| Computer Simulations of Voting Behavior / William R. Shaffer |
[Dec. 23rd, 2006|10:22 pm] |
Computer Simulations of Voting Behavior / William R. Shaffer, New York: Oxford University Press, 1972
Nothing very interesting in this book, but there's one interesting concept:
"tendency toward consistency" - even people who are "inconsistent" in their different pressures will tend to vote their "own" party. This shows the strength of habit-forming in the process of vote decision making. |
|
|
| What Voters Do / Redlawsk |
[Nov. 8th, 2006|09:21 am] |
What Voters Do: Information Search During Election Campaigns / David P. Redlawsk, Political Spychology, 25 (4), 2004, p. 595-610
This study sets out to find out how voters find information about candidates in a political campaign. There are two general systems of systematically collecting information: (1) compensatory rules - the voter seeks out a variety of bits on each of the candidates and compares them methodically, creating an index based on the comparative strengths and weaknesses of each one, to finally come up with the (rationally) best option. This method is difficult and taxing. (2) non-compensatory rules - the voter seeks spurious information on the candidates and drops off candidates that don't stand up to scrutiny by themselves, until only one is left. This system is less rational, because a candidate may be rejected based on one bit of information even though he may be preferred based on many other bits.
Also available are non-structured searches, although these are ignored - to the detriment of the study. The study assumes all candidates are on a level playing field from the outset, whereas in fact, a voter may be biased from the start, and therefore look for information only on a subset of the candidates, and only if they are rejected, will he attempt to seek information on other options.
According to the study, cognitive resources should influence how the decision is taken - the more resources, the more the voter would tend towards structured, compensatory rules of searching.
In truth, the major factor is the number of candidates - that is, how complex the decision is, the less one would be able to use structured, compensatory rules, and one turns to less structured options.
The study also found that the more structured your search is, the more likely you are to rate the rejected candidates higher. "These results support the notion that as coters learn more about candidates, they may find more to like about their less preferred options. Those who choose to focus mostly on their preferred candidate never learn much to like about other choices. But as they learn more about their preferred choice, they either do not learn disliked information or they fail to take into account any negatives that they do learn" (p. 606)
"cognitive misers"
"People often settle for "good enough" once they learn an adequate amount about the choices they face. Value maximizing behavior simply does not occur in complicated decision environments" (p. 596)
"When alternatives are many or indistinct, information is overwhelming, and the decision is important, decision makers may run up against their cognitive limits" (p. 596) |
|
|
| Retrospective Voting in American Nat. Elections / Fiorina - Chapter 1 |
[Oct. 2nd, 2006|03:51 pm] |
Morris P. Fiorina, Retrospective Voting in American National Elections (Chapter 1: Therories of Retrospective Voting)
In this introduction, Fiorina explains the distinction between prospective and retrospective voting -- (1) The different significance of electoral outcomes (moral support for the winner's suggested policy vs. favouring or rejecting the status quo); (2) the different interests of the voters (policy issues vs. actual results - and all that that entails); (3) normative implications for the concept of "responsible voter". He also shows the problematic nature of the then-pervailing Michigan School perception of politics (see also preface).
Finally, he discerns between "classical" retrospective voting (Key), which is all about reward and punishment, and "Downsian" retrospective voting, which merely views the results of the last administration as a good sign of what's to come from this same administration if it is reelected, and compared to what the opposition might have done given the same circumstances (i.e., some vestige of prospective voting remains).
Things to think about: Is this at all applicable to multi-party systems? Is it specifically applicable to the highly politicized Israeli public, with its greater concerns regarding specific policies? Is this applicable to present day with ever increasing coverage of the day-to-day politics and policy shifts?
Fiorina suggests that politicians aim to please the public by the end of their term - how does this interact with ever decreasing terms, as in Israel over the last two decades? Are politicians expected to change how they work in light of the fact that the average term has decreased significantly during this time period? Has the public changed its way of appraising politicians in light of this? |
|
|
| On Party Attachment in Western Europe/ Schmitt |
[Sep. 26th, 2006|07:57 pm] |
Schmitt, Herman (1989), "On Party Attachment in Western Europe and the Utility of Eurobarometer Data", West European Politics, 12, pp. 122-139
Schmitt offers a comparative study of stability of party ID in European countries between 1975-1987. He finds four types of behaviors in different countries: Stability (i.e., no change in the size of the population attached to a party, or to the intensity of the attachments), post-realignment (no change in the size of the population attached, but with less intensity), peripheral dealignment (the size of the population attched to a party lessens, but those whose attachment is strong are stable), and full dealignment (the size of the attached population grows smaller, and those that are still attached are less intensly so). The trend in the EC in general, during these years, mostly resembles peripheral dealignment, but the different countries show different patterns.
