Which matters most? Comparing the impact of issues and the economy in American, British and Canadian elections

British Journal of Political Science

Volumn 34, Issue 3, 2004, Pages 555-563

  Blais, André ^a   Turgeon, Mathieu ^b   Gidengil, Elisabeth ^c   Nevitte, Neil ^d   Nadeau, Richard ^a  

^a UNIVERSITÉ DE MONTRÉAL   (Canada)

^b UNIVERSITY OF TEXAS AT AUSTIN   (United States)

^c MCGILL UNIVERSITY   (Canada)

^d UNIVERSITY OF TORONTO   (Canada)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 3042822086     PISSN: 00071234     EISSN: None     Source Type: Journal    
DOI: 10.1017/S0007123404220178     Document Type: Note

References (39)

1. See, most recently, Anders Westholm, 'On the Return of Epicycles: Some Crossroads in Spatial Modeling Revisited', Journal of Politics, 63 (2001), 436-81; Stuart Elaine Macdonald, George Rabinowitz and Ola Listhaug, 'Sophistry versus Science: On Further Efforts to Rehabilitate the Proximity Model', Journal of Politics, 63 (2001), 482-500.
2. See, most recently, Anders Westholm, 'On the Return of Epicycles: Some Crossroads in Spatial Modeling Revisited', Journal of Politics, 63 (2001), 436-81; Stuart Elaine Macdonald, George Rabinowitz and Ola Listhaug, 'Sophistry versus Science: On Further Efforts to Rehabilitate the Proximity Model', Journal of Politics, 63 (2001), 482-500.
3. The authors of the American Voter argued that 'many voters fail to appreciate that an issue exists, others are insufficiently involved to pay attention to recognized issues, and still others fail to make connections between issue positions and party policy.' See Angus Campbell, Philip Converse, Warren Miller and Donald Stokes, The American Voter (Chicago: The University of Chicago Press, 1960), p. 183.
4. The fact is that for the great majority of respondents the two models make identical predictions about which party will be preferred. For instance, Blais et al. find more support for the proximity model, and report that each model separately explains the same proportion of variance. See André Blais, Richard Nadeau, Elisabeth Gidengil and Neil Nevitte, 'The Formation of Party Preferences: Testing the Proximity and Directional Models', European Journal of Political Research, 40 (2001), 81-91.
5. Gerald H. Kramer, 'The Ecological Fallacy Revisited: Aggregate vs Individual-Level Findings on Economics and Elections, and Sociotropic Voting', American Political Science Review, 65 (1983), 131-43; Raymond Duch, Harvey D. Palmer and Christopher J. Anderson, 'Heterogeneity in Perceptions of National Economic Conditions', American Journal of Political Science, 44 (2000), 635-52.
6. Gerald H. Kramer, 'The Ecological Fallacy Revisited: Aggregate vs Individual-Level Findings on Economics and Elections, and Sociotropic Voting', American Political Science Review, 65 (1983), 131-43; Raymond Duch, Harvey D. Palmer and Christopher J. Anderson, 'Heterogeneity in Perceptions of National Economic Conditions', American Journal of Political Science, 44 (2000), 635-52.
7. Michael B. MacKuen, Robert S. Erikson and James A. Stimson, 'Peasants or Bankers? The American Electorate and the American Economy', American Political Science Review, 86 (1992), 597-611; Helmut Norporth, chapter entitled 'The Economy', in his Comparing Democracies: Elections and Voting in Global Perspective (Thousand Oaks, Calif.: Sage, 1996); Robert S. Erikson, Michael B. MacKuen and James A. Stimson, 'Bankers or Peasants Revisited; Economic Expectations and Presidential Approval', Electoral Studies, 19 (2000), 295-312.
8. Michael B. MacKuen, Robert S. Erikson and James A. Stimson, 'Peasants or Bankers? The American Electorate and the American Economy', American Political Science Review, 86 (1992), 597-611; Helmut Norporth, chapter entitled 'The Economy', in his Comparing Democracies: Elections and Voting in Global Perspective (Thousand Oaks, Calif.: Sage, 1996); Robert S. Erikson, Michael B. MacKuen and James A. Stimson, 'Bankers or Peasants Revisited; Economic Expectations and Presidential Approval', Electoral Studies, 19 (2000), 295-312.
