The Power Debate by Joseph M. Ellis

Against Empiricism: The ‘Enduring Nature’ of the Power Relationship

Introduction

The power debate dominated political science for much of the 1950s, 1960s and 1970s. Coinciding with this debate was the influence of more positivistic political science, particularly the behavioral revolution and economic models of political rationality that were at the forefront of the discipline. These new models shaped the nature of the debate. Those that supported positive political science —Robert Dahl, Peter Bachrach, Morton Baratz, and Steven Lukes— painted the power argument with their own methodological brush.[1] Others – Jeffrey Isaac in particular — reacted against these more “scientific” arguments from a structuralist perspective. In this paper, I seek to flesh this debate out and describe in detail the various conceptions of power put forth by Dahl, Bachrach and Baratz, Lukes and Isaac, as well as analyzing their different methodological approaches. Although, as David Ricci critiqued, “nothing definitive emerged from the field of community power,” the discussion proved to be a powerful teaching tool about the competing methodologies of the discipline (Ricci 1984: 273). More importantly, I argue that Isaac’s structural approach best explains power, especially his conception of the “enduring nature” of it. I suggest that the first three faces of power were chock full of relaying various details on how power was exercised, but did little in the way of providing a comprehensive understanding of how power was possessed or how it endured.

The Faces of Power: From Dahl to Lukes

Some of the initial views of power stemmed from the early ruling-elite models of power. Advanced most notably by C. Wright Mills, the elite theory highlighted the undemocratic nature of power, whereas prominent citizens in the community shaped governmental influence: bankers, party machines, businessmen and those in City Hall. One of the main problems with the ruling-elite model, as Dahl explained, was its emphasis on overt and covert spheres of influence, making it almost impossible to prove (Dahl 1958: 463). On the one hand, some forms of power were observable. On the other hand, if power was covert and unobservable, this posed a problem for a maturing discipline interested in observation and positivism. Moreover, this elite model countered pluralist ideas that viewed government as an interactive process among various groups. The elitist model was undemocratic. Dahl hoped to prove the democratic aspects of the system.

In Dahl’s, Who Governs? a classic tome on New Haven, Connecticut’s pluralist politics, he observed that power is not dominated by one group. Instead, when asking rhetorically, “Who governs?” Dahl answers that it is a multitude of groups. While some groups have certain ‘direct’ influence, politicians for example, others have a degree of ‘indirect’ influence, such as constituents. Since, “only a small number of persons have much ‘direct’ influence,” and even this influence was constrained by the electorate, to suggest that power existed in some conspiratorial, covert manner seemed odd (Dahl 1961: 91). The bourgeoning behavioral movement also weighed on Dahl’s mind. Concerned as he was with observation and positivist political science, the elite model proved an unfruitful way to analyze variables, much less power. Dahl’s formula for power, how A makes decisions that effect B, or how one group makes decisions that affect another, became known as a first face view of power.

This first face — how overt decisions were observed — seemed to be a great improvement on the elitist model, especially for those interested in empiricism in the discipline. However as Bachrach and Baratz argued, this ‘first face’ argument had its limitations. Because behavioral methodologists were interested in what could be observed, Dahl’s notion left little room for the types of power that were unobservable. Although “specific and visible” power was important and better than the “pure speculation” of other models, Bachrach and Baratz asserted that there was a ‘second face’ of power (Bachrach and Baratz 169). To discard immeasurable elements in the power discussion was great folly and missed one key component. What E.E. Schattschneider called “limiting the scope of conflict,” Bachrach and Baratz noted that power was not only A making decisions that affected B, but also A making decisions which reinforced preconceived social and political values and institutional practices that limited or narrowed the scope of the debate. As they note: “To the extent that A succeeds in doing this, B is prevented, for all practical purposes, from bringing to the fore any issues that might in their resolution be seriously detrimental to A’s set of preferences” (Bachrach and Baratz 170).

The two face nature of Bachrach and Baratz stemmed from the idea that power involved both decision-making and nondecision-making. Although some decisions were potentially suppressed and certain others deemed unobservable, conflict itself could be observed, and this, for Bachrach and Baratz, made observation important. While this second face of power seemed to be an improvement on its pluralist predecessor, Bachrach and Baratz did not discard the empiricist approach altogether. Rather, they broadened their methodological lens to include those “immeasurable” events related to power. This could include power struggles that are not apparent to the human eye, or, potential conflict and interaction that limited one agent or another.