[However, even in the countries with more or less stable party ID, the numbers of those who are "fairly close" or "very close" to a party are rather small, and the majority of so called "attached" voters are merely sympathisers.
This brings us to an insight not fully alluded to by the study: the definition of "Party ID" has been diluted over the years, from a strong affinity between a voter and a political organization to something more ephemeral, to the point where it hardly means anything more than "support for party X at time T" (p. 138: "we find the Michgan style of measurement of Party ID heavily confounded with the Party Choice variable"). This variable, already not completely clear conceptually to political scientists, is even less clear to the subjects of polls, and may therefore bring about questionable results. The Eurobarometer questions themselves have been amended a number of times to allow for better comparability, seemingly with the aim of achieving results as close to the EC average as possible - again, a questionable end for a cross-national poll.
Two causes within the literature may be cited for this increasing confusion. One is the tendency to seek out "Party ID" in short intervals (see other studies documented here), and continue to refer to "identification" that changes considerably and numerous times over a period of several years as something other than temporary preferences.
The other is also exemplified in the present study - considering party ID not as a dichotomous variable (either you identify or you don't), but as a scale ranging from "sympathisers" to "very close" identifiers. This reinforces the perception of party ID as no more than support for the party, a mere quantitative degree above "indifferent", and in studies such as the present ones, creates a false impression of stability (because, after all, people are still "sympathisers", they didn't lose their party identification).] |
|
|
| "The Flow of the Vote in Israel" / Zuckerman |
[Sep. 25th, 2006|08:55 pm] |
Alan S. Zuckerman, "The Flow of the Vote in Israel: A Reconsideration of Stability and Change" in Arian & Shamir's Elections in Israel 1988
In this study, Zuckerman examines accepted truths about the stability of the Israeli electorate, and proves them false. Namely, he shows that between any two given elections, about a third of the voters change their chosen party, and more than a half change parties in the course of three consecutive elections. Zuckerman also shows that this holds true when comparing switching between blocs (unfortunately, the left bloc is seperate from Labor and the right from the Likud).
These numbers are very similar to most other western countries.
It can be assumed that these numbers have risen since the article was written, given the aggregate volatility of Israeli elections during the 90's.
The findings of this study pose a question whether the social factors are as predictive as they are thought to be over a longer period of time - if each voter changes his vote so often, how can any statements be made about the influence of steady social factors on his voting choice?
Interesting bits:
p. 190: "data based on voters' recollections of the previous election exaggerate the level of stability by about 10 percent"
p 192: quote: "One can doubt whether the traditionally popular research which aims to discover differences in social-structural and spychological characteristics between stable and changing voters makes much sense." (and more)
ibid: "There is no evidence that these are recent patterns or that voters used to be more stable."
p. 193: coding of broken-up parties as same parties (same as I did)
p. 195-197: "31 percent of the switchers jumped to the other side of the political spectrum" (about 3/5 of decided switchers"
p. 197: lots of points of interest about switching patterns between 1984-1988.
p. 199: "Voting appears to be one of those activities in which people tend to repeat what they did the last time they were in similar circumstances. Efforts to analyze electoral behavior should be able to eplain vote choice at least as well as a model built around the simple extrapolation of the previous vote." |
|
|
| Demographic Change and Partisan Support/ Abramson |
[Sep. 23rd, 2006|03:28 pm] |
Abramson, Paul R., "Demographic Change and Partisan Support", in Arian and Shamir's (eds.) The Elections in Israel 1988
Abramson tried to show in this article that, beyond hawkishness and SES, two factors served as primary agents in the shift of electoral power in Israel: Ethnicity, and generational replacement. He shows how merely the passing away of older voters and entry of new voters into the electorate accounts for between one quarter and two-thirds of the fall of Labor between 1969 and 1988. He also shows that most of this can be attributed to the entry of young Sepharadi voters (whereas Labor support among young Ashkenazim remained quite high).
