Weingast B.R., Wittman D. The Oxford Handbook of Political Economy

Подождите немного. Документ загружается.

982 politics and scientific enquiry

how they may be predicted to act under diﬀering circumstances; and how analytical

models of state action might generate scientiﬁcally falsiﬁable hypotheses to be tested.

An empirically based macropolitics might be constructed in some such fashion and

with some claim to genuinely scientiﬁc standing.

A science limited to the exposing of regularities in patterns of state action, without

some grounding of these regularities in the behavior of those persons who act as

agents of and for the collective unit, however, would be too truncated for the modern

investigative mindset. The organic state, as a metaphysical unit, is simply not philo-

sophically meaningful, and not only to the would-be scientist. To be sure, elements

of the “state as God” mentality may have been force fed during the socialist epoch,

and residues of such mentality may remain descriptive over some spectra of public

attitudes, but a science of “politics without politicians” will be almost universally

rejected. Any political science must be constructed on microfoundations.

History is not helpful here. With some notable exceptions, political units have

evolved from origins that locate authority in identiﬁable autocratic leaders, whose

choices are made without necessary or formal inputs from those who are externally

aﬀected by such choices. In this setting, the monolithic political agent, the embod-

iment of the state, is, quite simply, acting as an isolated individual who might be

modeled as maximizing an objective function subject to whatever constraints are

faced. Micro- and macropolitics merge into one. “The prince” becomes the ideal

type for any analysis, and Machiavelli becomes the direct precursor of any scientiﬁc

enquiry.

It is perhaps not surprising that those who ventured into enquiries about poli-

tics shied away from the approach taken by Machiavelli and his Italian peers. The

ultimate normative purpose or motivation for science is control aimed at reform

or improvement, and Machiavelli’s advice to the prince did not sit well with those

whose interests were not coincident with those of the prince. One avenue of escape

involved the rejection of the scientiﬁc approach itself, a withdrawal of eﬀorts aimed

at explaining how princes do, in fact, behave, and intensiﬁcation of discourse about

how princes “should” behave.

3 The “Ought”Replaces the “Is”

.............................................................................

Ethical norms for the guidance of the behavior of those who make decisions for the

collectivity became center stage in discourse on politics, and Plato was recognized

as the fountainhead of ideas in the political theory that was to dominate enquiry

until the middle of the last century. Such elevation of normative discussion, to the

comparative neglect of positive analysis, might possibly have been justiﬁed so long

as political units were best described as principalities, under the authority of single

¹Fortwoeﬀorts that could be classiﬁed in this category, see de Jasay 1985,andvanCreveld1999.

james m. buchanan 983

rulers. Surprisingly, however, the mindset that conceives the whole exercise as that

of proﬀering advice to a benevolent despot was retained as autocratic regimes were

being replaced by democratic structures of governance. Those persons elected in

democratic processes were implicitly modeled as “temporary princes” immunized

from constituency pressure. There seems to have been a generalized failure, on the

part of those who concentrated scholarly attention on politics, to recognize that

decisions emergent from the processes involving multiple participants are simply not

amenable, even on some “as if” basis, to unitary models of choice behavior. This

unwillingness to factor down the observed outcomes of complex political processes

in which many persons make “public choices,” and in varying agency roles, was

strengthened, and especially in English-language scholarship, by the dominance of

utilitarianism, throughout political economy. This normative structure appeared to

provide a common measuring standard which, if adhered to by agents in diﬀering

roles, would allow even the most complex institutions to generate outcomes that

would be equivalent to those emergent from some “behind the curtain” force that

aimed to maximize “social welfare” or “the public interest.” It continues to be amazing

that, even into this new century, welfare economists have never ceased their search for

“social welfare functions.”

Note that in the several settings mentioned here, whether that of the organic state,

the single ruler, or the existence of a common standard for all political agents, discus-

sion and analysis could proceed, either positively or normatively, based on a model

of a single or unitary mind. A methodological individualism, of sorts, could provide

the psychological framing for eﬀorts aimed at understanding political actions. And,

within this structure, precepts of rationality may or may not be considered applicable

to the choices that are analyzed.

