Ventriloquism or an Echo Chamber? Measuring the Strength of House Members’ Rhetoric
in Local Newspapers.

Michael K. Romano
Department of Political Science
Georgia Southern University
P.O. Box 8101, Carroll Building 2205
Statesboro, GA 30460
(912) 478-0838 (phone)
(912) 478-5348 (fax)
mromano@georgiasouthern.edu

Paper prepared for the 2015 Midwest Political Science Association Meeting.
Chicago IL.
April15-19, 2014

 Ventriloquism or an Echo Chamber? Measuring the Strength of House Members’ Rhetoric
in Local Newspapers.
Political elites have been shown to excel at manipulating the information environment. Theories
about press control by elites vary. Some argue that actors manipulate the media like a
ventriloquist, using the press to restate their rhetoric verbatim, while others argue that journalists
act with more agency when deciding the news. I examine the ability of members of the House of
Representatives to manipulate news about their work in Washington. Members have an incentive
to promote themselves in order to control the public's perception and attempt to leverage the
local press through mediated homestyles; cultivating news agencies as a constituency using press
releases that are ready-made for news consumption. The extent to which news articles echo press
statements will vary depending on the complexity of the issue and the clarity of a member's
expressed agenda. To test this relationship, data on press releases is gathered and is matched with
corresponding district news coverage. Press releases and matched articles are examined using
plagiarism detection software to evaluate overlapping use of representatives' rhetoric.

 The media system in the United States is tasked with taking the millions of signals,
events, actions and non-actions of political actors and condensing them down into meaningful
and understandable stories that can be absorbed easily by the populace. Even with advances to
media technology that allow individuals to customize and filter the news they receive, the
prevailing wisdom is still that politics is a mediated experience that requires an intermediary to
transmit information about events that are largely “out of sight, out of mind” (Lippmann, 1922).
Since politics is chiefly a mediated experience (Edelman, 1995; Delli Carpini and Williams,
2001), proponents have argued that the effects of this sort of news-producing and disseminating
process has become more vital because of its impact on citizens and their perceptions of
government (Delli Carpini and Williams, 2001). Essential to understanding this process then, are
the factors which affect the decisions to focus on particular events over others in the news media.
Beyond the concept of gatekeeping (Soroka, 2012; Shoemaker, 1991), the news media has the
ability to influence the political agenda by focusing attention and framing particular events as
newsworthy while ignoring others. But who is actually controlling the news agenda?
This same agenda setting power can be influenced, however, by political elites; many of
whom work to influence the frame used by journalists by providing their own accounts of the
news. Agenda setting, the “successful transfer of salience from the mass media agenda to the
public agenda” (McCombs, 2004: pg. 37) is based primarily on the belief that individuals have
an interest in public affairs and a need for information about the world they live in but may not
have direct access to. Since the early studies performed on undecided voters in North Carolina
(McCombs and Shaw, 1972), the theoretical boundaries of agenda setting have expanded to
include examinations of rhetoric, symbols and their importance to the successful adopting of the
media’s agenda (Entman, 1993; Comstock and Scharrer, 2005), the psychological effects of

 agenda setting on individuals (Soroka, 2003; Kinder and Iyengar, 1987), and the impact of
advances to communication technology on the ability of the press to influence the public (Zeller,
2005; Hamilton, 2004). Political elites are argued to excel at manipulating and controlling the
information environment that media work in, however empirical examinations have returned
mixed results. The common belief is that political actors can manipulate the media like a
ventriloquist, using the press to restate their rhetoric word-for-word. Journalists are autonomous
actors, however, and one could argue that the world of journalists is more like an echo chamber,
with various messages bouncing off the walls in differing degrees of clarity. We therefore have
to ask the question: to what extent can representatives exert control over the content of media
reports about their behavior in such an environment? That is, how successful are congressional
press operations at controlling the message distributed to the public utilizing their own rhetoric?
Here, I focus attention on the production side of news making, specifically examining
how the relationships that members of Congress develop with journalists affect patterns of news
coverage that define the media’s agenda. I am most interested here in examining the abilities of
members of the House of Representatives to manipulate and control the agenda and topics
concerning their work in Washington in the local press. I argue that Representatives have an
incentive to manipulate local news outlets in order to control their public profile and that local
press outlets are constrained by the need for a localized angle for discussing national issues.
Members of Congress attempt to leverage local press outlets through the use of press releases
sent out to local news agencies that are ready-made for news consumption and thus lower the
costs of providing news to the public. In this way, member of Congress develop a mediated
homestyle, in which they attempt to cultivate journalists as a constituency in order to use them as
access points for the purpose of influencing public opinion in the district. One way in which

 representatives cultivate journalists in this way, I argue, is through the use of press releases
tailored to be used by local press outlets. Press releases are beneficial to members of Congress
since they provide a frame for news stories that can positively affect a member’s image in the
eyes of their constituents.
In order to understand the level of success representatives have at controlling the message
of news articles written about them in the local press, I analyze the content of press releases and
corresponding coverage using publicly available cheating detection software in order to calculate
the percentage of echoing of representative’s press statements that occur in the press. Data on
press releases was gathered utilizing members own web-pages on House.gov and matched with
corresponding news coverage within their local district using the Newsbank research collection.
I expect that press outlets will likely exhibit a significant amount of echoing of Representative’s
statements overall, and that the ability of news agencies to overcome economic and information
constraints will mitigate the level of echoing that is exhibited overall.
Legislative Strategy and the Media Agenda
The choice in what events journalists focus on and write about on a day-to-day basis is
the result of the norms and traditions of news making, interactions among news organizations
and with specific agencies, and interactions with sources, which have their own agendas that they
hope the news will focus on (McCombs, 2004; Soroka, 2012; Entman, 1993; 2007). News
organizations, acting as strategic entities, decide the news agenda in order to maximize
readership and reputation (Mullainathan and Heifer, 2005; Hamilton, 2004; Gentzkow and
Shapiro, 2006). To this end, smaller and more regional newspapers rely on news about local
representatives in order to stay competitive in the market and provide consumers with

