ArticlePDF Available

Forecasts of the 2012 US Presidential Election based on Candidates? Perceived Competence in Handling the Most Important Issue

Authors:

Abstract and Figures

The Big-Issue Model predicts election outcomes based on voters’ perceptions of candidates’ ability to handle the most important issue. The model provided accurate forecasts of the 2012 U.S. presidential election. The results demonstrate the usefulness of the model in situations where one issue clearly dominates the campaign, such as the state of the economy in the 2012 election. In addition, the model is particularly valuable if economic fundamentals disagree, a situation in which forecasts from traditional political economy models suggest high uncertainty. The model provides immediate feedback to political candidates and parties on the success of their campaign and can advise them on which issues to assign the highest priority.
Content may be subject to copyright.
For Review Only
Forecasts of the 2012 U.S. presidential election based on
candidates’ perceived competence in handling the most
important issue
Journal:
Political Science Research and Methods
Manuscript ID:
PSRM-RN-2012-0026.R3
Manuscript Type:
Research Note
Date Submitted by the Author:
n/a
Complete List of Authors:
Graefe, Andreas; LMU Munich, Center for Advanced Studies; LMU Munich,
Department of Communication Science and Media Research
Armstrong, J. Scott; University of Pennsylvania, Wharton School
Keywords:
Elections and Campaigns, Linear Regression < Quantitiative Methods,
Voting Behavior
Cambridge University Press
Political Science Research and Methods
For Review Only
1
Forecasts of the 2012 U.S. presidential election based on
candidates’ perceived competence in handling the most important issue
Dr. Andreas Graefe
Department of Communication Science and Media Research
LMU Munich, Germany
Email: a.graefe@lmu.de
Dr. J. Scott Armstrong
Wharton School
University of Pennsylvania, Philadelphia, USA
Email: armstrong@wharton.upenn.edu
Abstract. The Big-Issue Model predicts election outcomes based on voters’ perceptions of
candidates’ ability to handle the most important issue. The model provided accurate forecasts
of the 2012 U.S. presidential election. The results demonstrate the usefulness of the model in
situations where one issue clearly dominates the campaign, such as the state of the economy
in the 2012 election. In addition, the model is particularly valuable if economic fundamentals
disagree, a situation in which forecasts from traditional political economy models suggest
high uncertainty. The model provides immediate feedback to political candidates and parties
on the success of their campaign and can advise them on which issues to assign the highest
priority.
Keywords. Election forecasting, model accuracy, economic perceptions, issue-handling
competence, political economy models, Iowa Electronic Markets, IEM, prediction markets,
FiveThirtyEight, PollyVote
Acknowledgments. Jennifer Kwok did editorial work.
Page 1 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
2
1. Introduction
The state of the economy is commonly considered a determining factor for the
outcome of U.S. presidential elections. Voters reward incumbent party candidates when the
economy is doing well and punish them otherwise (Tufte 1978). Since the late 1970s,
economists and political scientists have built upon this idea and developed so-called political
economy models for forecasting election results. These models use one or more economic
variables, usually along with political variables. There is general agreement on the integration
of political variables, such as the incumbent’s popularity and the time he or his party has
controlled the White House. The main difference among models lies in the selection of
economic variables. Some authors rely on economic growth (measured in terms of GDP or
GNP), others use perceptions of personal income (either retrospective or prospective), and
still others measure job growth (see Holbrook 2010). Since economic indicators often
correlate with one another, the selection of specific economic variables does not seem crucial.
As a result, there has mostly been consensus among the different models’ forecasts in
historical elections.
When economic indicators disagree, however, uncertainty increases, as was the case
at the outset of the 2012 U.S. presidential election. While income growth was at a historical
low, leading economic indicators and perceived business conditions were on par with
historical averages (Erikson and Wlezien 2012). In such a situation, forecasts from different
political economy models diverge depending on the selection of the economic variables.
Table 1 shows forecasts from fourteen models that were published in the October 2012 issue
of PS: Political Science & Politics, plus Fair’s model (2009). Fair’s model is the
“grandfather” of all presidential election forecasting models grounded on “fundamentals”. All
but one of these models use at least one economic variable among a set of other predictors in
order to forecast the incumbent’s two-party popular vote share several months prior to
Election Day.
