Information Has No Role In Explaining Individuals’ Vote Choices. Discuss.

Abstract

The question this essay will try to answer is an empirical one. It seeks to analyse whether information has a role in individual vote choices, not whether this is good or bad. Although we will see some normative analysis within the essay, it is mainly driven by empirical literature and therefore forms an empirical argument. The consensus in the political science profession is that voters are rational but this isn t the case in reality. This essay will look into the validity of this argument and what really explains an individuals vote choice. We will begin with the Columbia studies followed by the Michigan studies, where we will see that voters do in general lack information and political knowledge. This section will also look into the works of influential authors such as Paul Lazarsfeld, Angus Campbell, Warren Miller, Donald Stokes and Philip Converse. The Michigan Model will also be analysed and used in this section. Once we see that information is almost non-existent in vote choices, we will begin section two where we look into what actually does explain an individuals vote choices. Here I will look into information shortcuts and the miracle of aggregation . Recent studies do suggest that information can play a role with the increasing numbers of issue voting, which the essay will look briefly into, but issue voting still remains illusionary and only a small contributor to vote choices. Information can play a role but it is often with supplement to economic evaluation. Therefore, the third and final section of this essay will look at how the economy, retrospective voting and pocket book voting can explain vote choices. In this section we will use literature from very influential authors like Anthony Downs, Samuel Popkin and Morris Fiorina. The essay is rich in literature, using so many other influential texts on the field too, providing a very rich insight into information and its role in vote choices. By the end of the essay, we will have seen that information cannot explain an individuals vote choices the only time information can play a role is when it is in connection with economic evaluation.

Keywords: Michigan model information shortcuts economic voting pocket book voting issue voting party identification brand loyalties .

Introduction

The first study on voting behaviour was by Columbia University, followed by the second, which were the Michigan studies. Both of these were unintentional (Bartels and Leighley 2010). Both studies showed that information has no role in vote choices. The Michigan University later on created the Michigan Model , which did incorporate little elements of information, such as issue opinions and campaign activity, but nevertheless the model was used to show how there was a lack of political knowledge in American voters (Bartels 2008). Information shortcuts and retrospective voting explain vote choices more than information. Although there are claims that voters do use information and vote based on issues, it is very difficult to prove the validity of these claims and issue voting is often illusionary (Bartels 2008). Voters are definitely not fools, but neither are they rational (Downs 1957). Vote choices are usually explained by the economy because it is the number one issue usually the only issue most voters care about. Information can work alongside in connection with economic evaluation but an individual s vote choices are definitely not explained by information alone (Achen and Bartels 2013). If anything, vote choices are explained by partisanship and economic evaluations with little elements of information to support it.

Section One: The Columbia Studies The Michigan Model

According to Rick Shenkman, in his influential book just how stupid are we? information and political knowledge is very important (Bartels 2008). Information can play a beneficial role when voters make their decision on which candidate or party to vote for, but unfortunately voters lack this rationality (Bartels 2008). Shenkman says, Public ignorance is the most obvious cause of the foolishness that marks so much of American politics (Shenkman 2009). Information plays no role in individual vote choices but it should and we need to strive for this civic improvement, answering both the normative and empirical side of the argument (Bartels 2008). According to Shenkman, the consensus in the political science profession is that voters are rational but unfortunately this isn t the case in reality (Bartels 2008:45 Przeworski, Stokes and Manin 1999). Vote choices are influenced more by faith and ideology than a careful prediction of consequences identified by using information and rationality. Due to this, voters don t make sensible choices (Bartels 2008). According to Shenkman, voters can only make sensible choices if they are rational and informed but information doesn t play a role in individuals vote choices so voters lack the ability to make sensible choices (Bartels 2008 Shenkman 2009).