 information that is relevant to their interests. Arnold (2004), in his study of local news coverage
and political accountability, supplies evidence that local newspapers do provide a sufficient
volume of coverage to voters about incumbent legislators, however there is wide variation
between individual papers. The media environment has been shown to be affected in some ways
by member’s own press operations, specifically during election years and in influencing the
frequency of neutral stories written about a representative and their actions (Cook, 1989;
Gershon, 2012; Romano, 2014).Substantive coverage of representatives and Congress tends to
focus more attention to distributive benefits that legislators bring to the district rather than policy
stances, however, since these types of messages are often more appealing to readers interested in
the local impact of legislation (Zaller, 1992; Tidmarch et al, 1984; Grimmer, 2010; Grimmer et
al, 2012).
In order to maximize attention and influence the public agenda, journalists must provide a
frame that consumers can understand and accept. Entman (1993: pg. 52) offers the most widely
accepted definition of a frame, stating that, “to frame is to select some aspect of a perceived
reality and make them more salient in a communicating text, in such a way as to promote a
particular problem definition, causal interpretation, moral evaluation, and/or treatment
recommendation” (emphasis in original). For local journalists and news agencies, the prevailing
frame is one of localization, where journalists focus on the impact of news items in a more
limited scope and focus on how stories will affect readers “at home” rather than painting a broad
picture about issues under discussion and allowing local papers to remain competitive with the
national press (Vinson, 2002; Schaffner and Sellers, 2003; Arnold, 2004; Larson, 1992). A
localized frame only provides a single part of the various angles journalists utilize when making
news decisions, the aggregation of which becomes the eventual story the public sees. From the

 press perspective, the overarching concern when framing a story is whether it will provide
readers with something they find interesting, that will stimulate readers to continue reading other
stories and continue patronizing a particular outlet (Hamilton, 2004). This leads reporters to
focus on the widespread news values of focusing on conflicts, attention-grabbing narratives,
easily simplified stories, and events with some sense of novelty and timeliness (Gans, 1979;
Graber, 2010; Larson, 1992).
There is a growing body of literature that provides evidence in support of the argument
that political actors work with the media to produce the news by focusing on the interactions that
occur between the two (Arnold, 2004; Gans, 1979; Fogarty, 2009; Vinson, 2002; Larson, 1992;
Schaffner, 2006). While it can be reasonably argued that the final news product is created in part
through interactions between representative A and news agency B, we cannot be certain of just
how much of an interaction there is, and how successful representatives are at incentivizing
journalists to focus on their work and manipulating the story to meet their goals. Entman (2007),
in his work examining media bias, argues that while bias itself may be difficult to fully grasp,
researchers can measure media slant by examining the differences in the skills and decision
making strategies of journalists and media staffers of political actors, and the interactions
between each entity. In this way, we can imagine the production of a news product as a type of
political game in which actors utilize a series of incentives to coerce journalists to produce
favorable news, and journalists respond by using their abilities as gatekeepers and agenda-setters
to coerce representatives to give them greater access. Fogarty (2009) develops several models of
news agency-legislator interactions in which signals sent by the electorate and investigated by
the media are picked up on by legislators to varying degrees, and constrains a legislator’s voting
preferences in order to remain in office. Previous work by Sellers (2010) has shown that political

 parties can be successful in manipulating the frame of news coverage so long as members
maintain a single, coordinated media message. Likewise, Domke et al. (2006) finds that
members of executive agencies have similar success in the media so long as they remain “on
message” with the President. However, how does this sort of result translate when we focus
attention on local media sources and the messages of individual members of Congress?
Ventriloquist Dummies and Echo Chambers
The ways in which representatives present themselves via press statements has a
significant impact on the ability of a legislator to forge an electoral connection (Grimmer, 2010).
Variations in the types of messages transmitted to the public by representatives have been shown
to have a significant impact on constituent opinions (Grimmer et al, 2012). In order to influence
constituent opinion, legislators must act outside of their official capacity as decision makers and
interact with constituents through various means, the most well-known of which is the
development of personal “homestyles” when directly communicating with the public (Fenno,
1978). In order to assess the success of these efforts, one can measure how strongly constituents
believe that legislator A is at standing for their interests via the analysis of “fit,” or congruence –
the level of correspondence between the actions of legislators and the desires of the public
(Miller and Strokes, 1963; Golder and Stramski, 2010). Implicit in these theories is the
assumption that a representative’s actions are known to the public, and that this knowledge was
developed independently. This neglects the fact that representatives allocate significant amounts
of time and resources to ensure that constituent reactions to their behavior is framed in a way that
will bolster support in the district (Arnold, 1992; 2004; Fenno, 1978; Mayhew, 2000; Grimmer,
2010). Representatives cannot interact personally with constituents on a daily basis, however,
and must also rely on the media in order to successfully control public support (Schaffner and

