Election laws and representative governments: Beyond votes and seats

British Journal of Political Science

Volumn 36, Issue 2, 2006, Pages 291-315

  Powell Jr , G Bingham ^a  

^a UNIVERSITY OF ROCHESTER   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 33644914224     PISSN: 00071234     EISSN: 14692112     Source Type: Journal    
DOI: 10.1017/S0007123406000160     Document Type: Article

References (102)

1. Douglas Rae, The Political Consequences of Electoral Laws (New Haven, Conn.: Yale University Press, 1967);
2. landmarks in this large literature include Gary Cox, Making Votes Count: Strategic Co-ordination in the World's Electoral Systems (Cambridge: Cambridge University Press, 1997);
3. Michael Gallagher, 'Proportionality, Disproportionality and Electoral Systems', Electoral Studies, 10 (1991), 33-51 ;
4. Arend Lijphart, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945-1999 (Oxford: Oxford University Press, 1994);
5. and Rein Taagepera and Matthew S. Shugart, Seats and Votes: The Effects and Determinants of Electoral Systems (New Haven, Conn.: Yale University Press, 1989).
6. See G. Bingham Powell Jr and Georg Vanberg, 'Election Laws, Disproportionality and the Left-Right Dimension', British Journal of Political Science, 30 (2000), 383-411;
7. and Michael D. McDonald, Silvia M. Mendes and Ian Budge, 'What Are Elections For? Conferring the Median Mandate', British Journal of Political Science, 34 (2004), 1-26, esp. 18-20.
8. For rare exceptions, see P. J. Taylor and A. Lijphart, 'Proportional Tenure vs. Proportional Representation', European Journal of Political Research, 13 (1985), 387-99;
9. John D. Huber and G. Bingham Powell Jr, 'Congruence between Citizens and Policymakers in Two Visions of Liberal Democracy', World Politics, 46 (1994), 291-326;
10. McDonald et al., 'What Are Elections For?' pp. 21-3.
11. Hanna F. Pitkin, The Concept of Representation (Berkeley: University of California Press, 1967), pp. 232-4.
12. The language is that of Cox, Making Votes Count, pp. 226-7.
13. Arend Lijphart, Democracies: Patterns of Majoritarian and Consensus Government (New Haven, Conn.: Yale University Press, 1984), chap. 1;
14. G. Bingham Powell Jr, Elections as Instruments of Democracy: Majoritarian and Proportional Visions (New Haven, Conn.: Yale University Press, 2000).
15. Cox, Making Votes Count, esp. chap. 12.
16. Cox, Making Votes Count;
17. Powell and Vanberg, 'Election Laws, Disproportionality and the Left-right Dimension'.
18. However, in these particular countries during these elections, the citizen means and medians are quite similar and the results do not depend on this choice of measure.
19. Gary W. Cox, 'Centripetal and Centrifugal Incentives under Alternative Voting Institutions', American Journal of Political Science, 34 (1990), 903-35;
20. Cox, Making Votes Count, chap. 12.
21. Rae, The Political Consequences of Electoral Laws, pp. 74-6.
22. Maurice Duverger, Political Parties: Their Organization and Activity in the Modern State (trans. B. North and R. North) (New York: John Wiley, 1954);
23. William H. Riker, 'The Two-party System and Duverger's Law: An Essay on the History of Political Science' ,American Political Science Review, 76 (1982), 753-66;
24. Cox, Making Votes Count.
25. Huber and Powell, 'Congruence between Citizens and Policymakers'.
26. Cox, Making Votes Count, chap. 12.
27. Anthony Downs, An Economic Theory of Democracy (New York: Harper and Row, 1957).
28. Cox, Making Votes Count, chap. 12.
29. Downs himself noted the potential role of abstaining extremists in limiting convergence (Economic Theory of Democracy, pp. 118-20.) Also see the 'Mays' Law' line of research, and the analysis of internal party incentives, arguing that patterns of recruitment will generally produce party representatives more extreme than party voters:
30. J. D. May, 'Opinion Structure of Political Parties: The Special Law of Curvilinear Disparity', Political Studies, 21 (1973),136-51;
31. Herbert Kitschelt, 'The Internal Politics of Parties', Political Studies, 37 (1989), 400-21;
32. and Pippa Norris, 'May's Law of Curvilinear Disparity Revisited', Party Politics, 1 (1995), 29-47. For a review of studies showing implications of modifying one or more of the many theoretical assumptions of the Downsian model
33. see Bernard Grofman, 'Downs and Two-Party Convergence', Annual Review of Political Science, 7 (2004), 25-46.