1
Based on these fifteen forecasts, any outcome seemed possible in the 2012
election. Eight models predicted Obama to win and seven models had him losing to Romney.
The forecasts ranged from an almost eight-point victory for Obama to a six-point victory for
Romney. On average, the models predicted a virtual tie in the popular vote, with a 0.2
percentage point lead for Romney. In reality, Obama won the election with a lead of four
points and achieved 52.0% of the two-party vote.
The results suggest that one should be cautious with forecasts from political economy
models if economic indicators disagree. In such situations, the models’ implicit assumption
1
The exception is the model by Norpoth and Bednarczuk (2012), which used three non-economic predictor
variables: the candidates’ performance in the primaries, the historical cycle of presidential election outcomes, and
an adjustment for partisanship.
Page 2 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
3
that voters can accurately observe the state of the economy and infer how these changes affect
their future well being may not hold. In addition, political economy models are subject to
other limitations, which arise from their focus on economic fundamentals as predictor
variables. First, the use of such variables makes political economy models retrospective in
nature. Most models lack prospective components and evaluations of the out-party, which
makes them particularly less useful in open-seat elections.
2
Second, economic indicators can
be affected by large measurement errors. Initial estimates of economic figures that are used to
calculate the forecasts a few months prior to the election often differ widely from the revised
figures that the forecasters use to calibrate their models. Third, the electorate may perceive
the same economic figure differently, depending on the electoral context. For example,
incumbents may strongly benefit from a GDP growth rate of three percent after a recession.
However, the same growth rate may help them less in a booming economy. Fourth, due to the
focus on economic variables, most models ignore non-economic issues.
3
This is problematic
because the economy is not the only issue of concern to voters. Depending on the electoral
context, many other issues influence the voting decision. Oftentimes, non-economic issues
may be even more important than economic ones. Fifth, political economy models provide
little decision aid to those involved in political campaigns. See Graefe (2013) for a detailed
discussion of these issues.
The Big-Issue Model (Graefe and Armstrong 2012) was developed in an effort to
address some of these limitations. The model relies on take-the-best, a simple heuristic for
choosing between alternatives based on a single piece of information. Take-the-best is
particularly valuable in situations in which one variable is clearly more important than all
other variables (Gigerenzer and Goldstein 1996).
The Big-Issue Model differs from traditional political economy models in that it does
not use economic fundamentals as predictor variables. The model also does not solely focus
on economic issues. Instead, the model uses a single predictor variable that measures people’s
perceptions of which candidate will better handle the most important issue facing the country.
Through this variable, the model captures retrospective and prospective candidate
evaluations, considers economic and non-economic issues depending on their importance in a
particular election, and avoids measurement errors and misperceptions of economic
2
Exceptions include the model by Norpoth and Bednarczuk (2012), which captures candidate evaluations through
primary performance, and by Lockerbie (2012), who uses a question from the Index of Consumer Sentiment that
asks people whether they think they will be better off financially, worse off, or about the same, in a year from now.
While the latter is clearly a prospective measure, it does not link perceived economic conditions to the
responsibility of the government. Finally, the models by Campbell (2012b) and Erikson and Wlezien (2012) are to
some extent prospective, since they include information from trial-heat polls.
3
The exception is Hibbs (2012), who uses U.S. military fatalities in foreign conflicts as a second predictor besides
growth of per capita real disposable income.
Page 3 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
4
fundamentals. In addition, the model can provide quick advice on which issues candidates
should assign the highest priorities on their campaign agenda.
The Big-Issue Model builds on Downs’ (1957) classic model by assuming that voters
are most interested in delivery (i.e., who can do the job). In addition, the Big-Issue model
builds on the work of political psychologists who found that people use simple heuristics and
cognitive shortcuts when deciding for whom to vote (Redlawsk 2004). With the focus on
issue evaluations, the Big-Issue Model relies on one of the three variables of the valence
politics model of individual voting behavior. Valence politics assumes that a person’s vote
choice is determined by partisanship, party leader images, and the candidates’ (or parties’)
issue-handling competence (Clarke, Sanders, Stewart and Whiteley 2009). An in-sample
comparison of rival models of electoral choice showed that the valence politics model did
best in explaining individual vote choice in the 2008 U.S. presidential election. The model
correctly “predicted” vote choice for 97% of respondents. However, a simple model based
solely on the parties’ perceived issue-handling competence also performed well and yielded
88% correct predictions (Clarke, Kornberg, Scotto, Reifler, Sanders, Stewart and Whiteley
2011).
