crp finished
TRANSCRIPT
Anthony Malky, Catie Davis, Michael Herman, Sarah HuffmanPSCI 207Professor Bos 30 July 2016
Location, Location, Location: The Connection Between National Party Convention
Location and Party Benefits
Literature Review
Many studies have noted that the Democratic Party receives more electoral
benefits from the Democratic National Convention (DNC) than Republicans enjoy from
the Republican National Convention (RNC). One area of research that needs further
study, as will be shown, can be supplemented through research of the following question,
why does the Democratic Party receive a greater bump from its National Convention
than the GOP? Our particular research question seeks to further understand why there is
a difference between parties and hopes to uncover answers through surveys of the local
population. First we will analyze the debate about whether or not the strategic location of
national conventions affect election results and then we will note the variances in results
and methodology regarding convention bumps, media attention, and the demographics of
mobilized voters.
Overall, the literature is unable to agree whether the strategic location of a party’s
National Convention produces electoral benefits. Powell (2004) notes that voters seem to
be relatively fixed regarding vote choice, an outcome of party identification and social
and demographic characteristics, leaving little room to be influenced by campaigns and a
National Convention (Powell 2004, 120). The very fact that conventions are held towards
the end of the nominating season weakens their impact on the electorate, most voters
have already made up their mind by the time the DNC or RNC is held. Furthermore, as
regional differences in public opinion and voting have diminished over time, conventions
have shown to have less of an impact on host states (Powell 2004, 121). Moreover,
Powell discovered that parties received no electoral boost from states where the
convention was held between 1932 and 2000 and that the site selection process is
irrelevant from support within a state hosting the convention (Powell 2004, 121). A
critique of the study is that the conclusions are drawn without further examination:
Powell concludes that National Convention location does not have an impact; however,
the state level effects are all that is observed. In order to properly refute conventional
wisdom of the impact of strategic convention location, multiple units of analysis, such as
demographics and resource allocation, are necessary. Specific demographics, such as low
resource voters, may help to explain the difference in benefits each party receives from
their convention.
On the contrary, Atkinson, et al. found that the strategic location of a National
Convention is significant and creates electoral benefits for the host party. This
discrepancy lies in the units of analysis as Powell examined states as a whole and the
effects of a National Convention on the host state, while Atkinson et al took a closer look
of the effects at the city and county level on specific designated media markets (DMAs).
This study was interested in populations that reside within the DMAs. The latter surmised
there is no reason to expect that the heightened information flow from a convention
within a local area would extend throughout the entire state. Multiple and differing media
markets limit the ability for the effects of a National Convention to spread across an
entire state (Atkinson et al., 2014, 1046). Thus, they argue that campaigns target toss-up
counties or cities as opposed to swing states when strategically choosing their location
(Atkinson et al., 2014, 1046). In accordance with Powell’s theory, convention location
does not have an impact on the host state, but locally, within the host city or county:
location matters.
The effects and benefits parties enjoy from the strategic location of their National
Convention can be a result of the media. Media coverage and the media markets are
important; people within the local area of a convention are overloaded with convention
and election coverage. The frenzy of media attention around a local area and its voters
receive from hosting a convention opens the door for parties to turn voters into supporters
and increases the importance of convention location. Fine provides the historical
narrative for how the haves and the have-nots have debated convention rules to tip the
scales in their favor. However, Atkinson et al. fails to acknowledge the way in which
media coverage coincides with the National Conventions. The literature simply notes the
impact without taking into account the ways in which the media could influence the
results of the study.
Again, the literature agrees on one aspect: the Democratic Party receives more
benefits from convention location than the GOP. While there are local advantages for
strategically choosing the National Convention site for both parties, Democrats enjoy
greater electoral boots. Powell demonstrates that Democrats have the upper hand when it
comes to home-state advantage, meaning Democratic candidates showcased in their
home state receive more of a boost than Republicans in their home state. Of the
nominees showcased at a convention as home-state advantage for Democrats produced a
7.4% vote increase, opposed to just .4% for the GOP (Powell 2004, 127). Therefore, if
the DNC is held in the home state of one of its candidates, the party is likely to enjoy an
increase in support. There is no evidence provided for why the party enjoys such an
advantage. Although Powell rejects the overall impact of conventions, he does
acknowledge that Democrats fared better than Republicans when it comes to an electoral
boost from the state hosting a convention. Democrats experienced a 1.8% increase in
votes within the state the convention was held, Republicans, just .3% (Powell 2004, 127).
