Electoral Systems and Intra-Party Candidate Selection Effects on Legislators’ Behavior Yael Shomer Tel-Aviv University

,

Research Question

How do electoral systems and intra-party candidate selection processes affect legislator’s behavior?

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

2/31

Jacob Shefi

I see myself, first and foremost, loyal to the workers of the Israeli Aerospace industries and not the party.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

3/31

Why Do We Care? legislators’ behavior specifically, and political parties more generally, influence: Democratic governance (Bowler, Farrell and Katz 1999; Cox and McCubbins 1993; Katz 1987; Ranney 1954) Representation style and relations between voters and their representatives (Bawn and Thies 2003; Crisp et. al. 2004; Kitschelt 2000; Sartori 1976) Accountability and corruption (Bowler and Katz 1998; Chang 2005; Heidar and Koole 2000; Powell 2000; Powell and Whitten 1993; Sartori 1976, Stokes 1997; Strøm 1997) Policies (Person and Tabellini 2003; Strøm 2000; Thies 2000) Government stability, formation and duration (Giannetti and Beniot 2009; Pulzer 2006; Rose 1983; Strøm 2000) Legislative efficiency (Aldrich 1995; Cox 2006; Hazan 2006) ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

4/31

Electoral Systems

Electoral Systems Electoral Systems create constraints and incentives

The incentives to emphasize a personal or a party reputation are captured by four aspects of electoral systems (Carey and Shugart 1995; Shugart 2001; Shugart, Valdini and Suominen 2005): Ballot Pool Vote District magnitude (M)

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

5/31

Candidate Selection Candidate Selection Selection processes create constraints and incentives (Bowler, Farrell and Katz 1999; Faas 2003; Gallagher and Marsh 1988; Rahat and Hazan 2001).

Restrictive selection processes encourage incentives to behave in a party-centered manner. Democratized selection processes encourage incentives to emphasize personal reputation (De-Luca et al. 2002; Faas 2003; Gallagher and Marsh 1988; Hazan 2000; Mainwaring and Shugart 1997; Rahat and Hazan 2001).

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

6/31

The Current State of Theory

Amalgamation of Electoral systems and selection processes (Carey 2009; Hallerberg and Marier 2004; Shugart et. al. 2005). Additive combined effect (Crisp 2007; Depauw and Martin 2009; Mitchell 2000).

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

7/31

Modifying the Theory: Why?

Differentiating electoral systems from candidate selection processes Electoral systems and candidate selection processes are distinct institutions that might induce contradictory incentives for legislators’ behavior. Electoral systems operate (mostly) at the country level, whereas selection processes operate (mostly) at the party level.

Combined effect of electoral systems and candidate selection processes Legislators’s behavior is influenced by electoral systems and selection processes simultaneously.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

8/31

Modifying the Theory: Why?

Differentiating electoral systems from candidate selection processes Electoral systems and candidate selection processes are distinct institutions that might induce contradictory incentives for legislators’ behavior. Electoral systems operate (mostly) at the country level, whereas selection processes operate (mostly) at the party level.

Combined effect of electoral systems and candidate selection processes Legislators’s behavior is influenced by electoral systems and selection processes simultaneously.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

8/31

Modifying the Theory: Why?

Differentiating electoral systems from candidate selection processes Electoral systems and candidate selection processes are distinct institutions that might induce contradictory incentives for legislators’ behavior. Electoral systems operate (mostly) at the country level, whereas selection processes operate (mostly) at the party level.

Combined effect of electoral systems and candidate selection processes Legislators’s behavior is influenced by electoral systems and selection processes simultaneously.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

8/31

Modifying the Theory: Why?

Differentiating electoral systems from candidate selection processes Electoral systems and candidate selection processes are distinct institutions that might induce contradictory incentives for legislators’ behavior. Electoral systems operate (mostly) at the country level, whereas selection processes operate (mostly) at the party level.

Combined effect of electoral systems and candidate selection processes Legislators’s behavior is influenced by electoral systems and selection processes simultaneously.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

8/31

Modifying the Theory: Why?