This research note reviews the performance of the Big-Issue Model in forecasting the
2012 U.S. presidential election. In general, the model can be expected to provide accurate
forecasts if there is one issue that is regarded as clearly more important than others (Graefe
and Armstrong 2012; Gigerenzer and Goldstein 1996). Thus, the political environment of the
2012 election favored the Big Issue Model; the economy was the dominating issue during the
whole campaign.
2. Materials and method
Starting early 2011, we collected polls from pollingreport.com that asked people
which candidate they expected to do the best job in handling the issue that was seen as most
important at that time (i.e., the economy).
4
For example, “Regardless of who you support,
which (2012 presidential election) candidate do you trust to do a better job handling the
economy—(Barack) Obama or (Mitt) Romney?” (ABC News/Washington Post Poll,
November 1-4, 2012). By Election Eve on November 5
th
of 2012, we had obtained a total of
61 polls. For each poll, we calculated the two-party support for Obama. This Big-Issue score
4
Monthly Gallup surveys that asked voters in open-ended questions to name the most important problem facing
the country regularly found that economic issues were mentioned by 65% to 72% of respondents. In comparison,
non-economic issues were mentioned by 38% to 46%. Source: http://www.gallup.com/poll/1675/most-important-
problem.aspx
Page 4 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
5
(S) represents the single variable in the vote equation of the Big-Issue Model: V= 27.0 + 0.50
* S, where V is the two-party popular vote share of the incumbent party’s candidate.
This vote equation was derived by simple linear regression of V on historical Election
Eve values of S, using data from the ten prior elections from 1972 to 2008.
5
For details on the
exact procedure see Graefe and Armstrong (2012).
6
The first forecast of the Big-Issue Model
was published at pollyvote.com on January 10, 2011, almost two years before the election.
The forecasts were continuously updated whenever new polls were released. The complete
data and calculations are available in the PSRM Dataverse.
7
3. Results
Since its first release, the Big-Issue Model predicted Obama to win the popular two-
party vote, except for a short period in late August when Romney was expected to gain 50.2%
(vs. Obama at 49.8%).
8
As shown in Figure 1, since late May 2012, more than half a year
before the election, the forecast was within a narrow range of about two percentage points,
with a maximum of 52.2% for Obama in mid-July.
9
The final forecast calculated on Election
Eve predicted Obama to gain 51.4% and thus, missed the final election outcome by 0.6
percentage points.
Figure 1 also shows the daily forecasts of three benchmark approaches: (1) the Iowa
Electronic Markets (IEM) vote-share prediction market, (2) Nate Silver’s New York Times
blog FiveThirtyEight.com, and (3) PollyVote.com, which combines forecasts from different
methods that use different data (Graefe, Armstrong, Jones and Cuzán 2014). The forecasts
from the Big-Issue Model, FiveThirtyEight, and the PollyVote were stable and relatively
close to each other for most of the forecast horizon. In comparison, the IEM forecasts were
much more volatile from about mid-July to Election Day and tended to predict a large victory
for Obama until two weeks prior to Election Day.
5
One reviewer made an interesting and well-founded suggestion to update the vote equation after each day. We
tested this approach for the last 100 days prior to the election and found that it would have reduced the error of the
current model by 6%.
6
Instead of the three-day poll average in the original model specification, the 2012 Big-Issue Model used
exponential smoothing to aggregate polls over time. The smoothing factor for weighting the most recent poll was
0.3, which is roughly equivalent to a six-day moving average.
7
http://dx.doi.org/10.7910/DVN/22949.edu/dvn/dv/PSRM
8
A possible explanation for this result is that Romney benefitted from the Republican convention, which took
place from August 27 to 30. Candidates usually benefit from their conventions, since these events tend to unite the
party, create favorable media coverage, and thus increase people’s enthusiasm for their party and candidate
(Campbell, Cherry, Wink1992). As a result, polls might have difficulties to accurately measure issue perceptions
around conventions, since people might be strongly influenced by party identification and candidate evaluations.
The forecasts seem to reflect this, as Figure 1 reveals a convention bump for Romney.
9
Figure 1 covers forecasts from May 31, 2012 to coincide with Nate Silver’s first published forecast.