Shenkman s work was very influential but the origins of voting research began in 1940 at Columbia University (Bartels and Leighley 2010). Paul Lazarsfeld and his colleagues from Columbia University surveyed 600 voters from the Erie County, Ohio region, in the 1940 presidential campaign (Bartels and Leighley 2010). The University then conducted a second study in 1948 at Elmira, New York, providing the basis of the influential book Voting: A Study of Opinion Formation in a Presidential Campaign (Bartels and Leighley 2010). The Columbia studies found out that mass media had little direct impact on vote choices it was mainly personal influences that played a role when an individual decided to vote (Bartels and Leighley 2010). Information has no role in explaining vote choices, instead it s explained by brand loyalties , rooted in religion and social class (Bartels and Leighley 2010). Lazarsfeld et al argued that, for many voters political preferences may better be considered analogous to cultural tastes in music, literature, recreational activities, dress, ethics, speech, social behaviour both have their origin in ethnic, sectional, class, and family tradition (Bartels and Leighley 2010:4). These are then reinforced and strengthened by interactions with like-minded acquaintances through homogenous social networks (Popkin 1991). This relates back to what Shenkman said previously, vote choices are characterised more by faith than by conviction, more by wishful expectation rather than careful prediction of consequences (Shenkman 2009 Bartels and Leighley 2010). In addition, Lazarsfeld et al found out that voters often ignored or misperceived their favourite candidate s policies when these were in conflict with their own, claiming voters only accept the information that supports and fits into their brand loyalties (Bartels and Leighley 2010). The Columbia studies demonstrated in considerable detail the extent to which Truman s late surge represented the reactivation of latent democratic loyalties as the salience of traditional class issues came to the fore over the course of the fall campaign showing us that information cannot break loyalties and partisanships (Bartels and Leighley 2010:5). In summary, the Columbia studies showed us that information could not explain vote choices because people fall short from the democratic expectations of a well-informed citizen and break the general requirement of the political science profession (Thomassen 2005). Vote choices are better explained using brand loyalties and partisanship (Bartels 2008).

Lazarsfeld and his colleagues at Columbia University demonstrated the rich potential of election surveys as valuable data for understanding voting behaviour (Bartels and Leighley 2010). The next, and even more advanced election studies, emerged in the University of Michigan. Angus Campbell and Robert Kahn conducted a national survey on foreign policy attitudes but also asked respondents whether they planned to vote in the upcoming presidential elections and for which party (Campbell and Kahn 1952:3, cited in Bartels and Leighley 2010). After the much-publicized failure of the Gallup Poll to foretell Truman s come-from-behind victory they decided to re-interview the same respondents to analyse the crystallization of the vote and assess the various factors that affected individuals vote choices (Bartels and Leighley 2010:6). The Michigan studies are one of the longest-running research projects in the history of academic social science (Bartels and Leighley 2010). The university conducted national surveys in 1952, 1954, 1956, and 1958. Angus Campbell, Warren Miller, Donald Stokes and Philip Converse, later then produced the most important landmark in electoral research, The American Voter (Campbell at al 1960, cited in Bartels and Leighley 2010:7). The data from The American Voter was from the Michigan surveys in 1952 and 1956. The findings showed that voters had attachments to parties but this isn t just a reflection of party loyalty or group membership or of any other factors that may lead to perceptual distortion attitudes toward the objects of politics, varying through time, can explain short-term fluctuations in partisan division of the vote, whereas party loyalties and social characterises, which are relatively insert through time, account but poorly for these shifts (Campbell et al 1960:65, cited in Bartels and Leighley 2010:8). Nevertheless, the Michigan studies are in agreement with Columbia studies that information has no role in explaining vote choices. Individuals are influenced by other factors, which is more than just brand loyalties , when making political choices (Jakobsen 2013).

Campbell et al later on created the Michigan Model (also known as the funnel of causality) to show all these factors that explain an individual s vote choice (Jakobsen 2013). The model links all the factors together in a V-shape (Appendix one). At its narrow end are located the factors that the authors believe are the immediate determinants of the vote whilst at the more wider end are the remote underlying factors, usually the brand loyalties that was identified in the Columbia studies (Jakobsen 2013). The model shows us that the economic structure, social divisions and historical patterns, create our value orientations, which then lead to party attachments, forming our brand loyalties . These are further reinforced by government actions, influence of friends, media influences, candidate image and economic conditions, which are more immediate determinants of individuals vote choices (Appendix one). The model is a blueprint to explain individuals vote choices and it doesn t give much role to information (Jakobsen 2013).