He further shows that this cannot merely be attributed to the change in the ratio of Ashkenazim to Sepharadim in Israel - the generational gap is an important factor.
As an aside, he also notes that the political socialization of a voter happens in the first few elections he participates in, which explains why younger voters are more prone to vote Likud than those who came of age under Labor's dominant rule. |
|
|
| The Intifada and Israeli Voters / Arian & Shamir |
[Sep. 19th, 2006|10:38 pm] |
"The Intifada and Israel Voters: Policy Preferences and Performance Evaluations", Shamir and Arian in Arian & Shamir's "The Elections in Israel 1988"
In this study, Arian & Shamir compared two panel surveys conducted just at the beginning of the Intifada in 1987, and shortly before the elections in 1988, to see how the violence in the territories has influenced voting behavior. Their results show that there is a clear influence of issue voting (particularly long-term considerations, which is exactly oppposite to Barzilai's conclusions). Whatmore, these effects go beyond the influence of voting intentions in 1987, meaning the events of that year effected voting behavior, through a change in perception of the desired ends, as well as the evalution of means. These results also withstand control of ethnic and SES factors (p. 86-87).
The study also shows that evalution of the different parties' abilities plays a significant part in the decision on how to vote.
Nonetheless, it should be emphasized that most people who were affected by the events have merely strengthened their tendencies - hawks became more hawkish and doves became more dovish. a movement towards the center (or to the opposite bloc) was far more rare. As a result, party-switching was mainly within the party bloc, which explains the growth of the smaller parties in these elections.
Specific things to note:
The panel included 416 people who answered in both waves. Of these, only 265 indicated their voting intentions in both waves. The remaining panelists (a full 151 persons, or more than a third) did not indicate their voting intentions at least once. This significant group may very well represent less "hardcore" voters, meaning either less informed voters or voters with less party ID. Their decision making patterns may make a serious difference from the conclusions of this study.
Find out the reference given in note 20.
"Valence issues' partisan or confluctual relevance is created by the connection between the parties and the goals which the electorate desires -- which party is better able to achieve goals. These images may be retrospective or prospective; they often are both" (p. 83)
p. 84 - table 5.2 includes only prospective factors (relating to evaluation of the performance of Likud and Labor), but retrospective factors were also checked. On page 85 it is revealed that though they are not as strongle correlated with voting as prospective factors, they still ranged between .41 to .58 (bloc vote). Could I possibly get the raw data on this?
Possible methodological problem: when dividing the party system according to a left-right scale, it is no surprise that it correlated best with left-right tendencies. If we divided it by secular-religious (for example, moving tzomet to the same group as Ratz and Shinui), religiosity would most likely show up as an extremely useful predictor. Dummy variables should be used to disqualify such explanations. |
|
|
| "National Security Crises and Voting Behavior" / Gad Barzilai |
[Sep. 18th, 2006|10:18 pm] |
National Security Crises and Voting Behavior: The Intifada and the 1988 Elections, Gad Barzilai, in Arian & Shamir The Elections in Israel 1988
Barzilai sets out to show that, in the 1988 elections at least, ideology mattered more than other factors such as ethnicity or religion. In particular, he shows that attitude towards the use of force to quell the Intifada is highly correlated with voting on the left-right scale, even when controlling for SES factors. |
|
|
| "Realignment in the Isreli Part System"/ Shamir |
[Sep. 16th, 2006|02:46 pm] |
"Realignment in the Israeli Party System", Michar Shamir (in Arian & Shamir, Elections in Israel 1984)
This article addresses the question of the causes of the realignment which took place in Israel between 1969 and 1984. Two possible factors are examined: ethnic-voting and issue/ideology voting. The results show that the two factors worked together to create the change. At the outset, the distribution of the vote indicated a more issue-based voting (although there was a dominant party which received a plurality of the vote among all groups), and in progressive elections, different groups of Israelis elected to move to and from the Likud, with most switchers earlier on moving due to ideology, while in the early 80's the switching was due more to ethnic cleavages.
The study also included detailed tables of the switching patterns between Alignment and Likud over the period.