Page 5 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
6
Figure 1: Daily forecasts of the Big-Issue Model and benchmarks
(from May 31
st
to Election Eve 2012)
So how would one have done by simply relying on one of the four methods for each
day of the five-month period? Figure 2 answers this question. At each point in time, Figure 2
shows each method’s mean absolute error (MAE) for the remaining days in the forecast
horizon. That is, each data point in the chart shows the average error that one would have
achieved when relying on the method for the remaining days prior to Election Day. For
example, across the last 50 days prior to the election (i.e., from September 17
th
to
Election
Eve), the MAE of the Big-Issue Model was 0.35 percentage points, compared to 0.48 for the
PollyVote, 0.76 for FiveThirtyEight, and 1.60 for the IEM. That is, the Big-Issue Model
reduced the errors of the PollyVote, FiveThirtyEight, and the IEM by 27%, 54%, and 78%,
respectively. However, the advantage of the Big-Issue Model fades as the forecast horizon
increases. Over the full 159-day period from May 31
st
to Election Eve, the respective errors
were 0.79 for the Big-Issue Model, 0.35 for the PollyVote, 0.72 for FiveThirtyEight, and 1.11
for the IEM. That is, across the whole forecast horizon, the error of the Big-Issue Model was
29% below the respective error of the IEM, but 10% larger than the error of FiveThirtyEight,
and more than twice the error of the PollyVote.
10
10
FiveThirtyEight is the only approach that predicts the election outcome within each state. While this would
generally also be possible with the Big-Issue Model, the model is currently limited to predictions of the national
49.50
50.50
51.50
52.50
53.50
54.50
55.50
56.50
5/31 6/21 7/12 8/2 8/23 9/13 10/4 10/25
Obama two-party popular vote
PollyVote FiveThirtyEight IEM Big-Issue Model
Page 6 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
7
A possible reason for the lower accuracy of the Big-Issue Model when making long-
term forecasts is that issues play a less important role early in the campaign. Prior research
has shown that party identification and candidate evaluations are at least as important as issue
evaluations when making long-term forecasts. However, at around forty days prior to Election
Day, which is about the time of the presidential debates, the importance of issues as a
predictor of election outcomes increases sharply (Graefe 2013). This is also about the time
when the forecasts of the Big-Issue Model became highly accurate and stable (cf. Figure 1).
Figure 2: Mean absolute errors of daily forecasts of the Big-Issue Model and three
benchmarks across the remaining days of the forecast horizon.
Finally, one might want to compare the accuracy of the Big-Issue Model to the one-
shot forecasts of political economy models. Table 1 reports the forecasts of the Big-Issue
Model from the same day each of the fifteen political economy models was published. The
Big-Issue Model provided more accurate forecasts than nine models and was less accurate
than six models. On average, the Big-Issue Model’s forecast error was 1.2 percentage points
lower than the error of the typical political economy model, which corresponds to an error
reduction of 52%. That is, if one had relied on the forecasts from the Big-Issue Model instead
of the forecast of one of the political economy models, one would have reduced the error by
more than half. In addition, the Big-Issue Model’s error was 0.8 percentage points below the
error of the combined forecast from the fifteen models, calculated 57 days prior to Election
popular vote. The extension of the Big-Issue Model to the state level is up to future research and is expected to
improve accuracy given previous tests on the bottom-up approach (Armstrong 2006).
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
5/31 6/21 7/12 8/2 8/23 9/13 10/4 10/25
MAE
PollyVote FiveThirtyEight IEM (original) BigIssue
Page 7 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
8
Day. This corresponds to an error reduction of 60%. It is interesting to note that, except for
the model by Holbrook (2012), models that use both economic indicators and polls as
predictor variables were among the most accurate. This result conforms to prior research,
which showed that adding public opinion variables to models that are solely based on
structural variables increases forecast accuracy (Erikson, Bafumi and Wilson 2001).
