Newsmaker

and yes anyone who reads this and takes me up on this will get a free beer if I am wrong…I am that convinced…

Barack Obama will win the US presidential election with 54.5% of the popular vote, because of efforts to overcome voter disenfranchisement, the increase in Democratic party identification, the theory of economic voting, the theory of the directional spatial voting, and because of the polling trends and the trends of the political betting markets. On all important categories that political scientists have identified as predictive of election outcomes, Obama holds an advantage. Above all else, it is the political betting markets that have been found to most accurately predict election outcomes, and with the Iowa market currently trading at Obama receiving 54.5% of the popular vote, it is apparent that all available information is indicating an Obama victory on November 4^th.

It is important to look at who is likely to vote in this election and whom this will benefit. Beyond individual and social reasons people will choose to vote, there are institutional rules that govern voting. Much fuss has been made about the issue of voter disenfranchisement, either legal bars on felons or practical bars such as long voting lines during lunch time. Manza and Uggen concluded in their paper Punishment and Democracy that felon voting laws disproportionately affects minorities. They also noted, “…states with larger proportions of nonwhites in their prison populations were more likely to pass restrictive laws…”[1] They also found that this provided a small advantage for Republicans, since minorities are more likely to be Democrats, looking at data from 1972 to 2000.[2] While this research seems to indicate an advantage for McCain, the consistency of this advantage makes it seem unlikely to truly determine this election’s come.

Far more important than felon voting laws is the Republicans’ attempts to purge voters whose homes have been disclosed in swing states. But Obama’s legal team has effectively nullified this practice.[3] Obama’s team has gone even further by trying to help those voters, who traditionally have found it harder to vote,: such as students, nurses, firemen, and others who work long shifts or are traditionally mobile. Their biggest success has been in Ohio, where early voting allows an entire week of same day registering and voting.[4] These targeted registration campaigns are important because research has shown that they are the most effective way of increasing turnout, even beyond simply liberalizing registration laws.[5]

Beyond who will vote, it is important to evaluate the theories of how people determine their vote. The research on vote choice has found party identification and the state of the economy as the most important factors in determining vote choice. But there are also issues associated with the campaigns and political ideology that could affect the outcome of this election. The theories on vote choice are broken down into the importance of party, issues, and candidates.

According the Michigan Model of vote choice Party identification is the most important factor affecting vote choice, and since the Democratic Party has been enjoying a huge increase in party registration this would indicate an Obama victory. In Miller and Shanks’ paper Multiple –Stage Explanation of Political Preferences the found that when holding all other variables constant Party ID’s coefficient of correlation was still statistically significant at 0.16.[6] This is to say, when all else is accounted for, 16% of the vote is still determined by party ID. The authors explain the effect as such, “It acts as a ‘tie breaker’ because the effects of other political attitudes are balanced, or as the only influence for individuals, who do not have any other basis for choice…”[7] So Party ID is heuristic. It either tells people what their best choice is, when all other information leaves them with tied impressions of the candidates, or it is the only deciding factor, when the voter has no information on the candidates.

More importantly, this same paper found that party ID was the most important factor determining vote choice. Issues were not as relevant. For example, holding all other factors equal government assistance to the disadvantaged had a coefficient of 0.04, homosexuals in the military a coefficient of 0.05, and abortion -0.01. What this indicates is, not that these issues are necessarily important, but that they tell us almost no new information about vote choice when party ID is known. Knowing someone’s party identification is the best way to determine how they will vote. So the rise in Democratic registration should equal an increased support for Obama.

Another way to look at the importance of Party ID in this election is to observe its effects on the campaigns. It has been shown in studies that campaigns do not have a huge effect on the election. Campbell in his paper How Campaigns Matter hypothesized that this is because campaigns are really about convincing members of the candidates party not to jump ship. He explains,  “…reinforcement is one in which  the voter at the outset of the campaign intends to defect from their predispositions (usually partisanship) but change their intention and remain loyal as a result of the campaign’s reinforcement…”[8] This is consisted the Michigan Model which says that your perceptions of the candidates are influenced by your party ID. The classic example from this election was John McCain’s choice of Sarah Palin as his VP choice. Facing large defections from the evangelical base McCain needed Palin to secure these votes. But the fact that this was even a problem for McCain means he is campaigning from a disadvantage.