Interesting bits: Key suggeted two routes to electoral and party realignment: critical and "secular". (p. 268)
p. 272 - Arian's estimates of individual volatility over the period (25% in 1969, peaking at 50% in 1977 and returning to 25% by 1984)
Also: under "notes" (p. 293), an explanation why it doesn't matter that the poll doesn't correctly reflect the final results in the elections - "we are not interested in the actual vote distribution, but rather in its group base and conversion patterns". |
|
|
| Fiorina, Abrams and Pope, "The 2K US Pres. Elections" British J of Pol. Sci. 33(2) 163-(+25) |
[Aug. 4th, 2006|02:09 pm] |
Fiorina, Morris, Samuel Abrams, and Jeremy Pope. "The 2000 US Presidential Election: Can Retrospective Voting Be Saved?". British Journal of Political Science, 33(2), 2003, pp: 163-188.
Fiorina et al try to come to terms with the results of the 2K elections in the US, in which the incumbent (sorta) Gore, coming from a position of power as the VP of a highly successful (if morally dubious) prez, lost what should have been, by conventional Retrospective wisdom a landslide victory, and achieved a mere tie which led to his ultimate defeat in the elections, faced with unexperienced G.W.Bush.
The article examines several hypothesis for such an occurance:
1. Retrospective factors ("the fundamentals", in their phrasing) were somehow weaker for this particular election. (variations: A. VPs don't get the credit; B. people were happy for too long, so they didn't credit Clinton-Gore for the good economy, but rather discounted it from their factors for voting; C. The credit for the booming economy went to Greenspan and entrepreneurs, and was not associated with the administratio; D. Gore didn't want credit because he didn't want to be associated with Clinton morally.)
2. Personality-Voting: Al Gore was an unattractive candidate personally. (variations: A. Oprah - personality is more important now than in the past; B. Media bias.)
3. Ideological (prospective) - Al Gore went too far to the left.
4. Clinton Fatigue - elections were an anomaly, because there was a discrepancy between Clinton's work ratings and personal ratings.
Results: Personality - very little effect. All in all, the public didn't view Gore and Bush as so far apart in that regard.
Ideology - some effect, however, not as much effect as could have been, as voters seem to have compensated for the larger distances between candidates by giving this factor less weight. (possibly interpretaion: voters decide who is closer, but they don't care specifically for the actual distance).
Clinton Fatigue - substantial effect. Fiorina at el put most of the blame for the miss of the model on this factor, going as far as suggesting that a "dummy-variable" for 2000 elections would suffice to "save" the model.
Fundamentls - the weaker relationship in this election had an impact on Gore's success, but this is but an anomaly, and unless there is further proof in the future of a trend, there's no reason to change the model. |
|
|
| Toka, Gabor (1998), Party Appeals and Voter Loyalty in New Democracies |
[Jul. 29th, 2006|11:59 pm] |
Toka, Gabor (1998), Party Appeals and Voter Loyalty in New Democracies, Politicals Studies 46, 589-610
This article asks what types of voting behavior promote a stable voting choice - i.e., low (gross) volatility. It focuses on East European countries, but could easily apply to more long-lived democracies as well. It examines three types of voting: value-based, structure-based, and cleavage-based, which is a combination of both. It also tests a different model which includes organizational encapsulation, that is - the voter's immersion in a partisan organization. The study show a strong effect of value based voting on voting stability, and a somewhat lesser effect of structural voting, and that too only when this includes organizational encapsulation.
Some interesting bits in the article:
1. the defition of net and gross volatility (macro-level and voter-level changes in voting behavior);
2. high volatility signals to politicians that working through established parties is not necessarily the best option, and other venues may be more profitable, electroally (p. 590, 592);
3. on the other hand - very low volatility may be a sign of a rigidly polarized society. Plus - an overly stable electorate means little political accountability;
4. p.592-596 - common explanations of volatility (esp. 595);
5. party ID is strengthened when the people one meets and the media one consumes are his copartisans - hence the common use of structural (or cleavage) voting as explanation;
6. value-voting is more likely when the voters have easy access to (and the resources to account for) policy commitments, competence and values of the competing parties. When that is lacking, voters resort to "heuristical" voting based on socio-demographic and identity factors rather than values (p. 598, Klingmann and Watterberg);
7. given a high-level of political fragmantation, structural voting becomes less likely because it is less likely that a single group would be wholly contained within a single party - implication: value voting reduces structural voting (beacuse of 3) (p. 600).
