Fighting over Numbers

About: Isabelle Bruno and Emmanuel Didier, Benchmarking. L’État sous pression statistique (Benchmarking: The State under Statistical Pressure), Paris, 2013 ; Isabelle Bruno, Emmanuel Didier and Julien Prévieux, eds., Statactivisme. Comment lutter avec des nombres (Statactivism: How to Fight with Numbers), Paris, 2014.

The use of numbers to define the contours of public action is nothing new, as the etymology of the word “statistics” suggests. [1] Nor is the invocation of figures in public debates a recent phenomenon: thus Theodore M. Porter traces back to the 1830s the “tidal wave of numbers that has swamped so many aspects of social, governmental, and scientific life” (S, p. 249). [2] In recent years, a number of sociological studies of quantification have appeared, examining the concept and consequences of this “numbering” of social life which often accompanies its “ordering. Published a year apart in the same collection and initiated by the same authors, Isabelle Bruno and Emmanuel Didier, who coauthored the first book and coedited (with Julien Prévieux) the second, Benchmarking. L’État sous pression statistique (Benchmarking: The State under Statistical Pressure) and Statactivisme. Comment lutter avec des nombres (Statactivism: How to Fight with Numbers) constitute a diptych that offers a series of empirical studies and a toolkit for activists, [3] introducing them to various approaches in the social sciences, notably those of Alain Desrosières and Luc Boltanski, [4] both of whom contributed to the later book.

The distinctive character of recent times lies in the fact that the proliferation of numbers has extended to the social world in its entirety: every aspect of society, having becoming quantifiable in ever increasing detail, can now be the object of comparative assessment, including states themselves, which “no longer appear above the fray, protected from competition by [their] sovereign incommensurability” (B, p. 10). The case of “benchmarking” is remarkable in that, over the course of a history that is far bumpier than its triumphant mythology would suggest, the dogma of “total quality” coalesced as it shuttled back and forth between Japan and the United States, fanning out from the private sector (and companies such as Xerox [5]) and endowing itself with quantitative measures, in keeping with the now well documented logic of the “new public management,” on the basis of which the authors coined the term “new public quantification” (B, p. 28). Initially, benchmarking’s approach consisted in describing every aspect of the activity that one was seeking to assess. This frequently resulted in an “inflationary growth in indicators” B, p. 139), which were too numerous to allow for synthetic measurement. [6] Once the latter could be determined, it became the basis for classifications and rankings (of universities, hospitals, and so on), which were particularly appreciated for distributing features once deemed incommensurable across a single axis, making an ordinal approach possible. [7]

A New Kind of Number

Not only were the numbers traversing social space growing; they were are also of “a new variety” and a “new kind,” ultimately generating “new quantities” (B, p. 17). Their originality can, in particular, be located in two characteristics. First, these techniques, exemplified by benchmarking, conflate the evaluatee and the evaluator, since the indicators are reported by the very people to whom they apply. Second, since the goal of these techniques is continued performance enhancement, they initiate a “race with no finishing line” (B, p. 72-88) and trigger “indefinite cycles of comparative evaluation” (B, p. 13). By connecting these two features, benchmarking becomes a tool both for observing and transforming the world: the description of existing practices leads to the prescription of new practices, which are presumed to be more efficient. Yet the quest for total quality is, by definition, never ending: “benchmarking creates a quantified reality yet it does not, however, consecrate that reality; it renders it always disappointing, unsatisfactory, and perfectible. This process of de-valuing reality by comparing it with itself is crucial to inciting actors to change reality. Herein lies benchmarking’s strength and its most distinctive feature: it does not limit itself to translating reality into statistics as the basis for action, it stimulates this action and channels it towards something ‘better,’ the definition of which eludes the actor’s rational self-determination” (B, p. 27).

Assigned to evaluatees and evaluators who cannot set their own goals or timeframes, the endless quest for quantitative perfection has two main consequences, both of which have been documented in studies of the “new public management.” [8] First, the aspiration to meet numerical goals imposed from above leads to manipulation. The striking example of the ways in which the police do this is discussed in both volumes, notably by way of an amusing series of photograms drawn from a film produced by the Association Pénombre (S, p. 105-111). The latter pokes fun at the so-called “numbers policy” that leads the police to engage in frank abuses of power (B, p. 140-143). Furthermore, management by numbers, in demanding participation in perpetual cycles of assessment that result in the setting of ever higher goals, is premised on increased psychological pressure, an effect that is generally confined to the shadows, as is workplace suffering, which Ivan du Roy discusses in his contribution (S, p. 169-182).