Differentiating electoral systems from candidate selection processes Electoral systems and candidate selection processes are distinct institutions that might induce contradictory incentives for legislators’ behavior. Electoral systems operate (mostly) at the country level, whereas selection processes operate (mostly) at the party level.

Combined effect of electoral systems and candidate selection processes Legislators’s behavior is influenced by electoral systems and selection processes simultaneously.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

8/31

Modifying the Theory: How? The Separate effect of electoral systems and selection processes Differentiating electoral systems from selection processes Ballot Type: The degree to which party leaders versus voters control the ballot at the general election stage Ballot Access: The manner by which legislators “gain permission” to use the party’s ballot.

Hypotheses Separate effect of electoral systems on legislators’ behavior The more permissive an electoral system is (Ballot type, Pool and Vote encourage personal vote seeking), the less discipline we should observe (and vice versa). Separate effect of selection processes on legislators’ behavior The more permissive the selection process is (more democratized), the less discipline we should observe (and vice versa). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

9/31

Modifying the Theory: How? The Separate effect of electoral systems and selection processes Differentiating electoral systems from selection processes Ballot Type: The degree to which party leaders versus voters control the ballot at the general election stage Ballot Access: The manner by which legislators “gain permission” to use the party’s ballot.

Hypotheses Separate effect of electoral systems on legislators’ behavior The more permissive an electoral system is (Ballot type, Pool and Vote encourage personal vote seeking), the less discipline we should observe (and vice versa). Separate effect of selection processes on legislators’ behavior The more permissive the selection process is (more democratized), the less discipline we should observe (and vice versa). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

9/31

Modifying the Theory: How? The Separate effect of electoral systems and selection processes Differentiating electoral systems from selection processes Ballot Type: The degree to which party leaders versus voters control the ballot at the general election stage Ballot Access: The manner by which legislators “gain permission” to use the party’s ballot.

Hypotheses Separate effect of electoral systems on legislators’ behavior The more permissive an electoral system is (Ballot type, Pool and Vote encourage personal vote seeking), the less discipline we should observe (and vice versa). Separate effect of selection processes on legislators’ behavior The more permissive the selection process is (more democratized), the less discipline we should observe (and vice versa). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

9/31

Modifying the Theory: How? The Separate effect of electoral systems and selection processes Differentiating electoral systems from selection processes Ballot Type: The degree to which party leaders versus voters control the ballot at the general election stage Ballot Access: The manner by which legislators “gain permission” to use the party’s ballot.

Hypotheses Separate effect of electoral systems on legislators’ behavior The more permissive an electoral system is (Ballot type, Pool and Vote encourage personal vote seeking), the less discipline we should observe (and vice versa). Separate effect of selection processes on legislators’ behavior The more permissive the selection process is (more democratized), the less discipline we should observe (and vice versa). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

9/31

Modifying the Theory: How? The Separate effect of electoral systems and selection processes Differentiating electoral systems from selection processes Ballot Type: The degree to which party leaders versus voters control the ballot at the general election stage Ballot Access: The manner by which legislators “gain permission” to use the party’s ballot.

Hypotheses Separate effect of electoral systems on legislators’ behavior The more permissive an electoral system is (Ballot type, Pool and Vote encourage personal vote seeking), the less discipline we should observe (and vice versa). Separate effect of selection processes on legislators’ behavior The more permissive the selection process is (more democratized), the less discipline we should observe (and vice versa). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

9/31

Modifying the Theory: How? Combined Effect of Electoral Systems and Selection Processes The combined effect of electoral systems and selection processes is a conditional one. Electoral systems and selection processes are substitutive disciplinary tools. Under restrictive electoral system that encourage party-centered behavior the effect of selection processes is redundant. Under permissive electoral systems that encourage personal-vote seeking behavior the effect of selection processes should be apparent. Under restrictive selection processes that encourage party-centered behavior the effect of electoral systems is redundant. Under permissive selection processes that encourage personal-vote seeking behavior the effect of electoral systems should be apparent.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

