The Public Policy of Skepticism

Steven Teles; Matthew Kaliner

doi:10.1017/S1537592704000611

The Public Policy of Skepticism

Published online by Cambridge University Press: 10 June 2004

Steven Teles and

Matthew Kaliner

Show author details

Steven Teles: Affiliation:
Is an assistant professor of politics at Brandeis University (teles@brandeis.edu)
Matthew Kaliner: Affiliation:
Is a senior research assistant at the Murray Research Center, Radcliffe Institute of Advanced Study (kaliner@radcliffe.edu)

Article contents

Abstract
The Politics of Faith and Skepticism
Unintended Consequences
Policy Implications
Exceptions
Conclusion
References

Rights & Permissions

Abstract

Drawing on the thought of political philosopher Michael Oakeshott, we present a framework for thinking about public policy rooted in a balancing of what Oakeshott called the poles of “faith” and “skepticism.” This approach takes seriously the maintenance of institutional forms in policy making, seeks to preserve order while maintaining social diversity, is sensitive to the likelihood of unintended consequences, and eschews the attempt to produce ideal end states, emphasizing instead remedying specific defects in existing arrangements. While it has some affinities with libertarianism and modern conservatism, this approach places less emphasis on limiting the scope of governmental interventions and more on shaping their character. Specific case studies of education, public health, and Social Security flesh out this argument, demonstrating that it differs from contemporary conservatism, libertarianism, and liberalism.

Information

Type: Articles
Information: Perspectives on Politics , Volume 2 , Issue 1 , March 2004 , pp. 39 - 53

DOI: https://doi.org/10.1017/S1537592704000611 [Opens in a new window]
Copyright: 2004 by the American Political Science Association

The battle for Michael Oakeshott is on. On the left, Adam Gopnik of The New Yorker claims Bill Clinton's “lack of consistent principle and a ‘bottom line’” rendered him the very model of an Oakeshottian statesman.¹

Gopnik 1996, 196.

On the right, Andrew Sullivan argues: “On Social Security, [Oakeshott] would prefer to maximize the wisdom of millions, by giving people more control over their investments, rather than buttress a one-size-fits-all solution devised decades ago.”²

Sullivan 2001, 6.

Traditionally, Oakeshott is associated with the political right, and not without reason. During his lifetime, he was out of sympathy with many programs of leftist governments and, more broadly, with the mind-set that drove them, which ranged from utopian to (at its most bloodless) technocratic. It would be easy, and there would be sources within Oakeshott's own writing, to suggest an overall hostility to the welfare state and a sympathy with libertarianism—a term he himself some-times used.

Despite Oakeshott's reputation for libertarianism, we ask what advice he might have provided nonlibertarians who believe that there is legitimate room for a redistributive and regulatory welfare state.³

We should note early on that this paper has been written to be comprehensible both by those concerned with concrete matters of public policy and by experts in Oakeshottiana. The former may find the discussion of particular policy issues too superficial, and the lat ter may think we fail to plumb the depths of Oakeshott's thought. This paper is intentionally a “mongrel,” as Oakeshott might have called it, and we hope that it is read in that spirit.

Oakeshott's thought, like that of all serious thinkers, has a value that transcends ideology. In this paper, we utilize it as a corrective even to projects he might have viewed with disfavor.

In brief, the key components of our Oakeshottian approach are as follows. First, we place as much emphasis on the institutions and processes of policy making as on its ends. Second, with respect to ends, we hold that the ultimate goal of government is to maintain order while preserving social diversity and complexity. Third, we emphasize the ubiquity of unintended consequences, regardless of whether the stated goal is to shrink or expand the state. Finally, and as a result, we do not attempt to produce imagined, ideal end states, recognizing that such action tends to devalue procedure, threatens social diversity, and is particularly prone to unintended consequences. Instead, by seeking to facilitate spontaneous, independent social action, an Oakeshottian approach focuses on remedying specific defects in present arrangements in ways consistent with existing institutions and practices.

In building this framework for analyzing public policy, we draw in particular upon the distinction that Oakeshott made in one of his posthumously published books, which we restate as the public policy of faith and skepticism.⁴

The Politics of Faith and the Politics of Skepticism is not one of Oakeshott's best-known works, and so our heavy use of it in this article might be controversial. What is more, PFPS was not published in Oakeshott's lifetime—a fact that raises the question of whether Oakeshott intended that it be read as an accurate state ment of his beliefs. It is our judgment that this latter question is not terribly relevant, because the application of political theory depends not on what the author thought of the text, but on what use we can make of it. Additionally, there is a good reason to use this text rather than some of his other works—especially On Human Conduct. While others may disagree, we find Oakeshott's distinction between faith and skepticism in PFPS to be somewhat clearer and easier to use than the similar distinction he makes in On Human Conduct between civil and enterprise associations. Finally, we believe that the primacy we give the text is justified by its being, in the words of Timothy Fuller (who, along with Shirley Letwin, was Oakeshott's literary ex ecutor), “as close to a book of advice for the practice of modern politics as Oakeshott ever produced.” Oakeshott 1996, x.

The Politics of Faith and Skepticism

For Oakeshott, the politics of faith and the politics of skepticism represent the magnetic poles of modern politics. He does not describe these two poles as choices to be made—as though political action would be improved if, for example, the politics of skepticism were to defeat the politics of faith. On the contrary, he argues that “each is not less the partner than the opponent of the other; each stands in need of the other to rescue it from self-destruction, and if either succeeded in destroying the other, it would discover that, in the same act, it had destroyed itself.”⁵

Ibid., 92.

Crucial to any application of Oakeshott's thought to public policy is his belief that the proper relationship between the politics of faith and the politics of skepticism is tension.

The central objective of the politics of faith is the ideal of human perfection, not to be sought in the afterlife, but “to be achieved by human effort.” Human perfection is not simply an individual human striving, for “the chief agent of the improvement, which is to culminate in perfection, is government. Thus it comes about that the activity of governing is understood as the control and organization of human activity for the purpose of achieving human perfection.”⁶

Ibid., 23, 24.

Human perfection is thus both definite and unitary, and the public interest inheres in the eradication of alternative visions of the human good, which are understood to be immoral because untrue. Achieved by planning, the essential act of statesmanship is the divining of truth; it is the discovery and enforcement of comprehensive, predetermined outcomes, arrived at by reason apart from—and “uncorrupted” by—life as it is actually lived. It is utopian precisely because its preferred state of affairs is derived from logic, religious doctrine, aesthetics, science—that is, anything other than an extrapolation from what already exists.

The politics of faith judges rules, institutions, processes, and procedures purely by their perceived ability to produce the desired outcomes. As Oakeshott notes, “[I]t would appear that a high degree of formality in the activity of government would be out of character. Governing in this style is a godlike adventure, and a nice observance of rules and constitutions will readily be felt to hinder its impetus.” The politics of faith is one of constant improvisation and expansion, since the range of objects necessary to reach the desired outcome constantly expands in the face of negative feedback from previous interventions. As a result, “the politics of faith understands governing as an endlessly proliferating activity, integrating all the activities of the subject and (if it is faithful to its charge) always at the end of its tether.”⁷

Ibid., 29, 28.

This relentless expansion is the most visible characteristic of the politics of faith. The conflict between this vision and a recalcitrant society dictates that new activities be drawn into the ambit of government. Defeated at one level of intrusiveness, it increases its intensity of control and the scope of its planning, only to find that its new interventions create effects that themselves must be compensated for, generating yet further opportunities for growth. As a consequence, “in the politics of faith, governing is the minute and comprehensive control of all activities.”⁸

Ibid., 93.

The politics of faith is therefore propelled toward an anti-institutional governing mode, in which all political parties, the separation of powers, and the rule of law become inconvenient annoyances, and politics is reduced to administration.

The politics of faith is anti-institutional also because it arises from, and reinforces, a state of affairs in which plurality of interests is highly attenuated, thereby reducing the need for, and requiring the elimination of, mediating institutions. This state is one characterized by “the gradual disappearance of the intermediate authorities which had formerly stood between a then weak central government and the subjects, leaving them naked before a power which in its magnitude was becoming comparable to a force of nature.”⁹

Ibid., 49.

