Electoral Engineering & Turnout Pippa Norris ~ UNDP Democratic Governance Details:www.undp.org/governance
Electoral engineering 2 Structure I. Theoretical framework: Multilevel model of electoral turnout II. III. Research design & evidence: Macro-level data worldwide - International IDEA Micro-level CSES Module 1 Conclusions: Rules matter for turnout, with direct and indirect effects Important for public policy reforms Yet their impact is constrained by levels of societal modernization, the role of mobilizing agencies, cultural attitudes, and structural resources
I. Theoretical framework
Electoral engineering 4 Context: why of interest? Renewed interest in the limits and capacities of electoral engineering and rule design For transitional and consolidating democracies For established democracies debating electoral reform Major revision Eg NZ, UK, Italy, Israel, Japan, Venezuela etc. Modifications of procedures Eg e-voting, voter registration procedures, etc. But can electoral engineering boost turnout in the shortterm?
Electoral engineering 5 Context: why of interest? Increased popular concern about turnout as an indicator of the health of representative democracy Yet no consistent fall in turnout across all established democracies since 1945 (Democratic Phoenix) Turnout has eroded modestly in established democracy during the last decade: reasons unclear Alternative types of activism have expanded (demonstrations, consumer politics, petitions etc)
II: Research design & evidence
Electoral engineering 7 Research design Classify constitutions, laws, procedures Analyze aggregate data and cross-national election surveys Compare formal rules Role of political actors Compare behavior of citizens Party campaigns, candidate strategies, electoral appeals, party organizations and members
Electoral engineering 8 Evidence Macro-level turnout worldwide: International IDEA database Voter Turnout Since 1945 www.idea.int Micro-level: CSES dataset: surveys of 32 nations in Module 1 (1996-2001) Countries in the CSES Module 1 Excluded (159) Included (32)
Electoral engineering 9 Electoral Systems Worldwide Nation States 191 Majoritarian 91 Combined 29 PR 64 No direct elections 7 Majority 26 Plurality 65 Independent 21 Dependent 8 STV 2 Party List 62 AV 2 2nd Ballot 24 FPTP 54 Block Vote 9 SNTV 2 Dual-ballot Preference-ballot Party-ballot 35 Preference-ballot 27 Candidate-Ballot Candidate-ballot Candidate-ballot Party-Ballot or Preference-ballot Preference-ballot
Electoral engineering 10 CSES Elections (37) Majoritarian electoral systems (14 elections) Legislative Australia (1996) Britain (1997) Canada (1997) United States (1996) Presidential larus (2001) Chile (1999) Israel (1996) (i) Lithuania (1997) Mexico (2000) Peru (2000) Romania (1996) Russia (2000) Taiwan (1996) United States (1996) Combined electoral systems (10 elections) Germany (1998) (l,c) Hungary (1998) (l,c) Japan (1996) (l,c) Korea, Republic of (2000) (c) Mexico (1997) (c) New Zealand (1996) (l,c) Russia (1999) (l) Taiwan (1996) (c) Thailand (2001) (c) Ukraine (1998) (l) Proportional electoral systems (15 elections) Belgium (1999) Czech Republic (1996) Denmark (1998) Iceland (1999) Israel (1996) Netherlands, The (1998) Norway (1997) Peru (2000) Poland (1997) Portugal (2002) Romania (1996) Slovenia (1996) Spain (1996, 2000) Sweden (1998) Switzerland (1999).