Table 1: Two-party popular vote forecasts of political economy models
and the Big-Issue Model for the 2012 election
Forecaster Name of model
Days
prior to
Election
Day
Predicted
two-party
popula r
vote for
Obama
Big
issue
model
forecast
at same
day
Error
reduction
due to
Big-Issue
Model
Jerome & Jerome-Speziari State level political economy model 142 51.6 50.6 -1.0
Lockerbie Expectations model 130 53.8 51.1 1.0
Klarner State level presidential forecast model 114 51.3 52.2 0.5
Berry & Bickers State level economic model 111 47.1 51.4 4.3
Fair Economic voting model 102 49.5 50.8 1.3
Hibbs Bread and peace model 102 47.5 50.8 3.3
Erikson & Wlezien Leading economic indicators and the polls* 99 52.6 50.8 -0.5
Cuzan Fiscal model I 97 46.9 50.8 3.9
Cuzan Fiscal model II 97 45.5 50.8 5.3
Abramowitz Time for change model* 69 50.6 49.8 -0.8
Lewis-Beck & Tien Jobs model* 69 48.2 49.8 1.6
Lewis-Beck & Tien Proxy model 69 52.7 49.8 -1.4
Holbrook National conditions and incumbency* 67 47.9 49.8 1.9
Campbell Trial-heat model* 57 52.0 50.7 -1.2
Campbell Convention bump model* 57 51.3 50.7 -0.6
Average of all 15 model forecasts 57 49.9 50.7 0.8
* Models that us e economic indicato rs and polls as predictor variables.
- Forecasts ordered by days prior to Elec tion Day.
- Source for politica l economy models is C ampbell (2012a), except for the model by Fair, which was derived f rom
fairmodel.econ.yale .edu. The forecast by Klarner was incorrect in Campbell’s original table and thus corrected above.
- The error reduc tion is the difference betwee n the absolute error o f the political econom y model and the absol ute e rror o f the Big-
Issue Model forecas t of the same day. Positive values indicate that the Big-Issue Model was m ore accurate, negative values indicate
that the Big-Issue Model was less accurate.
4. Discussion
The 2012 U.S. presidential election was the first real test case for the Big-Issue
Model to provide ex ante forecasts. The model provided more accurate forecasts than both the
majority and the average of fifteen established political economy models. In addition, the
model outperformed the PollyVote, FiveThirtyEight, and the IEM when making daily
updated forecasts during the hot phase of the campaign. The Big-Issue Model was only less
accurate than FiveThirtyEight and the PollyVote when making long-term forecasts.
This performance conforms to a prior analysis that showed that the model also
provided competitive forecasts for historical elections. Ex ante forecasts from the Big-Issue
Model for the three elections from 2000 to 2008 reduced the error of the typical of eight
Page 8 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
9
established political economy models by five percent. In addition, the model outperformed
the IEM in predicting who would win (but was less accurate than the IEM when predicting
vote-shares). This performance was achieved even though the conditions for the 2000 and the
2004 elections were unfavorable to the Big-Issue Model. The reason is that during these
elections, voters’ perceptions of which issue was most important often changed (Graefe and
Armstrong 2012). Not so in 2012. Throughout the 2012 campaign, economic issues overrode
all other issues. As mentioned earlier, such a situation is favorable for the Big-Issue Model, as
the model relies on the take-the-best heuristic (Gigerenzer and Goldstein 1996). In addition,
the results suggest that the Big-Issue Model is particularly valuable in situations in which
economic indicators disagree and thus paint an unclear picture of the state of the economy. In
such situations, the theory that economic fundamentals provide a proxy for people’s
expectations of the state of the economy appears to be of limited value. Forecasts from
traditional political economy models of the 2012 election outcome differed widely and
suggested high uncertainty about the election outcome. Finally, the Big-Issue Model can be
expected to provide valuable forecasts even if the conditions are not ideal. The reason is that
the big-issue score likely captures information that goes beyond the candidates’ competence
in handling the most important issue. In particular, it is long known that issue evaluations are
influenced by other factors such as partisanship and candidate evaluations (Asher 1992). For
example, a large literature argues that voters’ perceptions of the economy are influenced by
partisanship (e.g., Wlezien, Franklin and Twiggs 1997; Evans and Andersen 2006; Evans and
Pickup 2010), although the direction of causality remains a matter of debate (Lewis-Beck,
Nadeau and Elias 2008). The Big-Issue Model does not aim at contributing to this theoretical
debate. Rather, part of the reason why the Big-Issue Model predicts well is that its
explanatory variable serves as a proxy in picking up much information about the electoral
context.