 Sellers, 2003). Whether direct or indirect, when interacting with constituents, “legislators use
tools of communication to explain to constituents why their representative is effective.
Legislators use this communication to define the type of representation they provide, through the
creation of homestyle” (Grimmer, 2010: 627).
In order to influence the perception of the public, representatives have increasingly
allocated resources to media operations within their offices (Born, 1982, Romero, 2006). This is
done with the hope that press operatives can incentivize media agencies to accept and broadcast
a representative’s message with little editorial framing. While representatives have their own
style of influencing the perceptions of the public at a personal level, it is the mass media largely
controls the regular transmission of messages from public figures to the mass public. In order to
engage the electorate, therefore, members of Congress must engage in a form of impersonal
influence (Mutz, 1998), relying on indirect contact and associations in order to influence the
public’s perception about their actions and behaviors while in Washington. News agencies are
autonomous agencies with their own interests and beliefs on what constitute a “good story,” and
have the ability to conduct their own investigations into the behaviors of representatives in the
legislature. This could potentially lead to stories that portray a representative acting out of step
with the political beliefs of the district they serve. The interaction between news agencies and
representatives, therefore, creates a type of mediated homestyle, in which representatives attempt
to brand themselves using their own unique styles and rhetoric while appealing to the journalistic
norms and standards of the news. Essentially, representatives and their press operations attempt
to “cultivate a constituency” in the press in order to incentivize journalists to convey their
behavior in a positive light. The new “weapons of combat” (Zaller, 1999: pg. 1-1) are the press
releases, statements, press conferences and “spin” that representatives put on their work in order

 to reassure the public that they are acting in the best interests of the district and the nation.
Members of Congress and their staffs are afforded several opportunities to strategically insert
themselves into the process of perception building. Representatives work with the media by
developing personal relationships with journalists, editors, and news agencies (Gershon, 2012;
Cook, 1989; 2005), managing media gatekeeping by varying the amount of statements coming
from their offices (Soroka, 2010; Shoemaker, 1991), and inserting their own narratives into the
frames of stories to manage the agenda setting function of the media in the district (McCombs,
2004; McCombs and Shaw, 1972; Entman, 2007).
Here, I focus on the concept of ventriloquism and echoing that occur in the final news
product by analyzing the level of congruence that exists between the press releases of members
of the House of Representatives and corresponding local news agencies. In order to insert
themselves into the news production process, members of Congress utilize a number of press
releases which are designed specifically for press consumption (Cook, 1989). Newspapers,
especially local papers, are often subject to economic constraints that make it difficult to dedicate
time directly to investigating and observing the actions of members of Congress in order to cover
what representatives do in the legislature (Vinson, 2002). To fill this gap, editors and journalists
often rely on press releases sent out from congressional offices (Cook, 1989; Arnold, 2004;
Vinson, 2002; Schaffner, 2006; Grimmer, 2010). Press secretaries commonly believe that they
will have a high degree of success generating news coverage with press releases (Cook, 1989;
Gershon, 2012). Cook (1989) notes, however, that press releases are far more successful at
generating attention to specific topics, rather than controlling the language used by journalists to
describe members’ actions. The use of press releases as a source in journalists’ articles is often
referred to as “ventriloquism” by scholars (Grimmer, 2010: pg. 5) due to the fact that releases

 can often be duplicated as news stories almost verbatim, “much like a ventriloquist’s dummy.”
Printing of press releases in this way is commonly believed to be done more often by smaller
newspapers, and newspapers that are constrained economically in other ways. Table 5.1 presents
an example of ventriloquism of Sam Farr’s (D-CA) press statements by the Santa Cruz Sentinel.
The press release and subsequent article focus on a planned workshop for immigrants looking to
gain citizen status in the United States. While the language of the September 3, 2010 article is
rearranged slightly, it bears a striking resemblance to the previously released statement made my
Farr’s office on September 2, 2010.
[Insert Table 1 about here]
The factors that lead to press ventriloquism may vary depending on certain characteristics
of the news agency and characteristics of a district’s representative, however, and the concept of
ventriloquism may not be the norm for many news outlets. Representatives who work with the
press and develop a mediated homestyle can have some influence over the eventual frame of a
news story through the use of press releases. While this influence over the frame can be
beneficial by leading journalists to write more positive articles, representatives must still
overcome the hurdle presented by norms of newsworthiness in order to affect the press agenda.
Contrary to the idea of direct copying of press releases and ventriloquism, I propose here that
legislative press releases – especially those sent out by members of the House of Representatives
– exist inside what can be described as an “echo chamber.” The concept of an echo chamber is
based on the idea that statements made by members of Congress are often picked up by the press
to varying degrees based on characteristics of the news agency – such as economic constraints
and the ability of an agency to provide independent investigations – as well as characteristics of
the representative and the district they serve. Some legislators are more adept at creating

 statements that the press will utilize, and thus are better at controlling journalists like
“ventriloquist dummies.” The relationship between the frequency of press releases distributed
and news articles produced is rarely perfectly correlated, however, and members of the House are
often not the only source of information used by news editors and staff when generating a story.
The use of a member’s press releases and the variation in the level of congressional rhetoric that
ends up in a final news product, therefore, is often more disbursed, leading the press to echo the
sentiments of a legislator without explicitly becoming a representative’s mouthpiece, or “partner
in propaganda” (Bagdikian, 1979).
Briefly, Table 5.2 presents an example of press echoing of a statement made by Paul
Tonko (D-New York) by the Albany Times-Union, which serves Tonko’s district. As the side-byside comparison shows, the level of clarity between the press release distributed by Tonko’s
office on September 2, 2010 and the subsequent news article printed by the Times-Union on
September 3, 2010 is focused specifically on a few key quotes in Tonko’s press release. The
Times-Union focuses specifically on Tonko’s call to increase the level of oversight by public
safety officials after a gas leak evacuated several homes in New York’s 21st district. The article
falls short of emphasizing the potentially catastrophic impact that could occur if the cause of the
leak is not discovered, an opinion that is emphasized by Tonko’s statement. The author of the
Times-Union piece does use the press release and statements made by Tonko as the main subject
of the final article; however the final story does not completely copy the statement and utilizes
the release to create a concise, short article that explains the representative’s feelings regarding a
local event.
[Insert Table 2 about here]