9. Michael B. MacKuen, Robert S. Erikson and James A. Stimson, 'Peasants or Bankers? The American Electorate and the American Economy', American Political Science Review, 86 (1992), 597-611; Helmut Norporth, chapter entitled 'The Economy', in his Comparing Democracies: Elections and Voting in Global Perspective (Thousand Oaks, Calif.: Sage, 1996); Robert S. Erikson, Michael B. MacKuen and James A. Stimson, 'Bankers or Peasants Revisited; Economic Expectations and Presidential Approval', Electoral Studies, 19 (2000), 295-312.
10. Norporth, 'TheEconomy'; Michael S. Lewis-Beck and Martin Paldarn, 'Economic Voting: An Introduction', Electoral Studies, 19 (2000), 113-21.
11. Norporth, 'TheEconomy'; Michael S. Lewis-Beck and Martin Paldarn, 'Economic Voting: An Introduction', Electoral Studies, 19 (2000), 113-21.
12. Morris Fiorina, Retrospective Voting in American National Elections (New Haven, Conn.: Yale University Press, 1981), p. 5; see also Samuel Popkin, John W. Gorman, Charles Phillips and Jeffrey A, Smith, 'Comment: What Have You Done for Me Lately? Towards an Investment Theory of Voting', American Political Science Review, 70 (1976), 779-805.
13. Morris Fiorina, Retrospective Voting in American National Elections (New Haven, Conn.: Yale University Press, 1981), p. 5; see also Samuel Popkin, John W. Gorman, Charles Phillips and Jeffrey A, Smith, 'Comment: What Have You Done for Me Lately? Towards an Investment Theory of Voting', American Political Science Review, 70 (1976), 779-805.
14. Michael R. Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election', American Journal of Political Science, 39 (1995), 714-44; Michael R. Alvarez and Jonathan Nagler, 'When Politics and Models Collide: Estimating Models of Multicandidate Elections', American Journal of Political Science, 42 (1998), 55-96; Michael R. Alvarez and Jonathan Nagler, 'Economics, Entitlements, and Social Issues: Voter Choice in the 1996 Presidential Election', American Journal of Political Science, 42 (1998), 1349-63; Michael R. Alvarez, Jonathan Nagler and Shaun Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections', American Political Science Review, 94 (2000), 131-49; Michael R. Alvarez, Jonathan Nagler and J. R. Willette, 'Measuring the Relative Impact of Issues and the Economy in Democratic Elections', Electoral Studies, 19 (2000), 237-53.
15. Michael R. Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election', American Journal of Political Science, 39 (1995), 714-44; Michael R. Alvarez and Jonathan Nagler, 'When Politics and Models Collide: Estimating Models of Multicandidate Elections', American Journal of Political Science, 42 (1998), 55-96; Michael R. Alvarez and Jonathan Nagler, 'Economics, Entitlements, and Social Issues: Voter Choice in the 1996 Presidential Election', American Journal of Political Science, 42 (1998), 1349-63; Michael R. Alvarez, Jonathan Nagler and Shaun Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections', American Political Science Review, 94 (2000), 131-49; Michael R. Alvarez, Jonathan Nagler and J. R. Willette, 'Measuring the Relative Impact of Issues and the Economy in Democratic Elections', Electoral Studies, 19 (2000), 237-53.
16. Michael R. Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election', American Journal of Political Science, 39 (1995), 714-44; Michael R. Alvarez and Jonathan Nagler, 'When Politics and Models Collide: Estimating Models of Multicandidate Elections', American Journal of Political Science, 42 (1998), 55-96; Michael R. Alvarez and Jonathan Nagler, 'Economics, Entitlements, and Social Issues: Voter Choice in the 1996 Presidential Election', American Journal of Political Science, 42 (1998), 1349-63; Michael R. Alvarez, Jonathan Nagler and Shaun Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections', American Political Science Review, 94 (2000), 131-49; Michael R. Alvarez, Jonathan Nagler and J. R. Willette, 'Measuring the Relative Impact of Issues and the Economy in Democratic Elections', Electoral Studies, 19 (2000), 237-53.
17. Michael R. Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election', American Journal of Political Science, 39 (1995), 714-44; Michael R. Alvarez and Jonathan Nagler, 'When Politics and Models Collide: Estimating Models of Multicandidate Elections', American Journal of Political Science, 42 (1998), 55-96; Michael R. Alvarez and Jonathan Nagler, 'Economics, Entitlements, and Social Issues: Voter Choice in the 1996 Presidential Election', American Journal of Political Science, 42 (1998), 1349-63; Michael R. Alvarez, Jonathan Nagler and Shaun Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections', American Political Science Review, 94 (2000), 131-49; Michael R. Alvarez, Jonathan Nagler and J. R. Willette, 'Measuring the Relative Impact of Issues and the Economy in Democratic Elections', Electoral Studies, 19 (2000), 237-53.