In this vision, a multiplicity of governments within a political space—and a variety of institutions, groups, and organizations—is seen not as a sign of social vigor and health, but as a series of obstacles to the imposition of patterns from above. Formal institutions are designed precisely to protect social complexity, to require that certain interests be considered and reasons heard so that some degree of continuity among past, present, and future action can be maintained. The politics of faith, however, encourages formlessness so as to facilitate rapidity of action and an absence of constraint by vested interests.¹⁰

Ironically, this formlessness actually makes institutions more easily controlled by vested interests—this was the point of Lowi 1979.

Finally, the politics of faith is aimed at a society whose parts are consistent with a whole. “All human activity should be integrated,”¹¹

Oakeshott 1996, 63.

meaning that it should be rendered visible and capable of manipulation from the center, so that it can be made to reflect a pattern or ideal. It is this aesthetic preoccupation with patterned consistency that James C. Scott has called “seeing like a state.”¹²

Scott 1998. A classic case: the urban renewal plans of the 1950s–1970s that destroyed existing neighborhoods and substituted in their stead buildings that looked “right.”

Beauty inheres not in the details of life, but in the place those details have within a larger schema, one comprehensible only from the planner's standpoint. What is more, that vision is basically static: “everything in this style of government is essentially built to last…. [W]hat is unshakable is preferred to what is recognized as ephemeral.”¹³

Oakeshott 1996, 96.

The resulting mind-set claims to be modern and yet is hostile to the spontaneity and decentralized change that are modernity's hallmarks.

This planner's-eye view runs contrary to a second feature of modernity: “the characteristic of the communities of modern Europe is the multiplicity of their activities and directions of activity”—a multiplicity that gives the planner's insistence on uniformity an “incongruous appearance.” The consequence of this conflict of visions is that the politics of faith “demands an exclusive direction from subjects who do not readily recognize the relevance of the demand.” Faced with such a conflict, governance generates resistance and incomprehension, spawning popular cynicism and alienation from government. This leads to an intensification of force (and a consequent increase in the fervor of supporters) on the part of government at the same time that its ability to legitimate itself wanes. Uncorrected, governance in this mode tends to write its own epitaph.¹⁴

Ibid., 95, 97.

The alternative pole of modern politics—the politics of skepticism—is the one that Oakeshott associates himself with. But it would be a mistake, as we argued earlier, to extrapolate from this a pure choice for skepticism over faith. As Oakeshott observes early on in The Politics of Faith and the Politics of Skepticism, “[I]f there is any conclusion I wish particularly to avoid, it is the fruitless conclusion that a virtuous politics would seek simplicity and ‘shun ambiguous alloy,’ that what we ought to aim at is a resolution of the ambivalence and ambiguity of our politics or at least a formula under which they can be vanquished [italics added].”¹⁵

Ibid., 20.

The art of statesmanship, within the skeptical tradition, is to preserve enough faith to prevent the pathologies of skepticism while avoiding an overabundance of faith. Oakeshott's statesman is a “trimmer” who keeps the ship afloat between the crosswinds of faith and skepticism—meaning that even a policy maker of skeptical tendencies might need recourse from the instruments of faith.

With this background in mind, what are the characteristics of the politics of skepticism? Drawing on Hobbes (as Oakeshott did in all of his writings), we can view politics as a mode of action concerned neither with the specification of a particular vision of perfection nor with the application of truth. Instead, “its chief office is to lessen the severity of conflict by reducing the occasions of it.”¹⁶

Ibid., 32.

The special character of politics inheres in maintaining order while preserving diversity. To facilitate this goal, the potential power of government must be unlimited—since the challenges to order are potentially limitless—but the actual use of power is closely controlled. This approach is quite different from that of adherents to the politics of faith, who believe, to use the words of Lyndon Johnson, that “having the power, we have the duty.”¹⁷

Johnson 1964.

Government is always, in Oakeshott's phrase, “at its tether.” Against this tendency, “it is clearly in character for the skeptic to consider economy in the use of power in governing.”¹⁸

Oakeshott 1996, 33.

Just as the practitioner of the politics of faith is constantly looking for—or at least propelled toward—expansion of his activities, the essence of skepticism is to avoid new commitments and to limit the obligations that any particular decision might require. The skeptical statesman conserves resources—ensuring that his potential power is unlimited while avoiding open-ended engagements over which he will have no control. He is ceaselessly aware of the possibility of mission creep, the inexorable ramping up of obligations as a consequence of overly broad initial commitments.¹⁹

Mission creep is our term, not Oakeshott's. The term was originally designed to describe how humanitarian relief can slowly melt into peacekeeping, which may soon turn into outright war making. The so-called Powell Doctrine was a not entirely satisfactory attempt to cope with this problem.

The skeptical statesman avoids mission creep in two ways. First, he explicitly shuns sweeping goals for social transformation based on a preferred pattern of social conduct. Whereas the statesman of faith has a preference for idealized patterns, the skeptical statesman is “suspicious of a great love of symmetry and an overbearingly eager impulse to abolish anomalies.” On the contrary, he recognizes that “for ordinary use a mongrel [that is, a creature that internalizes anomalies] is recognized to have some advantages: it is more adaptable to circumstances and more fertile than a thoroughbred, and it takes less looking after.”²⁰

Oakeshott 1996, 35, 10.

The absence of an aesthetic preference for tidiness is itself an instrument in the economy of power—this allows the skeptical statesman to distinguish between problems that can be allowed to sort themselves out and ones that require the intervention of state power. Because he does not seek to impose a pattern, deviations from a pattern do not call for the depletion of his resources.

This lack of an aesthetic standard of order is itself derived from an understanding of the character of modernity. Unlike the politics of faith, which understands the central feature of modernity as the capacity of government to engineer social outcomes on the basis of centrally determined plans, the politics of skepticism sees the essential features of modernity as complexity and diversity. Modernity unleashes a wide range of social projects, groups, and interests, and the skeptical politician sees this phenomenon as a condition to be conserved and facilitated, not as an obstacle to be overcome. Whereas the politics of faith threatens to “impoverish mankind by reducing all activity to that which could be easily controlled by government”²¹

Ibid., 74.

—what Scott calls rendering society “legible”—the politics of skepticism does not pretend that society as a whole can, or should, ever be susceptible to comprehensive apprehension. As in The Federalist Papers, “[T]he regulation of these various and interfering interests”—the complexity that Oakeshott identifies with modernity—“forms the principal task of modern legislation.”²²

Hamilton et al. 1961 [1788], No. 10, 79.

That is, the statesman acts not by rendering the social landscape flatter in order to make it more susceptible to his control (a task that would constantly keep his power “at its tether”), but by attempting to remedy specific defects of order as they arise.

Another way the skeptical statesman exercises an economy of power is through formalism. Oakeshott associates formalism with the impersonality of office, for the modern statesman acts not in his own name but in the name of the state, limited by the definition of the office he holds.²³

Oakeshott 1996, 75; and see Mansfield 1989.

This association of skepticism and formalism leads Oakeshott to call the U.S. Constitution “the most profoundly skeptical constitution of the modern world.” He says that formalism and skepticism produce “the conduct of government (that is, the maintenance of order) by means of known and settled laws and a system of rights which is bound up with (which, indeed, is historically derived from) known and easily operated means of redress.”²⁴

Oakeshott 1996, 80.

This formalism is both consequence and cause of the more modest objectives that the skeptical mode of government imposes. It is a consequence because remedying specific defects and maintaining a tolerable level of order do not entail the ever-expanding need for power—or the need for a form-less administrative politics—that results from open-ended commitments to social reconstruction. It is a cause of those modest objectives because the “conduct of government…by means of known and settled laws” makes impossible the comprehensive administrative control necessitated by planning.

That is not to say that the politics of skepticism implies or necessitates libertarianism—a charge that Oakeshott goes out of his way to deny²⁵

He puts this point in another way at the end of the book, when he observes: “The opposite of current ‘communism’ is not ‘capitalism’ (which is not a manner of government at all) but ‘skepticism.’” Ibid., 38.

—any more than it totally precludes a role for the politics of faith. The critical question is, what constitutes a problematic condition worthy of state intervention—and how should the state intervene? In a telling passage, worth quoting at length, Oakeshott observes that

protection against some of the vicissitudes of fortune is recognized to be among the activities of government. Here the inspiration is the observation of actual miseries suffered; and “security” is understood as the assurance of relief. Nevertheless, the range of this assurance is not determined by the magnitude of the misery, but by a perception of the displacements consequent upon its removal. Any “protection” involves government taking charge of some of the activities of the subject; but the limit here is the “protection” which can be supplied without imposing a comprehensive pattern of activity upon the community. When a man is defended against misfortune in such a way as to deprive him of the authority to defend himself, the limit is passed [emphasis added].²⁶

Ibid., 99–100.