The attraction of voters and conversion of the opposition is largely responsible for
the electoral benefits the Democratic Party enjoys as a product of the DNC. Atkinson et
al. finds that in Democratic counties within the convention’s media market, there is a
large increase in support for the Democratic nominee of up to seven percentage points
from being exposed to the DNC (Atkinson et al., 2014, 1050). Also, within the DNC, the
Democratic Party has the ability to persuade or flip voters who are in slightly Republican
areas. Therefore, the location of the Democratic convention is significant as the party can
rally not only staunch supporters, but also voters from the opposing party. Highlighting
even further the exposure effect being stronger to the DNC for local areas exposed or
hosting it; for those not already supporting the Democratic candidate, being exposed to
the convention increased their chance of support by around 6% (Atkinson et al., 2014,
1053). For the GOP, exposure to the RNC actually decreases the chances of support
within the locale. Although Democrats are actually able to gain votes within the area
hosting the DNC, the GOP is not able to do the same with the RNC. The patterns are
consistent within the literature; local exposure to the DNC for people in the host area
moves a broad range of voters towards supporting the Democratic candidates and in turn
creates an electoral boost for the party.
Other scholars have also focused on media coverage and its significant effects on
voter mobilization and interest in a particular election. In chapter two of Rewiring
Politics, Panagopoulos discusses the impact of the campaign bump on the DNC and
RNC. They discovered that there has been a decrease in election coverage in the
traditional media between 1964 and 2004. They also found that convention bumps for the
Democratic Party have varied more in size, meaning that it is dependent on the
convention and candidate while the RNC is less varied. They discover that Democratic
voters care more about the convention, which can be presumed to be a result of the
excitement from the media and the candidates; this has conflicting results with other
studies. Another article, The Impact of Television Market Size on Voter Turnout in
American Elections notes that the size of local television markets affects the amount of
political election coverage, which will in turn affect the voter turnout rate. Oddly, they
found that the size of the television market is indirectly related to the voter turnout rate,
which is unexpected. This means the larger the television market does not necessarily
mean more voters tend to turn out in their area. They test their argument by examining
four election cycles between 1984 and 1990, and looking at aggregate voting data in
nearly every sub-county in the continental United States. The author notes that their
research could be supplemented by further research on variance in rural, urban, and
suburban communities and how this affects voter turnout, which could be included within
our future study and this demographic effect.
Panagopoulos (2007, 19-22) claims that the amount of convention coverage and
viewership for both conventions in traditional media has declined. However, the 24-hour
news cycle has devised a culture where people have greater access to coverage than ever
before due to innovations in media. The rise of social media content surrounding politics
could also attribute to the decline in viewership and coverage. In terms of location, the
lack of traditional media coverage could alter how people are excited by the election
within their community. While Panagopoulos fails to recognize specifically who is being
affected by the convention bump in each party, another study, Battleground States versus
Blackout States: the Behavioral Implications of Modern Presidential Campaigns focuses
on a campaign’s ability to mobilize low resource voters such as minorities, the lesser
educated, or lower income citizens through an analysis of American National Election
Studies for the 2000 and 2004 elections. They examined the effects of political
involvement in contested states where both political parties directed a large number of
their resources. This study discovered that political interest levels for low-income voters
in contested states did increase. However, in some instances, likely due to lower levels of
party contact resource voters received less party contact with 13% higher contact for
battleground states in 2000 and 36% in 2004 (Gimpel et. al. 2007,791). Finally, the study
did find a slightly increased likelihood of political involvement in contested areas for
low-income voters in its 2004 study, while controlling for contact levels. This study will
contribute to our question when we seek to understand whom each national convention
appeals to in contact levels, and how this might affect the interest levels and involvement
of the local population. Interestingly, this particular study found that contested states do
not increase voter involvement for the affluent, “but the interaction of battleground status
and low-income, notably enhances the engagement of the have-nots” (Gimpel et. al.
2007, 792). Another way people are engaging with their political environment is through
the use of mass media. Fine (2007, 166) claims that mass media allow citizens to interact
with elites during political events where traditionally they are closed to the public.