Philip Converse in his 1964 article, The Nature of Belief Systems in Mass Publics , argued that the majority of the voters lacked the ability or desire to understand information that did not relate to them as an individual, which links back to Shenkman s idea that voters lack the ability to make sensible choices based on careful prediction of consequences (Converse 2008). The Michigan data further highlights this as it suggests, many people know the existence of a few if any of the major issues of policy, and that major shifts of electoral strength reflect the changing association of parties and candidates with general societal goals rather than the details of legislative or administrate action (Campbell et al, 1960:170, cited in Bartels and Leighley 2010:9). Thus, individuals are influences by partisan loyalties developed early in life but also influenced by short-term forces before elections, such as a candidate s popularity or policy agenda. Here information can play a role but it s a very limited one. In summary, the Columbia studies and the Michigan Model shows us that information cannot explain individual vote choices. The Michigan Model does give a slight small role to information in its model, as it does gives a small role to issue opinions and campaign activity (Appendix one), but it is still a very limited and vague one (Popkin 1991). In 2004, a few scholars replicated the Michigan studies and found out that things have not changed much and the model is still very much accurate in explaining vote choices in the 21^st century (Bartels 2008).

Section Two: Rational Ignorance

Further scholars also argued that information does not play a role in explaining vote choices, nor does it need to play a role. Information doesn t play a direct active role because it is substituted with other mechanisms used when voting (Bartels 2008). Does it really matter if you can name the secretary of defence or how long a senates length of term is? because voters can make good choices without all of this unnecessary information (Bartels 2008). In fact, this is usually the case. All voters need is basic information, not all the detailed knowledge of candidate policies and platforms so information isn t totally irrelevant but nor is it the factor that explains voting behaviour (Page and Shapiro 1992). Voters can use limited information from partisan stereotypes, endorsements from interest groups and personal narratives to vote, which is referred to as information shortcuts (Bartels 2008). In the 1976 presidential campaign, many Mexican-American voters were unsure to vote for then-president Gerald Ford. Gerald Ford showed on several occasions in the media that he did not know how to eat tamale, which the Mexican-American voters then viewed him as unable to relate to the Mexican culture (Popkin 1991). This is an example of where voters used small information shortcuts to decide on whom to vote (Popkin 1991). In reality, president Gerald Ford was not an anti-Mexican leader, in fact he never made any offensive statements towards Mexicans, however the Mexican-Americans did not have enough political information to know this and assumed he was not going to be a good president for Mexican-Americans due to the fact that he could not eat properly a Mexican dish (Popkin 1991). If voters were more informed and did not use information shortcuts in the 1976 election, President Gerald Ford may have not lost the election to Jimmy Carter (Popkin 1991). Therefore, the 1976 presidential election is a good indicator of how a lack of information can skew elections (Popkin 1991). Richard Lau and David Redlawsk analysed the same election using individual s partisanship, policy position and voters views with the information provided in the election campaign. They found out that only 70% of the respondents voted correctly . These 70% would have voted the same even if they were informed (Lau and Redlawsk 1997). Using ANAS data, on average, 75% vote correctly (Niemi, Weisberg and Kimball 2010:81). Therefore, according to Lau and Redlawsk, 25-30% voted incorrectly and this is enough to skew entire elections, although many political scientists used the miracle of aggregation to counter this point.

Revisionist streams in the literature argued that individuals use different means of shortcuts to cast their vote, and that collective decisions can still be a good reflection of the population, even if individual voters are suspect (Popkin 1991 Page and Shapiro 1992 Lupia and McCubbins 1998 Niemi, Weisberg and Kimball 2010). They believe individual errors due to a lack of information will be random, thus cancel out when votes are totalled. This means election outcomes do usually reflect the public mood (Page and Shapiro 1992). They dubbed this the miracle of aggregation . In a nutshell, the principle is that, the bigger the sample the less impact small outliers will have on the overall data. Nevertheless, revisionists also agree with the view that most voters have little interest and knowledge about politics and so information is irrelevant when explaining vote choices. So the theory of miracle of aggregation does not say information plays a role in vote choices, it is just a theory to show why a lack of information is not a democratic problem, thus still agreeing with the position that information has no role in explaining vote choices.