34. There are differences within even the five single-member district systems, of course, with the alternative vote in Australia and majority run-off elections in most French elections. The eleven proportional representation systems, in ten countries plus France 1986, have a variety of 'effective' electoral thresholds, ranging from the Netherlands at 0.67 to France 1986 at 12 (Lijphart, Electoral Systems and Party Systems).
35. In an overlapping dataset covering forty-five elections in the SMD countries over twenty-five years, Powell reports that 'only in Australia in 1975 and in France in 1981 did a party or preelection coalition win a clear voter majority; Canada 1984 and Australia 1972 are very close' (Elections as Instruments of Democracy, p. 129.)
36. Lijphart, Electoral Systems and Party Systems, p. 108.
37. An alternative method of linking voters, legislative medians and governments was proposed and operationalized by HeeMin Kim and Richard Fording, 'Voter Ideology in Western Democracies', European Journal of Political Research, 33 (1998), 73-97;
38. and HeeMin KiM and Richard Fording, 'Government Partisanship in Western Democracies 1945-1998', European Journal of Political Research, 41 (2002), 165-84;
39. and adapted by McDonald et al., 'What Are Elections For?' The advantage of their approach, building on the party-manifesto data and assuming voters support a party close to their preferences, is that it allows cross-national comparisons of substantive positions, while the distance approach used here can only compare deviations from citizen positions. Moreover, our approach assumes citizens and expert placement of parties in the same country have in mind similar scales and is dependent on availability of citizen and elite surveys. However, the advantage of the current approach is that it builds directly on citizen self-placements, without assuming that citizen votes correspond to their underlying preferences. Votes are in part shaped by party alternatives presented to the voters; using them to impute citizen preferences imposes greater structure on the analysis of their correspondence to those parties in the legislature. Given these respective advantages and disadvantages, it is extremely reassuring that the substantive conclusions about legislative and government deviation from the (imputed) voter median in SMD and PR systems reported by McDonald et al., pp. 18-23, are quite consistent with the results reported below.
40. See Matthew Gabel and John D. Huber, 'Putting Parties in their Places', American Journal of Political Science, 44 (2000), 94-103;
41. John Huber, 'Values and Partisanship in Left-right Orientations: Measuring Ideology', European Journal of Political Research, 17 (1989), 599-621;
42. Ronald Inglehart, The Changing Structure of Political Cleavages in Western Society', in Russell Dalton et al., eds, Electoral Change in Advanced Industrial Societies (Princeton, N.J.: Princeton University Press, 1984), pp. 25-69;
43. Ronald Inglehart and Hans Klingemann, 'Party Identification, Ideological Preference and the Left-right Dimension among Western Mass Publics', in ian Budge et al., eds, Party Identification and Beyond (London: JohnWiley, 1976), pp. 243-73;
44. Powell and Vanberg, 'Election Laws, Disproportionality and the Left-Right Dimension' ; and the elite surveys mentioned in fn. 24 below.
45. See, e.g. Soren Holmberg, 'Issue Agreement', in Peter Esaiasson and Knut Heidar, Beyond Westminster and Congress: The Nordic Experience (Columbus: Ohio State University Press, 2000), pp. 155-80;
46. Herbert Kitschelt, Zdenka Mansfeldova, Radoslaw Markowski and Gabor Toka, Post-Communist Party Systems: Competition, Representation and Inter-Party Cooperation (New York: Cambridge University Press, 1999), chap. 9;
47. and Ian McAllister, 'Party Elites, Voters and Political Attitudes: Testing Three Explanations of Mass-Elite Differences', Canadian Journal of Political Science, 24 (1991), 237-68.