4 Unitary and Multiple Sovereigns

.............................................................................

A categorical distinction must be made between all such models that, either directly

or indirectly, analyze political choices as if they emerge from a monolithic unitary

mind and those models that involve multiple participants in the generation of any

result. A major element in the continuing confusion in political analysis has been the

failure to recognize even the existence of this distinction.

In any unitary-mind model, the alternative outcomes are within the choice set of

the entity that makes the choice. There is a one-to-one relationship between “that

which is chosen” and “that which emerges from choice.” The chooser is uniquely

responsible for the result or outcome of the choice that is made. This relationship

cannot exist in any model that involves more than one participant in the selection

process.

Consider the simple case with only two persons, A and B, locked in a collective unit,

who are confronted with alternatives that must be jointly available (say, dormitory

984 politics and scientific enquiry

roommates who must live with a single thermostat setting). In any unitary model,

either A or B must be the ruler or dictator. If, however, both are participants in the

decision process, neither A nor B can uniquely determine which alternative is to be

chosen. The alternatives, as such, are not within the choice set of either participant.

The participants can, of course, order the alternatives and agreement may be reached

on some process or rule for making a selection.

It seems evident that the disjuncture between a person’s ordering of the alternatives

and that same person’s expressed “choices” (votes) in processes in which other per-

sons are known to participate becomes a possible source for diﬀerences in behavior.

Further, it seems clear that characteristics or attributes of a person’s ordering of

alternatives need not necessarily correspond to the characteristics or attributes of that

same person’s “choice” behavior in the process of selection. In diﬀerent words, there is

no necessary mapping between the individual’s evaluative orderings of the collective

alternatives and the ranking of the choices that would be made as participation in the

institutions of collective selection, including voting, takes place. There is a logically

derived attenuation between these two orderings that stems from the absence of full

responsibility for choice.

Methodological clarity, if nothing else, would seem to dictate that these categor-

ically diﬀering models of state or collective action be separately analyzed in parallel

tracks, rather than incoherently intermingled. In this respect, only the Italian ﬁscal

theorists who developed their primary contributions over the course of a century,

roughly, 1850–1950, are deserving of special commendation. These scholars, among

whose names would be placed Ferrara, Pantaleoni, Mazzola, de Viti de Marco, Ein-

audi, Puviani, Fasiani, Cosciani, Barone, along with others, did not neglect attention

to the model of the collective unit subject to analysis. De Viti de Marco and Fasiani, in

particular, wrote treatises in which the two models, that of the unitary or monopoly

state and that of the cooperative or democratic state, were presented side by side.

Interestingly enough, Duncan Black and I, who were early contributors to the “new”

approach to the analysis of politics in mid-century and later, were both exposed, early

in our own research programs, to this Italian tradition in ﬁscal theory.

5 Theory of “Democratic”Politics

.............................................................................

As noted, residues of the organic state model of collective action remain in modern

analyses, and especially in the sense that much enquiry seems to be informed by a

longing for patterns of collective action that would, in fact, be comparable to those

² For early discussion, see Buchanan 1954.Notethatthedisjunctureheredoesnotdepend,inanyway,

on the logical grounds for expressive voting that emerge only in large-number electoral settings. For

discussion of expressive voting, see Brennan and Buchanan 1984 and Brennan and Lomasky 1993.

³ For a general summary overview of the whole Italian tradition in public ﬁnance, see Buchanan 1960.

The only major work in English translation is de Viti de Marco 1936. In Italian, see Fasiani 1951.

james m. buchanan 985

that would emerge from a unitary decision-maker. For the most part, however, mod-

ern analyses commence from the presumption that politics is a process, or processes,

that involve multiple participants, who may inﬂuence, variously, the collective out-

comes that are observed. To this extent, the models introduced are almost exclusively

“democratic.” Attention is focused on the behavior of the individual, as a participant

in the process of collective decision-making and on the ways through which the

separate individual behaviors are combined or aggregated so as to generate deﬁnitive

collective outcomes.