10/31

Modifying the Theory: How? Combined Effect of Electoral Systems and Selection Processes The combined effect of electoral systems and selection processes is a conditional one. Electoral systems and selection processes are substitutive disciplinary tools. Under restrictive electoral system that encourage party-centered behavior the effect of selection processes is redundant. Under permissive electoral systems that encourage personal-vote seeking behavior the effect of selection processes should be apparent. Under restrictive selection processes that encourage party-centered behavior the effect of electoral systems is redundant. Under permissive selection processes that encourage personal-vote seeking behavior the effect of electoral systems should be apparent.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

10/31

Modifying the Theory: How? Combined Effect of Electoral Systems and Selection Processes The combined effect of electoral systems and selection processes is a conditional one. Electoral systems and selection processes are substitutive disciplinary tools. Under restrictive electoral system that encourage party-centered behavior the effect of selection processes is redundant. Under permissive electoral systems that encourage personal-vote seeking behavior the effect of selection processes should be apparent. Under restrictive selection processes that encourage party-centered behavior the effect of electoral systems is redundant. Under permissive selection processes that encourage personal-vote seeking behavior the effect of electoral systems should be apparent.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

10/31

Combined Conditional Effect

Hypothesis Combined effect of electoral systems and selection processes The effect of candidate selection processes on legislators’ behavior should be greater under permissive than under restrictive electoral systems. Similarly, the effect of electoral systems should be greater when selection is democratized than when selection is permissive.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

11/31

90 85 80 70

75

Party Cohesion (Rice Scores)

95

100

Hypothesized Combined Effect of Elections and Selections

Line Colors restrictive selection democratized selection

Restrictive electoral system, party−centered incentives

Permissive electoral system, candidate−centered incentives Electoral Systems

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

12/31

Research Design

The differing levels at which elections and selections operate. Party level model: two-level varying intercept varying slope hierarchical model, where parties are nested within countries. Individual level model: three-level varying intercept varying slope hierarchical model, where legislators are nested with parties and parties are nested within country-sessions.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

13/31

Personal and Party Vote: Definition and Operationalization Definition Personal reputation: Legislators’ personal reputations distinct from those of their party. Party reputation: Parties’ collective shared reputation.

Operationalization Cohesive voting record bolster a party’s collective shared reputation (Carey and Shugart 1995; Cox and McCubbins 1993; Gallagher and Marsh 1988; Norris 2002; 2004). Deviation from party leaders’ directions as to how to vote bolster legislator’s personal reputation (Bowler, Farrell and Katz 1999; Giannetti and Benoit 2009; Hazan 1999; 2000). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

14/31

Personal and Party Vote: Definition and Operationalization Definition Personal reputation: Legislators’ personal reputations distinct from those of their party. Party reputation: Parties’ collective shared reputation.

Operationalization Cohesive voting record bolster a party’s collective shared reputation (Carey and Shugart 1995; Cox and McCubbins 1993; Gallagher and Marsh 1988; Norris 2002; 2004). Deviation from party leaders’ directions as to how to vote bolster legislator’s personal reputation (Bowler, Farrell and Katz 1999; Giannetti and Benoit 2009; Hazan 1999; 2000). ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

14/31

Outcome variable

Party level: Rice Scores (Rice, 1928) RICEpj =

|AYEpj − NAYpj | , for party p on vote j. AYEpj + NAYpj

(1)

where we average across j to get party level Rice score index. Individual level: Distance of ideal point from the median of the party |Di | = |Xi − median(Xp )| (2)

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

15/31

Main Explanatory Variables: Ballot Type and Ballot Access Ballot Type: the degree to which party leaders versus voters control the ballot at the general election stage. It differentiates between fixed ballots (0), weak preferential (1) and strong preferential (2) ballots. Ballot Access: the manner by which legislators “gain permission” to use the party’s ballot. Ballot Access’ Operationalization