 Data
To be able to come to some conclusions how representatives shape the news agenda, a
sample dataset of sixty members of the House of Representatives and corresponding local news
outlets was created in order to test what factors impact the amount of ventriloquism that local
news agencies exhibit when writing stories involving legislators work in Washington.
Measurement of legislative success at getting their message to the news media was done by
collecting a unique database of over ten thousand press releases issued between 2009 and 2010.
Press releases were chosen since these documents are designed specifically for press absorption
in order to manipulate news frames in a way that focuses the public’s attention on issues and
events that representatives want their constituents to hear about – what Grimmer (2010) refers to
as a legislator’s “expressed agenda” – and have been shown to be a useful tool to measure
legislators’ strategies with the media (Kaplan, Park and Ridout, 2006). Press releases have also
recently been utilized by in order to systematically evaluate the level of attention that legislators
allocate to particular topics of political importance, and have been shown to be a unique form of
homestyle that focuses on journalists as the target audience (Grimmer, 2010).
[Insert Table 3 about here]
In order to measure the saturation of a legislator’s press releases in a local news outlet,
each member was matched first to a local newspaper that services the representative’s district.
Afterwards, news articles were scraped from each newspaper utilizing keyword searches of the
Newsbank Digital Archive, which provides a comprehensive list of several thousand local,
regional, and national newspapers within the United States. The final sample contained 10,276
news articles, collected based on whether a representative’s name was mentioned in the article.

 To determine if a press release from a representative’s office was used in a newspaper, I used
publically available cheating detection software to analyze whether a press release and
newspaper had similar content (Bloomfield, 2008). Table 5.3 presents some brief descriptive
information about the initial results of the detection analysis. Overall, the software was able to
detect 897 newspaper-press release dyads which contained a percentage of quotes, paraphrasing,
or explicit plagiarism in newspapers from congressional press releases. Of the 10,087 press
releases sampled, only 8.81% were found to be used in a corresponding news article. The
average number of press release by a news agency included in the sample was approximately
fifteen total. In column six of Table 5.3, the average percentage of press releases that was copied
by news articles for each member of the House. The range of use is between 0.0% - indicating
that there were no measurable similarities between a member’s press releases and news articles –
to 57.7%, indicating a high level of copying. On average, members’ press releases were copied
by news agencies, either in quotes, paraphrasing, or direct plagiarism, a total of 18.36% of the
time.
One way in which representatives may cultivate journalists in order to influence the news
is by having media operations tailor press statements in a way that is easily transformed into a
story by journalists. Among the subjective factors that journalists consider when writing the
news is the level of readability that a story has. Readability, the ability of a text to be understood
by its audience, has received a great deal of attention in the fields of literacy and education, and
has previously been used in political science in studies concerning ballot question readability
(Reilly and Richey 2011), understanding and absorbing information in newspaper articles (Johns
and Wheat 1984), policy making strategies in divided government (Colomer, 2005), and the
clarity of jury instructions (Charrow and Charrow, 1979). The first measures of readability were

 developed in the middle of the 20th century in an attempt to better quantify and evaluate the
literacy of students. While several measures of readability now exist, the baseline of each
measure is the difficulty an individual has with reading and assessing a piece of writing. Basic
models, such as the Flesch-Kincaid Grade Level Formula translate scores into a “grade level” for
a piece of writing, which allow evaluators to better assess whether a piece of prose can be
understood by a particular audience. Measures of readability are calculated utilizing a mixture of
discrete features in a text, such as the number of words, characters, sentences, or syllables. Each
of these features are easily observable and quantifiable characteristics of the text, the underlying
theory being that the shorter and simpler a word or sentence, the easier it is for an individual to
understand. Here, I rely on several factors that may influence the decision to replicate the
language of a press release in order to examine the level of skill a representative has at providing
journalists with news products they can use. Specifically, I include the traditional FleschKincaid Grade Level (FKGL) index and the Dale-Chall Readability Score (DCRS), which
measure readability utilizing two distinct formulas, as well as the number of syllables and
difficult words in a press release to act as controls for the length of a document and the amount
of “jargon” in the text respectively.
Findings
The average percentage of a member’s press releases that was copied by corresponding
news articles was analyzed using an OLS regression. Two models were tested in order to
examine the impact of including factors that may be significant to mediated homestyle. The first
model focuses primarily on variables that have previously been shown to impact the likelihood
of press attention in the media. Factors believed to be significant to causing variations in the
level of press echoing are based primarily on characteristics of the representatives themselves,