What is “an individual” for purposes of providing the starting point for analysis

of collective action? For comparison, think of the comparable question as put to the

economist who aims to analyze private behavior in the idealized marketplace. The

individual is readily deﬁned as a two-set combination of preferences and endowments

(capacities). In stylized collective action settings that are “democratic,” individuals are

equally endowed with the authority over ﬁnal outcomes. The deﬁnition of the indi-

vidual reduces, therefore, to the set of preferences possessed. Hence, the abstracted

person that becomes the basis for analysis of collective action is a set of preferences,

nothing more. It is perhaps worth noting that, for the pure theory of politics at this

level of enquiry, a presumption that persons have equal preferences eliminates totally

any problems of aggregation.

The individual, deﬁned as an abstract set of preferences, became the explicitly

recognized starting point for the pioneering work of Duncan Black (1948, 1958), for

the somewhat earlier and much neglected work of Howard Bowen (1943), and for the

more general analysis of Kenneth Arrow (1951). It is noteworthy that all of these early

contributors were economists by training, and the concentration of attention on the

behavior of the individual, as a participant in collective action, presumably stemmed

from the methodological individualism that was, and remains, characteristic of eco-

nomic analysis.

Thesecondstepinanyconstructionofanabstracttheoryofpoliticsrequiresthe

speciﬁcation of the structure within which individuals, as participants in the process

through which collective outcomes are generated, are presumed to act. Here it was

perhaps understandable that majority voting was placed at center stage, despite the

institutional complexities that seem to remove observed politics far from any simple

majoritarian models. Duncan Black commenced his research program in an eﬀort

to analyze majority voting in small committee settings, and Howard Bowen seemed

to incorporate majority rule as a natural element in any collective decision process.

Kenneth Arrow’s approach was much more general, and majority voting was only

one among the many processes that failed to pass the test imposed in the general

impossibility theorem. As Arrow’s analysis was interpreted in the scientiﬁc commu-

nity, however, concentration was centered on the failure of majoritarian processes,

on the possible instability of democracy itself, translated as the institutionalization of

majority rule.

⁴ The Stigler–Becker 1977 setting in which persons have equal preferences as one basis for analyzing

behavior in markets could not be carried over into politics.

986 politics and scientific enquiry

6 The Objects of Collective Action

.............................................................................

What are the alternatives among which collective “choices” are made? This area of en-

quiry has been surprisingly neglected.

Analysis has proceeded “as if” the alternatives

for collective action are exogenously determined outside the processes which generate

the results.

Duncan Black was interested in explaining how persons, primarily in small com-

mittees, make collective selections among alternative motions, candidates, or party

platforms. His attention did not extend to the larger questions as to the domain of

collective options. Kenneth Arrow’s seminal analysis oﬀered mathematical tractability

but at the expense of speciﬁcity. The domain for collective choice includes the whole

set of alternative “social states,” each member of which embodies a complete descrip-

tion of the allocation and distribution of value among all members of the polity. By

inference, Arrow’s domain for collective or political action becomes all inclusive.

Politics, as observed, is not, however, the inclusive domain for social interaction.

At some base level of choice, persons enter into collective action when they consider

that in such action they can secure more of that which they value (utility) than they

can secure in non-collective interactions. The ultimate logic of collective action ﬁnds

its raison d’être in the potentiality for shared gains. These strictures apply, of course,

only to consensual politics, at least at the level of constitutional structure. They do not

apply to regimes of governance that are non-democratic in the sense that, for some

sub-sectors of the membership, the “game” is negative sum over the whole anticipated

sequence of play. The scientiﬁc developments in our understanding are almost totally

concentrated on the presumption that political regimes are broadly “democratic,” at

least in this limited sense.

Analysis was greatly clariﬁed by Samuelson’s articulation of the concept of pure

“public goods” at mid-century (Samuelson 1954). In the stylized limiting case, a

pure public or collective-consumption good or service is non-rivalrous in usage

or consumption over large numbers of potential users. The costs of provision to

any user-consumer remain unchanged as others are added to the using-consuming

group. In any such setting, the value-enhancing potential for collective, as opposed

to non-collective, action in providing such a good is evident, made especially so if

the exclusion of users-consumers is itself costly. In the early analyses derivative from

Samuelson’s contribution, emphasis was placed primarily on why markets tend to

fail in producing eﬃcient quantities of goods that embody publicness characteristics.