A long line of research shows that there are huge levels of variations in political knowledge among the public. There is a large majority of voters that do not even understand basic political terms such as liberal or conservative (Lewis-Beck et al 2008 Niemi, Weisberg and Kimball 2010). Most voters do not even know the basic information about political leaders, parties, and political institutions (Zaller 1992 Carpini, and Keeter 1996 Converse 2008 Lewis-Beck et al 2008). Therefore, information is non-existent in voting behaviour and voters use other mechanisms to vote, information shortcuts being just one of these mechanisms. The question is, then, whether information shortcuts allow uninformed voters to behave as if they were fully informed and is it a good substitute for information? Although it is difficult to document exactly how many voters use information shortcuts , it is safe to say that information cannot explain vote choices as its non-existent (Niemi, Weisberg and Kimball 2010). Lau and Redlawsk find that almost all voters are uninformed and vote choices are mostly explained by heuristics based on endorsements and partisanship. However, they do also find out that more informed voters do generally produce correct votes, linking back to Shenkman s argument that less informed voters make less sensible vote choices (Lau and Redlawsk 1997). Usually ones ideological orientation is a stronger predictor of vote choices than political information and partisanship is arguably the most common voting heuristic and a stronger vote predictor (Niemi, Weisberg and Kimball 2010). In summery, voters only spend as much time in politics as they need to. Voters are cognitive misers they only need some information - rational ignorance (Downs 1957). Hence why, information has no role in explaining individuals vote choices.

However, recent studies do claim that political campaigns in the last fifty years became more informative as a result public knowledge has been increasing. Kam (2006) argues that political campaigns in recent years have encouraged voters to be more open-minded and reduced pro-incumbent biases among voters (Niemi, Weisberg and Kimball 2010). In fact, Geer (2006) argues that the growing frequency of negative campaign ads in elections is a positive development for informing voters. Therefore, saying information has no role in explaining vote choices may be unfair since many more scholars are arguing in recent years that information does have a role. There are many people in the 21^st century who voted based on the information they obtained on certain issues issue voting. Issue voting is a large part of recent British politics. With the rise of UKIP and its recent four million votes in the last election, we can see many people switching their partisan loyalties elsewhere due to new information and new issues on party manifestos (Strong 1977). It is not just the UK, Lewis-Beck et al, in American Voter Revisited , claimed that the level of education, on average, has increased among the American electorate since the 1950s, which is reflected in the more frequent issue voting and greater overall clarity in the structure of mass issue attitudes (Strong 1977 Lewis-Beck et al 2008). Yes, politics does remain peripheral to the everyday concerns of voters and this will probably be the case forever. Not all voters vote based on information and issues. Nevertheless, American voters are not fools , the American Voter is not applicable to the 21^st century voter and information does play a role in explaining vote choices. It may not be the main contributor but that does not mean it s totally non-existent, hence why V.O. Key s proclaimed that, voters are not fools (Lewis-Beck 2008:425).

Section Three: it s the economy, stupid!

Other recent voting studies, such as those by Gomez and Wilson, say that individual vote choices are often explained by the economy. Based on a theory of attribution of responsibility, they argue that well-informed people vote more on their own financial situation, while less informed voters look at the information on the national economy to cast their vote (Gomez and Wilson 2001). So information can have a role in explaining individual vote choices in connection with the economy, meaning individuals have more interest in information when it is based on economic interests (Gomez and Wilson 2001). In fact, some argue that information does play a role in explaining vote choices, it is simply that there is a focus on what voters don t know and cant do misses much of what they do know and can do (Niemi, Weisberg and Kimball 2010:80).