Further research could examine why Democrats received more electoral benefits
from the strategic location of their National Conventions as well as the factors causing
the difference in boosts for both parties regarding convention sites. Additionally, the
literature displays the impact the media has on influencing voters and the ability for
parties to mobilize support; therefore, the media’s role should be further investigated in
order to best understand the electoral benefits. It is possible that the effects of a National
Convention on a local area are the products of the media and its work, as opposed to the
conventions themselves or the respective parties or campaigns. Also, further investigation
is necessary in determining whether or not mass media integrates the three merits of
democratic theory: accessibility of leadership, voter goals, and participation (Fine 167,
2007).
Our future study will provide an updated understanding of these factors, as many
of these studies are not recent enough to keep up with technological advancements. We
also hope to synthesize the various elements of these studies to understand the disparity
in effectiveness between the DNC and the RNC, applying the findings from Battleground
states versus blackout states: The behavioral implications of modern presidential
campaigns more specifically to conventions. The study analyzed political interest and
involvement of lower resource voters geographically, but did not compare data between
party lines. If we were to study and determine if there is a major gap between parties, it
appears likely the differing demographic mobilization within lower resource voters could
explain the greater variance in the DNC bumps and the greater effectiveness of their
campaigns found in studies by Powell et al. and Panagopoulos.
Methodology
The effectiveness of the Democratic National Convention (DNC) and the
Republican National Convention (RNC) in mobilizing voter turnout regionally is largely
debated. Our question seeks to answer why does the Democratic Party receive a greater
bump from its National Convention than the GOP? Based off of previous scholarship
which has found that political parties are more likely to contact more educated, wealthier,
and older individuals (Gershtenson 2003), we hypothesize the following: The
Democratic Party and the Democratic National Convention targets more lower resource
voters than the Republican Party and Republican National Convention. Thus, we should
see more of a bump generated by the DNC than the RNC from reaching out to this often-
untapped group. By targeting we mean reaching out to this demographic; through
mailers, face to face contact, phone calls, email, or any way in which the party/candidates
tries to get into contact or direct their message to low resource voters. We will be
surveying voters within each convention city, Cleveland and Philadelphia, hosting the
RNC and DNC respectively, in order to collect our data. The goal of our research is to
determine why the Democratic Party receives a greater bump from its National
Convention than the GOP. More specifically, our study hopes to reveal that the
Democratic Party and DNC contacts low resource voters at a higher rate than the RNC.
The increased contact of low resource voters in contested areas should increase their
political engagement and excitability (Gimpel et al. 2007), leading them to support the
Democratic Party nominee and providing a greater bump. Conducting survey research is
the best choice of methodology in order to compare differences in demographic contact
and mobilization between the two party conventions.
The idea of the “bump” following a National Party Convention is essential to our
study. Panagopoulos (2007) defines the “bump” that tends to follow each convention as
the increase in support that a candidate typically receives at the conclusion of the party’s
convention. Holbrook supports the claim that voters who are more exposed to campaign
rhetoric on behalf of the candidate are more likely to support the candidate. For our
research, we will define low resource voters as, “individuals who have a high cost entry
to voting due to social, political, or economic inhibitors and thus are less likely to vote in
the general election.” Matsusaka (2001, 415) claims that in order to become a habitual
voter an individual has to overcome the high cost of entry into political participation.
These individuals are often from low income and younger voters who are disenfranchised
by the democratic process. Plutzer (2002, 42) claims that, “As young citizens confront
their first election: all the costs of voting are magnified.” We hope to find that low
resource voters who reside in the metropolitan area of a national convention are more
likely to become involved in the democratic process. For the purpose of our study,
demographic factors that we will observe for low resource voters include younger, lesser-
educated, low-income, and racial minority citizens. If an individual is below thirty, did
not receive a college degree, has an income of below $25,000, or if they are another
ethnicity besides white then they have factors towards low resource voting. Each of the
aspects of “low resource” will be operationalized through questions in our survey: we
will ask voters questions about their age, education, income, and race through clearly
demarcated groups. For the purposes of this survey, if an individual has two or more
“low resource” traits, then they will be defined as a low resource voter. This measure is
valid in that the variables we will control, directly measure “low resource” voters as we
define them.