Lukes, however, argued that this ‘second face’ suffered from similar problems as the ‘first face.’ In a more thoroughgoing critique of behavioralism, Lukes noted that decision-making and nondecision-making still relied on a supposition that power, at least the aspects of it related to conflict, could be observed. In his three-dimensional view, he casts off the empirical notion that power was observable. For Lukes, A exercised power over B when A effected B in such a way that negated B’s preferences (Lukes 34). This conception of power also became what Lukes described as a‘radical’ notion. A self-described Marxist, Lukes was convinced that power was conceptualized in value-laden ways. The liberal and reformist, for instance, witnessed power differently than the radical, who saw man’s wants and interests constrained systemically.

While it probably goes without saying, Lukes is right to argue that each conception of power was “value-laden” in some way. This was, as it appears to be, a more underhanded critique of the positivist, “value-neutral” literature than anything else. If political science is about constructing scientific models, not about seeking normative ends, then the positivists failed. What should be clear at least is that methodology greatly affects a political scientist’s description of something as seemingly simple as the term power. Rebelling against the hunches of elite-theorists, Dahl promoted a pluralist, empiricist view of power. In what Lukes described as a “qualified critique” of the earlier method, Bachrach and Baratz asserted that there was in fact a second face of power, arguing that Dahl was too concerned with observation and not aware of the potential conflict outside of the scope (Lukes 25). Lastly, Lukes discarded the behavioral methodology and applied his own ‘radical’ conception of power.[2]

Isaac’s Critique and Power’s ‘Enduring Nature’

Although it appeared that the discipline fleshed out the power debate in all conceivable directions, Isaac proved that the debate had been cursed by a “misconception that the purpose of social science [was] to document empirical regularities” (Isaac 4). In other words, empiricist political science missed the boat. Because observation details the exercise of power, and not how it is possessed or how it endures, the empiricist approach did nothing more than collect instances of “A affecting B” in some way. But what did this tell us about power? Dahl’s first face argument merely examined what one agent did to another. His case study of New Haven was rich in describing how Mayor Lee’s power was constrained “indirectly” by the constituency or other pressure groups, but said little about the “srtuctural determinants” inherent in these decisions (Isaac 15). Bachrach and Baratz improved on Dahl’s model with their insights on the scope of conflict and mobilization biases, but still behavioral in the sense that observing conflict was crucial to understanding power (Isaac 10). Lukes ‘radical’ theory argued that power existed in the absence of empirical conflict, hoping to trump the behavioral obsession with observation. Nevertheless, Lukes still relied on “behavioral regularities” to explain power (Isaac 13).

Isaac was not only interested in reexamining the ramifications of empiricism, however. Seeking to advance a new structural, realist argument of power, Isaac’s model definitively explains and clears up previous misconceptions empiricists had. To do so, he argued from a realist perspective. Realist theorists question observation because of its interpretive nature. For realists, causation and empirical regularities were certainly open to debate and to declare such analysis “scientific” seemed almost fraudulent. Isaac suggests that power is not just causation between agents A and B. Rather, causation is a red herring, leading the social scientist onto the wrong trail. What was more important to the realist was the “causal mechanisms” that operated in unpredictable and undetermined ways (Isaac 17). To clarify somewhat, realists do not just take at face value what is observed but understand that certain laws, customs and structures produce varied results. For example, the power of Mayor Lee in New Haven could be very different if he was in, say, Atlanta.[3]

Moreover, as Isaac noted, “the primary object of theoretical analysis [should] not be behavioral regularities, but the enduring social relationship that structures them” (Isaac 20). Although Isaac was not interested in methodological battles,[4] bringing structure back into the argument was an improvement on his behavioral counterparts. The structure of which Isaac speaks, however, is very different from overly deterministic Marxism or crude institutionalists. He is more interested in preconceived relational structures, like for instance, the teacher-student relationship or husband-wife relationship. Indeed, the essence of power for Isaac is how social agents maintain power based solely on the relations in which they participate (Isaac 22). In some respects, power is based on ritual and myth rather than causation in the sense Dahl or Bachrach and Baratz presume. Power survives when these relationships endure.[5]

John Meyer and Brian Rowan propose a good example of this in their argument about the perceived legitimacy of organizations and institutions. Rejecting the causation and empiricism of prior conceptualizations of legitimacy, Meyer and Rowan argue that legitimacy is merely socially structured (Meyer and Rowan 21). Although certain institutions, schools for example, could be terribly inefficient and costly, taxpayers continue to support them because laws, customs, and rituals reinforce their importance. As long as this legitimacy is perceived, the relationship between taxpayer and institution endures.