8 point selection index (Shomer, Comparative Political Studies, 2009) composed of: Selectorate: who can participate in the selection process De-centralization: selection exclusively by a national party selectorate with no procedure that allows for territorial and/or functional representation versus selection solely by local-party organizations ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

16/31

Main Explanatory Variables: Ballot Type and Ballot Access Ballot Type: the degree to which party leaders versus voters control the ballot at the general election stage. It differentiates between fixed ballots (0), weak preferential (1) and strong preferential (2) ballots. Ballot Access: the manner by which legislators “gain permission” to use the party’s ballot. Ballot Access’ Operationalization

8 point selection index (Shomer, Comparative Political Studies, 2009) composed of: Selectorate: who can participate in the selection process De-centralization: selection exclusively by a national party selectorate with no procedure that allows for territorial and/or functional representation versus selection solely by local-party organizations ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

16/31

Covariate: Selection Processes

Party Centered

1

1: 2: 3: 4: 5: 6: 7: 8:

2

Candidate Centered

3

4

5

6

7

8

Selection Index national party leadership up to 10 selectors national party delegates local party leadership up to 10 selectors, subject to approval of the national party local party leadership up to 10 selectors (at the local level) local party delegates, subject to approval of the national party local party delegates primaries subject to approval of national party primaries

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

17/31

Control Covariates

Country level: Pool, Vote, Average district magnitude, Presidential versus parliamentary, Unitary versus federal Party level: Coalition-opposition, Party size Individual level: Seniority, Minister, Deputy , Committee Chair, Gender.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

18/31

Data: 251 parties nested in 24 countries Soc People’s party(Denmark 94-95) Cent Dem(Denmark 94-95) Cons(Denmark 94-95) Lib Tasm(Australia 96-98) Raam(Israel 03-06) Soc Dem(Sweden 94-95) Green Party(Switzerland 99-03) Coastal Party(Norway 01-05) Workers Party(Ireland 89-92) Progressive Party(Ireland 89-92) Green(Ireland 92-97) Prog Dem(Ireland 02-07) Green(Ireland 02-07) Hadash(Israel 99-03) Labour(Ireland 89-92) Labour(Ireland 02-07) FF(Ireland 97-02) FG(Ireland 89-92) FG(Ireland 02-07) FG(Ireland 92-97) Labour(I eland 92-97) Labour(New Zealand 02-05) Social Dem(Denmark 94-95) Progress Party(Denmark 94-95) Social Dem(Austria 96-97) FF(Ireland 89-92) FF(Ireland 02-07) FG(Ireland 97-02) Labour Capital(Australia 96-98) Labour(Ireland 97-02) Dem Left(Ireland 97-02) Communist Party(France 62-67) Green(Ireland 97-02) Labour Queen(Australia 96-98) National Vic(Australia 96-98) New Zealand First( New Zealand 02-05) Dem Left(Ireland 92-97) Liberals(Denmark 94-95) SPD(Germany 57-61) People’s Union(Belgium 95-99) Social Dem.(Iceland 95-96) Socialist(France 62-67) Populist Right(Austria 96-97) Bloc Quebecois(Canada 97-00) Bloc Quebecois(Canada 94-97) Liberal WA(Australia 96-98) Agalev(Belgium 95-99) Christian Social(Belgium 95-99)

