Weingast B.R., Wittman D. The Oxford Handbook of Political Economy
Подождите немного. Документ загружается.
172 coalition government
immediate payoff) parties could wait for political-economic circumstances associated
with a higher expected government duration.
OneoftheappealingfeaturesofboththeMorelliandMerloandWilson
approaches is their close connections to the data. Morelli’s approach, for example, is
motivated by the so-called Gamson’s Law which states that cabinet portfolios among
the members of the ruling coalition should be allocated in proportion to their (nor-
malized) seat share (Gamson 1961).
22
Models based on Baron and Ferejohn (1989)
cannot account for this regularity since they imply a proposer premium regardless
of the distribution of seats. Morelli’s model, however, does imply Gamson’s Law if
proposer selection is done proportionally to seat share.
23
Merlo and Wilson’s (1998) model also implies portfolio distributions close to
Gamson’s Law provided the negotiating parties have low discount factors—i.e. if they
are very patient. If parties are impatient, however, proposers are able to capture higher
payoffs.
24
Adifferent variant of efficient bargaining third was recently proposed by Baron and
Diermeier (2001). Their model focuses on an environment where parties bargain not
only over distributive benefits, but also over (multidimensional) policy. The model
proceeds as follows. A proposer party (“formateur”) is selected proportionally to seat
share. That party then selects a “proto-coalition” (Axelrod 1970), i.e. a list of parties
that agree to negotiate under unanimity rule. If the proposed coalitional agreement
fails to win a majority in the chamber, a caretaker government implements an ex-
ogenously given status quo policy. Bargaining within a proto-coalition is efficient
and leads to a unique policy outcome for each proto-coalition.
25
The intuition is
that during proto-coalitions parties can make side payments to each other using
the distributive benefits. This ensures that no matter how the bargaining process is
structured, the parties must agree on the same efficient policy.
26
In contrast to previous bargaining models, the policy location does not depend
on the details of the bargaining process such as recognition probabilities or discount
factors. It only depends on the ideal points of the members of the proto-coalition.
This allows Baron and Diermeier to embed their coalition formation model into
a more complicated, full equilibrium, model that includes an election state under
proportional representation
27
as well as analysis of policy dynamics across legislative
periods.
²² See also Browne and Franklin 1973; Browne and Fendreis 1980;SchofieldandLaver1985;Laverand
Schofield 1990.
²³ Diermeier and Merlo 2004 provide evidence that formateurs (i.e. proposers) are indeed selected
proportionally.
²⁴ The Morelli and Merlo–Wilson approach has led to an empirical re-evaluation of the data
underlying Gamson’s Law. The key question is whether the data show a proposer premium. See Warwick
and Druckman 2001; Ansolohabere et al. 2003;Frechette,Kagel,andMorelli2004.
²⁵ For an empirical test of the Baron–Diermeier model and other coalition models see Carrubba and
Vo lde n 2000.
²⁶ If preferences are assumed to be quadratic, that policy is the centroid of the parties’ ideal policy
locations.
²⁷ See Austen-Smith and Banks 1988 for an electoral model under proportional representation that
uses a simplified version of sequential bargaining.
daniel diermeier 173
6 Non-cooperative Bargaining:The
Question of Cabinet Stability
.............................................................................
After having tackled the problem of government formation, non-cooperative models
were then applied to the issue of cabinet stability. A first connection was established
by Lupia and Strom (1995) who proposed a game-theoretic model of cabinet termi-
nation by re-evaluating the key notion of “critical events” (Browne, Frendreis, and
Gleiber 1984, 1986, 1988). Lupia and Strom suggested that external events should be
interpreted as the input of a coalition bargaining process, representing each party’s
outside option if the current government falls. Whether a cabinet terminates depends
on the attractiveness of this outside option compared to each party’s payoff from the
current government.
This interpretation of the impact of random events marks a potentially important
shift in the study of cabinet terminations. Cabinets are interpreted as equilibria of
an underlying bargaining process that must be sustained over time as the political
environment and thus the distribution of bargaining power among governing parties
changes.
The long-term impact of the Lupia–Strom model was mainly conceptual. The ac-
tual bargaining model is too simplistic to capture dynamics of coalition bargaining.
28
Their approach demonstrated that what is needed is a multistage bargaining approach
with an explicitly modeled stochastic process to capture the changing political envi-
ronment that may lead to critical events.
