keeping tabs on partners the logic of delegation in coalition governments michael f. thies...
Post on 19-Dec-2015
229 views
TRANSCRIPT
Keeping tabs on partnersThe logic of delegation in Coalition Governments
Michael F. Thies
• Governments require delegation:
• PARLIAMENT GOVERNMENT
• In multiparty coalitions, parties don’t enjoy full control over ministers from other parties
Governments and delegation• Governments require delegation:
PARLIAMENT GOVERNMENT
• In multiparty coalitions, parties don’t enjoy full control over ministers from other parties.
• Multiparty governments face greater delegation problems than single-party government for at least 2 reasons
1. Divergence of preferences2. Difficulty of sanctions
Two options to adress delegation problems:
1. MINISTERIAL GOVERNMENTEach minister enjoys virtual dictatorship over his/her peculiar jurisdiction (or at least the right to propose policies in that area; see Laver & Shepsle coalition theory).
2. MANAGED DELEGATIONDelegation without abdication:
a) Monitoring what agents are doingb) Check the ability of each agent to act
Ministerial government: disadvantages• Risks due to uncertainty: 1. Misperceived ideal points2. Unforeseen issues3. Change of the relevance of an issue
In these circumstances the net cost of having ceded competence to another party rises
• Opportunity costs: why instead of “abdication” a mutual deference and a compromise?Compromises are unstable
Managed delegationIf Delegation can be managed at relatively low cost to the coalition parties, then the coalition might be able to implement the compromise package making all coalition parties better off.
• EX POST CHECKS• They are insufficient, since ministers might be able
to deceive their cabinet colleagues into accepting proposals that make their colleagues worse off.
• EX ANTE CHECKS• Interministerial committees• Overlapping jurisdictions• Junior ministers
HYPOTHESESGeneral hypothesis• Mechanisms of control are used more extensively
where the risks and opportunity costs of abdication are higher.
• H1: Parties are more likely to monitor each other’s ministers the less contrained ministers are by other “institutional checks” (federal systems, asymmetric bicameralism, role of the Prime Minister).
• H2: Parties are more likely to monitor partner parties’ ministers who hold portfolios of greater salience.
• H3: Parties are more likely to monitor partner parties’ ministers whose preferences are more extreme.
• H4: Parties are more likely to monitor each others’ ministers when uncertainty is greater.
When the general salience is low, namely when each party cares much more about the jurisdiction it controls and much less about the jurisdiction controlled by its partners there is not much difference between an issue by issue compromise (x) and a ministerial government.
Empirical analysis• Dependent variable: appointment of Junior Ministers
(JM)• Only 2 hypotheses are tested • H1: Parties are more likely to monitor each other’s
ministers the less contrained ministers are by other “institutional checks” (federal systems, asymmetric bicameralism, role of the Prime Minister).
• H2: Parties are more likely to monitor partner parties’ ministers who hold portfolios of greater salience.
Five cases:• Italy 1970-1989• Germany 1966-1990• Netherlands 1971-1994• Japan 1993-1996• Japan 1965-1990
H1: monitoring vs institutional checksLess monitoring through hostile JM when Ministers are
otherwise constrained.
EXPECTATIONS• Germany: less monitoring as
• Strong prime minister (“chancellor democracy”)• Federal system• Inconguent bicameralism• Strong legislative committees
• Italy, Netherlands, Japan: more monitoring as • Weaker prime ministers• Unitary states• All bicameralism but “incongruent” only in Japan• Weak committes in Japan, slightly stronger in the
Netherlands, strong in Italy but with weak oversight powers
Evidence
• N. of ministers who have an hostile JM• N. of JM allocate to an hostile minister• Chances taken to allocate an hostile JM• Wasted JM
If a party in the coalition has few JM to allocate, it can’t cover all the hostile ministries,
so we have to look at the chances taken.
H2: salience of ministries vs hostile JMs
Parties are more likely to monitor partner parties’ ministers who hold portfolios of greater salience.