Relatively little attention was focused on the implications of the analysis for the

domain for political action, as such.

Critics suggested, correctly, that goods exhibiting the pure publicness character-

istics are rare indeed, and that the dividing line between the market and collective

choice domains was likely to be determined by other considerations. In particular,

⁵ For an eﬀort to remedy the neglect, see Buchanan 1995.

⁶ Economists may be charged with a comparable neglect in analyses of market choices. “Goods” are

more or less presumed to be deﬁned exogenously to the processes of market interaction.

james m. buchanan 987

distributional attributes alongside shared value potential are omnipresent in those

goods and services actually supplied and ﬁnanced through collective auspices. Collec-

tively supplied goods, like all others, cost something, and when the taxing side of the

public account is reckoned with, distributional conﬂict may swamp any underlying

collective action logic.

Further, public and social choice analysts were slow to recognize that the alter-

natives for collective “choice” are themselves critically dependent on the processes

and rules through which such “choices” are made. Two elementary examples: (1)the

institutions of majority rule introduce a necessary discriminatory bias toward alter-

natives that beneﬁt members of majority coalitions at the expense of those persons

in the minority; (2) the absence of party supremacy and discipline, contrasted with

the presence in parliamentary systems, tends to generate logrolling among special

constituency coalitions, exempliﬁed by the “pork-barrel” politics of the United States.

7 Toward an Operational Science

of Politics

.............................................................................

By concentrating attention on individual behavior in stylized majoritarian voting

processes, with some understanding of both the institutional complexity of structures

and the origins of collective action alternatives, analysts have developed an opera-

tional science of politics or collective action. We do, indeed, understand more about

how politics works than we did at the half-century. Problems arise and criticisms

emerge when analysis is extended to eﬀorts to develop a predictive science, which must

impute speciﬁc objectives to participants. What do persons seek to achieve through

political action?

Direct transposition of the motivations imputed to persons in their private market

behavior to behavior in politics allows for the generation of readily falsiﬁable hy-

potheses. But is net wealth, in some objectively measurable sense, comparably plau-

sible as a motive for behavior when participants recognize both that the alternative

ﬁnally selected is “public,” and hence applicable to all members of the polity, and,

further, that the individual who votes in the processes of selection is one among

many whose action determines the ﬁnal outcome? Overly zealous extension of the

economic models of behavior to political settings has provided soft targets for critics

of the whole modern enterprise, often described under the “rational choice” rubric.

These critics do not advance eﬀective explanatory alternatives, and they fail to

understand that their arguments apply only to the predictive elements in the posi-

tive science. The arguments do not undermine the basic rational choice postulates,

which do not require the elevation of net wealth to a position of dominance among

the motives for political actors. Practitioners in public choice–social choice should,

however, acknowledge the limits of their explanatory models. The rational choice

988 politics and scientific enquiry

perspective, whether or not net wealth is included as an important motivating force,

is one among several perspectives, each of which may yield some understanding of

the very complex reality that is “politics.”

8 Toward Normative Relevance

.............................................................................

As noted, science is aimed at understanding what is. But one ultimate purpose of any

such understanding is improvement; and, primarily, we seek to understand in order

to “make things better.” For the linkage here to function, however, we require a scalar

that will tell us what “betterness” is. We have little or no diﬃculty with many ordinary

scientiﬁc endeavors. Medical science has as its ultimate purpose the improvement of

the health of women and men. Physical science (with some exceptions) aims at im-

provements in our ability to utilize natural sources of energy more fully. Information

science and technology allow us to transmit information more quickly.

There is no comparable criterion for improvement in the social sciences. What,

then, is the ultimate normative purpose of achieving scientiﬁc understanding in

economics or political science, or, inclusively, in political economy? By understanding

more about how social interaction processes work, we can say that improvements or

reforms in these processes can allow participants in the polity to secure more value.

But there is no independent criterion for what value is. We are reduced to saying that

reform involves change that will allow persons to generate more of whatever it is that

they value.