AfirststepinthisdirectionistheDiermeier–Merlomodel(DiermeierandMerlo
2000). As in Lupia and Strom, Diermeier and Merlo conceptualize coalition gov-
ernments as sustained equilibria in an underlying bargaining game. Bargaining is
assumed to be efficient proto-coalition bargaining, similar to Baron and Diermeier
(2001), but is now embedded in a multistage bargaining game with random shocks
both to electoral prospects and to the parties’ reservation values. Such shocks may
lead to the termination of government before the end of the game. In that case,
a new cabinet would take office. Cabinets may avoid termination by redistribut-
ing (“reshuffle”) distributive benefits among the incumbent parties in response to
external events.
Government formation consists of two steps. First, a formateur is selected propor-
tional to seat share. Then that formateur chooses a proto-coalition which is assumed
to bargain efficiently. Any proto-coalition agreement needs to be approved by a
chamber majority.
Diermeier and Merlo are particularly interested in the study of minority gov-
ernment. They identify a government with an allocation of cabinet portfolios: a
party that supports a minority government on critical votes but does not hold any
portfolios is not part of the government, but is only part of the supporting coali-
tion. Government policy is negotiated in the cabinet subject to majority approval
²⁸ See Diermeier and Stevenson 2000; Diermeier and Merlo 2000.
174 coalition government
in parliament. This assumption is justified by the observation that in parliamentary
democracies ministries, not independent agencies, implement government policies.
Any policy bargain that involves parties without a seat at the cabinet table therefore
would lack credibility.
29
The inability to bargain credibly over policy with parties
outside the cabinet is a key feature of the model that is then used to study minority
cabinets and cabinet stability. Governments also have control over perks (e.g. seats
on public boards) that can be distributed to any party in the legislature. These are the
keys to maintaining support for a minority government.
The first important result in the Diermeier-Merlo model is that both minority and
super-majority cabinets can occur as equilibrium phenomena. Moreover, minority
and surplus coalitions are not rare exceptions, but may form for all parameter values.
The traditional literature had viewed minority and surplus governments as an ex-
ception that needed to be explained (e.g. Strom 1990). The Diermeier–Merlo model,
however, suggests exactly the opposite: in a single-period model, minimal winning
coalitions would never form. The underlying reason is that formateurs can choose
from many viable coalition governments. They will thus choose the government that
yields the highest payoff. In the case of risk-averse preferences this implies that if the
formateur party ever benefits from policy compromise, it is even better to include
all other parties into the bargain rather than just one. This case corresponds to a
super-majority government. If policy compromise is too expensive for the formateur,
a minority government results.
So, why do we observe minimal winning coalitions?
30
Surprisingly, minimal win-
ning coalitions only occur because of dynamic considerations. The key insight relies
on the importance of bargaining inefficiency for cabinet termination. Diermeier and
Merlo show that as long as bargaining is efficient, governing parties will have an
incentive to reshuffle distributive benefits in response to external shocks rather than
taking the risk of being left out of the ensuing new government. If efficient bargaining
is not possible, however, as in the case of bargaining between the minority govern-
ment and the parties in its supporting coalition, governments may fall. It follows that
during initial cabinet formation formateurs need to anticipate the possible costs of
early termination and reshuffles. Diermeier and Merlo show that for some parameter
values minimal winning cabinets are the best deal. On the one hand, they avoid early
termination; on the other hand, renegotiations occur only with one party, not with
two. For other parameter values surplus or minority coalitions are optimal. So all
observed cabinet types can be recovered as equilibrium phenomena.
One important consequence of the Diermeier and Merlo framework is that for-
mateurs anticipate the benefits and costs from maintaining a given cabinet over time.
Both costs and benefits depend on the cabinet type. So, depending on the parameters,
formateurs will choose different cabinet types, including some minority cabinets that
are likely to terminate early. The stability and the relative occurrence of different types
of governments are thus jointly determined in equilibrium. As we will see in the next
²⁹ A similar argument was made by Laver and Shepsle 1990.
³⁰ They make up about half of all observed coalitions.
daniel diermeier 175
section, this insight has important consequences for the empirical study of coalition
government.
7 The Non-cooperative Approach:
Consequences for Empirical Research
.............................................................................