In contrast to libertarianism, the politics of skepticism accepts the necessity of social intervention; in contrast to the politics of faith, it intervenes only in response to “actual miseries”—problems that are visible defects in present arrangements, not simply deviations from a putative ideal. Poverty, unclean air, traffic jams, and crime are all policy problems that can be detected without recourse to an alternative overall scheme of social organization, but envisioning a society without them requires precisely such an alternative.

From a libertarian point of view, even a readily acknowledged misery may not justify public action. From a politics of faith perspective, it not only justifies action but demands nothing less than total victory. A politics of skepticism finds that an acknowledged misery provides a rationale and direction for state action, but that victories are more easily imagined than won and can come at too high a price. Its warrant for action is limited by the disruptions that any intervention will cause—which, as the following section of this article will demonstrate, may be substantial—and by the diminution in the capacity for spontaneous action that an intervention may bring about. The skeptical politician will thus prefer to respond to actual miseries by facilitating social action rather than substituting for it. He will try to determine why spontaneous action has failed to remedy the misery and seek to increase the capacity of social actors to respond on their own. In this way, the skeptical politician attempts to remedy social ills while maintaining the society's capacity for undirected, spontaneous action to respond to future miseries—and calls for government to respond, if possible, in such a manner as to increase that capacity. This presents a substantial agenda for government and at least the outlines of a skeptical welfare state.

As we suggested earlier, although Oakeshott's sympathies are clearly with the politics of skepticism, he recognizes this mode's inherent limits. Understanding these limits, and the interaction between the politics of faith and the politics of skepticism, will help us pull together Oakeshott's insights regarding the actual conduct of politics and prepare us for his second set of lessons, about unintended consequences.

Consider the relationship between the American Declaration of Independence and the Constitution. Abraham Lincoln once observed that the Declaration was an “apple of gold in a frame of silver,” the silver frame being the Constitution. The Declaration pronounces high principles, such as equality and liberty—which by themselves could lead to a moralistic and informal politics. However, the Constitution sets out a governance of dispirited formalism, lacking in moral seriousness and tending toward a relativistic politics of interests. With the apple/frame metaphor, Lincoln asserted that the Declaration and the Constitution complete each other: “The picture was made, not to conceal, or destroy the apple, but to adorn and preserve it [italics in original].”²⁷

Lincoln 1953, 168.

It is this sort of political reasoning that, in the later parts of The Politics of Faith and the Politics of Skepticism, Oakeshott holds up as a desirable model for public policy making. This reminds us of his earlier admiration for the “mongrel.” Purebreds have a certain beauty, but years of inbreeding tend to make them weak in all but the characteristics bred for and therefore incapable of adapting to new circumstances and environments. The purer they become, the more susceptible they are rendered to sudden shocks or even gradual changes in the field of play.

For Oakeshott, each pole of modern politics has its “nemesis”—a set of pathologies that is revealed when either style is allowed to rule independent of the other.²⁸

Oakeshott 1996.

The chief nemesis of the politics of faith is its tendency to cease to be a form of politics at all: “In short, government in the service of ‘perfection’ appears not as a style of politics, but as a manner of abolishing politics.” By “politics,” Oakeshott means the conciliation of interests and the balancing of the past, the present, and the future. When the future becomes the only touchstone, what man and society are promised to become can overwhelm what they are. Since politics always starts with men as they are—with tangible commitments, values, beliefs, and prejudices—exclusive commitment to a future pattern places government in an adversarial relationship with its own people, as politics gradually is transformed into evangelism: “The enemies of the regime will be identified not as mere dissidents to be inhibited, but as unbelievers to be converted.” It is for this reason that “gratitude is turned into hatred,”²⁹

Ibid., 94, 97.

and the politics of faith inevitably creates an ever larger army of enemies.

Another nemesis of the politics of faith is its tendency to overestimate its own intelligence, to construct edifices built for eternity that are eaten away or rendered irrelevant by the ubiquity of change that characterizes modernity. Both nemeses render the politics of faith susceptible to unpleasant surprises. In one of his few precise criticisms of an actual public policy, Oakeshott observes that the British “Education Act of 1944 is deeply veined with the politics of faith, and some of its provisions are being riveted upon us by school buildings which are exclusively appropriate to them.” Oakeshott is alarmed by the possibility of reforms that—even if generally well meaning and intelligent—preclude the possibility of tactical adjustment or even strategic retreat. Even as faith's grandiosity of purpose calls for large displacements that require constant adaptation and expansion, its actual institutions—because purpose-built and not adaptable—make such reworking impossible. “It requires what it cannot command, and needs what its character prohibits.”³⁰

Ibid., 96, 99.

The politics of faith produces unintended consequences that it can neither comprehend nor respond to, a subject that we will attend to at greater length in the next section of this article.

The politics of skepticism has its own nemeses: excessive formalism and fetishizing of procedure. Its great ends are the maintenance of order and the preservation of social complexity and diversity. Skeptical politics provides the pre-conditions for progress but does not itself produce change or attempt to direct it. Because change is its goal but not its practice, the politics of skepticism is likely to be taken unawares:

In short, in the absence of a larger enterprise, the skeptical office of keeping the system of rights and duties relevant to the current activities which compose the community may be expected to be sluggishly performed. Without the pull exerted by faith, without the “perfectionism” which we have seen to be both an illusion and a dangerous illusion, itself evoking a nemesis, government in the skeptical style is likely to be overtaken by a nemesis of political quietism.³¹

Ibid., 107–8.

While the politics of faith is forever warning of impending crises and emergencies, calling the people to “the moral equivalent of war,” the politics of skepticism often ignores real dangers to the well-being of the community. What is more, because it focuses exclusively on present evils, “it responds slowly to inspired imaginings of perils to come.”³²

Ibid., 109.

Weary of predictions that the sky will fall tomorrow, it fails to take reasonable precautions against legitimate future dangers. Even when it acknowledges present evils or future threats, it often lacks the inspirational qualities and capacity for emotional mobilization required to address them.

Skepticism's most dangerous weakness is its tendency to fetishize “existing means of redress” and institutional arrangements. At an extreme, this can create a dangerous political inwardness. When politics is reduced to mere procedure, it takes on the character of play—an appreciation of the activity for its own rewards and a consequent lack of seriousness. Once politics becomes a game, it is of interest only to the players; it loses the attention of its subjects, not to mention their willingness to contribute to its success. Great acts of statesmanship can be achieved only when faith's sense of high purpose corrects skepticism's penchant for ingrained passivity, because some problems are sufficiently severe to require innovation in existing procedures and norms of action—or different ones altogether.

Unintended Consequences

If this mongrelism is the overall manner of government that Oakeshott prescribes, there is still one large point we must elaborate on before attending to the argument's implications for public policy. Even if government may legitimately respond to “actual miseries,” it must do so cognizant of the unpredictable displacements that its action will cause. These unintended consequences apply to all attempts at comprehensive change—whether from the left or the right.

Oakeshott's fundamental insight is that any reform or program aimed toward “progress” or “improvement” necessarily brings with it many changes that are quite different from this initial aim. And such changes—no matter how well thought out and rigorously managed the plan—can swamp the expected outcomes, especially in large or visionary projects. To quote at length a famous passage from Oakeshott:

For, innovating is an activity which generates not only the “improvement” sought, but a new and complex situation of which this is only one of the components. The total change is always more extensive than the change designed; and the whole of what is entailed can neither be foreseen nor circumscribed. Thus, whenever there is innovation there is the certainty that the change will be greater than was intended, that there will be loss as well as gain and that the loss and the gain will not be equally distributed among the people affected; there is the chance that the benefits derived will be greater than those which were designed; and there is the risk that they will be offset by changes for the worse.³³

Oakeshott 1991a, 411. Oakeshott's point, while beau tifully put, was not wholly original. Consider Edmund Burke's observation that “[t]here is not, there never was, a principle of government under heaven that does not, in the very pursuit of the good it proposes, naturally and inevitably lead into some inconvenience which makes it absolutely necessary to counterwork and weaken the application of that first principle itself, and to abandon something of the extent of the advantage you proposed by it, in order to prevent also the inconveniences which have arisen from the instrument of all the good you had in view.” Burke 1949 [1780], 217.

