Office and policy payoffs in coalition governments

1
Office and policy payoffs in coalition
governments Marc Debus
Presentation by Jacopo Gandin
2
Research questions

• Which parties in a multiparty system have to be
  considered as key players in the coalition-
  building game?
• And, if they exist
• Can we make predictions about office and policy
  payoffs for these parties in the government
  formation?

3
Definitions

• Key parties are defined by many coalition
  theories having superior powers in terms of
  coalition formation.
• Dominant player (Peleg 1981) a party that should
  be included in the government because of its
  parliamentary seat strenght.
• Central player (van Deemen 1991 van Roozendaal
  1993) a party that holds a pivotal position due
  to the inclusion of median legislator.

4
Models

• The author uses two models, drawn from the
  spatial theory of voting tradition, to define key
  actors in the coalition building process.
• The first is the portfolio allocation model
  (Laver Shepsle 1990), assuming that every party
  in the coalition receives a portfolio according
  to its own issue salience.
• The second is the political heart model
  (Schofield 1986), where key parties are supposed
  to be the parties bounding the heart (space
  delimited by the medians.

5
Theories of Coalition Politics

• A starting point can be to analyse the motivation
  for a party to put itself in coalition.
• The first approach is called office- orientated,
  and takes into account as a variable only party
  parliamentary seat strenght.
• The second is the policy- orientated approach,
  considering party position before and after
  elections in order to define their
  multidimensional programmatic position.

6
Portfolio allocation

• This approach is based on the idea of ministerial
  government. Each party tries to get a minister in
  the issue(s)that it considers salient.Who gets a
  portfolio can be the policy setter with regard to
  that issue.
• A government can be only formed by strong or very
  strong parties, that, according to Laver
  Shepsle definition a strong party is present
  with at least one minister in all possible
  coalitions preferred by every actor to the status
  quo.

7
Political heart

• We have to take into account the distinction
  between dominant and peripheral parties. Dominant
  parties must be members of every possible
  government.
• Dominant a party located at the core of the
  intercepts between all medians. If a dominant
  party does not exists, the only possible
  coalitions are formed by parties located at the
  bound of the core.
• Example Belgium 1978- only possible coalitions
  according to the core PSB-PVV, PSB-CVP and CVP-
  PVV.

8
Institutional and behavioural constraints

• That hypotesis was not verified in the reality
  the coalition in Belgium 1978 was CVP-PSB-FDF-VU.
• That was because institutional rules ensured
  equal representation to Flemish and Walloon
  components, forcing FDF (Walloon) and VU
  (Flemish), to be included in the coalition.
• In this case, institutions represent a limit to
  the spatial theory of voting, as in other cases
  behavioural aspects do (ex. Pre- electoral
  agreements- Germany 2005 FDP with CDU and w/o
  SPD- SPD and Greens w/o Die Linke).

9
Effects of these constraints

• There are some differences among actors during
  the coalition building process. Parties that
  don't reject any a priori coalition would get a
  higher payoff in the game than parties avoiding
  some particular coalition possibililities.
• In the latter case they break the core and modify
  the political heart, favouring parties standing
  inside the new asset, that should increase their
  blackmail potential.
• Behavioural constraints can strongly modified
  each party's coalition potential.

10
Case selection

• Selected cases Austria 1983- 2002, Belgium
  1985-2003, Ireland 1982- 2oo2, Germany 1980-
  2002, the Netherlands 1977- 2003.
• Similarities
• 1) classifiable by Laver Schofield typology
  (1998) about party system unipolar, bipolar or
  multipolar.
• 2) frequent a priori coalition statements and
  alliance rejection.

11
Method

• Ideological competition is based on conflicts in
  economic and social policy, referring to Laver
  Hunt data (1992).
• A third country- specific dimension is included
• Austria, Germany, the Netherlands foreign policy
• Belgium decentralization
• Ireland position about Northern Ireland
• In this way the author checks the presence of
  possible modified heart parties.

12
Method

• Since expert surveys do not make difference
  between party and government position, the author
  refers to some policy documents the coalition
  agreements contained in the election manifestos.
• Party ideology is weighted both on position and
  on issue saliency. Values are assigned with the
  wordscore method.
• To evaluate pre- electoral statements the author
  analyses electoral campaigns, considering only
  potential coalitions.

13
Findings

• The differences between original and modified
  solutions are quite relevant, either by using
  strong party or political heart model.
• Strong party model does not seems to fit very
  well the modified solution corrects the original
  one in 8 of 28 cases. In 7/28 cases the model
  gives a wrong prediction.
• Original political heart model often includes an
  extremely wide range of parties after the
  correction every time the resulting party is
  included in the coalition

14
Findings

• In 10/39 cases all coalition members are exactly
  predicted by the modified political heart model.
• Just in 2 cases a party that was not predicted by
  any model appears in the coalition they are
  Belgium 1988 and Ireland 1994, where peripheral
  parties were included in the coalition.
• Behavioural reasons stand at the basis of the
  correction of FPOE strong party role in 1994
  OEVP in that year ruled out any alliance with the
  liberal formation, unlike 1990 and 1995.

15
What means to be a key player?

• A key player should be able to get a greater
  number of offices in government than its actual
  parliamentary seat share.
• Furthermore, a key party should be closer to its
  programmatic position mentioned in the coalition
  agreement.
• The latter expectation is tested against the
  assumption that large parties are more likely to
  achieve their policy goals.

16
Testing the hypoteses

• The author tests his hypoteses by means of the
  formula pointing out the Euclidean distance
  between programmatic positions contained in a
  coalition agreement and that of a strong
  party,weighted by issue salience.
• FORMULA
• He compares this measure with intra- coalition
  strenght (measure by of parliamentary seat
  share) and share of captured offices.

17
Office and policy findings

• According to the author,using modified strong
  party method the first hypotesis is weakly
  confirmed only in 4/21 cases a strong party gets
  a higher number of offices than its intra-
  coalition strenght.
• The correlation seems to be stronger with regard
  to the second hypotesis if consider it, we
  observe that in 14 cases over 21 key parties
  remain closer to their programmatic platforms.

18
Office and policy findings

• The modified political heart method seems to
  provide similar outcomes with regard to the
  first hypotesis, 8/24 cases (33.3) get a higher
  payoffs than their intra- coalition strenght.
• The second hypotesis on the contrary is slightly
  confirmed 83.3 of cases (20/24) show how we
  can argue with good approximation that be a key
  party means the possibility to remain close to
  the political position expressed at the moment of
  coalition agreement.

19
Conclusions

• The author states that by method of modified
  strong party (portfolio allocation) we can only
  find out the existance of strong parties (as in
  28 over 39 cases happens).
• By the modified political heart system, we can
  generalize about the partisan composition of next
  government, and observe how this method is
  reliable for what concerns policy payoffs, less
  for office payoffs predictions.