Once we model the formation and stability of governments as equilibrium phenom-
ena the methodological foundations for empirical studies of coalitions need to be
reconsidered. To see the problem it is useful to revisit existing empirical studies of
cabinet stability. From Strom (1985) to Diermeier and Stevenson (1999), all studies
had been regression models, i.e. a suitably defined stochastic process with sets of
covariates that influence termination probabilities. Since Strom (1985), all these mod-
els have contained institutional characteristics, both of the polity (e.g. constitutional
features) and of cabinet (e.g. its majority status). The main lesson, however, from
an approach that models cabinets as equilibria is that expectations about cabinet
duration influence the choice of initial cabinets; they are jointly selected in equi-
librium. Empirically, this implies that we cannot treat cabinet-specific features as
proper independent variables in a regression model. Rather, we are facing a selection
problem.
A series of recent papers (Merlo 1997; Diermeier, Eraslan, and Merlo 2003a, 2003b)
has tried to resolve these problems by using a structural estimation approach. The
approach consists in specifying a bargaining model of government formation, es-
timating the model’s parameters, assessing the ability of the model to account for
key features of the data, and then using the estimated structural model to conduct
(counterfactual) experiments of comparative institutional analysis.
The first example is Merlo (1997) who uses the Merlo–Wilson framework to con-
duct a structural estimation of government formation and cabinet stability in Italy.
Diermeier, Eraslan, and Merlo (2003a, 2003b) propose a model of proto-coalition
bargaining similar to Baron and Diermeier (2001) and Diermeier and Merlo (2000)
to study the effect of constitutional characteristics on cabinet formation and duration.
An instructive example is the analysis of governmental bicameralism (Diermeier,
Eraslan, and Merlo 2003b).
31
In countries with governmental bicameralism
(e.g. Italy), both chambers need to approve a new government, but a vote in either
chamber terminates the government. This may suggest that removing governmen-
tal bicameralism would significantly improve the average duration of governments.
Diermeier, Eraslan, and Merlo show that this intuition is flawed. Using data from
Belgium they find that eliminating bicameralism does not affect government dura-
bility, but does have a significant effect on the composition of governments leading
to smaller coalitions. Surplus governments are virtually eliminated, but are being
³¹ See Cultrone and McCarty, this volume.
176 coalition government
replaced not by minimal winning governments, but by minority governments. These
surprising results are due to an equilibrium replacement effect. Removing bicam-
eralism does indeed affect the relative durability of possible coalitions of different
type, but this then affects which cabinets are selected in equilibrium. That is, for each
coalition type (minimal winning, minority, etc.) expected duration may increase, but
under unicameralism formateurs may now select minority coalitions when (under
bicameralism) they would have selected coalitions with majority status. In the Belgian
case these two effects cancel each other out resulting in no change in average govern-
ment durability, but a sharp increase in the proportion of minority governments.
8 Conclusion
.............................................................................
The last fiften years have seen a fundamental transformation in the study of coali-
tion government. Institutionalist approaches increasingly dominate empirical and
theoretical research programs. Moreover, paralleling the development in other ar-
eas of formal political science, non-cooperative game theory has become the most
widely used methodology. On the theoretical dimension, this approach has led to
models that integrate government formation and government termination. These are
promising developments, even though the existing models can only be considered
first steps. For example, both the Lupia–Strom model and Diermeier–Merlo are
restricted to the three-party case.
On the one hand, this approach marks a promising shift in the study of coali-
tion government. In particular, the specification of explicit models of inter-party
bargaining is likely to focus attention on the role of institutional features of the
bargaining environment in determining cabinet formation and stability. On the other
hand, the bargaining approach creates various challenging methodological problems.
First, Diermeier and Stevenson (2000) showed that testing the Lupia–Strom model
requires careful specifications of the stochastic model that precludes the usage of
“off-the-shelf” event study methods. Second, Merlo (1997) pointed out that changing
expectations about government duration (due to external events such as the release of
macroeconomic data) may lead party leaders to delay the formation of a government.