One of the earliest models of power used a similar methodology. Not concerned with empiricism or even advancing a definition of power itself, Floyd Hunter asked prominent citizens in Atlanta to rank the forty most powerful people in the city (Hunter as cited in Ricci 266). Consistently the same people were proposed by each respondent, indicating that reputation alone might be a good indicator of power. If you think someone has power, they must! Of course, the survey was blasted as “worthless” by quantitatively-savvy scholars and a crop of young behaviorists entering the field. But Hunter as well as Meyer and Rowan share something in common with Isaac. One, that reputation and ritual alone can explain aspects of a concept. Observing power and perceiving power might be two different things, but both can have similar and lasting effects. Two, getting bogged down in defining something may not be as important as understanding the structure in which the concept exists. Meyer and Rowan offered no procedural definition for what a legitimate organization was, other than to indicate that legitimate organizations endured and illegitimate organizations failed. Hunter was not interested in defining power as much as he was interested in the hierarchical structure the respondents created. Similarly, Isaac’s definition was modest and very broad at best. On the one hand, the agents involved were important, but the structure of the relationship is what mattered most.

Isaac’s conception of power embodied more than just how the relationship endured, however. Like any good Marxist, Isaac was concerned with aspects of domination (power to) and subordination (power over). Exploring both ‘power to’ and ‘power over’ relations was a critique of behavioral methodology that failed to recognize the relations of the social agents, but this was also a reflection of Isaac’s own methodological bias in favor of Marxist structuralism.[6] Isaac’s power definition was normative in way that Dahl’s, Bachrach and Baratz’s, and Lukes’ was not. One gets the impression that power is, even in seemingly harmless cases like a teacher-student relationship, somewhat oppressive. Like the structure of the capitalist state with its bourgeoisie and proletariat, power indeed endures and replicates itself, but often times in the most dominating fashion.

Conclusion

The power debate is a good example how methodology effects the lens through which research is done. Dahl’s response to the undemocratic elite model was to advance a more democratic, pluralist conception helped out by the new positivist techniques of the behavioral revolution. Bachrach and Baratz in their “qualified critique” of behavioralism and pluralist politics argued that all power was not observable. By mobilizing some groups into the conflict and others out, not all instances of power were so overt. Lukes tried to discard the behavioral approach all-together with his ‘radical’ view that power existed anytime one agent’s interests were subverted by another, a Marxist-type argument. In any case, as Isaac would show, all of the conceptions were influenced by “behavioral regularities” and the empiricism of the day. One does not have to be a Marxist to agree with Isaac’s conclusions that power exists in the social fabric of society. To explain power, therefore, one may not have to agree that it is about ‘subordination’ and ‘domination’ as much as it is about laws, customs, and rituals that reinforce the structure of the relationships and cause them to endure. Isaac’s critique of the “empiricism” in the discipline is both important and on the mark. Rather than just noting instances and examples of power in action, Isaac created a useful model to understand its ever important ‘enduring nature’.

Works Cited

Bachrach, Peter and Morton Baratz. “The Two Faces of Power.” The Search for

Community Power. pg 167-177.

Dahl, Robert. “Selections from ‘Who Governs?’” The Search for Community Power.

Eds. Hawley and Writ. pg 87-107

Dahl, Robert. “A Critique of the Ruling-Elite Model.” American Political Science

Review. 52: 463-469, 1958.

Hunter, Floyd. Community Power Structure. New York: Doubleday, 1953.

Isaac, Jeffrey C. “Beyond the Three Faces of Power: A Realist Critique.” Polity. V. 10

N. 1. Fall 1997.

Lukes, Stephen. Power: A Radical View. London: Macmillan, 1974.

Meyer, John W. and Brian Rowan. Organization Environments: Ritual and

Rationality. Ed. Meyer and Scott. Newbury Park, CA: Sage Publications, 1992.

Ricci, David. The Tragedy of Political Science. New Haven, CT: Yale University Press,

1984.

---

[1] Certainly saying they supported “positive” political science is making a sweeping generalization. However, each author, as will be shown, was interested in various “behavioral regularities.”

[2] Lukes admits that his use of the term “behavioral” is very narrow. Of course, even his three-dimensional view has a behavioral component to it in the fact that behavior produces evidence.

[3] This, of course, seems obvious, but the behaviorists paid little attention to this detail.

[4] As he suggests: “This approach need not result in a form of hyper-determinism which reifies social structure” (Isaac 19).

[5] Isaac notes, however, that the relationship itself has the potential to break down when one of the social agents loses their authority. For example, this could happen if it was shown that the teacher lost control of the classroom.

[6] In Isaac’s defense, he seems to be a soft on Marxism, more content with criticizing empiricism than trumpeting determinism.