18. Michael R. Alvarez and Jonathan Nagler, 'Economics, Issues and the Perot Candidacy: Voter Choice in the 1992 Presidential Election', American Journal of Political Science, 39 (1995), 714-44; Michael R. Alvarez and Jonathan Nagler, 'When Politics and Models Collide: Estimating Models of Multicandidate Elections', American Journal of Political Science, 42 (1998), 55-96; Michael R. Alvarez and Jonathan Nagler, 'Economics, Entitlements, and Social Issues: Voter Choice in the 1996 Presidential Election', American Journal of Political Science, 42 (1998), 1349-63; Michael R. Alvarez, Jonathan Nagler and Shaun Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections', American Political Science Review, 94 (2000), 131-49; Michael R. Alvarez, Jonathan Nagler and J. R. Willette, 'Measuring the Relative Impact of Issues and the Economy in Democratic Elections', Electoral Studies, 19 (2000), 237-53.
19. Alvarez and Nagler, 'When Politics and Models Collide'. We use a simple binary logit model for the two American elections with only two parties.
20. For American elections, but not for British and Canadian elections, Alvarez et al. included opinions on issues such as abortion as well as distance between respondents' and parties' positions. To be consistent and to maximize cross-national comparability, we systematically use distance variables.
21. J. B. Lewis and G. King, 'No Evidence on Directional vs. Proximity Voting', Political Analysis, 8 (2000), 21-33. They demonstrate that the debate hinges on key assumptions which cannot be tested with existing methods and data.
22. MacKuen, Erikson and Stimson, 'Peasants or Bankers?'.
23. Taxes are set up by governments. Unemployment and inflation are not directly decided by governments even though they may be strongly affected by measures such as monetary policy. Note that views about taxes did not have an independent effect on vote choice in the 1987 British election. See Alvarez, Nagler and Bowler, 'Issues, Economics, and the Dynamics of Multiparty Elections'.
24. Information on these two variables were not available in the 1987 British Election Study.
25. Ian Budge, Ivor Crewe and Dennis Farlie, eds, Party Identification and Beyond: Representations of Voting and Party Competition (New York: Wiley, 1976).
26. See, in particular, Eric Schickler and Donald P. Green, 'The Stability of Party Identification in Western Democracies', Comparative Political Studies, 30 (1997), 450-83.
27. André Blais, Elisabeth Gidengil, Richard Nadeau and Neil Nevitte, 'Measuring Party Identification: Britain, Canada, and the United States', Political Behaviour, 23 (2001), 5-22.
28. Richard Rose and Ian McAllister, The Loyalties of Voters: A Lifetime Learning Model (London: Sage, 1990); James Adams and Samuel Merrill III, 'Modeling Party Strategies and Policy Representation in Multiparty Elections: Why Are Strategies So Extreme?' American Journal of Political Science, 43 (1999), 765-91.
29. Richard Rose and Ian McAllister, The Loyalties of Voters: A Lifetime Learning Model (London: Sage, 1990); James Adams and Samuel Merrill III, 'Modeling Party Strategies and Policy Representation in Multiparty Elections: Why Are Strategies So Extreme?' American Journal of Political Science, 43 (1999), 765-91.
30. Blais, Gidengil, Nadeau and Nevitte, 'Measuring Party Identification'.
31. For a thorough discussion of the specificities of multinomial choice models, see Alvarez and Nagler, 'When Politics and Models Collide'.
32. The optimal position is the one that would yield the most votes for a party.