48. The citizen medians on the left-right scale are computed from citizen surveys, primarily from the Eurobarometers and the World Value studies, using the closest survey, but no more than eighteen months permitted between survey and election. The positions of political parties are computed from the expert placements of the parties in the studies reported in Francis Castles and Peter Mair, 'Left-Right Political Scales: Some Expert Judgments', European Journal of Political Research, 12 (1984), 73-88;
49. and John D. Huber and Ronald Inglehart, 'Expert Interpretations of Party Space and Party Locations in 42 Societies', Party Politics, 1 (1995), 73-111, assuming these are accurate within five years of the time of the survey. Most left-right scales are ten-point scales; those that are not are converted for that purpose.
50. Deciding which party is the 'plurality vote winner', is slightly more complicated than it sounds because of the continuing election coalitions between some parties in some countries. The Liberal and National Parties in Australia and the CDU and CSU in Germany have been in continuing alliances over many elections; the latter do not compete directly against each other, as the CSU runs only in Bavaria and the CDU does not run in that Land. Similarly, in the period covered here in France, the UDF and Gaullists divided up the districts even on the first round and did not offer voters a choice between them. Slightly less compellingly, the French Communists and Socialists withdrew in each other's favour on the second round of legislative elections. In each of these cases, the spirit of electoral competition and, especially, the choices offered the voters seemed best captured by the party alliance, not the individual party. Where any of these party pairs together won a plurality of the votes (over the other pair in France), I counted them as the plurality winner, and estimated their position from the (seat-weighted) average of the two parties. However, the legislative median was always based on a single party (not an alliance) containing the median legislator.
51. Spain suffers from malapportionment across districts as well as a higher effective threshold. See Lijphart, Electoral Systems and Party Systems, p. 128;
52. Jorge De Esteban and Luis Lopez Guerra, 'Electoral Rules and Candidate Selection', in Howard R. Penniman and Eusebio M. Mujal-Leon,eds, Spain at the Polls 1977, 1979, and 1982 (Durham, N.C.: Duke University Press, 1985), pp. 52-3, 56.
53. In the SMD systems the fact that in Table 1 the legislative median is slightly less distant than the plurality winner is primarily the result of the alliances in Australia and France, where the legislative median was held by the closer party. There were also some cases where vote-seat distortion led the plurality vote winner to finish second in seats, as in Australia 1990, Canada 1979 and New Zealand 1981; however, in these cases the respective distances from the citizen median were rather similar between the plurality vote winner and runner-up, so the 'wrong' outcome did not make much additional difference.
54. As also seen in Powell and Vanberg, 'Election Laws, Disproportionality and the Left-Right Dimension', Fig. 2.
55. In these mature parliamentary democracies unified party voting can usually be assumed, as discussed in Michael Laver and Norman Schofield, Multiparty Government (New York: Oxford University Press, 1990), chap. 2 and Appendix A;
56. and Shaun Bowler, David M. Farrell and Richard S. Katz, Party Discipline and Parliamentary Government (Columbus: Ohio State University Press, 1999). The major exception among these countries is Italy.
57. Cox, Making Votes Count, chap. 12;
58. Laver and Schofield, Multiparty Government, 110-19;
59. Norman Schofield, 'Political Competition and Multiparty Coalition Government', European Journal of Political Research, 23 (1993),1-33.
60. Lanny W. Martin and Randolph T. Stevenson, 'Government Formation in Parliamentary Democracies', American Journal of Political Science, 45 (2001), 33-50.
61. Martin and Stevenson, 'Government Formation', p. 42.
62. Also see Huber and Powell, 'Congruence'.
63. Laver and Schofield, Multiparty Government, p. 113;
64. Ian Budge and M. J. Laver, 'The Relationship between Party and Coalition Policy in Europe', in M. J. Laver and Ian Budge, Party Polity and Government Coalitions, (New York: St Martin's Press, 1992), p. 416;
65. Wolfgang Mueller and Kaare Strøm, Coalition Governments in Western Europe (New York: Oxford University Press, 2003), p. 564.