Voters have core values and they make choices based on these core values. Information depends on your core values and voters choose the information that accepts and proves these (Zaller 1992). There are common core values, often divided between the working class and the middle/upper class. These core values are, economic liberalism and moral conservatism , respectively for the two classes. Goren argues that the working class and the middle class apply these values with respect to economic policies of each candidate (Niemi, Weisberg and Kimball 2010). Regardless of a voter s political knowledge and information level, it is these core values that play a role in vote choices (Niemi, Weisberg and Kimball 2010). To put this in context, voters have core values that lead to partisanship and party loyalties, which information can strengthen, which then leads to vote choices, similar to the Michigan Model of voting with a slightly bigger role given to the economy and information (Niemi, Weisberg and Kimball 2010). This is more of a middle ground, arguing that economic conditions and interests explain vote choices but voters also use information to assess these.

In addition, the emerging ration choice paradigm, which applied to the idea of utility maximisation developed in economics, further highlighted the importance of both economics and information in vote choices (Bartels and Leighley 2010). Anthony Down s, An Economic Theory of Democracy , is arguably the most influential book on this field. Downs says, for a great many citizens in a democracy, rational behaviour excludes any investment whatever in political information per se since their individual choices have almost no chance of influencing the outcome (Downs 1975:245 Bartels and Leighley 2010:19). Popkin called this low-information rationality. Not knowing too much about politics is a cost-cutting benefit meaning it s more rational to be ignorant. Due to this, information plays no role in vote choices instead individuals use retrospective voting to make up their choice (Fiorina 1978).

Retrospective voting is another mechanism voter s use, just like information shortcuts, but more focused on economic evaluation. Voters evaluate the economic conditions and performances, deciding whether the government should be rewarded by getting re-elected or punished by being voted out the theory of reward-punishment (Fiorina 1978). Retrospective voting provided further empirical support for Key s claim that voters, or at least a number of them, are moved by their perceptions and appraisals of policy and performances (Bartels and Leighley 2010:20). Fiorina (1981:5) argues that retrospective voters need not know the precise economic or foreign policies of the incumbent administration in order to see or feel the results of those policies (Achen and Bartels 2013:1). Information shortcuts and retrospective voting often go together because voters use the shortcuts to evaluate the economy (Popkin 1991). For example, voters in the California insurance referendum used the position of their insurance company as an information shortcut to evaluate the economic conditions (Lupia 1998). Many voters also look at candidate images and vote for the one who looks most competent in managing the economy (Evans 2000). Voters therefore make up their vote choices based on the economic conditions, with little knowledge in the form of information shortcuts, meaning information has almost no role in explain individual vote choices unless it is in connection with economic evaluation (Achen and Bartels 2013).

Furthermore, vote choices are explained more by social class and partisanship, than by, information (Bartels 2000). Many countries, like the UK, have a two-party system with the working class voting for the Labour party and the middle class voting Conservatives. Therefore, election outcomes are very easily predicted in the UK, with both parties always gaining over 200 seats. Constituencies such as Hackney, Bow and Islington all are safe Labour seats with strong left-wing voters. Despite how much anti-labour information are provided in these regions, the seats are very safe Labour seats because their votes are based on social class and partisanship over information (Evans 2000 Bartels 2000). The Michigan Model emphasised this fundamental importance of partisan loyalties but did not give it as important of a role as it should have. We previously talked about issue voting and how information can play a role in vote choices, but in 2000, social security privatisation was a huge issue, yet political analyst Gabriel Lenz found that voters did not change their vote on privatisation due to their strong social class and partisan loyalties. Meaning issue voting is illusionary compared to partisanship and class (Bartels 2008).

Finally I will discuss briefly pocketbook voting before concluding. Pocket book voting is a type of economic voting traditionally contrasted with sociotropic voting, which explains how individuals make up their vote choices. Voters are motivated by their own economic self-interest and so their vote choices are based on which candidate or government can put more money into their pocket (Elinder, Jordahl and Poutvaara 2015). In 2013, ScotCen Social Research asked Scottish voters if they would support independence if it meant they were £500 more well off or less well off. The survey saw that 52% of the non-independence voters would vote for independence if it meant they would be £500 more well off, 72% of the pro-independence voters would vote against independence if it meant they would be £500 less well off, showing that people vote rationally using utility maximisation (Currie 2014). There are a few surveys that disprove this point too, for example, in the October 2016 polling papers, majority of the respondents said that their vote choices would not change if another candidate supported school vouchers (Appendix two). It is difficult to assess the validity of pocket book voting because there are so many other factors that contribute to individuals vote choices, however, it is still a very influential theory explaining voting behaviour, which is why I wanted to talk about it briefly.