Although Oakeshott expects the unexpected to be pervasive, he does not posit the source of unintended consequences in any specific phase in the planning or implementation of change. He is more interested in identifying how the skeptical statesman should react to the ubiquity of the unintended.

Such a statesman would tend toward inaction, since—despite having its own unintended consequences—inaction has the advantage of preserving the capacity for future intervention if problems do not sort themselves out on their own. As discussed earlier, he would take action only in “response to some specific defect, one designed to redress some specific disequilibrium”—in other words, to “actual miseries suffered.” Such definitions are susceptible to limitation and, therefore, “least likely to be corrupted by undesired and unmanageable consequences.”³⁴

Oakeshott 1991a, 412.

This allows policy makers to limit their commitments and thereby maximize their ability to maneuver, adjust, and, when necessary, disengage and retreat.

While Oakeshott was most concerned with the policy implications, rather than the cause, of unintended consequences, social scientists have typically reversed Oakeshott's priorities, stressing the prevalence of unintended consequences in social life but severely de-emphasizing their normative impact. Following Robert Merton's classic article “The Unanticipated Consequences of Purposive Social Action,” much analysis has avoided Oakeshott's normative judgments by positing that, in Merton's words, “undesired effects are not always undesirable effects.”³⁵

Merton 1936, 895.

Typically, unintended consequences are understood to emerge from poor decision making or, more generally, from an absence of relevant knowledge. In this understanding, harried or rushed policy makers fail to think through the consequences of their actions; and as a result, they do not consider things that they should have—and this leads to mistakes. For instance, Merton emphasizes the problem of ignorance, especially given “that the exigencies of practical life frequently compel us to act with some confidence even though it is manifest that the information on which we base our action is not complete.” The most devastating form of this error is the historical fallacy: assuming that past experiences can linearly predict future behavior. Merton also points to the danger of universalizing the particular: “procedures which have been successful in certain circumstances need not be so under any and all conditions” (italics in original).³⁶

Ibid., 899–901.

A similar error occurs when one ignores the feedback effects that derive from local conditions and assumes that a program implemented under different circumstances will continue to affect outcomes in similar ways.³⁷

Jervis 1997 demonstrates this point at length.

Unintended consequences in this vein are viewed as “avoidable errors,” reparable through greater care and knowledge.

While there is certainly something to this, we assume (with Oakeshott) that even policy makers with all available information, and with no prejudices or blinkers, will still make decisions that do not work out as expected. There are inherent limits to what any policy maker can know, especially when doing novel things or when change is sufficiently sweeping that it will create chains of responses that are impossible to realistically model. Unintended consequences are not a case of risk, which can be calculated ahead of time, but one of uncertainty—the pervasiveness of unknown and therefore incalculable occurrences.³⁸

For a fuller consideration of this distinction, see Knight 1965.

If we assume, with Oakeshott, that unintended consequences put us in a world of uncertainty, rather than risk, one important conclusion follows: the impossibility of fully calculating the costs and benefits of alternatives to existing policies and institutions. In fact, any attempt to do so is inherently biased against what is in place, since existing policies have already incorporated the “inefficiencies” that result from having to respond to unpredictable political, social, and environmental feedback.³⁹

A fuller version of this argument can be found in Teles 2000.

Thus, all theoretical alternatives should be subject to an “uncertainty discount” relative to the policies they might displace. This is not necessarily a counsel of quietism, since some existing policies are sufficiently counterproductive or incomplete that a theoretically attractive alternative will still be worth trying. But the uncertainty hurdle should be high enough to force policy makers to appreciate what reform will cost—and to be skeptical of what it will gain.

Unintended consequences apply equally to government expansion and contraction. A libertarian utopia with minimal government responsibility is still a utopia; it still requires the disruption of existing arrangements with consequences that are necessarily uncertain. To put it bluntly, as Oakeshott did in his critique of Friedrich von Hayek: “A plan to resist all planning may be better than its opposite, but it belongs to the same style of politics.”⁴⁰

Oakeshott 1991b, 26.

Policy Implications

In sum, Oakeshott's take on policy making can be divided into two large categories of critique. First, standard policy justification is insufficiently political: it lacks a defensible understanding of the province of politics, and it implies—and to some sense requires—a formless, anti-institutional mode of governance. Second, standard policy making requires a knowledge of the consequences of political action that is not accessible to policy makers, and it imposes a concentrated agency on government that leads to illimitable expansion in an effort to correct for the displacements that political actors cannot predict but (because of institutional formlessness) feel obligated to correct.

If one accepts the force of the preceding argument, what is the positive lesson for the conduct of public policy? We would suggest that constitutionalism (versus formlessness) and the solving of discrete problems (versus the pursuit of ideal end states) form the basis of a skeptical approach to policy making. This is not the same thing as saying that governance in the skeptical mode is necessarily smaller than in the mode of faith, or that it will always prefer to facilitate spontaneous action rather than deliver services directly. Dogmatism on either of these questions would run counter to the guiding spirit of skepticism. As we will see in the case of Social Security, economy in the use of power will sometimes require large but simple and limitable interventions instead of ones that are initially smaller but less easily controlled or that promise less intervention in the future. And it is the capacity to control an intervention that matters. We will briefly sketch out the differences between faith and skepticism when applied to specific policy controversies, and then proceed to a more theoretical conclusion.

Education

From the politics of faith perspective, education should foster certain broad outcomes, such as enabling every child to excel and increasing American economic competitiveness. In practice, these goals translate into increasing graduation rates, higher scores on standardized tests, and the like.⁴¹

For an example of the politics of faith in education, consider the recently passed No Child Left Behind Act, with its overwhelming emphasis on testing and its minute specification of pedagogy from the center. See www.ed.gov/legislation/ESEA02/pg1.html.

Virtually all policy analysis assumes that schools should be held responsible for meeting such goals. From the skeptical viewpoint, though, this approach invites certain problems: mission creep, institutional overstretch, and measurement myopia.

First and foremost, educational achievement is the joint product of the school, the student, the family, and the larger society. If schools are held responsible for outcomes as measured by individual student performance, eventually they must either (1) try to assume the responsibilities of these other actors or (2) be held responsible for outcomes over which they have little control, leading to inevitable disappointment, cynicism, and recurrent attempts to change institutional leadership or organization. This in turn leads to a creeping and theoretically illimitable expansion of schools' missions and their associated tasks, and a watering down of the traditional functions over which schools have some control.

Accompanying mission creep is the related phenomenon of institutional overstretch. Schools that have taken responsibility for noneducational inputs in order to achieve specified outcomes end up spreading scarce resources over a wide range of objectives. Schools are responsible for teaching not only the traditional core curriculum—reading, math, science, and so on—but also violence prevention, sex education, diversity appreciation, and more. Further, schools are also responsible for ensuring children's overall well-being—their nutrition, home life, and physical and emotional health. We make this point not to argue that schools should never teach beyond the basics or offer social services, but to suggest that the costs and benefits of each program, and of expecting more and more from the schools, need to be balanced. And time and money are only the most obvious expenses. Teachers' talents and administrators' leadership are diluted by the attempt to manage numerous tasks that demand competing, and sometimes contradictory, skills. Given such a wide range of objectives, it becomes impossible to communicate a coherent organizational ethos; so schools become incoherent, and both management and teaching take on a coping, rather than an inspirational, attitude. Asked to do everything, schools find it hard to do anything particularly well.

This mode of action is likely to be highly unstable, since it sets public goals that are intrinsically impossible for a school system to meet. Therefore, as Frederick Hess has argued, school authorities make token attempts to demonstrate that they are “doing something” to improve the quality of schools.⁴²

Hess 1998.

This typically results in what Hess calls policy churn: school authorities announce dramatic innovations that must take place before other changes can be effectively made. Policy churn is often accompanied by high turnover of school administrators, since firing superintendents is the easiest symbolic gesture of commitment and sympathy toward frustrated parents. But the turnover of policies and leadership ensures that nothing fundamental ever changes, because teachers (whose skills and practices are most critical to schools' success) learn to dismiss commands from leaders who probably will not stay long enough to implement even promising innovations. This, of course, dooms new innovations and leaders to fail, causing a recurrence of the cycle of failure, but at a deeper level of cynicism.