 such as demographic features like race and gender, as well as terms in office and age. Members
own behavioral traits, such as ideological extremism based on the Poole-Rosenthal scores as well
as bill sponsorship and the number of press releases distributed by a congressional office are
included to examine whether behavioral characteristics are cause the press to adopt the language
of a representative more readily. Frequency of press releases is believed to have a curvilinear
relationship, however, based on previous research performed by Schaffner, Schiller and Sellers
(2003) and Romano (2014). Finally, characteristics of the news agency are included to test
whether factors such as daily circulation, chain ownership, and whether the newspaper has a
bureau in Washington significantly impact the ability of a press agency to generate news items
free from the manipulation of congressional press offices. Table 4 presents summary statistics
for each of the variables utilized in the model.
[Insert Table 4 about here]
The results of the OLS regression analysis are displayed in Table 5. While certain
demographic features - such as being a minority in Congress or being female - were believed to
spark more press attention and thus increase the likelihood that members of the press will echo
statements sent out by these representatives, the analysis here finds that there is no significant
effect of either two demographic characteristics on the level of press echoing in the model.
Likewise, terms in office is not found to have a significant impact on press echoing, however it is
signed appropriately based on previous theories of press attention to more senior members
(Gershon, 2012; Romano, 2013). The amount of bills sponsored by a member of Congress,
however, was found to increase the level of press echoing. Based on the results in Table 5, an
increase in the number of bills sponsored by a member of the House increases the average
percentage of press echoing by 0.241 points, all other variables held constant. Bill sponsorship is

 an activity that both representatives and members of the press find newsworthy, since
representatives can utilize sponsorship as a way to provide public goods to the district, and
journalist can easily craft stories that have a local angle which readers can understand. Adding in
variables to test the specific factors that lead to a more mediated homestyle does not significantly
affect the strength of bill sponsorship on press echoing, as well. Even accounting for mediated
homestyle variables, an increase in the number of bills sponsored by a representatives lead to a
0.252 percent increase in the level of press echoing by local newspapers.
[Insert Table 5 about here]
Looking at characteristics of the newspapers included in the models, the analysis in Table
5 shows that the daily circulation of a newspaper actually has the effect of decreasing the level of
press echoing that occurs overall.

In Figure 1, the daily circulation of a newspaper is plotted

along with the predicted level of press echoing in order to show how increases in the daily
circulation of a newspaper decrease the expected level of press release mirroring by news
agencies. Overall, holding other variables constant, an increase in the logged daily circulation of
a newspaper has the effect of lowering the percent average level of press echoing by -3.198
points. This amount is not significantly affected by including factors attributed to a
representative’s mediate homestyle, with the impact of an increase in daily circulation decreasing
the percentage of press echoing by -3.541 points. These results suggest that, following theories
put forward by Kaplan (2006) and others, news agencies exist in an open system which constrain
their ability to independently produce news on a daily basis, and the ability of a certain outlet to
overcome certain constraints – such as economic constraints caused by their level of circulation
or information constraints based on their ability to conduct independent investigations in the

 chamber – will significantly affect whether a journalist must rely on legislators as a primary
source of information.
[Insert Figure 1 about here]
Adding factors that can be attributed to a member’s mediated homestyle does have an
invaluable effect on our understanding of what causes press outlets to adopt the language of
congressional press statements. Including variables ascribed to a representative’s media tactics
(such as number of press releases and the readability of those press releases), allows us to
explain 18.7% more of the variance in the model itself (R2 = 0.447). Interestingly, the number of
press releases sent out by a member’s office does have a significant, positive effect on the level
of press echoing that occurs in local newspapers. Increasing the number of press releases has the
effect of increasing the level of press echoing by 0.080 points in the model. This is conditioned,
however, by the inclusion of the quadratic function to test whether press releases have a
curvilinear relationship with press plagiarism. Accordingly, it should be understood, based on
the results in Table 5, that while increasing the level of press releases does have a positive effect,
this effect can be mitigated slightly by sending out too many releases at one time. Controlling
for length of press releases as well as the number of difficult words in a release also had the
expected effect on press echoing. Increasing the length of a press release (measured based on the
number of syllables in a release), increased the amount of echoing in a news article by 0.024
points, while using more difficult words in a release had the effect of decreasing the level of
press echoing by -0.210 points per word used.
The key feature of any text is its readability, however, and here the results in Table 5
show some mixed results. According to the results in full model presented in Table 5, an

 increase in the difficulty of a text, according to the Flesch-Kincaid Grade Level (FKGL), has the
effect of decreasing the overall use of echoing by the local press by -3.263 points. This makes
sense, as studies have found that journalists often write at approximately a 9th grade reading level
to ensure the maximum number of readers comprehend a given news story. An increase in the
readability of a press release, therefore, should decrease the likelihood that it will be utilized
verbatim by a journalist, since they would need to rewrite the release in order to cater to a more
general knowledge audience. However, this finding is opposed to the results found when
examining press echoing using the Dale-Chall Readability Score (DCRS). Unlike the FleschKincaid Grade Level, the results in Table 5 show that an increase in the readability of a press
release when measured using the DCRS has the effect of increasing the amount of press echoing
by 8.439 points. This is likely due to the fact that the FKGL and DCRS both measure readability
in fundamentally different ways. Whereas the FKGL primarily takes into account the basic
attributes of a text, the DCRS accounts for the complexity of language used.
Conclusions
Despite changes in the business of news production (Pew Research Center, 2013),
constituents often still rely on local news media to keep them updated about the actions of their
elected officials. News about representatives is dependent on the interactions that journalists
have with a variety of sources, and legislators tend to be viewed as “experts” where the local
interests of national events is concerned. This leads news agencies to depend on representatives
to provide a localized angle to stories, which journalists can then use to inform the public more
fully. Variations in the constraints that limit the ability of journalists to conduct independent
investigations in the political matters eventually lead them to echo the statements made by