This will create an observational, mixed study, because we are not directly
manipulating the environment or changing the independent variables across participants.
Due to our choice of survey methodology, voter’s responses to our questions will
measured both qualitatively and quantitatively. Our independent variable will be low
resource voter contact while political engagement is our dependent variable. In low
resource voter contact, we are talking about contact from the campaigns or parties
themselves. Both campaigns and parties contact voters through face-to-face contact,
emails, phone calls, social media and print. Our aim is to measure this contact by asking
voters questions like, “Has your political party contacted you within the last six months?”
We define political engagement as, “The interactions of citizens with their society and
their government.” For the purpose of this study we will be operationally measuring both
political contact and political engagement levels by asking questions regarding electoral
indicators and indicators of political voice (Hill and Leighley 1996). These types of
questions will increase the validity of this measure by illustrating the political contact and
political engagement of the respondents. However, the reliability of measuring political
engagement may come into question when considering the variation in voters’ self
reported responses.
Our cases for this study will be the two National Party conventions: the DNC in
Philadelphia, Pennsylvania and the RNC in Cleveland, Ohio. Within each of the two
cases we will be conducting a random sample of voters located in each convention area
for the duration of the event. Voters for our purpose are those of voting age, eighteen and
older, and those who are registered or may register. We intend for our sample to be
representative of the voting population who reside within the local media market of the
convention site. Our sample will be determined by the amount of people who agree to
respond to our survey. Our analytic technique will utilize both Quick Response (QR)
codes and paper surveys. We will distribute both to the public within the area of each
convention site. QR codes are useful in that they are accessible to a wide range of people
with smartphones who are not in the immediate convention location. However, the use of
QR codes may be difficult for some low resource voters who may not have access to a
smartphone and thus will be more likely to respond to the paper surveys.
We intend to handout QR codes in high traffic areas around the convention site.
People will then be able to take a picture on their smartphone, which would then direct
them to our survey. We will also conduct surveys in person in close proximity to the
convention by developing an unbiased system: we will ask every third person we see in
public. Paper surveys may be more accessible to lower resource voters, but it will limit
our ability to interact with a wider range of participants because we will be confined to
the immediate convention location area. In order to remove bias and to conduct our
survey in the most random way possible, for paper surveys we will approach every third
person on the street and ask them to complete the survey. In this case, we will survey
every third person at the establishment. We have no control regarding those who respond
via the QR codes that we will post, therefore this removes a majority of the bias involved
with the technique. Ultimately, paper surveys, which we will conduct in person will be
our primary method of data collection.
Within our survey we will embed a “pre-test”. The pretest will be conducted by
asking questions at the beginning of the survey that will allow us to observe pre-
convention levels of each measure in order to compare the results to the responses
regarding post-convention levels of each measure. In addition, this part of the survey will
allow us to gain information about the respondent, including the demographic they fall
under, which is essential to our study. Survey responses for each measure will vary. For
example, depending on the type of question regarding income level, the responses may be
either interval or ratio. We may ask the voter to report their income in number form, or
we could ask them to select a range in which their income falls under. The level of
measurement we employ will be dependent on the type of question that we ask. We are
certain our questions will seek to measure the variables we are concerned with, contact of
low resource voters and political engagement. After conducting our surveys, we will
analyze our data and determine if there is a statistically significant relationship between
political party contact and political engagement, with a focus on comparing contact levels
by each of the parties towards certain demographics of voters. We will also analyze our
data and attempt to make the samples between the two conventions as close as possible
demographically so that our data can be compared. Finally, we hope to be able to depict
on a map the places from which our respondents reside, their level of political
engagement and contact they received from the parties or candidates. After plotting our
respondents on a map of each convention city we will be able to easily compare.
Particular strengths with using an observational survey include a greater level of
internal validity, which will allow us to determine the direct causal relationship in a more
controlled environment where we can test for any possible spurious factors. Another
strength is its feasibility of methods. We can easily gain participation from the local
population, rather than individuals who are more difficult to contact and there are little
obstacles to the completion of our study. One weakness is that our sample we collect
might not be representative of the populations of Cleveland or Philadelphia, or it might
be more difficult to receive survey data from low resource voters such as the Spanish-
speaking population. A specific concern is that the National Conventions are likely to
draw in the party elite and those more politically excitable; thus, our survey sample size
may run small and reduce our validity.