SALIENCE• Laver and Hunt 1992• Issue dimensions• Top, middle, bottom tiers
Alternative explanations about Germany ?
Two party coalition as a “rule” in Germany; A big party with a smaller ally
The major party doesn’t need to control the small one, which have few powers
The small is not interested in monitoring the other, being then unable to act.
Unsatisfying reasons:• Why the SPD didn’t try to monitor the FDP?• The bargainig power doesn’t correspond to
the party’s parliamentary representation.
Conclusions
• Presidential systems and multiparty governments are much in common:the conflict built into a multiparty coalition is very similar to the one inherent in a separation of powers.
• Once one recognizes the centrality of cross party compromise and managed delegation, multiparty parliamentarism emerges as distinct regime type from its single party cousin
Focus on legislatures in parliamentary democracies policymaking in periods of coalition government
Formal model of the interactions between coalition parties:• Multiparty governments involve delegation to cabinet
ministers • Delegation creates agency problems • Agency problems increase in severity on issues that
divide the coalition• Legislative review to overcome these tensions
Empirical evaluation: original dataset on 336 government bills from two parliamentary democracies
Results: legislatures play a more important role in parliamentary democracies than is usually appreciated
LITERATURE REVIEWAgency problems inherent to modern democratic
government:• citizens/government officials (Lupia and McCubbins
2000; Strøm 2000)• elected official/bureaucrats (Huber and Shipan 2002)• parties in a coalition:
– Cabinet-level institutions as a solution (Müller and Strøm 2000; Thies 2001)
– Legislative-level institutions as a solution:» They strenghten opposition parties (Powell
2000; Saalfeld 2000; Strøm 1990)» They strenghten government parties
“Principal-agent” frameworkTwo problematic features of policymaking by coalition
governments:
• Divergent preferences: policy and position-taking incentives of coalition parties continually put them in competition with one another
• Delegation to the cabinet provides considerable discretion to ministers in drafting legislation
Incentives for ministerial position-taking
Ministers, acting as agents of the cabinet as a whole, may not work fully in the interests of all their principals need for ex post oversight and correction (legislative scrutiny)
Modeling coalition policymakingCoalition composed of two parties (A and B) in a one-
dimensional policy space
A’s ideal point: 0 B’s ideal point: X > 0
Coalition compromise at X/t, where t ≥ 1
A
0
B
X>0
X/t
X/t is exogenously defined because M.V.are not interested in the bargaining process inside the government
As X increases, the policy divergence between the two parties increases
Policy divergence
Modeling coalition policymaking
A
0
B
X>0
X/t
Other elements affecting A and B’s Utility (different from policy outcomes)
c = resource and opportunity costs in challenging the ministerial draft
d = opportunity costs of drafting a proposal that ultimately does not become a law; audience costs for the drafting minister for appearing unable to implement the original version of the billa > 0 : the importance of the position taking for the minister’s party. As a increases party A is punished more severly for making proposal far from A’s ideal point.
Coalition policymaking game
A
0
B
X>0
X/t
A B
w
Choose w ()
accept
Amend at cost c
X/t
1. A minister belonging to A introduces a policy proposal w ϵ
2. In response, B can:a) accept A’s
proposalb) amend A’s
proposal at the cost c > 0 in order to have X/t implemented
3. The final policy p is :a) p = w if B acceptsb) p = X/t if B
amends. In this case A pays a cost d ≥ 0
Coalition policymaking gameWhat Party B will decide to do ?
to rewrite party A's bill if
- | - | - - | - |
- - -
-
B chooses
XX c X w
tX
X c X wt
Xc w w
t
' -
B chooses to accept party
XA s bill if c w w
t
Amendment threshold
Because rewriting the proposal made by party A is costly (c), B is willing to accept some deviation from the coalition compromise policy (X/t). As long as A’s proposal lies above this threshold, B is willing to accept it.