Hence, governments that actually form should be viewed as the result of strategic
selection by the members of the proto-coalition. Third, Diermeier and Merlo (2000)
and Diermeier, Eraslan, and Merlo (2003a, 2003b) have shown that expectations
about government duration may also influence which governments are chosen in
the first place. Together, these results imply that government type, formation time,
and government duration are all simultaneously determined in equilibrium. That
is, taking the concept of governments-as-equilibria seriously may require a radical
departure from existing empirical approaches that rely on regression-based reduced
form specifications. A shift to structural models offers one promising, but technically
challenging, new direction.
daniel diermeier 177
References
Ansolabehere,S.,Snyder,J.M.J.,Strauss,A.B.,andTing,M.M.2003. Voting weights
and voter advantages in the formation of coalition governments. Mimeo.
Austen-Smith,D.,andBanks,J.S.1988. Elections, coalitions and legislative outcomes.
American Political Science Review , 82: 405–22.
1990. Stable governments and the allocation of policy portfolios. American Political Sci-
ence Review, 84: 891–906.
2005. Positive Political Theory,ii:Strategies and Structures.AnnArbor:Universityof
Michigan Press.
Axelrod,R.1970. Conflict of Interest. Chicago: Markham.
1972. Where the votes come from: an analysis of electoral coalitions. American Political
Science Review, 66: 11–20.
Banks,J.S.,andDuggan,J.2000. A bargaining model of collective choice. American Political
Science Review, 94: 73–88.
2003. A bargaining model of legislative policy-making. Mimeo.
Baron,D.P.1989. A noncooperative theory of legislative coalitions. American Journal of
Political Science, 33: 1048–84.
1991. A spatial theory of government formation in parliamentary systems. American
Political Science Review, 85: 137–65.
1993. Government formation and endogenous parties. American Political Science Review,
87: 34–47.
1996. A dynamic theory of collective goods programs. American Political Science Review,
90: 316–30.
1998. Comparative dynamics of parliamentary governments. American Political Science
Review, 92: 593–609.
and Diermeier,D.2001. Elections, governments, and parliaments in proportional rep-
resentation systems. Quarterly Journal of Economics, 116: 933–67.
and Ferejohn,J.A.1989. Bargaining in legislatures. American Political Science Review,
83: 1181–206.
Binmore,K.1986. Bargaining and coalitions. In Game-Theoretic Models of Bargaining,ed.
A. Roth. Cambridge: Cambridge University Press.
Browne,E.C.,andFranklin,M.1973. Aspects of coalition payoffs in European parliamentary
democracies. American Political Science Review, 67: 453–69.
and Frendreis,J.P.1980.Allocatingcoalitionpayoffs by conventional norm: assessment
of the evidence for cabinet coalition situations. American Journal of Political Science, 24:
753–68.
and Gleiber,D.W.1984. An “events” approach to the problem of cabinet stability.
Comparative Political Studies, 17: 167–97.
1986. The process of cabinet dissolution: an exponential model of duration and stability
in western democracies. American Journal of Political Science, 30: 628–50.
1988. Contending models of cabinet stability: a rejoinder. American Political Science
Review, 82: 923–30.
Carrubba,C.J.,andVolden,C.2000. Coalitional politics and logrolling in legislative insti-
tutions. American Journal of Political Science, 44: 261–77.
Diermeier,D.1997. Explanatory concepts in positive political theory. Mimeo, Graduate
School of Business, Stanford University.
Eraslan,H.K.K.,andMerlo,A.2003a. A structural model of government formation.
Econometrica, 71: 27–70.
178 coalition government
Diermeier,D.2003b. Bicameralism and government formation. Mimeo.
and Krehbiel,K.2003. Institutionalism as a methodology. Journal of Theoretical Politics,
15: 123–44.
and Merlo,A.2000. Government turnover in parliamentary democracies. Journal of
Economic Theory, 94: 46–79.
2004. An empirical investigation of coalitional bargaining procedures. Journal of Public
Economics, 88: 783–97.
and Stevenson,R.1999. Cabinet survival and competing risks. American Journal of
Political Science, 43: 1051–98.
2000. Cabinet terminations and critical events. American Political Science Review,
94: 627–40
and van Roozendaal,P.1998. The duration of cabinet formation processes in western
multi-party democracies. British Journal of Political Science, 28: 609–26.
and Vlaicu,R.2005. Voting to propose. Mimeo, Northwestern University.
Eraslan,H.,andMerlo,A.2002. Majority rule in a stochastic model of bargaining. Journal
of Economic Theory, 103: 31–48.