The Big-Issue Model adds valuable information to existing models, as it uses
people’s assessments of which candidate will have the most potential in effectively dealing
with the most important problem facing the country to forecast election results. This is an
advantage of the Big-Issue Model over political economy models, as it can provide rapid and
inexpensive decision-making recommendations to those involved in political campaigns,
especially in terms of which issues to prioritize on campaign agendas. Furthermore, the model
can help candidates and parties acquire quick feedback on the effectiveness of their campaign
strategies.
Page 9 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
10
5. Conclusion
The accuracy problem in forecasting U.S. presidential elections has been solved in
the past decade. For the past three elections, the combined forecast at PollyVote.com has
provided highly accurate forecasts of the election outcome, starting months before Election
Day. On average, the PollyVote’s Election Eve forecasts missed the actual outcome by little
more than half of a percentage point (Graefe, Armstrong, Jones and Cuzán 2013). It seems
unlikely that individual models, such as the Big-Issue Model, can compete with the accuracy
of a combined forecast in the long term. However, the model uses a different method and
different data than established models and thus contributes to the accuracy of a combined
forecast. Finally, one area that has received relatively little attention in election forecasting is
the improvement of forecasts’ ability to aid decision-making of parties, candidates, and
voters. The Big-Issue Model is a step in this direction as it provides decision aid on which
issues to emphasize in the campaign.
References
Armstrong, J. Scott. 2006. "Findings from evidence-based forecasting: Methods for reducing
forecast error." International Journal of Forecasting no. 22 (3):583-598.
Asher, Herbert B. 1992. Presidential Elections and American Politics: Voters, Candidates, and
Campaigns since 1952. 5th ed. Belmont: Brooks/Cole Publishing Company.
Campbell, James E. 2012a. "Forecasting the 2012 American national elections - Editor's
introduction." PS: Political Science & Politics no. 45 (4):610-613.
Campbell, James E. 2012b. "Forecasting the presidential and congressional elections of 2012: The
trial-heat and the seats-in-trouble models." PS: Political Science & Politics no. 45
(4):630-634.
Campbell, James E., Lynna L. Cherry, and Kenneth A. Wink. 1992. "The convention bump."
American Politics Research no. 20 (3):287-307.
Clarke, Harold D., Allan Kornberg, Thomas J. Scotto, Jason Reifler, David Sanders, Marianne C.
Stewart, and Paul F. Whiteley. 2011. "Yes we can! Valence politics and electoral choice
in America, 2008." Electoral Studies no. 30 (3):450-461.
Clarke, Harold D., David Sanders, Marianne C. Stewart, and Paul F. Whiteley. 2009. Performance
Politics and the British Voter. Cambridge: Cambridge University Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Addison Wesley.
Erikson, Robert S., Joseph Bafumi, and Bret Wilson. 2001. "Was the 2000 presidential election
predictable?" PS: Political Science & Politics no. 34 (4):815-819.
Erikson, Robert S., and Christopher Wlezien. 2012. "The objective and subjective economy and
the presidential vote." PS: Political Science & Politics no. 45 (4):620-624.
Evans, Geoffrey, and Mark Pickup. 2010. "Reversing the causal arrow: The political conditioning
of economic perceptions in the 2000–2004 U.S. presidential election cycle." The Journal
of Politics no. 72 (4):1236-1251.
Evans, Geoffrey, and Robert Andersen. 2006. "The political conditioning of economic
perceptions." Journal of Politics no. 68 (1):194-207.
Fair, Ray C. 2009. "Presidential and congressional vote-share equations." American Journal of
Political Science no. 53 (1):55-72.
Gigerenzer, Gerd, and Daniel G. Goldstein. 1996. "Reasoning the fast and frugal way: models of
bounded rationality." Psychological Review no. 103 (4):650-669.
Graefe, Andreas. 2013. "Issue and leader voting in U.S. presidential elections." Electoral Studies,
doi: 10.1016/j.electstud.2013.1004.1003.
Page 10 of 11
Cambridge University Press
Political Science Research and Methods
For Review Only
11
Graefe, Andreas, and J. Scott Armstrong. 2012. "Predicting elections from the most important
issue: A test of the takethebest heuristic." Journal of Behavioral Decision Making no.
25 (1):41-48.
Graefe, Andreas, J. Scott Armstrong, Randall J. Jones, Jr., and Alfred G. Cuzán. 2014.