Hadash(Israel 92-96) Liberal Vic(Australia 96-98) Hadash(Israel 03-06) CVP(Belgium 95-99) SP(Belgium 95-99) PRONAL(Nicaragua 2000) Liberal NSW(Australia 96-98) Balad(Israel 03-06) Liberal SW(Australia 96-98) Progressive Party(Iceland 95-96) SPD(Germany 53-57) PRL(Belgium 95-99) Gesher(Israel 99-03) One Nation(Israel 99-03) Labour WA(Australia 96-98) Hadash(Israel 06-08) Balad(Israel 06-08) Tsomet(Israel 96-99) Ecolo Belgium 95-99) Labour NSW (Australia 96-98) Labour Tasm(Australia 96-98) National Party NSW(Australia 96-98) Independence Party(Iceland 95-96) Labour SW(Australia 96-98) National(New Zealand 02-05) FF(Ireland 92-97) Christian Dem(Austria 96-97) Labour Vic(Australia 96-98) Raam(Israel 06-08) National Party Queen(Australia 96-98) CSSD(Czech Republic 96-98) CDU-CSU(Germany 87-90) Liberal Queen(Australia 96-98) Israel Betenu(Israel 06-08) Prog Conservatives(Canada 97-00) SPD(Germany 61-65) Shas(Israel 06-08) NRP(Israel 92-96) Hadash(Israel 96-99) Meretz(Israel 06-08) Balad(Israel 99-03) Shas(Israel 96-99) PC(Canada 94-97) Agudat Yisrael(Israel 06-08) Liberals(Canada 97-00) Bloc Quebecois(Canada 00-04) Agudat Yisrael(Israel 03-06) Yisrael baAliyah(Israel 96-99)

Third Way(Israel 96-99) NRP(Israel 03-06) ps(Belgium 95-99) New Democrats(Canada 00-04) NZF(New Zealand 93-94) Yahadut-Hatorah(Israel 92-96) Kadima(Israel 06-08) NRP(Israel 06-08) Labour(UK 97-01) Meretz(Israel 03-06) PCN(Nicaragua 00) Likud(Israel 06-08) Labour(Norway 93-97) Yahadut-Hatorah(Israel 96-99) Yahadut-Hatorah(Israel 99-03) CCN(Nicaragua 00) NRP(Israel 96-99) Labour(New-Zealand 90-93) Raam(Israel 99-03) Liberal dem(UK 97-01) Liberal Alliance AL (Nicaragua 00) Meretz(Israel 96-99) Green(Austria 96-97) Conservatives(UK 97-01) Gaullist Party(France 62-67) Shinuy(Israel 03-06) NDP(Canada 94-97) Ecology(Sweden 94-95) Left Party (Sweden 94-95) Ind Dem Union(Chile 97-20) People’s Party(Sweden 94-95) Christian Dem(Sweden 94-95) Moderate Unity Party(Sweden 94-95) New Democrats(Canada 97-00) Center Party(Sweden 94-95) Christian People’s(Norway 93-97) Meretz(Israel 92-96) Centre Party(Norway 93-97) NRP(Israel 99-03) SPD(Germany 87-90) Conservatives(UK 01-05) Socialist Left(Norway 93-97) Shas(Israel 03-06) FDP(Germany 87-90) Conservatives(UK 92-97) Liberal Dem(UK 01-05) Labour(UK 01-05) Reform Party(Canada 94-97)

FDP(Germany 57-61) National Party(New-Zealand 90-93) Labour(Israel 06-08) Con Party(Norway 93-97) Socialist(Chile 97-00) Labour(Israel 96-99) Labour(Israel 92-96) Likud(Israel 03-06) SPS(Switzerland 99-03) Moledet(Israel 96-99) Reform(Canada 97-00) FSLN(Nicaragua 00) SPR-ESC(Czech Republic 96-98) Shas(Israel 92-96) Shinuy(Israel 99-03) ODA(Czech Republic 96-98) Christian Dem(Norway 01-05) Labour(UK 92-97) National Party(New Zealand 93-94) Socialist Left(Norway 97-01) CDU-CSU(Germany 57-61) Shas(Israel 99-03) Conservative(Norway 01-05) Likud(Israel 99-03) Likud(Israel 96-99) Labour(Norway 97-01) Liberals(Canada 00-04) Center(Israel 99-03) Meretz(Israel 99-03) EVP(Switzerland 99-03) Centre Party(Norway 97-01) Labour(New Zealand 93-94) Christian People’s(Norway 97-01) PPD(Chile 97-00) Worker’s Party PT(Brazil 02-07) PDC(Chile 97-00) Liberal Party(Canada 94-97) Ihud leumi(Israel 03-06) Centre Party(Norway 01-05) Progress Party(Norway 97-01) Canadian Alliance(Canada 00-04) Gil(Israel 06-08) Socialist Left(Norway 01-05) ODS(Czech Republic 96-98) Conservative(Norway 97-01) Radical Socialists(France 62-67) People’s Alliance(Iceland 95-96) Progress Party(Norway 01-05)