It is important to note that most voters are actually uninformed about their own and the country s national economy (Aidt 2000 Fern ndez-Albertos and Kuo 2015). Brendon et al studied a sample of US voters in 1996 and found out that only one in eight voters can correctly cite the rate of inflation and unemployment (Aidt 2000). Brendon et al also found out that those with college degrees are more informed about current economic performances but this group is often always a minority. In the 2012 election, more white non-bachelor voters voted Romney over Obama, whilst when it came to bachelor degree voters the vote was much more even (Appendix three). In addition, 65% of all no college degree voters in the 2016 election voted Trump compared to Clintons 29% (Appendix four). This does not mean Trump had a better economic plan but the uninformed majority was not able to assess it correctly and vote sensibility . Therefore, the majority of the voters are uninformed and base their vote choices on the economy. Information can play a role in explaining vote choices but it is only when it is in connection with the economy.

Conclusion

There is no doubt that voters are uninformed. Individuals use other methods and substitutes for information that make up their vote choices. They are not totally ignorant but neither rational. Bartels said, Individuals make candidate choices significantly better than they would by chance but significantly less well than they would with complete information , which might be the best explanation so far (Bartels 1996:24). It is the economy that plays a large role in explaining vote choices. Voters usually vote for the party that can manage the economy. In 1932, voters replaced the republicans with the democrats to improve the economy. In 2010, British voters replaced Labour with Conservatives to improve the economy. Swedish voters replaced conservatives with liberals during the same period. Germans even rejected mainstream parties and voted for the Nazi party during the 1930s-1940s so the economy improved after world war one. So in every year, the voters punished the party in power where a depression happened and voted accordingly (Bartels 2008). Voters do use information to evaluate the economy but it is information in connection to the economy, often in forms of shortcuts. In conclusion, information cannot explain individuals vote choices at most it is just a contributor that works alongside economic evaluation. Retrospective voting and partisanship is a better explanation for vote choices than information.

Bibliography

- Achen, C. and Bartels, L. (2013). Blind Retrospection: Why Shark Attacks Are Bad For Democracy. 1st edition. Vanderbilt University Press.

- Aidt, T. (2000). Economic voting and information. Electoral Studies, [online journal article] 19(2-3), pp.349-362. Available at: https://doi.org/10.1016/S0261-3794(99)00056-6 [Accessed: 2 December 2017].

- Bartels, L. (1996). Uninformed Votes: Information Effects in Presidential Elections. American Journal of Political Science, [online journal article] 40(1), pp.194-230. Available at: http://www.jstor.org/stable/2111700 [Accessed: 29 November 2017].

- Bartels, L. (2000). Partisanship and Voting Behaviour, 1952-1996. American Journal of Political Science, [online journal article] 44(1), pp.35-50. Available at: http://www.jstor.org/stable/2669291 [Accessed: 2 December 2017].

- Bartels, L. (2008). The Irrational Electorate. The Wilson Quarterly, [online journal article] volume 32(4), pp. 44-50. Available at: http://www.jstor.org/stable/40262068 [Accessed: 28th November 2017].

- Bartels, L. and Leighley, J. (2010). The Oxford Handbook of American Elections and Political Behaviour. 1st edition. Oxford University Press.

- Carpini, D. X, M. and Keeter, S. (1996). What Americans know about politics and why it matters? 1st edition. Yale University Press.

- Converse, P. (2008). The nature of belief systems in mass publics (1964). Critical Review: A Journal of Politics and Society, [online journal article] 18(1-3), pp.1-74. Available at: https://doi.org/10.1080/08913810608443650 [Accessed: 1 December 2017].