66. Martin and Stevenson, 'Government Formation', pp. 34, 36.
67. The theoretical formulation suggesting the importance of the plurality party in its formateur role in non-majoritarian situations was David Austen-Smith and Jeffrey Banks, 'Elections, Coalitions and Legislative Outcomes', American Political Science Review, 82 (1988), 405-22. We cannot test Martin and Stevenson's causal hypothesis directly with the current data.
68. Laver and Schofield, Multiparty Government, p. 113;
69. Muller and Strøm, Coalition Governments, p. 568.
70. For a discussion of the importance of party pre-election promises as constraints, see Kaare Strøm, Ian Budge and Michael. J. Laver, 'Constraints on Cabinet Formation in Parliamentary Democracies', American Journal of Political Science, 38 (1994), 303-35;
71. also see Mueller and Strøm, Coalition Governments, p. 563.
72. Michael Pinto-Duschinsky, ' Send the rascals packing : Defects of proportional representation and the virtues of the Westminster Model', Times Literary Supplement, 25 September, 1999, pp. 10-12;
73. Michael Pinto-Duschinsky, G. Bingham Powell, Arend Lijphart, Jack Vowles and Matthew Soberg Shugart, 'Send the Rascals Packing!' Representation, 36 (1999), 117-55.
74. Cox, Making Votes Count, p. 237.
75. However, see the interesting proposal of Anthony McGann concerning the skewedness of the citizen preference distribution, 'The Advantages of Ideological Cohesion a Model of Constituency Representation and Electoral Competition in Multi-Party Democracies', Journal of Theoretical Politics, 14 (2002), 37-70.
76. Laver and Schofield, Multiparty Government, p. 111.
77. France 1981 is an exception, as the Socialists had a legislative majority, but nonetheless gave about 10 per cent of the cabinet positions to the Communists, with whom they had had an electoral alliance.
78. Note that this same argument applies, exactly, to taking the median party within the government.
79. Kathleen Bawn, 'Money and Majorities in the Federal Republic of Germany: Evidence for a Veto Players Model of Government Spending', American Journal of Political Science, 43 (1999), 707-36.
80. Laver and Schofield, Multiparty Government, pp. 167-81 and references therein.
81. Michael Laver and Kenneth Shepsle, eds, Cabinet Ministers and Parliamentary Government (New York: Cambridge University Press, 1994).
82. The subsequent analysis of governments estimates the position of the first government formed after the election. As shown in Powell, Elections as Instruments of Democracy, p. 220, in cases where the government fell before the next election, later governments were usually closer to the median than the original government. These later governments were on average 0.36 closer to the median (averaged by country), consistent with the greater importance of the median party in post-election bargaining. Seven of the PR systems (Belgium, Denmark, Finland, Germany, Italy, Norway and Sweden) experienced such multiple governments, as did France and New Zealand. If we used the average government position, rather than that of the first government, the PR advantage would be slightly enhanced. (Beyond the cases reported in Powell, only Norway 1997 experienced multiple governments of different party composition before the next election; in that case the new Labour government was slightly further from the median than the original KFR-SP-V government formed after the election.)
83. Also see McDonald et al., 'What Are Elections For?' p. 22, showing that although government position is responsive to the parliamentary median in both SMD and PR systems, the average 'distortion' (the absolute difference between the left-right position of the parliamentary median and the government) is nearly twice as large in the PR systems.
84. The nature of the influence of the median party is perhaps further illuminated by noting that of the twelve coalitions that do not include it, the median legislative party is an explicit support party in four cases (in Denmark and Spain), is extremely close (0.18) to one of the government parties in another, and joined a replacement coalition before the next election in two more cases.
85. These were adapted, corrected and updated from Powell, Elections as Instruments of Democracy, pp. 73-6, 261-2, focusing on explicit pre-election pacts, rather than general expectations;
86. additional sources included Sona N. Golder, The Logic of Pre-Electoral Coalition Formation (Columbus: Ohio State University Press, forthcoming);
87. and Keesing's Contemporary Archives; West European Politics.