Classical utilitarianism circumvented this problem by elevating utility to the po-

sition of the criterion for improvement. The social sciences found their ultimate

purpose in discovering institutional changes to be made that would increase utility,

which was both objectively measurable and interpersonally comparable. Utilitari-

anism remained the dominating normative foundation for neoclassical economics

in particular and residues remain in modern analysis. By the 1930s, however, and

speciﬁcally with the powerful arguments from Robbins (1932), the emptiness of utili-

tarianism was laid bare. There was no ultimate criterion against which reforms might

be evaluated.

Economists seemed to be placed in the position where their science had no norma-

tive implications. Fortunately, however, Pareto was discovered, who had recognized

the problem early in the century and who had provided a basis from which a “new”

welfare economics could be constructed. This branch of enquiry, sometimes called

“theoretical welfare economics,” oﬀered the normative base for the economics of the

middle-century decades.

The Pareto criterion was profoundly individualistic in a basic philosophical sense,

making it categorically diﬀerent from utilitarianism. Since aggregation over separate

individuals was not possible, the criterion for improvement became the enhancement

of value for each and every person, considered separately, and value as determined

james m. buchanan 989

within the psyche of the person alone. Somewhat interestingly, however, this highly

restrictive condition did not totally eliminate normative implications from economic

science. Economists could infer that value enhancement for all parties becomes pos-

sible when exchanges are voluntary and when entry and exit into exchanges are free.

Economic policy aimed at the opening up of exchange or market opportunities was

normatively justiﬁed as value enhancing, even when values are only those expressly

shown by voluntary actions of separate persons. The science of economics, which

produced a better and better potential understanding of this relationship, was put on

all fours with the other sciences.

As noted earlier in connection with the mid-century development of public goods

theory, the welfare economists of those decades were primarily concerned with

demonstrating when and under what conditions voluntary exchange relationships, or

markets, fail to ensure increases in value, as assessed against the Pareto criterion. To

the extent that elements of “publicness” are technologically embedded in the qualities

of goods, or, to put the same thing diﬀerently, that extra-exchange externalities are

present, voluntary exchanges, as normally interpreted, will increase the values for

some individuals, but, at the same time, may reduce or increase values for others

outside the exchanges. Potential values may remain unexploited.

Modern public choice theory made it clear that the welfare theorists of mid-

century moved too quickly from their diagnoses of market failure to the challenges

involved in politicized or collectivized corrections for such failures, and without

explicit eﬀort to analyze the working of politics. There are two separate but related

avenues of enquiry to be followed up for those settings where voluntary exchanges,

as normally interpreted, are acknowledged to be incomplete in the exhaustion of po-

tential values. The ﬁrst, which we may associate with the work of R. H. Coase (1960),

involves an extension of the notion of voluntary exchanges to encompass more than

the simple two-party traders in goods and services conventionally considered. The

second, which we may associate with Knut Wicksell (1896), involves a transposition

of the logic of exchange to politics itself, precursory to the late-century development

of constitutional political economy or, possibly, labeled as welfare politics. These two

research programs may be discussed in turn.

9 Extending the Exchange Nexus

.............................................................................

If persons are free to enter into exchanges, one with another, and if exchanges oﬀer

prospects for mutual gains, such exchanges will occur. This elementary insight is the

starting point for Coase’s critique of neoclassical welfare economics, with its emphasis

on market failures. From the Coasian perspective, the existence of an externality is a

signal that unexploited gains from further possible exchange remain, and, if persons

are at liberty to enter such exchanges, externalities will be internalized. Markets will

not fail to exhaust potential values. If apparent externalities seem to be observed,

990 politics and scientific enquiry

failure to exploit these opportunities may be traced to the presence of prohibitive

transactions costs barriers.

The normative implications are clear; participants in the economy will be able to

secure more of that which they separately value if barriers to possible exchanges are

removed or reduced. Particular focus is to be placed on the facilitation of exchanges

in legal rights to undertake or to prevent activities, often treated as beyond the con-

ventional nexus of exchange in goods and services. The implications of the Coasian

analytical thrust for the development of a science of politics are largely indirect. To

the extent that the exchange or market nexus can be extended by the clariﬁcation

of rights and improved enforcement of contracts, any value-based argument for

politicization or collectivization becomes less convincing. On the other hand, the

argument for collective or political provision of protective state services, those that

involve guarantees of freedom to enter and exit exchanges, along with protection

of person and property and enforcement of voluntary contracts, is strengthened. In

the idealized setting that may be inferred from the Coasian analytical structure, the

protective state would be strong, and the productive state would be quite limited in

scope.