 legislators, and allow representatives to manipulate the news that is distributed to constituents
through local news outlets.
Here, I have focused attention on the ability of members of the House to successfully
manipulate the language and topics used by members of the press to describe political events as
well as member’s own actions in Washington. Using a unique sample of the press releases of
sixty representatives and corresponding news stories from local news outlets, I have shown that
there is a significant level of echoing that occurs between members of Congress and the press,
however the level of mirroring is not so high as to conclude that members of the press are
mouthpieces for representative’s to utilize. The findings here provide evidence that variations in
the political characteristics of representatives and the limitations of news agencies have a
significant impact on the amount of echoing that occurs in the local press, and thus the ability of
representatives to manipulate the language used to describe their actions to the electorate.
While the factors examined here provide evidence for the theory that variations in
congressional coverage can be explained by the constraints being placed on news agencies and
the behavior of representatives in general, future research into the topic still needs to draw out
variations that may occur due to more transient factors as well. Factors such as the perceived
hostility of a district to their member, the quality and quantity of challengers during an election
season, and the amount of pork members bring back to the district may further our understanding
of the variation that exists in congressional coverage. Developing a more accurate understanding
of the effect of the environment that the media and members of Congress exist within, both
politically and structurally with regard to their organizations, will help us better understand how
constituents develop opinions about their elected officials, political knowledge, as well as voter
turnout and vote choice.

 Table 1: Side-by-Side Comparison of Sam Farr (D-California) and the Santa Cruz Sentinel Newspaper
Sam Farr Press Release (September 2,
2010)
The Santa Cruz Sentinel (September 3, 2010)
Citizenship workshop Saturday
Congressman Sam Farr, D-Carmel, will host his first
citizenship workshop from noon-3 p.m. Saturday at

WASHINGTON, D.C. - Congressman Sam Farr (DCarmel) on Saturday will host his first citizenship
workshop, to take place in Watsonville. The goal of the
event is to increase the number of citizenship applicants

Table 2: Side-by-Side Comparison of Paul Tonko (D-New York) and the Albany Times-Union Newspaper
The Albany Times-Union (Sept. 3, 2010)
GAS LEAK BRINGS CALL FOR OVERSIGHT
Times Union, The (Albany, NY) - U.S. Rep. Paul Tonko sent
letters to state and federal agencies on Thursday, calling for
more oversight after a propane gas leak last week in Schoharie
County.

Paul Tonko Press Release (Sept. 2, 2010)
Congressman Paul Tonko today sent letters to U.S. Department
of Transportation Secretary Ray LaHood and NYS Public
Service Commissioner Garry Brown calling for answers and
oversight in the wake of the Texas Eastern propane pipeline leak
in Gilboa last week, which led to an evacuation of the area and

 Table 3: Measuring the Coverage Rate of House Press Releases
Number of Press
Name
Newspaper
Releases
Aaron Schock
Peoria Journal Star, The (IL)
Adam Putnam
Ledger, The (FL)
Al Green
Houston Chronicle
Alcee Hastings
Sun Sentinel (FL)
Anna Eshoo
Daily Post/Daily News (Palo Alto)
Bart Stupak
Traverse City Record-Eagle (MI)
Portland Press Herald/Main Sunday
Chellie Pingree
Telegram (ME)
Chris Van Hollen
Baltimore Sun, The (MD)
David Wu
Oregonian, The (OR)
Donna Edwards
Baltimore Sun, The (MD)
Duncan D. Hunter San Diego Union-Tribune
Dutch
Ruppersberger
Baltimore Sun, The (MD)
Ed Perlmutter
Denver Post, The (CO)
Eddie Bernice
Johnson
Dallas Morning News
Elijah Cummings Baltimore Sun, The (MD)
Erik Paulsen
Twin Cities Star-Tribune, The (MN)
Frank Wolf
Winchester Star
Fred Upton
Kalamazoo Gazette, The (MI)
Gary Miller
San Gabriel Valley Tribune (CA)
Ginny BrownWaite
Tampa Bay Times (FL)
Glenn Thompson
Centre Daily Times (State College, PA)
Gregg Harper
Starkville Daily News (MS)
Hank Johnson
Atlanta Journal-Constitution (GA)
James Langevin
The Pawtucket Times
Jim Costa
Fresno Bee
Jim McDermott
Seattle Post-Intelligencer
Jim Moran
Alexandria Gazette Packet (VA)
Joe Pitts
Intelligencer Journal-Lancaster New Era
Joe Wilson
The State (Columbia, SC)
John Mica
St. Augustine Record (FL)

Number of News
Articles

Frequency of Press
Release Use

114
140
81
106
171
369

459
158
143
201
73
98

6
9
0
54
7
25

Average Percent
Usage
20.2%
24.6%
0.0%
29.4%
18.4%
23.2%

373
367
76
219
74

228
51
108
31
53

18
1
1
0
0

27.6%
10.0%
5.0%
0.0%
0.0%

159
241

100
149

7
7

9.0%
7.4%

214
244
120
212
79
39

302
35
94
93
236
70

4
0
6
18
7
13

11.8%
0.0%
24.5%
22.4%
20.9%
17.0%

81
162
83
168
182
208
135
144
121
206
153

212
58
57
161
43
339
126
96
188
316
146

24
5
10
8
14
13
16
6
12
14
11

25.6%
6.6%
39.4%
29.6%
41.3%
14.4%
44.0%
32.2%
25.2%
15.4%
24.2%

 John Tierney
Judy Chu
Kathy Castor
Kenny Marchant
Kurt Schrader
Lois Capps
Lynn Jenkins
Mac Thornberry
Marcia Fudge
Marsha Blackburn
Mike Doyle
Mike Pence
Mike Simpson
Niki Tsongas
Patrick McHenry
Patrick Tiberi
Paul Tonko
Ralph Hall
Raul Grijalva
Richard Neal
Robert Brady
Rosa DeLauro
Sam Farr
Sander Levin
Spencer Bachus
Steve Scalise
Tim Ryan
Virginia Foxx
Wally Herger
Zoe Lofgren