Coalition policymaking gameWhat Party A will decide to do ?
if B chooses to rewrite party A's bill then for A's minister
it is better to propose its ideal point in order to save costs
in terms of position taking
In other terms w =0
| | - | | | |
if B c
A
X XU a w d d
t t
hooses to accept party A's bill then for B
-
in other terms for A
(1 ) | | (1 ) | - |A
Xc w w
t
XU a w a c
t
Coalition policymaking gameWhat Party A will decide to do ?
For A's minister it is better insisting with its original bill even if
B chooses to rewrite party A's bill if
| | (1 ) | - |
1 ˆ
X Xd a c
t tX X aX
d c act t tdt ct aX act
aX dt ct act
t d c aX X
a
Position taking threshold
1) If the cost of challenging the minister's draft is so large relative
to policy divervence between A and B that - 0,
the minister can implement his ideal policy setting w = 0 (the case of
Xw c
t
ministerial autonomy in Laver and Shepsle 1996)2) If w>0, namely cost c for B are not so high to make
Xcounterproductive an amendment to . Party A' s minister
t makes a proposal that is acceptable to B and is closer to A' s ideal point than the
coalition compromise (ministerial drift), setting w = w
3) Under certain circumstances , even when 0, A can introduce a bill that maximizes position-taking benefits (setting w = 0) but that will be amended to the coalition compromise by B: when the poli
w
cy distance between the parties is sufficiently large
1ˆ t d c a
X Xa
ExpectationIf the coalition is sufficiently divided, ministers face strong incentives to deviate from coalition compromises to “take a position”, thus triggering attempts by other coalition members to challenge and amend draft legislation to bring it in line with the coalition compromise
Main Hypothesis
The incidence of amendmends to ministerial drafts should become more pronounced as the level of policy disagreement between parties increases
Why legislative review
Coalition partners possess resources that allow them to challenge and amend the draft legislation introduced by ministers
Resources: • Cabinet-level institutions:
Thies 2001: JMsMüller and Strøm 2000: JMs, cabinet commitees and “inner
cabinets”
• Legislative-level institutions: legislative review– Cost of monitoring from within the cabinet– Formal rules regarding policymaking– Systems of strong standing committees
THE DATA
Original dataset: all the modifications made to government-sponsored legislation between introduction and final resolution in the lower chambers of Germany (1983-1994) and Netherlands (1982-1994)
THE DATAGermany and Netherlands (“control by design” strategy):
• Policy differences between coalition parties • Similar coalition-specific features
– Governed by MW coalitions
• Similar institutional rules1. Nonmajoritarian democracies with proportional electoral system2. Legislative institutions that strenghten the role of legislators in policymaking:
a. Large number of specialized committees (composition, jurisdiction)b. Agenda for debate determined by (a committee of) the chamberc. Committee powers not restricted by a prior decision of the plenaryd. Similar voting procedure at the final voting stage (simple majority, quorum,
“successive” voting method)
• Representative of other European parliamentary democracies (institutional features and coalition-specific features –MWC)
• Drawback: unable to assess the impact of alternative institutional arrangements and coalition features on the extent of bill modification
• Advantages of the time frame chosen:– Relatively small number of years less prone to unmeasured temporal effect– Recent expert survey (Laver and Hunt 1992)
Unit of analysis: government bill
Excluded bills: – budget bills, budgetary adjustment bills, constitutional
bills– Bills non classifiable into Laver and Hunt’s eight policy
dimensions
336 government bills
• Delegation to cabinet ministers poses a principal-agent problem for coalition governments
• This problem is particularly pronounced on issues that divide the coalition internally
• Parties will take advantage of the opportunities presented by the legislative process (legislative review) to counteract the negative effect of delegation to ministers
• Implication: we should see more such “corrective” activity on bills that divide the coalition
OPERATIONALIZATION:the dependent variable
How to measure the degree to which ministerial drafts of government bills are changed in the course of legislative review?