The flip side of institutional overstretch is measurement myopia—when authorities react to the failures of over-stretched institutions by “getting back to the basics,” most often by establishing a centralized curriculum and high-stakes testing. In such cases, the purpose is to get schools to achieve their goals by narrowly defining the purpose of education and forcing teachers and administrators to reduce their activities to those included in centrally determined tests. Certainly there are times when restricting tasks may be better than the alternative—teaching to the test is certainly better than teaching nothing coherent at all. But institutional understretch has its own pathology—eviscerating the core mission of education by defining it too narrowly—and this pathology is exacerbated by the fact that imposing such standards from above involves reconstructing, in the words of Diane Ravitch, “almost every aspect of education—including teacher education, testing of students and teachers, professional development, textbooks, preschool instruction, and financing of schools.”⁴³

Ravitch 2000, 1.

Such sweeping changes are necessary simply for putting standards in place, to say nothing of how the results of testing will actually be used to reform the educational system.

The central problem with this version of the public policy of faith is “teaching to the test”; testing is transformed into a goal in and of itself.⁴⁴

This phenomenon is quite widespread in other areas of public policy, such as “getting tough” on drugs. Police may devote their time to highly public drug busts—in the inner city, for instance, where dealing often occurs on street corners—and ignore dealing that is more effectively concealed, as in suburbs and on university campuses. One consequence is a racial imbalance in drug arrests and a related insensitivity to the price of drug enforcement in the most politically powerful parts of the community.

There are few high-stakes tests that are not to some degree “coachable.” In other words, the skills needed for passing can be transmitted through facility with the test itself, rather than with the body of knowledge that the test was designed to measure. More broadly, testing on a mass, impersonal scale can measure only a portion of the specific information and capacities that most people think schooling should impart. But where all incentives are directed toward the passing of a test, the conduct of schooling is likely to adapt accordingly. Skills not measured will be de-emphasized, substantial time will be spent on test-specific learning (consider, at the extreme, SAT-preparation classes), and teachers uncomfortable with a wholly goal-oriented mode of instruction will gradually be dissuaded from teaching altogether or flushed into educational institutions (such as private schools) without testing. As in most cases of the public policy of faith, planning of this sort reaches its goals only by reshaping the world into a form subject to central control. Where schools have become effectively anarchic, such strong measures may be worth their costs. But this argument suggests that they should be applied only to districts that are persistently dysfunctional.⁴⁵

An excellent overview of the problems with designing high-stakes testing measures can be found in Kane and Staiger 2002.

The opposite mode, the public policy of skepticism, does not see education as a wholly goal-oriented activity, but admits that the ambiguity of education's mission is, far from being a flaw, an intrinsic part of the activity at its best. While some of the sciences may be reduced to a method, the mastery of which can be readily measured from a distance, the goals of history, literature, and civic education are essentially ambiguous and contested—the content of what needs to be learned is neither infinite nor definite. Although certain basics (such as literacy and numeracy) are needed for the effective teaching of all subjects, beyond that minimum there are a great number of worthwhile things to learn, a finite amount of time to learn them in, a pool of teachers whose abilities do not stretch across the whole body of potential knowledge, and a group of students whose interests and abilities vary even more widely.

Faced with these constraints, and knowing the problems with the public policy of faith, skeptics reject the idea that schools should be held responsible for certain closely specified educational outcomes, because the specification of those outcomes and the attribution of institutional responsibility both have effects on the conduct of education that skeptics are unwilling to accept. In this spirit, Daniel Patrick Moynihan likened the federal goals of the early 1990s—achieving a 90 percent graduation rate and making American students first in the world in math and science, all within six years—to Soviet grain production quotas, and he anticipated similar outcomes.⁴⁶

Referenced in Will 2003. Other observers have independently drawn similar analogies. Consider Michael Winerip's assessment of Houston's dropout projections, which have recently created more than minor embarrassment for the current secretary of education: “Some people compare it to working under the old Soviet system of five-year plans. In January, just before the scandal broke, Abelardo Saavedra, deputy superintendent, unveiled Houston's latest mandates for the new year. ‘The district-wide student attendance rate will increase from 94.6 percent to 95 percent,’ he wrote. ‘The district-wide annual dropout rate will decrease from 1.5 percent to 1.3 percent.’” Winerip 2003, B7.

As an alternative, skeptics see education as the delivery of a service, not as the attempt to produce an outcome.

Focusing on inputs rather than outputs diffuses agency, deflecting some of the credit or blame for the quality of schooling onto students, parents, and other social actors. This approach avoids mission creep by assuming that some of the concomitant elements of education, and the inputs that produce them, are the responsibility of other institutions.

Teachers in particular and schools as a whole can be held accountable for the education they present to students, but whether students can take full advantage of that education (and thereby produce specified outcomes) is partly in the hands of actors beyond the school (some of whom may be governmental, some not).

Diffusion of agency also helps preserve at least some of the multiplicity of ends, as well as the diversity of approaches and institutions, rather than flattening everything down in the pursuit of a single larger goal, governed by a single over-arching principle. Schools are understood to be responsible only for teaching. This is what we will call the principle of institutional limitability: institutions are accountable for their performance to the degree that their goals are modest and internal, rather than grand and external. Beyond this, teachers and administrators have the flexibility and freedom to determine the mission of their school and how best to achieve it.

The foregoing argument would seem to call for parental choice. If diversity of supply is essential to the bounded ambiguity of education, and schools are not to be held responsible for inputs beyond their control, then it is the duty of the parents to select a school whose ethos they accept and whose educational approach fits the character of their child. If a child is not learning in a particular school, the parent or guardian should either correct the deficiency that prevents the child from learning in that context or transfer the child to a school that is more appropriate. Each school cannot be expected to be right for every child.

That said, Oakeshott's concern for unintended consequences cuts against pursuing choice with excessive enthusiasm or speed. In schooling, as with much of the rest of society, we are not faced with a tabula rasa, so the question is how to inject more of the skeptical spirit into already existing institutions. This means that the unintended consequences of moving from one system to another may, in some cases, weigh heavily against comprehensive choice proposals. It does not take a terribly fertile mind to imagine what those unintended consequences might be. Choice, by using private entities to deliver what would still be seen as a public service, would inevitably bureaucratize existing private schools, bringing in its train much of the regulation that currently imposes a stultifying sameness on public schools. In addition, most Americans are not accustomed to exercising educational choice, beyond the influence that perceived school quality plays in residential decisions. Making a market in any area previously characterized by monopoly is always difficult, since competition can become counterproductive if those asked to make choices lack clear criteria on which to base their decisions. One way to solve this problem, as some choice proponents have recommended, is to combine choice with testing. But this may have the effect of indirectly imposing sameness on institutions by encouraging parents to choose on the basis of a narrow, and often deceptive, metric.

Proponents of choice, then, need to pay much more attention to the supply side of education (schools), rather than emphasizing only the demand side (parents), as almost all choice plans do. If we follow Oakeshott's insights, the most important question is not public or private, but how to allow for real diversity of schools and expand the ability of parents to match a school's goals with a child's character. An apparent solution is to deregulate public schools and the process by which children are assigned to them. This perspective also implies a need for injecting diversity into the supply of teachers—perhaps by allowing for different means of entry into teaching (other than standard certification and course work in education schools). Charter schools, which expand diversity of provision within the public school system, may serve Oakeshottian ends better than full-scale private school choice, particularly if public regulation undermines the diversity of private schools. Better to begin by deregulating a portion of public schools, perhaps en route to the more comprehensive reform represented by vouchers.

Public health

Public health mirrors education in certain respects, especially regarding the politics of faith's vulnerability to mission creep. Additionally, the specific case of tobacco control shows that the politics of faith is likely to justify institutional formlessness and to corrode procedural and constitutional norms. As we will demonstrate, tobacco policy was best when it reflected a standoff between crusaders for health on one side and, on the other, judges and legislators committed to procedural—especially legal—norms; it has grown steadily worse as the skeptical tendency in politics has weakened. This downturn will have consequences not just for health policy but for the conduct of law and politics generally, since the wearing away of procedure always provides precedent for formlessness in other areas.

Consider again the central feature of the public policy of faith: the idea that the purpose of government is to produce certain desired end states—in this case, “health.” Health is not a particular, stable state, but at most a direction toward which progress might always be made if the opportunity costs are not considered. Health cannot be produced solely by institutions for which it is the overriding objective. It is a joint product of the medical profession, community norms, economic opportunities, and individual decisions. Just as students will not learn if they refuse to study, patients may not heal if they disobey a doctor's orders. Institutions responsible for “health” will inevitably find medical care in and of itself no guarantee of success.