Fréchette,G.R.,Kagel,J.H.,andMorelli,M.2004. Behavioral identification in coalitional
bargaining: an experimental analysis of demand bargaining and alternating offers. Mimeo.
Gamson,W.A.1961, A theory of coalition formation. American Sociological Review, 26: 373–82.
Kalandrakis,T.2004. Genericity of minority government: the role of policy and office.
Mimeo.
King,G.,Alt,J.,Burns,N.,andLaver,M.1990. A unified model of cabinet dissolution in
parliamentary democracies. American Journal of Political Science, 34: 846–71.
Kramer,G.1972. Sophisticated voting over multidimensional choice spaces. Journal of Math-
ematical Sociology, 2: 165–80.
Laver,M.2003. Government termination. Annual Review of Political Science, 6: 23–40.
and Schofield,N.1990. Multiparty Government. Oxford: Oxford University Press.
and Shepsle,K.A.1990. Coalitions and cabinet government. American Political Science
Review, 84: 873–90.
(eds.). 1994. Cabinet Ministers and Parliamentary Government. Cambridge: Cam-
bridge University Press.
1996. Making and Breaking Governments: Cabinets and Legislatures in Parliamentary
Democracies. Cambridge: Cambridge University Press.
1998. Events, equilibria and government survival. American Journal of Political Sci-
ence, 42: 28–54.
Lupia,A.,andStrom,K.1995. Coalition termination and the strategic timing of parliamentary
elections. American Political Science Review, 89: 648–69.
McKelvey,R.D.1976. Intransitivities in multidimensional voting models and some implica-
tions for agenda control. Journal of Economic Theory, 12: 472–82.
1986. Covering, dominance, and institution-free properties of social choice. American
Journal of Political Science, 30: 283–314.
Martin,L.W.,andStevenson,R.T.2001. Government formation in parliamentary democ-
racies. American Journal of Political Science, 45: 33–50.
and Vanber,G.2003. Wasting time? The impact of ideology and size on delay in coalition
formation. British Journal of Political Science, 33: 323–32.
Merlo,A.1997.Bargainingovergovernmentsinastochasticenvironment.Journal of Political
Economy, 105: 101–31.
and Wilson,C.1995. A stochastic model of sequential bargaining with complete infor-
mation. Econometrica, 63: 371–99.
daniel diermeier 179
1998.Efficient delays in a stochastic model of bargaining. Economic Theory, 11: 39–55.
Mershon,C.1996. The costs of coalition: coalition theories and Italian governments. Ameri-
can Political Science Review, 90: 534–54.
Morelli,M.1999. Demand competition and policy compromise in legislative bargaining.
American Political Science Review, 93: 809–20.
Plott,C.1967. A notion of equilibrium and its possibility under majority rule. American
Economic Review, 57: 787–806.
Schofield,N.J.,andLaver,M.1985, Bargaining theory and portfolio pay-offsinEuropean
coalition governments. British Journal of Political Science, 15: 143–64.
Sened,I.,andNixon,D.1998. Nash equilibria in multiparty competition with “stochas-
tic” voters. Annals of Operations Research, 84: 3–27.
Selten,R.1992. A demand commitment model of coalition bargaining. In Rational Interaction
Essays in Honor of John Harsanyi, ed. R. Selten. Berlin: Springer Verlag.
Shepsle,K.A.1979. Institutional arrangements and equilibrium in multidimensional voting
models. American Journal of Political Science, 23: 27–60.
Strom,K.1985. Party goals and government performance in parliamentary democracies.
American Political Science Review, 79: 738–54.
1988. Contending models of cabinet stability. American Political Science Review,
82: 923–30.
1990. Minority Government and Majority Rule. Cambridge: Cambridge University Press.
Warwick,P.1992a. Ideological diversity and government survival in western European par-
liamentary democracies. Comparative Political Studies, 25: 332–61.
1992b. Rising hazards: an underlying dynamic of parliamentary government. American
Journal of Political Science, 36: 857–76.
1992c. Economic trends and government survival in west European parliamentary
democracies. American Political Science Review, 86: 875–87.