Salem News, The (MA)
San Gabriel Valley Tribune (CA)
Tampa Bay Times (FL)
Fort Worth Star Telegram
Oregonian, The (OR)
Ventura County Star (CA)
Topeka Capital-Journal, The (KS)
Wichita Falls Times Record
Plain Dealer, The (OH)
Commercial Appeal, The (Memphis, TN)
Pittsburgh Post-Gazette (PA)
Journal Gazette, The (IN)
Idaho Statesman, The (ID)
Sun, The (MA)
Hickory Daily Record (NC)
Columbus Dispatch, The (OH)
Times Union, The (NY)
Dallas Morning News
Arizona Daily Star
Milford Daily News (MA)
Philadelphia Inquirer (PA)
New Haven Register (CT)
Santa Cruz Sentinel (CA)
Detroit News, The (MI)
Birmingham News
Times-Picayune, The (LA)
Vindicator (Youngstown, OH)
Winston-Salem Journal (NC)
Chico Enterprise-Record (CA)
San Jose Mercury News (CA)

143
77
171
51
133
344
137
115
120
98
69
531
178
159
122
93
179
50
337
23
56
559
170
251
222
206
167
83
146
56

300
183
276
66
122
230
196
206
145
178
198
178
215
539
111
204
248
61
233
80
27
170
261
152
211
379
92
214
263
114

5
17
10
2
4
58
6
15
6
8
0
7
9
47
54
15
8
0
23
0
2
28
61
26
51
49
19
13
27
3

30.0%
14.7%
12.7%
8.0%
13.5%
12.7%
38.0%
24.7%
18.3%
16.0%
0.0%
19.3%
27.2%
21.7%
57.7%
13.4%
8.6%
0.0%
13.2%
0.0%
20.0%
15.7%
23.7%
23.6%
13.2%
22.0%
24.4%
11.9%
16.0%
12.3%

 Table 4: Summary
Statistics
Variable
Mediated
Homestyle Factors
Number of Press
Releases
FKGL
DCRS
Difficult Words
Syllables
Environmental
Factors
Daily Circulation
Bill Sponsorship
Chain Ownership
Ideological Extremity
Age
Minority
Representative
Female
Terms in Office
Washington Bureau

Mean

S.D

Median

168.10
12.44
9.35
103.89

108.70
1.03
0.47
30.00

145.00
12.40
9.43
99.01

640.30

243.87

572.10

133,39
0.00
20.12
0.67
0.52
57.22

109,83
5.00
13.96
0.46
0.14
11.33

88,776
.00
15.00
1.00
0.52
57.50

0.17
0.25
5.50
0.45

0.38
0.44
3.93
0.50

0.00
0.00
5.00
0.00

Min.

23.00
10.29
7.93
61.88
352.4
0

Max

559.00
16.35
10.95
254.68
1,888.3
0

3,700. 448,271
00
.00
1.00
83.00
0.00
1.00
0.22
0.84
28.00
86.00
0.00
0.00
1.00
0.00

1.00
1.00
15.00
1.00

 Table 5: News Release Echoing by Representative Style
(Traditional and Mediated Homestyles)
Trad.
Mediated
Homestyle
Homestyle
Coef.
(S.E.)
Coef.
(S.E.)
Political and Environmental Factors
-3.198** (1.490) 3.541* (1.384)
Daily Circulation
*
0.252*
0.241** (0.109)
(0.103)
Bill Sponsorship
*
-5.177* (3.033) 7.359* (2.715)
Chain Ownership
**
Ideological Extremity
(10.07
17.063* (9.963) 12.224
(Median)
5)
-0.118
(0.173) -0.097 (0.152)
Age
-1.13
(3.385) 0.254 (3.043)
Female
-0.145
(0.488)
-0.13
(0.466)
Terms in Office
-2.293
(3.212) -3.892 (2.965)
Washington Bureau
-2.771
(4.012) -3.177 (3.733)
Minority Rep.
Mediated Homestyle
Num. Press Releases
Num. Press Releases^2
FK-Grade Level
DCRS
Syllables

--

--

--

--

---

---

--

--

--

--

Difficult Words
53.572* (19.09
**
1)
60
0.373
0.26
10.156 (df=50)
3.299*** (df=9;
F-Statistic
50)
***p < .01; **p < .05; *p < .1
Constant
N
R2
Adj. R2
Residual St. Error

0.080*
*
0.0002
**
-3.263*
8.439*
0.024*
*
0.210*
**
22.321

(0.038)
(0.000
1)
(1.799)
(4.275)
(0.011)
(0.077)
(38.52
8)

60
0.587
0.447
8.780 (df = 44)
4.176*** (df =
15; 44)