Gesher(Israel 96-99) PPS(Brazil 02-07) Labour(Norway 01-05) Labour(Israel 99-03) Yisrael baAliyah(Israel 99-03) Ihud leumi(Israel 99-03) CSSD(Czech Republic 98-02) Swedish People’s(Finland 95-96) Labour(Israel 03-06) National Coalition(Finland 95-96) CDU-CSU(Germany 53-57) KDU-CSL(Czech Republic 98-02) CDU-CSU(Germany 61-65) Center Party(Finland 95-96) ODS(Czech Republic 98-02) Liberal Party(Brazil 02-07) Greence Leagure(Finland 95-96) National Renewal(Chile 97-00) Prog Dem(Ireland 92-97) Likud(Israel 92-96) PSB(Brazil 02-07) KDU-CSL(Czech Republic 96-98) Tsomet(Israel 92-96) PTB(Brazil 02-07) Progressive Cons(Canada 00-04) Left-wing alliance(Finland 95-96) Social Dem(Finland 95-96) FDP(Germany 61-65) Progress Party(Norway 93-97) Liberal Dem(UK 92-97) CVP(Switzerland 99-03) PMDB(Brazil 02-07) PDT(Brazil 02-07) PV(Brazil 02-07) FDP(Germany 53-57) Popular Republican(France 62-67) Progressive Party(Brazil 02-07) Democratic Party(USA 93-95) ROP(Poland 97-99) Republican Party(USA 93-95) PSDB(Brazil 02-07) Democratic Party(USA 63-65) PFL(Brazil 02-07) Peasants(Poland 97-99) Republican Party(USA 63-65) AWS(Poland 97-99) Freedom Union(Poland 97-99) SLD(Poland 97-99) Progressive Party(Ireland 97-02)

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

19/31

Party-level model: a two-level varying-intercept varying-slope hierarchical model

Ricep yˆp 

αc

∼

N yˆp , σy2

 ∼

α ˆc

(3)

= αc[p] + βselecc [p] ∗ Selectionp + βopp ∗ Oppp + βsize ∗ Sizep



βselecc




N

α ˆc βˆselec

c

  2 σαc , ρ1 σαc σβselecc

ρ1 σαc σβselecc σβ2selecc



= γ0α + γ1α ∗ Ballotc + γ2α ∗ Poolc + γ3α ∗ Votec + γ4α ∗ DistMc + γ5α ∗ Regimec + γ6α ∗ Unitaryc

βˆselecc

= γ0β + γ1β ∗ Ballotc ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

20/31

Individual-level model: a three-level varying-intercept varying-slope hierarchical model   2 yˆi , σy

Distancei

∼

N

yˆi

=

αp[i] + βseniority ∗ Seniorityi + βminister ∗ Ministeri

+

βdeputy ∗ Deputyi + βchair ∗ Chairi + βgender ∗ Genderi

αp

∼

N

α ˆp

=

γ0

γ0α

α c

α

γ ˆselec

c

α

cp

 N

γ

α

+ γselec

 ∼

γ ˆ0

  2 α ˆ p , σα

α cp

!

α c γselec c

(4)

γ ˆ0α αc γ ˆselec

α

∗ Selectionp + γopp ∗ Oppp + γsize ∗ Sizep

!