In order to find evidence for our proposed relationship, we hope to see an increase
in contact of low resource voters by the two parties. More specifically, we expect to find
the Democratic Party and the DNC reach out and target low resource voters substantially
more than the Republican Party and the RNC. In order gain feedback from a substantial
amount of low resource voters, we are considering surveying in an area with a high
density of these types of voters. For instance, we may survey at a community center or a
local food bank. In addition we plan to survey people at various protests and rallies where
“low resource voters” may be common. Further, our proposed theory suggests that as
these low resource voters are mobilized, they will experience more political or
convention related content in their lives. Therefore, if we were to observe that these low
resource voters had experienced an increase in political or convention coverage through
their local news outlets or if their social media content had become more politically
focused, thus our proposed relationship would be confirmed. In addition, the increase in
contact of low resource voters should increase their political engagement. In order to see
the proposed increase in political engagement we would hope to observe low resource
voters who begin to support the Democratic candidate or candidates, whereas they may
not have done so before receiving increased contact from the party or convention. We
expect the increase in political engagement of low resource voters to be the source of the
bump created by the Democratic Party and DNC. In conclusion, our proposed
relationships should develop within the conditions and locale of the DNC, but not
necessarily within the RNC. We suggest the possibility that the RNC does not target as
many low resource voters and thus will not receive as significant of a bump post-
convention.
Results and Data Analysis
Overall, 92 people completed our survey between the two cities and convention
locations. There were 45 respondents for the survey conducted near the RNC in
Cleveland. 47% of the respondents in Cleveland were low resource voters and 44% of
them had a local Cleveland zip code. In total, 47% of respondents were contacted by their
political party within 30 days of the RNC, while of the low resource voters only 42%
were contacted by their political party, which is a slightly lower percentage. However,
when looking at low resource Democratic voters a staggering 85.7% of them were
contacted by the Democratic party which is a substantially higher percentage. According
to the results, only 4% of respondents supported Donald Trump, the Republican nominee;
however, our survey only gathered five Republicans in the Cleveland area which could
help explain this lower percentage. There was a huge increase in engagement levels for
low resource voters, with 95% reporting an increase in their political engagement within
the thirty days around the convention. Within Cleveland, 52% of low resource voters
were local to the convention area and none of these low resource voters were Republican.
During the time of the DNC in Philadelphia there were 47 respondents to our
survey. There was a higher number of low resource voters in Philadelphia, with 72.3% of
those surveyed falling into this category, 72.3% of respondents resided in a local
Philadelphia zip code within the immediate DNC area. In total, 62% of respondents
reported being contacted by their political party within 30 days of the convention. 59% of
low resource voters were contacted by their political party, which is a marginally lower
rate. However, 75% of local low resource Democratic voters, our group of interest,
identified that they were contacted by the Democratic party. This is a higher rate of
contact than average voters of other political parties receive, including the Republican
party. 58% of low resource voters noted an increase in their engagement level once the
convention was underway in the area. Within Philadelphia, 82.3% of low resource voters
were local to the convention area and of these voters 64.2% were Democrats. Overall,
83% of local low resource Democratic voters supported the Democratic nominee, Hillary
Clinton. The effort by the Democratic Party to target low resource voters in the DNC area
of Philadelphia was substantial.
While in the Cleveland area 91% of respondents identified themselves as being at
least rarely engaged, overall individuals identified themselves to be more engaged in the
Philadelphia area with 66% of respondents identifying as ‘engaged a lot’ or ‘engaged a
great deal.’ One problematic aspect of our survey was that the randomized survey
collection gathered far more Democrats than it did Republicans. We gathered 92
responses but only 10 were Republicans. The sample demographics in Philadelphia were
as expected with a good number of low resource Democrats near the DNC area who
reported high levels of contact. However, in Cleveland there were no low resource
Republican voters, causing insufficient data to analyze how the Republican party
responds to its low resource constituents.