Electoral engineering 11 Voting Turnout, 1990s 3 0 4 0 5 0 6 0 7 0 8 0 9 0 Ic e la n d 8 8 Is r a e l 8 3 S w e d e n 8 3 C z e c h R e p 8 3 B e lg iu m 8 2 D e n m a r k 8 2 C h ile 8 2 A u s t r a lia 8 1 N e w Z e a la n d 8 0 K o r e a, R e p 7 9 S p a in 7 9 R o m a n ia 7 8 S lo v e n ia 7 6 P o r t u g a l 7 6 N o r w a y 7 6 N e t h e r la n d s 7 5 G e r m a n y 7 3 T a iw a n 7 2 U k r a in e 7 2 U k r a in e 6 7 L it h u a n ia 6 6 B e la r u s 6 4 S w R u s s ia P e r u J a p a n C a n a d a M e x ic o H u n g a r y T h a ila n d P o la n d U S it z e r la n d 3 8 4 7 6 3 6 2 6 1 6 0 5 9 5 8 5 8 5 7 Note: Mean Vote/VAP is measured as the number of valid votes as a proportion of the Voting Age Population in parliamentary elections during the 1990s held in the 32 nations in the CSES dataset under comparison. Source: International IDEA database Voter Turnout from 1945 to 2000. www.idea.int
III: Results
Electoral engineering 13 Electoral systems Source: Calculated from International IDEA database Voter Turnout from 1945 to 2000. Type of Electoral System Mean Vote/VAP 1990s Mean Vote/Reg 1990s N. MAJORITARIAN Alternative Vote 65.5 92.9 2 2 nd Ballot 58.5 65.0 21 First-Past-The-Post 61.2 67.7 43 Single Non-Transferable Vote 52.6 59.8 2 Block Vote 56.5 70.9 9 All majoritarian 60.4 68.3 77 COMBINED Combined-Dependent 66.6 71.9 7 Combined-Independent 63.5 69.0 19 All combined 64.0 70.4 26 PROPORTIONAL List PR 70.0 74.7 59 Single Transferable Vote 83.4 81.7 2 All PR Systems 70.0 74.6 68 All 65 0 70 8 164
Electoral engineering 14 Compulsory Voting Mean Vote/VAP Mean Vote/Reg N. Of Nations Older democracies Compulsory 79.4 86.9 7 Non-Compulsory 71.7 72.7 32 Difference +7.7 +14.2 39 Newer democracies Compulsory 67.7 75.8 9 Non-Compulsory 69.3 73.9 31 Difference -1.6 +1.9 40 Semi-democracies Compulsory 53.9 60.6 5 Non-Compulsory 56.6 67.0 40 Difference -2.7-6.4 45 Non-democracies Compulsory 40.9 70.6 2 Non-Compulsory 61.8 67.8 38 Difference -20.9 +2.8 40 All Compulsory 65.9 75.4 23 Non-Compulsory 64.2 70.0 140 Difference +1.9 +5.4 163
Electoral engineering 15 Party Competition 90 80 Mean Vote/Reg % Turnout 70 60 Mean Vote/VAP 50 L ess than 29.9% 30 to 39.9% 40 to 49.9% 50 to 59.9% M ore than 60% % Vote Share Winning Party
Electoral engineering 16 Social characteristics of turnout 60 65 70 75 80 85 90 95 All 77 Managerial & Prof Lower prof Skilled white collar Skilled manual Unskilled manual 74 77 76 83 84 Highest income High Moderate Low Lowest income 74 77 77 80 82 Older Middle aged Younger 74 79 81 Men Women 76 78 University Technical Secondary Primary 68 77 82 82 Union member 83 Urban Suburbs Small town Rural 77 77 79 81 Never attend church Attend service weekly 78 80 Strong Party id Moderate party id Weak party id 76 87 91
Electoral engineering 17 Age % Voted by Age Group 100 90 80 70 60 50 18-25 26-35 36-45 46-55 56-65 66-75 75+
Electoral engineering 18 Voter participation in legislative elections Source: CSES Module 1 1996-2002 Pooled N.24,413 b S ig. SOCIETAL MODERNIZATION Hum an developm ent 4.59 *** IN S T IT U T IO N A L C O N T E X T Electoral system (1=M aj, 2=Com b,3=pr).493 *** District size -.001 *** Parliam entary executive 1.96 *** Frequency of national elections -.002 n/s Use of any com pulsory voting 1.50 *** Party com petition (% vote party 1 st ).094 n/s P arty fractio n alizatio n (E N P P ).124 *** SOCIAL STRUCTURE Logged Age 2.12 *** Gender (m ale=1).003 n/s Education.294 *** In c o m e.1 0 2 *** MOBILIZING AGENCIES Union membership.188 *** R e lig io sity.0 9 5 *** CULTURAL ATTITUDES L eft-rig h t id eo lo gy.019 ** P arty identification.929 *** E xtern al p o litical efficacy.154 *** C o n stan t -5.9 % C o rre c tly p re d ic te d 8 4.0 Nagelkerke R2.198
IV: Conclusions
Electoral engineering 20 Conclusions 1. Rules matter: voting participation is maximized in elections: Using proportional representation, With small electoral districts, With regular but relatively infrequent national contests, With competitive party systems, and In presidential contests. 2. Important for public policy and electoral design 3. Yet the effect of rules is conditioned by other factors, including levels of human development, mobilizing agencies, and the resources and cultural attitudes of citizens. 4. Therefore limits to the capacity of electoral reform to engineer short-term improvements in turnout