Although Party ID might be the most significant predictor of vote choice, the increase in the number of Americans not identifying with a political party might indicate the importance of issues on vote choice. The literature on the issue of vote choice identifies economic conditions as the most important variable. Additionally, people tend to vote sociotropicly, looking at business conditions as a whole instead of their own personal condition when they vote.[9] As a result, Nadeau and Lewis-Beck’s found that voters first and foremost attribute the state of the economy to the actions of the president despite that we often have divided government, where the president and the congress are not of the same party.[10] They conclude that when the incumbent president is running for reelections people evaluate the economy retrospectively, looking at how the economy has been However, when no incumbent is running (such as the current election) voters look at the economy prospectively. They explain the effects as follows, “The classic V.O. Key economic voter then compares an in-party to an out-party, rather than an in-candidate to an out-candidate (since both candidates are “out”).”[11] Since the future outlook of the economy is not good, John McCain is disadvantaged, because he is of the in-party.

The research did find that with no incumbent running economic voting is weakened, which would lead to the conclusion that the economic conditions will not automatically give the election to Obama. It seems that with the financial crisis and the potential of recession that there is no possible way that economic voting is weaker simple because no incumbent is running. The Obama campaign’s constant attempts to convince voters that McCain and Bush’s economic policies are the same is further evidence that prospective voting is tied to “in-party” status and this it is important in this election. Given the literature on economic voting and the large degree to which economics affects vote choice, one can say with a high probability that this issue alone should be enough to predict an Obama victory.

The final, variable that the literature of vote choice attributes to vote choice is the candidates. Candidates are personified by their campaigns and their ideology. In Campbell’s paper, he concluded that campaigns have about a 4% effect on the outcome on the election.[12] Clearly campaigns do not have the effect on vote choice that party ID and economics do. Campaigns do attempt to locate the candidates ideologically. The literature on the effect of a candidate’s ideology on voting is summarized in two spatial models,: the proximity and the directional models. The proximity model predicts that two candidates will split the vote by aligning themselves just on either side of the median voter. So McCain goes just to the right of the median voter and gets everyone to the right. The model further predicts that support for candidates should peak where the candidates are located and then taper off on either side, a fact not supported by the data. The proximity model would predict that John McCain should win because he has long been considered a centrist. There is a problem with this model, namely that it is hard to decided where candidates are located on either issues. In response, political scientists came up with the directional model, which says that politicians try to locate themselves as taking sides on the issues and then people vote for the candidate that is on the same side as they on the greatest number of issues. This model would support Obama because there are a greater number of registered Democrats and a larger number of people who think he would be better at fixing the economy. One way to examine the model’s validity is to look at the candidates’ voting records. We will look at the 2006 voting records for the two candidates because both men voted very little in 2007 and voted only when their party most needed their votes, meaning their ideological rating would be more extreme. According to Lewis and Poole, using a formula developed by political scientists, reporting for voteview.com they found that ideologically Obama was the 20^th most liberal senator and McCain was the 2^nd most conservative senator.[13] Now their actual rank does not say much because who knows what they were actually voting on and senators tend to be more moderate than other elected officials in the first place. What is important to note however is that they are not located in the middle in terms of ideology. They seem to be following the logic of the directional model by choosing sides without going to a punishable extreme. This makes sense because the candidates want to offer the voters actual choices on the issues instead of seeming like slightly different shades of gray. It seems that the data supports the directional model of voting which should help Obama win.

Beyond the theory, it is also important to take into account the polls that are predicting the outcome of the election. As of 10/4/08 the polls are predicting a large Obama victory. RealClearPolitics.com (RCP), which averages all of the polls, shows Obama winning nationally by 5.9%.[14] More importantly however, RCP’s electoral vote calculator shows Obama with 264 electoral votes either locked in or learning his way (where leaning means a RCP state average of 5%). That means Obama only needs six more electoral votes to win and RCP shows him up in many of the leftover toss-up states. Furthermore, fivethirtyeight.com which adjusts its polls given historical trends, a method found to be more effective[15], computes probability scenarios and finds that the probability that McCain loses either Ohio and Michigan but still wins the election at 1.51%[16] and McCain’s campaign announced a few days ago that they were giving up on trying to win in Michigan.[17] Simply put, Obama has far more ways of winning the Electoral College than McCain does. But polls are not even the best method for determining who will win. Far more effective are political betting markets.[18] The markets are so accurate because, just like actually stock markets, they represent all available information. This includes all of the above theories mentioned. The markets were accurately able to predict how every state would vote in the 2004 presidential election and predicted every senate race in 2006.[19] The Iowa University market as of 10/5/08 has Obama’s vote share price quote at 0.545 or 54.5% of the vote with a 73.7% chance of an Obama victory.[20] The intrade market as of 10/5/08 closed with a 66.7% chance that Obama will win.[21] Given the history of these political markets accurately predicting the outcome and the unfavorable Electoral College map for McCain, it seems extremely unlikely that he could win.