- Currie, M. (2014). Is £500 the referendum deal breaker? [online webpage] BBC News. Available at: http://www.bbc.co.uk/news/uk-scotland-scotland-politics-25846914 [Accessed: 5 December 2017].

- Downs, A. (1957). An Economic Theory of Political Action in a Democracy. Journal of Political Economy, [online journal article] 65(2), pp.135-150. Available at: http://www.jstor.org/stable/1827369 [Accessed: 3 December 2017].

- Elinder, M. Jordahl, H. and Poutvaara, P. (2015). Promises, policies and pocketbook voting. European Economic Review, [online journal article] 75(1), pp.177-194. Available at: https://doi.org/10.1016/j.euroecorev.2015.01.010 [Accessed: 5 December 2017].

- Evans, G. (2000). The Continued significance of class voting. Annual Review of Political Science, [online journal article] 3(1), pp.401-417. Available at: https://doi-org.libproxy.york.ac.uk/10.1146/annurev.polisci.3.1.401 [Accessed: 2 December 2017].

- Fern ndez-Albertos, J. and Kuo, A. (2015). Income Perception, Information, and Progressive Taxation: Evidence from a Survey Experiment. The European Political Science Association, Political Science Research and Methods, [online journal article] pp.1-28. Available at: http://journals.cambridge.org/abstract_S2049847015000734 [Accessed: 29 November 2017].

- Fiorina, M. (1978). Economic Retrospective Voting in American National Elections: A Microanalysis. American Journal of Political Science, [online journal article] 22(2), pp.426-443. Available at: http://www.jstor.org/stable/2110623 [Accessed: 28 November 2017].

- Gomez, B. and Wilson, M. (2001). Political sophistication and economic voting in the American electorate: A theory of heterogeneous attribution. American Journal of Political Science, [online journal article] volume 45(4), pp. 899-914. Available at: http://www.jstor.org/stable/2669331 [Accessed: 28th November 2017].

- Jakobsen, T. (2013). Research on Electoral Behaviour A History. [online webpage] Popular Social Science. Available at: http://www.popularsocialscience.com/2013/01/07/research-on-electoral-behavior-a-history/ [Accessed: 30 November 2017].

- Lau, R. and Redlawsk, D. (1997). Voting Correctly. The American Political Science Review, [online journal article] 91(3), pp.585-598. Available at: http://www.jstor.org/stable/2952076 [Accessed: 29 November 2017].

- Lewis-Beck, M., Weisberg, H. and Jacoby, W. (2008). American Voter Revisited. 1st edition. University of Michigan Press.

- Lupia, A. and McCubbins, M. (1998). The democratic dilemma: can citizens learn what they need to know? 1st edition. Cambridge University Press.

- Niemi, R. Weisberg, H. and Kimball, D. (2010). Controversies in Voting Behaviour. 5th edition. CQ Press.

- Page, B. and Shapiro, R. (1992). The Rational Public: Fifty Years of Trends in Americans` Policy Preferences (American Politics and Political Economy Series) (American Politics Political Economy). 2nd edition. University of Chicago Press.

- Popkin, S. (1991). The Reasoning Voter: Communication and Persuasion in Presidential Campaigns. 2nd edition. University of Chicago Press.

- Przeworski, A. Stokes, S. and Manin, B. (1999). Democracy, Accountability and Representation. 3rd edition. Cambridge University Press.

- Shenkman, R. (2009). Just How Stupid Are We? Facing the Truth About the American Voter. 2nd edition. New York: Basic Books.

- Strong, D. (1977). Issue voting and party realignment. 1st edition. University of Alabama Press.

- Thomassen, J. (2005). The European Voter: A Comparative Study Of Modern Democracies. Oxford University Press Online, [online journal article] pp.108-124. Available at: http://www.oxfordscholarship.com.libproxy.york.ac.uk/view/10.1093/0199273219.001.0001/acprof-9780199273218 [Accessed: 2 December 2017].

- Zaller, J. (1992). The Nature and Origins of Mass Opinion. 5th edition. Cambridge University Press.