  ,

c

2 σγ α

0c

ρ1 σγ α σγ α 0c

γ

γ

selecc

γ

ρ1 σ γ α σ γ α



2 σγ α

 

0c

selecc

selecc

γ

=

δ0 0 + δ1 0 ∗ Ballotc + δ2 0 ∗ Poolc + δ3 0 ∗ Votec + δ4 0 ∗ DistMc

+

γ δ5 0

=

δ0 1 + δ1 1 ∗ Ballotc

γ

∗ Regimec +

γ δ6 0

∗ Unitaryc

γ

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

21/31

Party-level model Parameter

Median

80% C.I

Fixed Effects Intercept Selection Coalition Size Pool Vote District magnitude Regime Unitary Ballot Selection∗Ballot

4.45 -0.14 -0.41 0.001 0.09 -0.98 -0.02 1.78 -0.15 0.23 0.02

2.387 : -0.257 : -0.667 : -0.002 : -0.704 : -1.986 : -0.038 : 0.488 : -1.278 : -0.810 : -0.117 :

6.658 -0.033 -0.127 0.004 0.878 -0.042 0.002 3.096 1.005 1.226 0.150

Variance Components Country Level Intercept Selection Residual

1.57 0.09 1.42

0.966 : 2.153 0.0163 : 0.229 1.339 : 1.519 ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

22/31

Expected Values in Presidential Systems

0.95 0.90

Rice

0.985

Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

Fixed ballot

0.85

0.970

0.975

0.980

Rice

0.990

0.995

1.00

1.000

Expected Values in Parliamentary Systems

Weak preferential

Strong preferential

Ballot Type

(a) Parliamentary Systems

Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

Fixed ballot

Weak preferential

Strong preferential

Ballot Type

(b) Presidential Systems

Expected Rice Scores in Parliamentary and Presidential Systems

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

23/31

Densities of First Difference in Parliamentary Systems, by Ballot Type

Densities of First Difference in Presidential Systems, by Ballot Type

15

80

60

Restrictive_Elections 40

Permissive_Elections

..density..

density

density

10 ..density..

Restrictive_Elections Permissive_Elections

5 20

0

0 −0.10

−0.05

0.00

0.05

0.10

First Difference:selection varies

(a) Parliamentary Systems

−0.3

−0.2

−0.1

0.0

0.1

0.2

0.3

First Difference: selection varies

(b) Presidential Systems

First Differences in Expected Rice Scores when Selection Varies in Parliamentary and Presidential Systems ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

24/31

Are Parties Selecting under Restrictive Processes Significantly More Cohesive than Parties Selecting under Permissive Processes?

The probability in a parliamentary system with restrictive electoral systems, that a party that selects via restrictive selection processes will have higher cohesion than a party that selects via permissive selections is 0.953. The probability in a parliamentary system with permissive electoral systems that a party that selects via restrictive selection processes will have higher Rice scores than a party that selects via permissive selections is 0.835.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

25/31

Are Parties Selecting under Restrictive Processes Significantly More Cohesive than Parties Selecting under Permissive Processes?

The probability in a parliamentary system with restrictive electoral systems, that a party that selects via restrictive selection processes will have higher cohesion than a party that selects via permissive selections is 0.953. The probability in a parliamentary system with permissive electoral systems that a party that selects via restrictive selection processes will have higher Rice scores than a party that selects via permissive selections is 0.835.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

25/31

Expected Values in Presidential Systems

0.95 0.90

Rice

0.985

Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

Fixed ballot

0.85

0.970

0.975

0.980

Rice

0.990

0.995

1.00

1.000

Expected Values in Parliamentary Systems

Weak preferential

Strong preferential

Ballot Type

(a) Parliamentary Systems

Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

Fixed ballot

Weak preferential

Strong preferential

Ballot Type

(b) Presidential Systems

Expected Rice Scores in Parliamentary and Presidential Systems

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

26/31

Substantive Results: The conditional effect of selections and elections on legislators’ behavior On a Rice Scale from 0-100, the difference in the gap between restrictive and democratized selection processes under permissive electoral systems and this gap under restrictive electoral system is 0.7%. On the range of Rice scores where the middle 90% of parties fall, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 2.9%. On the interquartile range of Rice Scale, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 9.5%.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