In the mode of faith, when government discovers that medical care itself cannot produce health, it will tend toward expanding the scope of its intervention. It will become interested in what citizens eat, how frequently they exercise, and whether they smoke or drink. What is more, once government takes responsibility for the public's health, it will find that virtually everything affects health. Once again, policy makers set on producing “health” find that they can either admit to having set goals beyond the ability of their available instruments to achieve or else attempt to expand those instruments. Lawsuits against tobacco companies (about which, more below), mayoral diet and exercise campaigns (as in Philadelphia, whose citizens were told to forswear cheese steaks), and discussions of crime as a public health problem are only a few examples of the mission creep that accompanies the public policy of faith.

Health is normally a state of being jointly produced by medical authorities and other actors—in particular, individuals themselves. Pursuit of the ideal of “public health” implies a concentration of agency onto the state such that it can no longer diffuse that responsibility back onto communities or individuals themselves. Once reducing fat in one's diet is shown to “reduce heart attacks by 8 percent,” or cutting smoking is shown to “decrease lung cancer by 20 percent,” it becomes difficult to question on any grounds other than efficacy (and often not even for that reason) measures to reduce fat or smoking.

In her study of the recent American tobacco settlement, Martha Derthick contrasts two eras of policy making, one from 1964 through 1993, and the other from 1994 to the present.⁴⁷

Most of this discussion is based on Derthick 2001.

The first era was characterized by increasing concern with smoking, repeated high-profile warnings of its dangers from prestigious government sources (such as the surgeon general), restrictions on advertising (such as bans on TV advertisements), warnings on cigarette packaging, and gradual increases in taxes. While all of these measures reflected a desire to pursue health as a social outcome, advocates of tobacco control considered them grossly insufficient. They made no effort to conceal their desire to bankrupt the cigarette industry and thereby eliminate smoking. But because policy had to be made within existing institutions and legal norms, what they got were impressively effective decreases in smoking through relatively mild interventions that required few innovations or extensions in accepted government power. Derthick judiciously notes that “even if the utility of some of the government's measures in this period were debatable, the outcome deserves to be applauded as one of government's major achievements. Smoking was very much reduced, and with measures far more consensual than not…. Much benefit was realized in the years 1964–93 with a relatively low cost in coercion.”⁴⁸

Ibid., chapter 11.

One could see this period as a model of balance between faith and skepticism—the former keeping pressure on government for action, and the latter keeping that action within bounds and preventing damaging precedents for the use of public power.

The period since 1993 presents a very different, and much more disturbing, picture. Up until the 1990s, efforts to go beyond a rough faith/skepticism balance through the use of exotic new legal doctrines had gotten very little purchase in the courts. No attempt to hold tobacco companies liable for the health effects of smoking on individuals were upheld on appeal. The law generally accepted that smokers assumed the risks of smoking. The health effects of smoking were known because of the surgeon general's well-advertised warnings on cigarette packs, despite the industry's arguments to the contrary. So tobacco companies could no more be held liable than other industries that produced inherently dangerous but legal products.

But then, stolen and legally produced documents from the tobacco companies showed that the companies had intentionally concealed their own studies showing the health effects of smoking and that they had manipulated the nicotine content of cigarettes. This information leak, along with highly publicized attacks on the industry, helped produce a political environment in which legal and institutional limits on antitobacco policy gradually fell away. State attorneys general and private lawyers collaborated to sue the tobacco companies—based on the extremely dubious legal principle of third-party liability—to, as they saw it, recover the costs that smoking created for their Medicaid programs. In order to get out from under the potentially ruinous costs of such litigation, the tobacco companies, state attorneys general, and private lawyers, along with some antismoking activists, negotiated a “universal settlement” of all these claims. Eventually (after a long process well described by Derthick) they agreed to the generically named Master Settlement Agreement, which, in addition to extracting more than $200 billion from the tobacco companies, introduced a long list of provisions that are almost comic in their deviation from standard governing practice.⁴⁹

Among other things, the MSA required the cigarette industry to disband its own research and lobbying organization, the Tobacco Institute; permitted the cigarette industry to collude in setting prices, in order to ensure that it would have sufficient profits to cover the settlement costs and to maintain prices high enough to dissuade smoking; extracted very large sums of money from a single industry without explicitly imposing a tax; transferred amounts of money well in excess of reasonable fees to the private lawyers who were party to the agreement; and was negotiated without the requirements of disclosure that would have been mandated in any traditional political process. In effect, many of the requirements of the settlement would have been unconstitutional had they been written into law.

What is worse, many of the provisions had only the most tenuous relationship to actually reducing smoking. In practice, the formlessness that the public health rationale justified facilitated a fairly naked transfer of resources from smokers to private lawyers and state governments, while creating, ironically enough, a codependent relationship between the tobacco industry and government.⁵⁰

Fairclough and O'Connell 2003.

The pursuit of justice unconstrained by constitutional norms is apt, in the guise of high-mindedness, to open policy making to the kind of transfer seeking that complex institutions are supposed to prevent. This outcome is about as pure an illustration as one could find of the difference between the creative tension that characterized pre-1993 policy making and the formlessness that characterizes a decisive victory for the politics of faith.

The tobacco settlement has had effects beyond tobacco. By legitimizing a formless politics and deconcentrating responsibility, the settlement has spawned efforts to do to fatty foods what the lawyers have already done to the tobacco companies. Rogan Kersh and James Morone have found that many of the same rhetorical strategies that were used to disarm the tobacco industry are now being deployed against the purveyors of fatty foods.⁵¹

Kersh and Morone 2002.

But using “public health” as a mechanism to suspend normal politics goes beyond the merely rhetorical. John Banzhaf, a George Washington University lawyer who was an early enemy of the tobacco companies, has warned, “What we're trying to do is use some of the same legal tactics that have been so effective against the public health problem of smoking against the other public health problem of obesity.”⁵²

Vallese 2002.

Those tactics, combined with the rhetorical template of the successful tobacco litigation and the now-deep pockets of trial lawyers enriched by those tobacco suits, make a large-scale legal assault on the fast-food and fatty-snack industry not only possible, but highly plausible. And so we see one of the most insidious elements of the politics of faith: it has a tendency to spread, for once institutional norms have been weakened in one area, a precedent has been set for their weakening in others.

Pensions/Social Security

Recall, from our introduction to this article, Andrew Sullivan's claim that Oakeshott would be an ardent privatizer of Social Security. Is Sullivan right? There are certainly reasons to think so. Clearly the central policy lesson from the earlier section on education was a strong bias toward choice and diversity in providers. Why shouldn't this be true for pensions as well? In fact, don't Oakeshott's arguments presume diversification and privatization of all government activities?

In essence, the argument of this section will be that the opponents of Social Security as it is currently constituted are in the grips of the politics of faith, while its defenders are on the side of skepticism. Privatization is based on highly speculative predictions of future problems with existing arrangements and on equally speculative (but significantly rosier) predictions of the value of alternative policies. What is worse, privatization dismisses the likelihood of substantial unintended consequences, since the policy, like all attempts to dislodge deeply embedded social institutions, is subject to large but incalculable transition costs. This case simply confirms Oakeshott's observation that “the programme of every party is written in the language of faith, the enterprise of every government is conceived in the idiom of faith.”⁵³

Oakeshott 1996, 126.

The argument for Social Security privatization has two basic parts. The first is that Social Security as currently constituted will go bankrupt before current workers retire. The second component is that even if the program were to remain solvent, its benefits would be substantially below those that workers could obtain in the private market.⁵⁴

There is another, less publicized, argument for privatization: that it will dislodge citizens from support of activist government. To the degree that this claim has been tested in Great Britain, under conditions remotely similar to those in the United States, it has turned out not to be accurate. See Teles 1999, an unpublished manuscript available from the author.

The first claim is based on projections of Social Security's future income and the benefits it must pay out. Neither of these figures is quite so easy to predict over a 75-year period (the current time frame within which Social Security benefits and income must be in balance). Future income depends upon the growth of the economy, average age of retirement, future fertility, immigration, and income inequality, among other things.⁵⁵

Higher growth would mean higher incomes, which would increase Social Security taxes, as would higher fertility, later retirements, and growth in immigration. Reductions of income inequality would mean higher Social Security taxes (since the tax is capped, increases in earnings at the higher levels do nothing for the health of Social Security, even as they swell other tax sources).