But the project underpinning both books consists not only in critiquing the way in which neoliberalism uses numbers as a tool for domination, but also in sketching possible forms of resistance. This struggle does not involve appealing to the virtues of the “qualitative,” which would allegedly allow for a more accurate description of social phenomenon than would a “quantitative” approach that is exclusively concerned with aggregates. The authors do not say we should abandon the advantages of enumeration (a position summarized in the words concluding the chapter by Association Pénombre: “when you get down to it, numbers are neat” [S, p. 113]). This position is also tied to the strategic imperative, which the authors readily acknowledge, of refusing to surrender quantitative tools to one’s opponents: “Our reaction is not that of those who reject [statistics] as a whole and cry: ‘Not to quantophrenia! Not to numbers! Yes to quality!,’ because in so doing, they allow the powerful to monopolize these tools.” Yet there is no reason for quantification to always be on the side of the state and capital” (S, p. 7). The goal of “turning around” statistics, transforming them into an emancipatory tool, underpins both books. Thus the first book, Benchmarking, ends with an appeal that some readers might find pompous (“An eye for an eye, a number for a number! Statactivism!,” B, p. 120). This final word becomes the title of the second book, in which one also finds the same invocation of a new law of the talion (S, Damien de Blic’s chapter, p. 186). “Statactivism” is a neologism that evokes the possibility of a “different kind of number” and seeks to be a “call for quantified objects that reconfigure as many things as possible for a purpose that is desired and, perhaps, favored by the greatest number” (S, p. 30). In this way, it is both a “slogan” and a “descriptive concept” (S, p. 7) [9].

Consequently, one of the central arguments of both books is the need to reaffirm the political nature of quantitative tools in order to reclaim them for emancipatory struggles. To borrow the title of Desrosières’ essay (S, p. 51-66), the goal is to turn a “tool of power” into a “tool of liberation.” Thinking about tools of quantification is thus as political as it is technical, when one seeks, for instance, to highlight the conventions on which numerical indicators are based, such as wealth indicators that are alternatives to GDP and that, to the contrary, bring attention to poverty (S, p. 199-211, p. 233-245). Recalling the political dimension of these questions prevents their technical character from becoming a smokescreen that shields many issues from contestation. Porter, in this spirit, proposes a theory of “boredom” as a way of orchestrating technical details in a way that averts any potentially critical gazes: “The fact that something is boring is a sign that it is uncontroversial, that it has no dark corners, or at least that no one realizes it does. Technical routines silence dissent. Boredom is orchestrated by budget offices, corps of engineers, and major international banks. Lurking right behind these actions one can see, if one looks, intense struggles surrounding the best way to quantify, struggles that radically undermine any credulity in the spectacle of objectivity that is directed at the general public” (S, p. 261-262).

Elements of Resistance to Numerical Domination

The authors explore various approaches to resisting domination by numbers. The first accompanies the pursuit of quantification goals that are seen as pointless tools referring to nothing other than themselves. The case of the police, whether in France or the United States, illustrates this problem with particular eloquence. Both books deal with it at length (B, p. 122-156; the chapters by Eli B. Silverman and the Association Pénombre in S). They describe, for instance, a strategy known as “chanstique,” which they define as a way of “getting around the rules of statistical reporting” by artificially inflating or deflating indicators of police activity: “statistical activity is thus carried out for its own sake, independently of police work and its real effects…. The so-called realism of the politics of numbers refuses to see that it, too, is irreconcilably constructed” (B, p. 145). Even so, it is surprising that the authors equate such practices exclusively with resistance, when it would be entirely possible to see them as the triumph of quantification, behind which reality, which these tools are supposed to describe, has vanished once and for all. In any case, one would expect the authors to interrogate this paradoxical form of resistance based on conformism. Another strategy consists in highlighting the absurdity of certain forms of quantification, such as the collective Superflex’s “number of visitors” project,” (S, p. 196-198), which displays to a museum’s visitors the amount of people who have attended a particular exhibit—a decisive criteria for allocating public financing.

A second way of fighting numerical domination consists in contrasting them with alternative forms of quantification. These can, depending on the case, bring new categories into play, following the classic principle that to fight domination one must first make it visible. If the means employed do not break with customary statistical practices, they are henceforth applied to new categories. Louis-Georges Tin (S, p. 155-166) reminds us that “diversity statistics” are a weapon that have long been used by antiracist movements. In France, they date back to Victor Schœlcher. This new indicators approach thus consists in shedding light on what has remained in the shadows. For instance, while the number of people in France without working papers who are expulsed is regularly announced, de Blic (S, p. 184-195) suggests relating them to their cost, a point on which the Interior Ministry has been far more discreet. The author places these costs at an average of 26,000 € per expulsed person. The revelatory effect of such numbers explains why governmental bodies have reclaimed this approach for their own purposes. Similarly, Ivan du Roy (S, p. 169-182) recommends quantifying problems of workplace suffering by examining the initial objectification that occurs prior to qualitative investigations that complete the definition of social phenomena. Finally, though it has been criticized, a tool such as the Baromètre des inégalités et de pauvreté (Barometer of Inequality and Poverty, or BIP 40), is seen by Pierre Concialdi (S, 199-211) as guiding the INSEE (the French statistical agency) when it was on the verge of publishing a quality indicator relating to employment and, finally, with the creation, in late 2005, of a workgroup on the theme “Standard of Living and Social Inequality” in the National Council on Statistical Information (CNIS).