1994. Government Survival in Parliamentary Democracies. Cambridge: Cambridge
University Press.
and Druckman,J.N.2001. Portfolio salience and the proportionality of payoffsin
coalition governments. British Journal of Political Science, 31: 627–49.
and Easton,S.T.1992. The cabinet stability controversy: new perspectives on a classic
problem. American Journal of Political Science, 36: 122–46.
Winter,E.1994a. The demand commitment bargaining: a snowball in cooperation. Economic
Theory, 4: 255–73.
1994b. Non-cooperative bargaining in natural monopolies. Journal of Economic Theory,
64: 202–20.
chapter 10
..............................................................
DOES
BICAMERALISM
MATTER?
..............................................................
michael cutrone
nolan m
C carty
1 Introduction
.............................................................................
Perhaps the most conspicuous variation in the form of modern legislatures concerns
the practice of granting legislative authority to two separate chambers with distinct
memberships. While the majority of national governments empower but a single
chamber, at least a third of national legislatures practice some form of bicameralism,
as do forty-nine of the fifty American state governments.
1
Bicameralism is also at the
center of many contemporary debates about the design of institutions in new democ-
racies. For example, many proposals for the post-Hussein constitution of Iraq have
identified bicameralism as an important ingredient in building a stable democracy in
a nation beset with strong religious and ethnic cleavages.
2
Scholars have made a number of arguments to explain the emergence of
bicameral legislatures. One of the most common arguments for the emergence of
bicameralism in Britain and its American colonies is that it helped to preserve “mixed
¹ Tsebelis and Money 1997, 15, define bicameral legislatures as “those whose deliberations involve two
distinct assemblies.” This definition, however, masks considerable variation in the roles of each chamber
in policy-making. We define bicameralism more narrowly, as the requirement of concurrent majority
support from distinct assemblies for new legislation.
² These proposals and the positions of different Iraqi groups are discussed in Public International
Law and Policy Group 2003.
michael cutrone & nolan mC carty 181
governments,” ensuring that upper-class elements of society were protected
(Wood 1969; Tsebelis and Money 1997). In such settings, bicameralism allowed the
upper chamber, dominated by aristocrats, to have a veto on policy. More generally,
an explicit role of some bicameral systems has been the protection of some minority
which is over-represented in the upper chamber.
A second rationale for bicameralism is the preservation of federalism. The United
States, Germany, and other federal systems use a bicameral system in order to ensure
the representation of the interests of individual states and provinces, as well as the
population of the country. Under “federal bicameralism,” the lower house is typically
apportioned on the basis of population, while the upper house is divided amongst
the regional units. Some countries, such as the United States, provide equal represen-
tation for the states regardless of their population or geographic size, while others,
like the Federal Republic of Germany, unequally apportion the upper chamber by
providing additional representation to the larger units.
Despite its prominence, the role of bicameralism in contemporary legislatures has
not received the scholarly attention that other legislative institutions have. In this
chapter, we review and analyze many of the arguments made on behalf of bicamer-
alism using the tools of modern legislative analysis—the spatial model, multilateral
bargaining theory, and a game of incomplete information.
3
This analytical approach
allows us to distinguish the effects of bicameralism from those of other institutional
features that are often packaged with it, such as super-majoritarian requirements,
differing terms of office, and malapportionment.
2 Spatial Models of Policy-making
.............................................................................
The spatial model of policy-making has become the workhorse model in the study of
legislative institutions. Its stark parsimony makes tractable the analysis of a number
of institutional arrangements. Such models have proven quite useful in studying the
consequences of bicameralism and other multi-institutional policy-making settings
(i.e. Tsebelis and Money 1997;FerejohnandShipan1990).
Before we draw out the implications for bicameralism, consider a baseline unicam-
eral model. Assume that a unicameral legislature has an odd number of members, n,
with ideal points, x
i
, arrayed along a single ideological dimension represented by the
real line. We index the ideal points from left to right so that x
1
is the leftmost member
and x
n
is the rightmost member. The legislature seeks to pass legislation to change a
policy with a status quo q, which is also represented by a point on the spectrum. We
assume that each legislator has single-peaked, symmetric preferences so that member
i weakly prefers y to z if and only if |y − x
i
|≤|z − x
i
|, (that is, if and only if y is
³ We will not review the arguments about the role of bicameralism in the formation and duration of
parliamentary cabinets (Druckman and Thies 2002; Diermeier, Eraslan, and Merlo 2003).