Future benefits are somewhat easier to predict: they depend on the average age of retirement combined with the age at which people begin to receive benefits and at what level (both of these factors are subject to political change), as well as average longevity.

Essentially, assertions that Social Security income will not cover its current obligations are based on extremely loose speculation about the future. No one can say with any certainty how much immigration America will experience over the next 75 years—that is a function of both immigration policy (over which the government has some control) and the relative economic fortunes of the United States and sending countries (which, for the most part, the government cannot control). Future economic growth could either be much higher or much lower than currently predicted, depending upon changes in technology, government policy, education, and myriad other factors; the same holds true for income inequality. The average age of retirement is dependent upon employers' needs, workers' preferences, social pressure, and government policy. Fertility has been in a slump in the United States for decades now, although it appears to be rebounding.

Despite the inevitable imprecision of predicting Social Security's income, we are not saying that it is unwise to act on the best available predictions. Yes, the politics of skepticism “responds slowly to inspired imaginings of perils to come.”⁵⁶

Oakeshott 1996, 109.

And yes, the contract among the past, the present, and the future requires that those in the present at least do something to reduce future displacements that will result from existing policy. But while paying attention to the general direction of these predictions, the skeptical statesman will limit his changes to those that can be achieved within current arrangements (more about this later); and he will distinguish between predictions about the near future and those 40 or 50 years out, treating the latter more as science fiction than as social science.

The other argument for Social Security privatization is the claim that future benefits, if they exist at all, will be less than those from individual accounts invested in a mix of private and public securities.⁵⁷

See, for example, Feldstein et al. 2001 for a technically sophisticated version of this argument.

From a skeptical point of view, a great number of things are wrong with this argument.⁵⁸

These are presented at length in Teles 2000.

It posits an untested alternative based on speculative assessments of future conditions (especially the rate of return on equities) against a tested existing institution that is currently serving its principal goals at least tolerably well. Such an approach raises the question of transition costs—the price paid in social dislocation when moving from one state of affairs to another. Usually advocates of privatization attempt to obscure the magnitude of these costs by treating them exclusively as the price one generation must pay to cover both its own retirement and that of existing retirees. They give the impression of a transition that may be politically painful, but whose consequences can be easily predicted and accounted for.

From an Oakeshottian perspective, this argument is the height of fallacy. No institution as woven into the fabric of a society as Social Security can be radically altered without costs that are impossible to calculate in advance. Social Security currently permits those saving for their retirement to be more aggressive in their private accounts; shifting some of Social Security into privately held equities would necessarily have unpredictable effects on savers' asset allocation, possibly canceling out the advantages of the change. In moving from a system where government collects taxes and distributes benefits to one where it mandates and regulates contributions to individual accounts, the U.S. government would be exchanging a relatively unobtrusive and highly limitable power for one highly obtrusive and difficult to limit. Government would acquire the authority to coerce private investment and, consequently, a much greater interest in the conduct of the financial markets (since they would be expected to serve broad social goals of creating large returns on small accounts, something financial markets have no interest in doing at the moment). This scenario clearly violates the principle of the economy of power. Social Security now avoids this by using known, controllable instruments that do not require large numbers of decentralized actors to behave in a predictable manner. Privatization—inevitably—would.

Social Security, whatever its flaws, is a classically skeptical program. It does not attempt to organize society for a large social goal, but delivers a service directly. Privatization, on the other hand, is cut directly from the cloth of faith. It promises to regulate society to meet or exceed certain specific goals (income available on retirement), not on average, but in every case. Because the instruments to achieve this goal will likely be lacking from the start, disappointment will set in when it turns out that there are winners and losers—and the losers will demand compensation, through regulations on financial providers or directly from the government itself. To get to the lofty returns, the government will often have to pressure citizens to invest in a riskier mix of stocks and bonds than they can tolerate, thereby setting in motion a cycle of aggressive buying when stocks are on the way up, and selling when they are on the way down (and in the process eliminating most of the benefits of investing, which come from buying and holding). These effects will necessitate additional government controls. The main consequence of Social Security privatization will be, paradoxically, to socialize markets.

The public policy of skepticism in Social Security will be to look for changes that are “intimated in and not merely imposed upon the situation.”⁵⁹

Oakeshott 1991a, 411.

Skeptics would make changes to mitigate future financial stresses on the program, while not assuming that those (or any) predictions in and of themselves justify fundamental change. They would encourage Social Security's supporters to accept increases in the retirement age, cuts in benefits for higher earners, and increases in the maximum taxable income (among other changes) to further reduce stress on the program, while also disarming advocates for fundamental change in the present.

Exceptions

As we suggested earlier, there are times when even the skeptical statesman may lean toward the pole of faith. Skepticism reaches its limit in two cases: when incremental change draws one deeper into the condition for which remedy is sought, and when social conditions are such that decency and order require subverting a tradition rather than seeking intimations from it. In both of these situations, some degree of action in the mode of faith is impossible to avoid.

Health insurance and traffic both illustrate the first case. The United States has repeatedly attempted to expand access to health insurance through a series of incremental programs targeted at specific groups (such as Medicaid and the Children's Health Insurance Program) or specific medical conditions, or through increasingly generous subsidies to private employers. Seeing that spiraling health insurance costs make it difficult for employers or individuals to afford coverage, the government encouraged Health Maintenance Organizations (HMOs) to control costs—but then began regulating them when they started doing just that. Each one of these incremental efforts to expand coverage has had little or none of the intended effect, since other parts of the health care system have responded to expansion by contracting. (For example, increasing coverage in Medicaid causes low-wage employers to eliminate or scale back coverage.) Despite decades of efforts to improve access through changes at the margin, and despite the very large sums of money spent and quite enormous central government regulatory apparatus, the number of uninsured continues to increase each year. Here, incrementalism is clearly pathological—it exacerbates or at least has no positive overall effect on the condition it is meant to alleviate. In such situations, even the skeptical politician is forced to ask whether the very “mongrel” character of current interventions (which blend public and private provision) might itself be at the root of the problem, and whether some “purification” of principles, in the form of national health insurance, might actually require a less minute and confusing regulatory state than our current hybrid system does.⁶⁰

An excellent analysis of the way that the hybrid quality of health care policy in the United States leads to continual increases in centralized control and regulation can be found in Ruggie 1992.

Traffic is another example of how incremental changes within a tradition can be self-undermining. The response of most municipal and state governments to traffic has been to build more roads: this is the American tradition in transportation policy. But it turns out that one of the few laws of social action that actually works is that road construction never leads to a net reduction in traffic; it simply induces more people to drive and encourages a housing pattern that makes walking or public transit either impossible or uneconomical. The second law of social action in transportation is that every municipal area eventually recognizes this pattern, but only after it has become impossible to fix, or where the only remedies involve the kind of minute planning (restrictive zoning, subsidies to encourage car pooling, punitive taxation of center-city parking, et cetera) that skeptics deplore.⁶¹

A textbook case of this is Atlanta, Georgia; see Going nowhere 2001.

Given the seemingly inexorable nature of this process, even a politician in the skeptical mode would want to consider large-scale public transit systems, which discourage sprawl by creating a positive inducement for concentration, without the need for intrusive and minute regulation.

The second case in which the balance leans to faith over skepticism is where the capacity for spontaneous action and independent association has been depleted, and where a culture of dependence and incivility has taken hold. For example, at least part of the problem of persistent poverty is rooted in a self-reinforcing neighborhood cycle of low work levels, crime and fear, benefit dependence, and political and social disorganization. Here, the kind of light, detached touch that Oakeshott prescribes is manifestly inappropriate. There are certain times when simply creating the background conditions for spontaneous action presumes a greater capacity for collective action than is culturally feasible. Under these conditions, government has to actively—and sometimes intrusively and minutely—reconstruct social patterns through fairly paternalistic means.⁶²

See, for example, the essays in Mead 1997.

The kind of civic association that Oakeshott believed to be (along with the rule of law) the foundation of social order presumes cultural competencies that are sometimes absent. In such cases, the capacity for spontaneous action must be seen by the government as the objective of policy rather than its presupposition. Certain kinds of welfare reform, drug treatment, and intensive policing policies—which attempt to reconstruct civility through the direct imposition of administrative authority—fall into this category.