"Combining forecasts: An application to elections." International Journal of Forecasting
no. 30(1), 43-54.
Graefe, Andreas, J. Scott Armstrong, Randall J. Jones, Jr., and Alfred G. Cuzán. 2013. "Accuracy
of combined forecasts for the 2012 Presidential Elections: The PollyVote." Forthcoming
in PS: Political Science & Politics, Available at: http://ssrn.com/abstract=2246259.
Hibbs, Douglas A. 2012. "Obama's reelection prospects under ‘Bread and Peace’ voting in the
2012 US presidential election." PS: Political Science & Politics no. 45 (4):635-639.
Holbrook, Thomas M. 2010. "Forecasting U.S. presidential elections." In The Oxford Handbook
of American Elections and Political Behavior, edited by Jan E. Leighley, 346-371.
Oxford: Oxford University Press.
Holbrook, Thomas M. 2012. "Incumbency, national conditions, and the 2012 presidential
election." PS: Political Science & Politics no. 45 (4):640-643.
Lewis-Beck, Michael S., Richard Nadeau, and Angelo Elias. 2008. "Economics, Party, and the
Vote: Causality Issues and Panel Data." American Journal of Political Science no. 52
(1):84-95.
Lockerbie, Brad. 2012. "Economic expectations and election outcomes: The Presidency and the
House in 2012." PS: Political Science & Politics no. 45 (4):644-647.
Norpoth, Helmut, and Michael Bednarczuk. 2012. "History and primary: The Obama reelection."
PS: Political Science & Politics no. 45 (4):614-617.
Redlawsk, David P. 2004. "What voters do: Information search during election campaigns."
Political Psychology no. 25 (4):595-610.
Tufte, Edward R. 1978. Political Control of the Economy. Princeton: Princeton University Press.
Wlezien, Christopher, Mark Franklin, and Daniel Twiggs. 1997. "Economic perceptions and vote
choice: Disentangling the endogeneity." Political Behavior no. 19 (1):7-17.
Page 11 of 11
Cambridge University Press
Political Science Research and Methods
Article
Full-text available
Classical theories of political representation claim that high levels of political knowledge and awareness across the electorate are crucial for a well-functioning democracy. In real life, though, the actual amount of political information possessed by voters is small and unevenly distributed. In addition, electoral decision making is complex and often characterized by severe time pressure. This study therefore experimentally tests the theoretical framework of “fast and frugal” voters and the “less-is-more effect” (Gigerenzer and Goldstein in Psychol Rev 103(4):650–669, 1996), which expects that in some instances low levels of political knowledge may actually result in equally good or even better outcomes than in the case of well-informed decisions. It is hence assumed that employing specific cognitive shortcuts can help voters make good and quick decisions, even with a lack of information. We have used a laboratory experiment to create an environment characterized by severe time pressure that allowed for only shallow information search and limited comparisons across alternatives. Subjects were able to base their decisions only on a small number of criteria operationalized by candidates’ positions on current political issues with various levels of perceived salience. Voters appear to simplify decision making by focusing mainly on those policy attitudes they consider important. Consequently, it seems that shifting the decision-making strategy from accuracy to efficiency has no effect on the final outcome.
Article
Full-text available
This article offers an assessment of the current state of US presidential election forecasting models. It pays attention to presidential forecasting models from the last three election cycles. It starts by exploring 'under the hood' and describes the specifics of the most widely known models from the 2004 election. In addition, the predictions made by these models are addressed and the determinants of forecasting accuracy from 1996 to 2004 are identified. Moreover, the article explores the lessons learned from the 2000 campaign and the alternatives to the dominant aggregate-national forecasting models: electronic markets, citizen forecasts, and state-level forecasting models. From a forecasting perspective, the 2008 election outcome was business as usual. Some models were more accurate than others, as is always the case, but the average error was somewhat lower than in the past two elections cycles.
Article
Full-text available
Humans and animals make inferences about the world under limited time and knowledge. In contrast, many models of rational inference treat the mind as a Laplacean Demon, equipped with unlimited time, knowledge, and computational might. Following Herbert Simon's notion of satisficing, this chapter proposes a family of algorithms based on a simple psychological mechanism: one-reason decision making. These fast-and-frugal algorithms violate fundamental tenets of classical rationality: It neither looks up nor integrates all information. By computer simulation, a competition was held between the satisficing "take-the-best" algorithm and various "rational" inference procedures (e.g., multiple regression). The take-the-best algorithm matched or outperformed all competitors in inferential speed and accuracy. This result is an existence proof that cognitive mechanisms capable of successful performance in the real world do not need to satisfy the classical norms of rational inference.
