarnished by being associated with government in the first place.

Social science in the service of the state was, as it were, covered by the accountability of the government. Now, and in the absence of that purpose, social science is like a loose fire hose: a strong and immensely valuable tool that has lost its handler; it flails dangerously, spraying its high-pressure findings in all directions and, in doing so, risks being subject to the same fate: an eventual cut-off.

Hence the concern for accountability in the social sciences. It is both a systemic concern about the world of the “state-in-eclipse” and a self-interested worry about justifying the social sciences themselves. The first issue is no doubt more important, the second probably more urgent.

How best to ensure that the supply of support for social science is not cut off in the absence of the clear utility and unimpeachably virtuous purpose of Madison’s good government? As a first cut, the social science community of the United States and Europe—or at least a good part of it—is electing to behave as if it were accountable to each other, even if it is not clear who else might care: hence the myriad efforts to ensure data access, replicability, and transparency. Perhaps not surprisingly, these efforts have revealed interesting and instructive limitations in the practice of social science as science. As one such effort puts it:

For example, novel and positive results are published more frequently than replications, null results, or perplexing outcomes. This leads to a biased and incomplete record of research […] To promote open, reproducible research practices within the scientific community, [we are] working across disciplines to identify the most useful strategies and tools for data collection, analysis, and reporting. This includes the use of study registries, pre-analysis plans, disclosure standards, version control and data sharing platforms, and replication projects.

Much of this work is fraught with complexity and controversy, particularly at this early stage: humanistic social scientists are not confident that their research methods conform to this kind of protocol at all and worry that the value of their work will be belittled or diminished by these efforts. Social scientists who work in authoritarian contexts, in arenas where data is scarce or bad, or on sensitive issues, are anxious about the dangers to their colleagues and research subjects that the application of these kinds of standards might pose. Nonetheless, the effort is proceeding apace and it is likely to change the practice of social science research tangibly and, on balance, positively.

As for whom besides other social scientists this research might be useful, that is, who will be responsible for the “happiness of the people” and what knowledge they may need in pursuit of that object, it is hard to see: we are obviously in the early days of major transformations in how human societies will manage themselves. As Joseph Nye has observed in politics, “increasingly, power is exercised in the diffuse domain of cyber interactions, where the very identity of an attacker is often ambiguous. This is a new world of politics with which we have less experience.” So too the instantaneous and complex transactions of modern finance, including the murky worlds of cyber-currencies, challenge conventional economics.

As we begin to understand our new world, “non-governmental organizations” and “non-state actors” will come to have positive descriptive labels that better convey their intents and purposes. After all, to know what they are not seeking to be gives us little purchase on what they may wish to do, and still less capacity to assess their success at doing it or their impact on the “happiness of the people.” Will they supplement or supplant the governments and states with which we are familiar? As Jessica Mathews warned twenty years ago, “For all their strengths, NGOs are special interests. The best of them […] often suffer from tunnel vision, judging every public act by how it affects their particular interest.” They are explicitly and intentionally designed to serve not the general welfare but only a small part, however important. So, too, non-state actors are typically partisan, often extremely so—accountable, like their non-governmental colleagues, not for the general welfare, but for the benefit of a far narrower religious, ethnic, or ideological community.

We thus continue to await the institutions that will serve the general good. And in the meantime, we are remarkably wary, reluctant to define the principles and objectives to which we aspire in managing our affairs; we shy away from the why of rule and instead concentrate on how we do it. Governance replaces government, means replace ends, metrics supplant ambitions. This too is why we focus on accountability; it is a welcome distraction from puzzles about the world we live in that we are hard put to engage.

Yet that is the challenge I read in Kenneth Prewitt’s plea that we endeavor to better understand “the space in which social science is used.” He is right, as always, but it is a project that demands imagination, ambition, and no small measure of providential fortune. We await the Madisons and Lincolns, the Mills and the Marxes of the twenty-first century.