This signifies that the Democratic party actively targeted low resource voters in
the local Democratic National Convention area. We can make few conclusions about the
Republican party contact levels towards lower resource voters based on the limited data
and incongruencies in the demographics between the two cities. Since we found no low
resource Republican voters in the RNC area we cannot draw any conclusions regarding
this group in respect to the RNC or GOP. Our data was collected randomly yet we tried to
speak with low resource voters. Seeing that the majority of our low resource respondents
were Democrats it may be the case that low resource voters tend to affiliate mostly with
the Democratic party or as Independents, as opposed to Republican.
Conclusion
In seeking to understand the question why does the Democratic Party receive a
greater bump from its National Convention than the GOP? The first part of our
hypothesis that the Democratic Party and the Democratic National Convention targets
more lower resource voters than the Republican Party and Republican National
Convention cannot be upheld because there are not enough low resource Republican
voters to draw a conlcusion. However, we did find that the Democratic Party more
effectively contacted low resource voters who were Democrats, and that low resource
voters who were independent or of a third party were not contacted to the same degree.
The second part of our hypothesis, that we should see more of a bump generated by the
DNC than the RNC from reaching out to this often-untapped group, was upheld in that
the Democrats sufficiently targeted low resource voters, especially low resource members
of the Democratic Party within the DNC area. The efforts of the Democratic Party to
target low resource voters were successful in exciting the demographic towards being
more politically engaged. The increase in political engagement among the low resource
voters moved a majority of the demographic to support the Democratic nominee, Hillary
Clinton and thus providing the “bump” we expected to see. Our results also show us that
the Democratic Party is making an effort to reach across the aisle and contact voters who
are not members of the Democratic Party. This is as expected, since conventional wisdom
tells us that the Democratic Party and the DNC has the ability to attract voters from the
other side or voters outside of its party in a way that the GOP does not.
Studies of convention location and its connection to party benefits are worthy of
further research. Future projects could incorporate larger sample sizes and also make a
greater effort to survey Republican voters to determine if low resource Republican voters
are contacted in future convention cities. If there is a reasonable balance between parties
of the survey respondents it makes for the ability to draw further conclusions. Future
studies could also measure more than two convention cities across multiple election years
to determine trends over time and to control for any circumstantial anomalies. In
addition, if a convention is hosted in a city that has hosted a different party’s convention
in a previous year, the impact of the two conventions could be compared. It will be
interesting to compare the results of our research to the final effects of each convention
on the 2016 Presidential election.
References
Althaus, S., & Trautman, T. (2008). The Impact of Television Market Size on Voter
Turnout inAmerican Elections. American Politics Research, 36(6), 824–856.
Atkinson, Matthew D. et al. 2014. “(Where) Do Campaigns Matter? The Impact of
National Party Convention Location.” The Journal of Politics 76(04): 1045–58.
Fine, Terri Susan. 2007. “Mass Media and the Democratization of Presidential
Nominating Conventions.” In Rewiring Politics: Presidential Nominating
Conventions in the Media Age, Louisiana State University Press, 165–88.
Gershtenson, J. (2003). Mobilization Strategies of the Democrats and Republicans, 1956
2000. Political Research Quarterly, 56(3), 293–308.
Gimpel, J. G., Kaufmann, K. M., & Pearson-Merkowitz, S. (2007). Battleground states
versus blackout states: The behavioral implications of modern presidential
campaigns. Journal of Politics, 69 (3), 786-797.
Hill, K. Q., & Leighley, J. E.. (1996). Political Parties and Class Mobilization in
Contemporary United States Elections. American Journal of Political Science,
40(3) 787–804.http://doi.org/10.2307/2111795
Matsusaka, John G., and Nolan M. McCarty. 2001. “Political Resource Allocation:
Benefits and Costs of Voter Initiatives.” Journal of Law, Economics, and
Organization 17(2): 413–48.http://jleo.oxfordjournals.org/content/17/2/413 (April 18,
2016)
Panagopoulos, Costas. 2007. “Following the Bouncing Ball: Assessing Convention
Bumps, 1964-2004.” In Rewiring Politics: Presidential Nominating Conventions
in the Media Age, 16–29.
Plutzer, Eric. 2002. “Becoming a Habitual Voter: Inertia, Resources, And Growth in
Young Adulthood.”Am. Pol. Sci. Rev. American Political Science Review 96(01).
Powell, Richard J. 2004. “The Strategic Importance of State-Level Factors in Presidential
Elections.” Publius: The Journal of Federalism 34(3): 115–30.