27/31

Substantive Results: The conditional effect of selections and elections on legislators’ behavior On a Rice Scale from 0-100, the difference in the gap between restrictive and democratized selection processes under permissive electoral systems and this gap under restrictive electoral system is 0.7%. On the range of Rice scores where the middle 90% of parties fall, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 2.9%. On the interquartile range of Rice Scale, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 9.5%.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

27/31

Substantive Results: The conditional effect of selections and elections on legislators’ behavior On a Rice Scale from 0-100, the difference in the gap between restrictive and democratized selection processes under permissive electoral systems and this gap under restrictive electoral system is 0.7%. On the range of Rice scores where the middle 90% of parties fall, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 2.9%. On the interquartile range of Rice Scale, the difference in the gap between restrictive and democratized selection under permissive and restrictive electoral systems is 9.5%.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

27/31

Individual-Level Models: Three-Level Varying-Intercept Varying-Slope Parameter

Median

Intercept Selection Coalition Size Pool Vote District Magnitude Regime Unitary Ballot Selection∗Ballot Seniority Minister Deputy Ministers Committee Chair Gender

-1.66 0.018 -0.010 0.0015 -0.106 0.173 0.005 -1.517 -0.261 0.036 0.015 0.0234 0.150 -0.077 0.018 0.006

80% C.I.

Fixed Effects -2.322 : 0.007 : -0.077 : 0.0009 : -0.455 : -0.380 : -0.002 : -2.031 : -0.648 : -0.150 : 0.008 : 0.019 : 0.104 : -0.124 : -0.020 : -0.024 :

-1.016 0.0414 0.0582 0.0021 0.193 0.774 0.012 -1.051 0.080 0.219 0.045 0.028 0.193 -0.030 0.06 0.035

Variance Components Country Level Intercept Selection

0.582 0.017

0.486 : 0.738 0.003 : 0.040

Intercept Residual

0.215 0.674

0.188 : 0.244 0.667 : 0.682

Party Level

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

28/31

0.30

Expected Values in Presidential Systems Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

0.20

0.045

Deviance

0.25

Lines Selection=1 Selection=2 Selection=3 Selection=4 Selection=5 Selection=6 Selection=7 Selection=8

0.10

0.030

0.035

0.15

0.040

Deviance

0.050

0.055

0.060

Expected Values in Parliamentary Systems

Fixed ballot

Weak preferential

Strong preferential

Ballot Type

(a) Parliamentary Systems

Fixed ballot

Weak preferential

Strong preferential

Ballot Type

(b) Presidential Systems

Expected Ideological Distances in Parliamentary and Presidential Systems.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

29/31

Conclusions

Find support for the new theory of the combined conditional effect of electoral systems, and intra-party candidate selection processes on legislators’ behavior. Intra-country cross-party variation in legislators’ behavior between those who are selected in democratized and those who are selected in restrictive processes are greater in countries that employ permissive electoral systems, than they are in countries with closed list PR.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

30/31

Conclusions

Find support for the new theory of the combined conditional effect of electoral systems, and intra-party candidate selection processes on legislators’ behavior. Intra-country cross-party variation in legislators’ behavior between those who are selected in democratized and those who are selected in restrictive processes are greater in countries that employ permissive electoral systems, than they are in countries with closed list PR.

,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

30/31

The effect of democratizing candidate selection processes Democratizing candidate selection procedures improves the party’s image as a democratic entity, increases its legitimacy and helps it attract more members. On the other hand, democratized candidate selection processes weaken party leaders’ ability to maintain a unified party record. The results presented in this paper demonstrate that in countries with a permissive electoral system party leaders are better off maintaining restrictive candidate selection processes: sustaining the control over candidate selections enables party leaders to maintain a cohesive, unified party record even under permissive electoral systems. ,

Motivation Electoral Systems and Selection processes Research Design Results Conclusion

31/31

Electoral Systems and Intra-Party Candidate Selection Effects on Legislators’ Behavior Yael Shomer Tel-Aviv University

,