• (ideally) Policy impact of changes But hard to determine for both legislators and political scientists
• Number of articles changesNumber of articles altered (or deleted) in the draft version of bill +
Number of new articles added to the draft version of bill
From 0 to…
Advantages of this measure:– Articles are the logical “policy subunits” of a bill (the substantial
impact of changes is captured)– Reliably coded by other researchers
The independent variables:• Key explanatory variable:
Government issue divisivenessHypothesis: as the coalition becomes more divided on policy, coalition partners are more likely to
make use of the legislative process to “correct” government bills
• Control variables:– Opposition issue divisiveness
Hypothesis: if the opposition does possess significant influence over policymaking, we should expect more changes being made to those bills dealing with issues that greatly divide opposition parties from the party of the minister proposing them
– Junior minister from the partner partyHypothesis: we expect the presence of a partner party in the ministry drafting a bill to reduce the
number of changes made to the bill in the legislature
– Number of committee referralsHypothesis: we expect to see more changes made to bills, other things equal, the greater the
number of committees involved in the review process (ie the greater the number of legislators from a party who are able to scrutinize a minitser’s proposal)
– Number of articles in the draft bill (logged)Hypothesis: we would expect a positive relation between the number of draft articles and the
number of articles changed
– Expiration of bill before the plenary voteHypothesis: bills that do not reach the final voting stage are not changed as extensively as bills
that receive a full review in the legislative process
– Country indicator variableHypothesis: since in Germany (but not in Netherlands) committees has redrafting authority, we
expect more extensive changes being made in Germany
– Policy area indicatorsBills dealing with certain types of issues might be always changed more or less extensively,
regardless of the policy preferences of coalition partners
METHODOLOGY
Event count model
Assumptions:- The events accumulating during the observation period are conditionally
independent - The rate of event occurrence is homogeneous across any given time period
Violated: positive contagion overdispersion
Solution: negative binomial analysisAllows the rate of event occurrence to vary across an observation periodAnd allows for estimation of the degree of overdispersion as a parameter from the
data
RESULTSGovernment issue divisiveness
Hypothesis: bills dealing with issues that are more divisive for the coalition government are changed more extensively in the legislative process ok
Control variables:– Opposition issue divisiveness
Hypothesis: if the opposition does possess significant influence over policymaking, we should expect more changes being made to those bills dealing with issues that greatly divide opposition parties from the party of the minister proposing them not supported
– Junior minister from the partner partyHypothesis: we expect the presence of a partner party in the ministry drafting a bill to reduce the
number of chenges made to the bill in the legislature ok
– Number of committee referralsHypothesis: we expect to see more changes made to bills, other things equal, the greater the
number of committees involved in the review process (ie the greater the number of legislators from a party who are able to scrutinize a minitser’s proposal) ok
– Number of articles in the draft bill (logged)Hypothesis: we would expect a positive relation between the number of draft articles and the
number of articles changed ok
– Expiration of bill before the plenary voteHypothesis: bills that do not reach the final voting stage are not changed as extensively as bills
that receive a full review in the legislative process ok
– Country indicator variableHypothesis: since in Germany (but not in Netherlands) committees has redrafting authority, we
expect more extensive changes being made in Germany ok, but be careful
– Policy area indicatorsBills dealing with certain types of issues might be always changed more or less extensively,
regardless of the policy preferences of coalition partners not displayed
Predictive results
How do different levels of coalition policy divergence affect the predicted number of article changes in government bills?
CONCLUSION• Role of legislatures in parliamentary democracies
policymaking • Central role of the legislative process in strenghtening
government (not opposition) parties• Delegation creates tensions between coalition parties
that grow particularly strong on issues that sharply divide them
• Coalition partners attempt to manage these tensions trough both cabinet-level and legislative-level institutions
• Ministerial drafts are likely to be changed more extensively in the legislative process the greater the level of policy divisiveness among coalition partners
• Parliament as an arena in which coalition parties interact
Possible extensions
1. Minority governmentsDo the preferences of opposition parties
matter? And to what degree?
2. Multiparty governments in majoritarian systems
Cabinet–level institutions and/or “ministerial autonomy”