Beyond the media punditry and the coverage of gaffes and counter gaffes the political science theories all indicate an Obama victory this November. Despite the effective disenfranchisement of voters through purging and other tactics by the GOP in years past, Obama’s legal team has more than a reason to fight this challenge. More importantly, they are engaging in rigorous targeted voter drives, a technique found to be far more effective at increasing turnout than liberalizing voting laws. The increased number of registered Democrats also strongly indicates an Obama advantage. The theory of prospective economic evaluations explains that simple by being a member of the “in-party” John McCain is at a disadvantage. The directional model of voting also favors Obama and his expanded Democratic party. All the polling is trending Obama’s. Most importantly, John McCain’s possible Electoral College maps to victory are becoming more and more impossible. Finally, the extremely predictive political betting markets are all predicting over a 65% chance of Obama winning. While campaigns do fluctuate and change, with only a few weeks to go John McCain will not find favorable news from the electoral theories.

---

[1] Manza, Jeff & Uggen, Cristopher. 2004. Punishment and Democracy: Disenfranchisement of Nonincarcerated Felons in the United States. Perspectives on Politics 2 (3): 293.[2] Ibid: 497

[3] Halloran, Liz. 2008. ACLU Sues Michigan Over Voter Purge Program. U.S. News and World Report. http://www.usnews.com/articles/news/national/2008/09/18/aclu-sues-michigan-over-voter-purge-program-saying-it-hurts-college-students.html. (accessed October 4, 2008).

[4] Majors, Stephen. 2008. Ohio election officials brace for early voting. AP Report. http://ap.google.com/article/ALeqM5gmJpgsrR27lwSUQ24_WSSrU0W-JwD93GUI9G1. (accessed October 4, 2008).

[5] Mitchell & Wlezien. 1995. Article sited in lecture on 9/11/2008.

[6] Miller, Warren & Shanks, Merrill. 1996. Multiple-Stage Explanation of Political Preferences. In Controversies in Voting Behavior, ed. Richard Niemi, 226. Washington, DC: CQ Press.

[7] Ibid: 227.

[8] Campbell, James E. 2008. The American Campaign. College Station: Texas A&M University Press. 197.

[9] Prof Peter Ens Cornell Government Dept

[10] Nadeau, Richard & Lewis-Beck, Michael. 2001. National Economic Voting in U.S. Presidential Elections. In Controversies in Voting Behavior, ed. Richard Niemi, 209. Washington, DC: CQ Press.

[11] Ibid: 208.

[12] Campbell: 191.

[13] Lewis, Jeff & Poole, Keith. 2006. 109^th Senate Rank Ordering. Carnegie – Mellon Unversity. http://voteview.com/sen109.htm (accessed October 4, 2008)

[14] RealClearPolitics. 2008. General Election: McCain vs. Obama. http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html (accessed October 4, 2008)

[15] Prof Peter Ens Cornell Government Dept

[16] Silver, Nate. 2008. Scenario Analysis. FiveThirtyEight.com. http://www.fivethirtyeight.com/. (accessed October 5, 2008)

[17] Chris Cillizza. 2008. McCain Moves out of Michigan: What It Means. Washington Post. http://voices.washingtonpost.com/thefix/2008/10/mccain_moves_out_of_michigan_w.html?nav=rss_blog. (accessed October 6, 2008)

[18] Wolfers, Justin. 2007. Best Bet for Nest President: Prediction Markets. The Wall Street Journal, December 31.

[19] Ibid

[20] Iowa Electronic Markets. 2008. Market Quotes. http://www.biz.uiowa.edu/iem/quotes.cfm. (accessed October 5, 2008)

[21] Intrade. 2008 Realtime Electoral Vote Predictor. http://www.intrade.com/. (accessed October 5, 2008)