These two books thus reveal a conflict between two types of numbers: some reproducing domination, others leading to emancipation. In this way, they offer a discussion of the ways in which power relations are enmeshed with issues of quantification. [10] One of the central questions thus becomes determining “who ultimately has the power to choose these indicators, rejecting some, retaining others” (B, p. 21). Statactivism’s basic method is thus an attack on the foundation of numbers’ authority, recalling the necessity of a constructivist approach if one is to contest the illusion of neutral numbers that merely reflect reality. The power that some numbers possess must thus be grasped in reference to the positions of their advocates and detractors, of those who initiate their use and those who are targeted by them in power struggles. Thus while benchmarking policies are spreading, they have yet to become generalized: as Didier and Bruno recall, the summits of power still manage to exempt themselves from these policies. A plan to evaluate the ministers in Prime Minister François Fillon’s government, launched at the request of the French presidency in January 2008, was, for example, quickly swept under the carpet (B, p. 63-64).

In this sense, opposition to numbers, and perhaps more than anything the proposal of alternatives, is tied to the legitimacy of those who throw themselves into such endeavors in addition to, as Desrosières notes, “institutions that provide the data upon which [statistical arguments] rely” (S, p. 56). As tools of struggle, numbers cannot be conceived in isolation from the social groups that use them. The role of the resources available to actors, and, by extension, their unequal distribution across social space is strikingly apparent in the peregrinations of inequality indicators through the state statistics system, which Bernard Sujobert recounts. A central feature of this lengthy process involving the CNIS is tied to the ambivalence of the demands of the CGT, one of France’s leading trade unions. The CGT simultaneously adopts the language of the statisticians and of activists, meaning that it can make demands that are seen as highly credible by its interlocutors: “in principle, only a national trade union with a local in the central state statistics office can bring these two exigencies together” (S, p. 222).

Keeping Statactivism Alive?

The discussion of power relations relates rather naturally to one of statactivism’s goals: showing that “academic critique, social critique, and artistic critique can converge” (S, p. 9). The idea, which is also well-known in activist circles, of a convergence of struggles nonetheless raises a number of critical questions, which, for some, come down to applying to each book its own categories of understanding. Both texts pay particular attention to the invention of alternative indicators and categories, which “is already, and should become even more so, a terrain for statactivism” (S, p. 20). After reading these books, particularly Statactivisme, the reader is left wondering if one social category that is perhaps in need of invention is precisely that of “statactivism” itself—or, in other words, if one of the points of the book is not to group together under a single heading a disparate range of practices. To put this point in more political terms: do these struggles really converge, or is the purpose of the book to create the conditions of possibility under which such a convergence could occur? In ways that are often implicit, Statactivisme seeks to be a bridge towards the gradual elaboration of an activist “statactivist” movement, the first stirrings of which occurred in the 1960s, in the academic realm, in the wake of Pierre Bourdieu’s sociology, and in the art world, through such work as Hans Haacke’s Visitor’s Profile (S, p. 67-72). This effort requires us to consider what is distinctive about statactivism, a practice that, in its broadest sense, is very commonplace, but which, in its more restrictive sense, is conceived as particularly “adapted to the form of power exercised by neoliberal governmentality” (S, p. 8). The distinction between these two possible meanings is not developed any further, which would have it possible to underscore the originality of a project that, otherwise, risks being seen as little more than a semantic invention.

More generally, it remains ambiguous to what epistemic category a concept such as “statactivism” belongs: is it a social category or a sociological one? Though it is a neologism, it is presented as both a descriptive category and a slogan, an intellectual construct and a label that actors might invoke themselves. Ultimately, “statactivist circles” are objects that exist prior to research and which research seeks to shed light on and a possible horizon towards which the authors seek to direct us. Thus if statactivism is a “banner” under which whomever seeks to use statistics for emancipation can rally, its modus operandi has been set: since the critical interrogation of statistics implies that one must be both “inside” and “outside,” “statactivists tend to work in pairs or in groups, the most typical case being a police officer working with a sociologist” (S, p. 19). The books’ relative ambiguity towards statactivism, which they seek to describe and to bring forth, echoes the contrast between the two forms of social organization studied by Cyprien Tasset, which overlaps to some extent with the classic distinction between groups “in themselves” and groups “for themselves” (S, pp. 117-132): the book, as a whole, seems to alternate between these two forms of organization.

If the group status of statactivists can at times seem hazy, the same can be said of its opponents. If their stated target is a particular way of using numbers, one that is characteristic of neoliberal domination as exercised in public institutions, it is often managerial discourse that is the focus of their critique, even when it is not employed by public services or based on quantified data. This is particularly evident in the chapter by Martin Le Chevalier (S, pp. 133-153), which parodies the language of a consulting firm in a report allegedly addressed to an artist seeking to apply managerial methods to his artistic career. While the forms of advice and the certitudes of the consultants are mocked [11] and though some numerical indicators are mentioned briefly in passing, it is not clear that the essay is directed against anything other than the consultants’ managerial discourse (particular the SWOT model [12]), nor that statistics are criticized from any other standpoint than that of an opposition between “quality” and “quantity.” This critique of quantification as reductive, and thus inherently opposed to any artistic pursuits, which in practice also informs the approach of the Superflex Collective, thus sits uneasily alongside the “statactivist” project insofar as the latter refuses to posit itself as straightforwardly opposed to “quantophrenia.”