10 Wicksell and the Unanimity

Benchmark

.............................................................................

Knut Wicksell’s seminal contribution was written well before the twentieth-century

developments in welfare economics and in public choice (Wicksell 1896). His in-

terest was in extending the analytical framework of neoclassical economic theory,

which he had been instrumental in constructing, to the public or collective sector

of economic interaction. Wicksell started from the presumption that members of

an organized polity, through their representative agents, will seek to secure values

potentially promised by shared goods and services. But how will the collective action

be organized when the costs of providing such goods and services necessarily involve

distributional considerations? Wicksell was under no delusion to the eﬀectthatrepre-

sentative agents seek out some general or public interest rather than that of their own

constituents. And he stressed that majority voting in legislative assemblies contained

no protection against exploitation of minorities and no guarantee that the aggregate

beneﬁts of collective action exceed the costs.

How could the decision structure be constructed so as to ensure that the values

promised by shared goods and services could be secured in such fashion to ensure

that all members of the sharing group receive net gains? How could both economic

eﬃciency and justice be guaranteed? These questions made it necessary for Wicksell

⁷ For an explicit discussion of the distinction between the protective and productive state, see

Buchanan 1975.

james m. buchanan 991

to challenge the normative legitimacy of majority voting, as the dominating instru-

ment of “democracy.” And, although he did not himself appreciate the implications,

Wicksell’s contribution becomes precursory to later developments in constitutional

enquiry.

If all members of a sharing group are to be ensured net gains, how must the costs

be distributed? Clearly costs must be distributed in some relation to the distribution

of gains. The ultimate test can be found only in agreement among all members of

the group. If all persons in the sharing group agree to a distribution of the costs, this

agreement in itself becomes the criterion for eﬃciency. The voting rule must be that

of unanimity, which becomes the benchmark institution.

Wicksell recognized that, even in a stylized limiting case, strict implementation

of the unanimity rule would be diﬃcult, since each member of the decision-making

group would be placed in a position of potential veto power. In practice, some fall-

back rule, between simple majority on the one hand and unanimity on the other,

would be the best that could be expected to work. Wicksell’s analysis establishes the

legitimacy of qualiﬁed majority rules. Majority rule is eﬀectively dethroned as the

be-all and end-all of democratic politics.

It is worth noting that Wicksell’s unanimity criterion for collective action is iden-

tical to the Pareto criterion for assessing the eﬃcacy of any change in the constraints

within which persons engage in exchanges. All members of the relevant political

group must be brought into agreement, or at the least acquiescence, on a proposed

change in order to ensure against discriminatory exploitation. Both eﬃciency and

justice dictate that consensus be attained before collective action, with its necessarily

attendant coercion, be taken.

11 The Calculus of Consent

.............................................................................

The Wicksellian analysis suggests that, absent agreement, there is no basic normative

legitimacy for collective action. Developments in scientiﬁc understanding of how pol-

itics work oﬀer little justiﬁcation for the whole enterprise. Clearly, general agreement

is rarely observed to be present in any collective action; almost always there are net

losersaswellasnetgainers.

Viewed in retrospect, the primary contribution of Buchanan and Tullock, in their

1962 book The Calculus of Consent,wastooﬀer at least a partial resolution of the

Wicksellian dilemma. If politics is conceptualized as a two-stage or two-level process

(the constitutional and post-constitutional), with a separation between the politics

of selecting the basic rules and the ordinary action taken within these rules, the

agreement criterion may be applied diﬀerently, with more acceptable implications.

The Wicksellian criterion may be shifted from the level of ordinary politics to that of

the rules that constrain such politics—to the constitution, inclusively deﬁned. And

while there may be net losers in almost any collective action taken at the second stage,