88. For a brief review of the older literature and a statement of the spatial explanation of viable minority governments at the legislative median, as well as important institutional features, see Laver and Schofield, Multiparty Government, pp. 72-81, 111.
89. The eight minority cases without the plurality winner in the government are, on average, the most distant from the citizen median (1.6) and difficult cases to explain. They are all found in Scandinavia, were often short-lived, and seem to be linked to both pre-election alignments and the presence in the legislature of 'extremist' parties on the far left and far right, which were not included in any governments, but whose support on confidence motions could sustain or defeat them. These cases seem mostly closely to approximate majoritarian ideas of the unpredictability of government bargaining in PR.
90. A detailed study showing opposition influence on education policy under these conditions in France can be found in Frank Baumgartner, Conflict and Rhetoric in French Policymaking (Pittsburgh: University of Pittsburgh Press, 1989);
91. and Frank Baumgartner, 'Parliament's Capacity to Expand Political Controversy in France', Legislative Studies Quarterly, 12 (1987), 33-54.
92. Kaare Strøm, Minority Government and Majority Rule (New York: Cambridge University Press, 1990), chap. 6.
93. Cox, Making Votes Count, p. 230.
94. Kaare Strøm, 'Minority Governments in Parliamentary Democracies', Comparative Political Studies, 17 (1984), 199-227;
95. Strøm, Minority Government and Majority Rule; Powell, Elections as Instruments of Democracy, chap. 2, based largely on Herbert Doering, ed.
96. Parliaments and Majority Rule in Western Europe (New York: St Martin's, 1995).
97. For example, Powell, Elections as Instruments of Democracy, pp. 103-9, 267-8.
98. Michael Laver and W. Ben Hunt, Policy and Party Competition (New York: Routledge, 1992). The Laver-Hunt data are quite attractive because the question addressed exactly our current concern, provided a clear base-line ('no impact'), covered the countries of interest to us here and are based on at least five experts in each country. The primary limitations are reliance on a single question, an uncertain meaning of the upper end of the scale ('high impact'), no specification of the time frame and absence of comparable data at other time points than 1989.
99. This analysis explicitly assumes that the experts were referring to the government - opposition relations at the time of the survey in the first half of 1989 (Laver and Hunt Policy and Party Competition, p. 36.). These were governments resulting from elections in 1986, 1987 or 1988 in all countries except Greece, whose government dated to the 1985 election. In 1986-88 there were two elections in Denmark and France; the later outcomes are assumed to be the experts' point of reference.
100. Lijphart, Democracies, p. 212.
101. One problem that emerges as we extend the Laver-Hunt based analysis to the full dataset involves 'supported' minority governments. These are governments that do not themselves command a parliamentary majority based on the parties sharing cabinet portfolios, but they have some kind of fairly explicit arrangement with a small 'support' party outside the cabinet that makes it possible to sustain a government. There were no cases like this in the year of the Laver-Hunt survey, but there are six others in the dataset. A close look at the governments in question (Denmark 1984, Italy 1979, Norway 1985, Spain 1979, Spain 1989, Spain 1993) suggests that they reflect a variety of conditions, but in general are neither covert strong majority governments nor as open to opposition influence as the usual minority government. In the absence of more precise information, I made the simplifying decision to treat them as intermediate between these two, giving their oppositions a boost similar to control of an outside institution. The representatives of the small support parties are given a weight of 75 per cent of the government representatives. None of these small adjustments greatly affects the final conclusions.
102. The magnitude of difference depends on a number of factors, including what we assume 'high impact' means relative to the influence of the government. Recall that the estimate in column 2 of Table 4 shows the influence of the opposition relative to 'high impact'. In the calculations leading to policy-maker distances in Column 5, it is assumed that 'high impact' is half of full government impact-the government parties are, after all, overseeing the conduct of the ministries with all that means for policy initiative and implementation. If we assume 'high impact' implies as much influence as the government, then the PR advantage is additionally enhanced, but the basic conclusions are not altered.