33. Let us take the American 1988 presidential election as an example. With the logit estimates reported in Table Al of the website version of this article, one can compute the probability of voting for Bush for each individual as follows: Prob(Vote for Bush) = 1/(1 + exp(- (- 2.72 × Democrat identifier + 1.22 × Republican identifier + 0.31 × Ideology for Dukakis + 0.26 × Government services for Dukakis - 0.25 × Defence for Dukakis + 0.14 × Health for Dukakis + 0.25 × Government jobs for Dukakis +0.19 × Russia for Dukakis + 0.33 × Women's rights for Dukakis - 0.13 × Ideology for Bush - 0.24 × Government services for Bush - 0.52 × Defence for Bush - 0.17 × Health for Bush - 0.35 × Government jobs for Bush - 0.20 × Russia for Bush - 0.22 × Women's rights for Bush + 0.83 × Economy (retrospective) - 0.53 × Economy (prospective) + 0.04 × Personal Finances (retrospective) + 0.64 × Personal Finances (prospective) + 0.10×East + 0.32 × South - 0.04 × West +0.13 × Age18-29 + 0.36 × Age30-44 - 0.06 × Age45-59 - 0.09 × Education - 0.06 × Female + 2.77))). For each Prob(Vote for Bush) ≥ 0.5, the individual is predicted to vote for Bush. The others are predicted to vote for Dukakis. We can now easily calculate predicted vote shares under the full model for each candidate (or party). To calculate predicted vote shares under the counterfactual models we proceed in the same way except that we fix either all the issue coefficients to 0 (for the issue counterfactual scenario), keeping constant all other coefficients, or all the economic coefficients to 0 (for the economy counterfactual scenario), keeping constant all other coefficients. We compute predicted vote shares for each scenario and compare them with those obtained from the full model.
34. As indicated in fn. 9, we have a predicted vote choice for all individuals and for all scenarios (i.e. full model scenario, issue counterfactual scenario and economy counterfactual scenario). To compute the gross effect, we simply count the number of individuals who are predicted to have a different vote choice under the full model scenario and each of the two counterfactual scenarios. This number is divided by the total number of individuals in the analysis. For example, out of the 649 individuals included in the analysis of the American 1988 presidential election, forty-two are predicted to vote differently under the full model scenario and when the issues are assumed to have no impact on their vote. This gives the figure of 6.5 points reported in Table 1. Of course, many of these movements cancel out since some Bush voters become Dukakis voters, and vice versa. The net effect indicates how total vote shares are affected. Here, all in all, twenty fewer individuals are predicted to vote for Dukakis (and twenty more for Bush) when the issues are set to have no impact (compared to the situation under the full model), yielding a net effect of 3.1 percentage points.
35. Appendix A of the website version of this article shows the multinomial probit (and binomial logit for two-party vote choice) estimates of the determinants of vote choice in each election. Appendix B presents a description of the variables.
36. Take for example the British 1987 election, The simulation indicates that the Conservatives gained 12.4 percentage points because of the issues, and the Labour and the Social Democratic Party (SDP)/Liberals Alliance lost 3.0 and 9.3 points, respectively. The net gains and losses add up to 24.7, divided by 2.
37. We thank one of the referees for suggesting that additional test. A case could also be made for calculating the net impact on the average probability of voting for each party. The mean probability of voting for a party is, however, quite different from the actual vote that a party gets. For example, for the 1987 British general election, our model predicts the following vote percentages for the Conservatives, Labour and the Alliance party respectively: 43.7 per cent, 35.0 per cent and 21.4 per cent. However, the mean probability of voting for each of these parties across all individuals is 38.5 per cent, 37.4 per cent and 27.2 per cent in the respective order. As a consequence, vote probability provides useful information about individual voters' inclinations, but mean vote probabilities do not adequately reflect actual vote support.
38. Richard Johnston, André Blais, Henry E. Brady and Jean Crête, Letting the People Decide: Dynamics of a Canadian Election (Montreal: McGill-Queen's University Press, 1992), p. 141.
39. Michael R. Alvarez and Jonathan Nagler, 'A New Approach for Modelling Strategic Voting in Multiparty Elections', British Journal of Political Science, 30 (2000), 57-75, p. 242.