Article
Full-text available
The Incumbency and National Conditions (presidential approval and aggregated personal finances) Model predicted President Obama would garner 47.9% of the two-party vote, whereas he ended up with 51.8% (based on available information on December 3, 2012). The error in this forecast (3.9 points) is somewhat higher than the average out-of-sample error from 1952 to 2008 (2.4 points). Although the forecast was off the mark, the addition of the 2012 result to the data set does little to change the slope estimates, and the overall fit of the model is only slightly worse (table 1).
Data
Full-text available
The Big-Issue Model predicts election outcomes based on voters’ perceptions of candidates’ ability to handle the most important issue. The model provided accurate forecasts of the 2012 U.S. presidential election. The results demonstrate the usefulness of the model in situations where one issue clearly dominates the campaign, such as the state of the economy in the 2012 election. In addition, the model is particularly valuable if economic fundamentals disagree, a situation in which forecasts from traditional political economy models suggest high uncertainty. The model provides immediate feedback to political candidates and parties on the success of their campaign and can advise them on which issues to assign the highest priority.
Article
Full-text available
The importance of the economy in US presidential elections is well established. Voters reward or punish incumbent party candidates based on the state of the economy. The electorate focuses particularly on economic change, not the level of the economy per se, and pays more attention to late-arriving change than earlier change. On these points there is a good amount of scholarly agreement (see e.g., Erikson and Wlezien 1996; Hibbs 1987). There is less agreement, however, on what specific indicators matter to voters. Some scholars rely on income growth, others on GDP growth, and yet others on subjective perceptions (see Abramowitz 2008; Campbell 2008; Holbrook 1996b; also see Campbell and Garand 2000). In our work, we have used the index of leading economic indicators, a composite of ten variables, including the University of Michigan's index of consumer expectations, stock prices, and eight other objective indicators.
Article
What matters most to voters when they choose their leaders? This book suggests that performance politics is at the heart of contemporary democracy, with voters forming judgments about how well competing parties and leaders perform on important issues. Given the high stakes and uncertainty involved, voters rely heavily on partisan cues and party leader images as guides to electoral choice. However, the authors argue that the issue agenda of British politics has changed markedly in recent years. A cluster of concerns about crime, immigration and terrorism now mix with perennial economic and public service issues. Since voters and parties often share the same positions on these issues, political competition focuses on who can do the best job. This book shows that a model emphasizing flexible partisan attachments, party leader images and judgments of party competence on key issues can explain electoral choice in contemporary Britain. © Harold D. Clarke, David Sanders, Marianne C. Stewart and Paul F. Whiteley 2009 and Cambridge University Press, 2009.
Article
Do our models of political behavior bear any resemblance to reality? Forecasting elections is one opportunity to assess whether our models of voting behavior are accurate. Over the past few decades, political scientists have been willing to put themselves out there to forecast elections. Explaining a past event allows us the ability to retrofit our models before we make them available to the broader community. In short, forecasting elections provides us the opportunity to develop humility. The forecasting community has done a reasonable job over the past few elections. Aside from 2000, forecasters have been largely accurate. Even in 2000, the forecasting community can claim a modest victory. The community was right about the popular vote winner; it just happened that the popular vote winner lost the election that counts—the Electoral College.
Article
President Barack Obama is going to defeat Republican challenger Mitt Romney by a comfortable margin. Obama has history on his side as well as the fact that he was unchallenged in the primaries. The PRIMARY MODEL, which formalizes these predictors, forecasts an Obama victory with 53.2 percent of the two-party popular vote. This forecast assures Obama’s re-election with 88-percent certainty. The forecast model relies on primary elections as well as an electoral cycle, using elections as far back as 1912, the first year of presidential primaries. The primary performance of the incumbent-party candidate and that of the opposition-party candidate enter as separate predictors. For elections since 1952, the primary-support measure relies solely on the New Hampshire primary. In the period since then, no other primary beats New Hampshire in predictive power. The Primary-Model forecast was posted January 12, 2012, on the Huffington Post.