References: Bartolini and Mair, Identity, Competition and Electoral Availability |
|
|
| Kiewiet and Rivers, "A Retrospective on Retrospective Voting", Pol. Behavior, 6(4), 1984, 369-393 |
[Jul. 25th, 2006|10:29 pm] |
A meta-analysis of the literature on retrospective voting. Places the gensis of the field in Gerald Kramer (1971). Notes that the majority of the literature in the field has focused on one particular model heralded by Kramer, although other models significantly different from it are possible and even likely.
The model followed by most scholars is based on three hypotheses:
(1) Voting is retrospective (i.e., voters decide according to measures of things that have already happened, not according to promises for future policies)
(2) incumbency-oriented (the decision is for or against the incumbent or his party, whereas the identity of the opposed candidate is of less importance)
(3) based upon the results of economic policies, and not upon the actual policies themselves, meaning a (national or private) pocketbook voting, and not an ideology-driven vote.
In other words: voters punish or reward incumbents according to how the economy is doing (nationally or privately) under their administration.
Two types of research was done in this field: time-series data and cross-section data. The first examines the aggregate response of the nation to national-level economic indicators, and the latter examines individual response to different types of economic cues.
The research has found that the most significant influence nationally is exerted by the economic indicators of the second quarter of an elections year. National indicators have found strong influence in both presidential and congressional elections (including midterms), and the most successful indicator is change in per capita income, estimated as between 0.5 to 1.0 decrease in votes for every percentage point decrease in income per capita.
Individual cues are less clear, with only prez elections appearing to be influenced significantly, and even then numbers range between 0.08 to 0.34 percentage points decrease for every point income per capita decrease. Sociotropic voting, that is - voting influenced by the individual's perception of the nation's economic well-being, are somewhat stronger, but still considerably less successful in explaining aggregate vote than the time-series data.
Pooling data from both types of research could bring further information to the fore, such as the possiblity that while national indicators drive and incumbency-oriented vote, but a more partisan approach is taken when dealing with redistributional policies, which effect different strata of society differently. |
|
|
| Bowler, S. and Lanour, D.J., "New Party Challenges and Partisan Change", Pol Behavior 18(4), 1996 |
[Jul. 22nd, 2006|10:07 pm] |
Bowler, S. and Lanour, D.J., "New Party Challenges and Partisan Change: The Effects of Party Competition on Party Loyalty", Political Behavior 18(4), 1996, pp. 327-343
This study looks into the alleged stability and robustness of Party ID, by challenging it with a simple change in the party system: the addition of a new substantial party. The setting used for the study is the Canadian elections of 1988, which introduced the Reform Party to the political system, but in which the Reform has not gained much ground as far as representative election is concerned.
The study shows that the mere introduction of a viable new political option (measured by the spending of the Reform candidate on the campaign) significantly lowered the Party ID of conservative supporters in those areas, compared to the previous elections. They also showed that the appearance of the Reform party had no influence on the party ID of left-wing party supporters.
These results show that party ID is possibly far less stable than previously thought. As a sidenote, the study also shows (table 6-8) that issue, economic and personal factors all had a significant influence on the reported party ID of voters. |
|
|
| Best, R., "Disloyalty, De-Mobilization or Diminishing Numbers?", 2005 (unpublished?) |
[Jul. 21st, 2006|03:50 pm] |
Looks into the purportedly diminishing tendency towards class-voting (views voting on issues in a negative light, see also Franklin 1985b, there). Cites three possible reasons for this trend: class groups are disloyal to their traditional parties (i.e., realignment, volatility), lower turnouts among distinct class groups (disenchantment), and class groups diminishing in size due to sociatal changes. Compares 6 European democracies over nearly 30 years.
Reaches the conclusion that class based voting is still strong, but that class lines themselves have blurred, and there are more people in a no-man's-land in the middle which could go both ways. However, for those who are distinctly within a group, class voting is still evident. This works against left, because their power centers are dwindling. |
|
|
| navigation |
| [ |
viewing |
| |
most recent entries |
] |
| |
|
|
![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small}
thesisiod's journal