Surprising Strategies in the Panopticon

The first approach to workplace surveillance is inspired by Michel Foucault s influential discussion of the panopticon. It is not surprising that Foucault s metaphor based on Jeremy Bentham s proposal for a prison (Bentham, 1969) should emphasize conflict and power. In this chapter, we shall move from metaphor to a more precise model of conflict over surveillance. Our model suggests that the panopticon metaphor exaggerates the power and one-sidedness of workplace surveillance.

We begin with Foucault s Panopticon paradigm:

[T]he major effect of the Panopticon: to induce in the inmate a state of conscious and permanent visibility that assures the automatic functioning of power. So to arrange things that the surveillance is permanent in its effects even if it is discontinuous in its action; that the perfection of power should tend to render its actual exercise unnecessary; that this architectural apparatus should be a machine for creating and sustaining a power relation independent of the person who exercises it. (Foucault, 1977,p.201)

Those who appeal to the panopticon paradigm stress the power relationship that surveillance reinforces and the lack of choice on the part of the surveilled in the face of technology. Botan writes , One of the effects of this relationship is that occupants have no choice but to act as if they are being watched, even when they are not (Botan, 1996,p. 298). However, this paradigm is misleading. Even on its own assumptions, the model plausibly underlying Foucault s account does not support the conclusion that he hints at and Botan makes explicit. Surveillance is neither as permanent nor the surveilled as lacking in choice, as they conclude.

A Model

Behind the panopticon paradigm stands a model of pure conflict, which we develop as a two-person game. A game is a model consisting of a set of roles, each with a set of strategies, leading to outcomes ranked by the players values. In our model (Figure 1), the roles and strategies are Inspector, who can Watch or Not Watch, and Prisoner who can Conform or Not Conform. Next we specify values to generate behaviors that Foucault describes. Inspector wants to control Prisoner, to force him to conform; Prisoner wants not to conform. In addition, Prisoner would prefer not to be observed not conforming, as this leads to punishment . Finally, Inspector would prefer not watching, as this is costly. (This is the motive for discontinuous surveillance in Foucault.) We stipulate outcome values in Figure 1 that meet these constraints. We emphasize that these numerical values are arbitrary, chosen to illustrate the logic of the situation that Foucault describes.

Figure 1: Panopticon game

This model explains much of what Foucault says via the crucial concept of strategy equilibrium. If Inspector watches , Prisoner prefers to conform. But in this case, Inspector prefers to not watch, so we drop to the SW quadrant of the game matrix. This shows that the strategy pair Watch/Conform is not an equilibrium. If Prisoner is conforming, Inspector will prefer to not watch. A strategy pair is only an equilibrium if neither player has a better strategy alternative, given what the other player chooses. That is roughly , only equilibria strategies are strategically consistent. Continuing our examination of possible strategies, since Watching is not an equilibrium strategy for Inspector, we try Not Watching. If Inspector is not watching, Prisoner will prefer to Not Conform, leading Inspector to prefer Watching. So, we see that the strategy pair Not Watch/Not Conform is also not in equilibrium. Indeed, this failure of all pure strategies was once thought to be paradoxical (Rapoport, 1967). A basic result of the theory of games of conflict is while no pure strategies are in equilibrium, every pure conflict game has an equilibrium in mixed, or probabilistic, strategies. Inspector should then watch intermittently. As Foucault notes, it is too much that the prisoner should be constantly observed by an inspector Bentham laid down the principle that the power should be visible and unverifiable : The inmate must never know whether he is being looked at at any one moment (Foucault, 1977,p. 201). Our model, however, allows us to continue the analysis of the interaction, which Foucault assumes ends here. Since Prisoner only prefers to Conform when Inspector is watching, once Inspector is following a mixed strategy, Prisoner should also only conform discontinuously.

There are three crucial points to notice. First the outcomes are not automatic but follow from the strategies preferred by the players, given the constraints of the technology and the social situation. Second, Prisoner has a choice, which he exercises to intermittently non-conform. Third, the outcomes are not totally asymmetrical ” discontinuous for Inspector, continuous for Prisoner ” as the panopticon metaphor predicts. The equilibrium of the game is for both players to use mixed strategies. This is not to say that the two roles are equal. As stipulated, the payoffs in the matrix make Inspector more powerful. This is reflected in the inequality in the equilibrium mixed strategy. He need only watch less than half of the time; Prisoner needs to conform more than 80% of the time. Nonetheless, 80% is not 100%, and the outcome is not automatic, simply flowing from the technology, but flows in part from the agents preferences and choices. Summing up, even those under surveillance by the more powerful have options, and the equilibrium includes outcomes not preferred by the more powerful player.

Lessons from the Model

Having sketched a rough model of the panopticon paradigm, we put it to use, asking what it can teach us about workplace surveillance and how it structures our ethical analysis of this problem area. Researchers working in the panoptic paradigm have identified several adverse effects of workplace surveillance (Botan, 1996). A loss of privacy increased uncertainty about an employee s job situation, reduced self-esteem, and lowered levels of workplace communication. Each of these effects is obviously bad for employees and for that reason alone warrant further study and ethical analysis. In most cases, we would expect them to also adversely affect organizational goals. Unfortunately, to the extent that it assumes that the social situation is one of pure conflict, the panoptic critique of workplace surveillance lacks normative leverage here. Granted, the effects are bad for employees , but assuming pure conflict, what is bad for them is good for their adversary. That is why the game is classified as constant-sum: A s loss is strictly equal to B s gain. Botan notes this problem:

The panoptic effects studied here may not be ones anticipated by those who plan and implement programs of surveillance. Indeed, since they do not directly address profitability, the panoptic effects studied here may be of little interest to some organizational leaders . (Botan, 1996,p. 309)

Too Bleak for Ethics

The panopticon is a landscape too bleak for ethics. On many accounts, ethics only has a place where all parties can gain (Danielson, 1992; Rawls, 1971), or at least where there are multiple equilibria (Binmore, 1994), unlike this simple conflict model. This is our model s first lesson for the ethics of workplace surveillance: avoid overemphasizing power and conflict to the exclusion of the possibility of ethical appraisal and action. Rediscovering a role for agents choices helps make a beginning in the search for an ethics appropriate to the situation, even in the unpromising social landscape of pure conflict. But before we can do this, we need to move to a normative model that has room for mutual advantage. There we will see that there are real ethical choices to be made. This we do in the next section.

The Importance of the Strategic

The second lesson tempers the optimism of the first. Foucault s metaphor reminds us, even if only impressionistically, of the strategic aspects of surveil-lance. Typically, one uses surveillance to influence the other agent s behavior. But strategic interaction is complex; as we have seen, even in our simple model, Inspector doesn t simply observe and get the conformity he aims at. Those subject to surveillance can take countermeasures. Speed control radar is a familiar example of this process. Providing radar guns to the traffic police doesn t automatically generate greater compliance to speed limits; one finds drivers investing in new radar detection technology, leading to the need to improve or alter the original technology, higher costs, and law enforcement unfairly exempting those who buy detectors. The resulting situation may be quite different from the original one and may offer new opportunities.

Ideally, we would like our model to capture the situation s main dynamics. In the simple example of the highly structured prison environment, perhaps conforming is the only available strategy. In more open situations, however, countermeasures can be quite innovative and surprising.