In all of the cases discussed above, some sort of planning—dependent upon an at least minimally well-elaborated vision of social outcomes—is inevitable. The skeptical statesman, before moving in this direction, would want assurance that incremental change had already been tried and had demonstrated its inefficacy, or that the capacity for spontaneous action in a particular community had been largely depleted. One can think of planning as akin to chemotherapy—a very powerful treatment appropriate only under exceptionally dire circumstances. In undertaking it, the skeptic would be constantly on the lookout for a form of planning that achieves its ends in the least intrusive way possible and that implies the fewest open-ended commitments. Even while acting in the mode of faith, economizing in the use of power would never be far from his mind.

Conclusion

We hope to leave you with two general conclusions. First, the crucial issue in the analysis of public policy is the manner, not the size, of government. Generally, skepticism erects a significantly higher hurdle for government action than do certain manifestations of the politics of faith. But it is critical to recognize, as our discussion of Social Security surely makes clear, that libertarianism—not just liberalism—can be a kind of faith-driven politics. Skepticism is different from libertarianism in that it is impressed by the practical and political constraints on dismantling existing government commitments, as well as the unintended consequences that will inevitably plague such projects. While libertarians avidly apply the law of unintended consequences to government expansion, skeptics recognize that this law applies to contraction as well.

In addition, the skeptical statesman is more willing to accept the need for government action, but shares many of the libertarian's criticisms of how government is actually conducted. There are cases in which a properly skeptical approach to public policy leads one to seemingly large governmental commitments—e.g., when they deconcentrate agency, use well-worn powers of government, and do not require potentially dangerous and open-ended innovations in institutional practice. It is possible—and we do not need to reach into the realm of theory to prove it—to have a welfare state that is skeptical in its orientation.

The primary critique that skepticism shares with libertarianism concerns the increasingly informal, anticonstitutional style of modern policy making. Devices such as separation of powers and federalism were designed, in part, to rein in government generically, but also to prevent private interests from using government as a forum for transfer seeking. As the tobacco case discussed earlier suggests, these constitutional norms now show signs of disrepair. Although skeptics disapprove of libertarians' blanket opposition to the modern welfare state, they accept the need for reinvigorating formalism as a way of constraining the welfare state's excesses.

The essence of skepticism is the necessity of balancing the authorization and legitimization of government with restraint. This problem was at the root of the “new political science” of America's framers, whose words still have resonance today: “In framing a government which is to be administered by men over men, the great difficulty lies in this: you must first enable the government to control the governed; and in the next place oblige it to control itself.”⁶³

Hamilton et al. 1961 [1788], No. 51, 322.

This tension, which inspired what Oakeshott called “the most profoundly skeptical constitution of the modern world,”⁶⁴

Oakeshott 1996, 80.

should also stand as the foundational principle for political science's contribution to public debate, what we would call constitutional policy analysis. Those seeking inspiration for such an undertaking could find no better primary source than the thought of Michael Oakeshott.

This paper was originally prepared for the Inaugural Conference of the Michael Oakeshott Society, held at the London School of Economics, September 3–5, 2001. The authors thank Leslie Marsh for inviting them to participate, as well as Martha Derthick, Jennifer Hochschild, Mark Kleiman, Steven Lenzner, Lawrence Mead, Martin Rein, Peter Skerry, and three anonymous reviewers for providing invaluable criticism. They are especially indebted to Melissa Bass, who thoroughly edited this paper and ensured that it was completed in a (relatively) timely manner.

References

Burke Edmund. 1949[1780]. The duration of parliaments. In Burke's Politics: Selected Writings and Speeches on Reform, Revolution, and War, eds. Ross Hoffman and Paul Levack. New York: Knopf, 216–24.

Derthick Martha. 2001. Up in Smoke. Washington, D.C.: Congressional Quarterly Press.

Fairclough Gordon, and Vanessa O'Connell. 2003. Co-dependents: Once tobacco's foe, states are hooked on settlement cash. The Wall Street Journal, 2 April, A1.Google Scholar

Feldstein Martin, Elena Ranguelova, and Andrew Samwick. 2001. The transition to investment-based Social Security when portfolio returns and capital profitability are uncertain. In Risk Aspects of Investment Based Social Security Reform, eds. John Campbell and Martin Feldstein. Chicago: University of Chicago Press, 41 87.

Going nowhere and certainly not fast. 2001. The Economist, 12 July, 49.

Gopnik Adam. 1996. A man without a plan. The New Yorker 21/28 October, 194–8.Google Scholar

Hamilton Alexander, James Madison, and John Jay. 1961[1788]. The Federalist Papers, ed. Clinton L. Rossiter. New York: Mentor Books.

Hess Frederick M.1998. Spinning Wheels: The Politics of Urban School Reform. Washington, D.C.: Brookings Institution Press.

Jervis Robert. 1997. System Effects: Complexity in Political and Social Life. Princeton: Princeton University Press.

Johnson Lyndon B.1964. Proposal for a nationwide war on the sources of poverty: Special message to Congress, 16 March. In Modern History Sourcebook: President Lyndon B. Johnson: The War On Poverty, March 1964. Available at www.fordham.edu/halsall/mod/1964johnsonwarpoverty.html. Accessed 3 October 2003.

Kane Thomas J. and Douglas O. Staiger2002. The promise and pitfalls of using imprecise school account-ability measures. Journal of Economic Perspectives 16: 4, 91 114.Google Scholar

Kersh Rogan, and James Morone. 2002. How the personal becomes political: Prohibitions, public health, and obesity. Studies in American Political Development 16: 2, 162–75.Google Scholar

Knight Frank. 1965. Risk, Uncertainty, and Profit. New York: Harper and Row.

Lincoln Abraham. 1953. A meditation on Proverbs 25:11. In The Collected Works of Abraham Lincoln, ed. Roy P. Basler. Vol. 4. New Brunswick: Rutgers University Press.

Lowi Theodore J.1979. The End of Liberalism: The Second Republic of the United States. New York: W. W. Norton.

Mansfield Harvey C., Jr. 1989. Taming the Prince: The Ambivalence of Modern Executive Power. New York: Free Press.

Mead Lawrence M., ed. 1997. The New Paternalism: Supervisory Approaches to Poverty. Washington, D.C.: Brookings Institution Press.

Merton Robert K.1936. The unanticipated consequences of purposive social action. American Sociological Review 1: 6, 894 906.Google Scholar

Oakeshott Michael. 1975. On Human Conduct. Oxford: Clarendon Press.

Oakeshott Michael. 1991a. On being conservative. In Rationalism in Politics and Other Essays, ed. Timothy Fuller. Indianapolis: Liberty Fund, 407–37.

Oakeshott Michael. 1991b. Rationalism in politics. In Rationalism in Politics and Other Essays. Indianapolis: Liberty Fund, 6 42.

Oakeshott Michael. 1996. The Politics of Faith and the Politics of Skepticism, ed. Timothy Fuller. New Haven: Yale University Press.

Ravitch Diane, ed. 2000. Brookings Papers on Educational Policy: 2000. Washington, D.C.: Brookings Institution Press.

Ruggie Mary. 1992. The paradox of liberal intervention-ism: Health policy and the American welfare state. American Journal of Sociology 97: 4, 919–44.Google Scholar

Scott James C.1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed. New Haven: Yale University Press.

Sullivan Andrew. 2001. Negatives: George Bush, meet Michael Oakeshott. The New Republic, 6 August, 6.Google Scholar

Teles Steven M.1999. Pensions, social citizenship, and the ubiquity of the state. Unpublished manuscript, Brandeis University.

Teles Steven M.. 2000. The paradoxes of welfare-state conservatism. The Public Interest 141 (Fall), 17 40.Google Scholar

Vallese Julie. 2002. After tobacco suits, lawyers target fast food. CNN.com, 19 August. Available at www.cnn.com/2002/LAW/08/19/fast.food.lawsuit/index.html. Accessed 3 October 2003.Google Scholar

Will George F.2003. A beautiful mind. The Washington Post, 27 March, A21.Google Scholar

Winerip Michael. 2003. The “zero dropout” miracle: Alas! Alack! A Texas tall tale. The New York Times, 13 August, B7.Google Scholar

Article contents

The Public Policy of Skepticism

Abstract

Information

The Politics of Faith and Skepticism

Unintended Consequences

Policy Implications

Education

Public health

Pensions/Social Security

Exceptions

Conclusion

References

Save article to Kindle

Save article to Dropbox

Save article to Google Drive

Reply to: Submit a response

Your details

You have entered the maximum number of contributors

Conflicting interests