 Reference
Arnold, D. R. (1992).The Logic of Congressional Action. New Haven: Yale University Press.
Arnold, D. R. (2004).Congress, the Press, and Political Accountability. Princeton: Princeton
Press.
Bagdikian, B. H. (1979). Congress and the Media: Partners in Propaganda. In R. O. Blanchard,
Congress and the News Media (pp. 388-398). New York: Hastings House.
Bloomfield, L. (2008, June 1). WCopyFind. Retrieved from The Plagiarism Resource Site:
http://plagiarism.bloomfieldmedia.com/z-wordpress/software/wcopyfind/
Chadwick, A. (2001). The Electronic Face of Government in the Internet Age: Borrowing from
Murray Edelman. Information, Communication, and Society, 4(3), 435-457.
Comstock, G., & Erica, S. (2005). The Psychology of Media and Politics. New York: Elsevier
Academic Press.
Cook, T. E. (1989). Making Laws and Making News. Washington, D.C.: Brookings Institute.
DelliCarpini, M. X., & Williams, B. (2001). Let Us Infotain You: Politics in the New Media Age.
In W. L. Bennett, & R. M. Entman, Mediated Politics: Communication in the Future of
Democracy (pp. 160-181). Cambridge: Cambridge University Press.
Domke, D., Graham, E. S., Coe, K., Lockett John, S., &Coopman, T. (2006). Going Public as
Political Strategy: The Bush Administration, an Echoing Press, and Passage of the Patriot
Act.Political Communication, 23(3), 291-312.
Edelman, M. (1995).From Art to Politics. Chicago: University of Chicago Press.
Entman, R. M. (1993). Framing: Toward Clarification of a Fractured Paradigm. Journal of
Communications, 43(4), 51-58.
Entman, R. M. (2007). Framing Bias: Media in the Distribution of Power.Journal of
Communication, 57, 163-173.
Fenno, R. F. (1978). Homestyle: House Members in their Districts. New York: Longman Press.
Fogarty, B. J. (2009). A Simple Model of Legislator and News Media Interaction.Journal of
Theoretical Politics, 21(1), 5-24.
Gans, H. J. (1979). Deciding What's News. New York: Vintage.
Gentzkow, M., & Shapiro, J. (2006).Media Bias and Reputation.Journal of Political Economy,
114, 280-316.
Gershon, S. A. (2012). Press Secretaries, Journalists, and Editors: Shaping Local Congressional
News Coverage. Political Communication, 29(1), 160-183.

 Golder, M., &Stramski, J. (2010).Ideological Congruence and Electoral Institutions.American
Journal of Political Science, 54(1), 90-106.
Graber, D. A. (2010).Mass Media and American Politics. Washington, D.C.: CQ Press.
Grimmer, J. (2010). A Bayesian Hierarchical Topic Model for Political Texts: Expressed Agendas
in Senate Press Releases. Political Analysis, 18(1), 1-35.
Grimmer, J., Messing, S., & Sean, W. (2012). How Words and Money Cultivate a Personal Vote:
The Effectof Legislator Credit Claiming on Constituent Credit Allocation. American
Political Science Review, 106(4), 703-719.
Hamilton, J. T. (2004). All the NewsThat's Fit to Sell: How the Market Transforms Information
into News. Princeton: Princeton University Press.
Kaplan, N., Park, D., &Ridout, T. (2006).Dialogue in American Political Campaigns?An
Examination of Issue Convergence in Candidate Television Advertising.American
Journal of Political Science, 50(3), 724-736.
Kaplan, R. L. (2006). The News about New Institutionalism: Journalism's Ethic of Objectivity
and its Political Origins. Political Communication, 23, 173-285.
Kinder, D. R., &Iyengar, S. (1989). News that Matters: Television and American Opinion.
Chicago: University of Chicago Press.
Larson, S. G. (1992). Creating Consent of the Governed: A Member of Congress and the Local
Media. Carbondale: Southern Illinois University Press.
Lippmann, W. (1922).Public Opinion. New York: Macmillan.
McCombs, M. (2004).Setting the Agenda: The Mass Media and Public Opinion. Cambridge:
Polity .
McCombs, M. E., & Shaw, D. L. (1972).The Agenda-Setting Function of the Mass Media.Public
Opinion Quarterly, 36(2), 176-187.
Miller, W. E., & Stokes, D. E. (1963).Constituency Influence in Congress.The American
Political Science Review, 57(1), 45-66.
Mullainathan, S., &Schleifer, A. (2005).The Market for News.American Economic Review, 95,
1031-1053.
PEW Research Center. (2013, March 18). The State of the News Media, 2013. Retrieved March
19, 2013, from http://www.stateofthemedia.org
Puglisi, R., & Snyder, J. M. (2011).Newspaper Coverage of Political Scandals.Journal of
Politics, 73(3), 931-950.
Romano, M. (2013).Determining Congressional Press Coverage: A Relational
Approach.Midwest Political Science Association (pp. 1-35). Chicago: MPSA.

 Schaffner, B. F. (2006). Local News Coverage and the Incumbency Advantage in the U.S.
House.Legislative Studies Quarterly, 31(4), 491-511.
Schaffner, B. F., & Sellers, P. J. (2003).The Structural Determinants of Local Congressional
News Coverage.Political Communications, 20(1), 41-57.
Sellers, P. J. (2010).Cycles of Spin: Strategic Communication in U.S. Congress. Cambridge:
Cambridge University Press.
Shoemaker, P. J. (1991). Gatekeeping. Thousand Oaks: Sage Publications, Inc.
Soroka, S. N. (2003). Media, Public Opinion, and Foreign Policy.Harvard International Journal
of Press and Politics, 8(1), 27-48.
Soroka, S. N. (2012). The Gatekeeping Function: Distributions of Information in Media and the
Real World. Journal of Politics, 74(2), 514-528.
Tidmarch, C., & Pitney, J. (1984).Covering Congress.Polity, 17, 446-483.
Vinson, D. C. (2003).Local Media Coverage of Congress and its Members: Through Local Eyes.
Cresskill: Hampton Press.
Zaller, J. R. (1992). The Nature and Origins of Mass Opinion. Cambridge: Cambridge University
Press.
Zeller, T. (2005, May 25). Are Bloggers Setting the Agenda? It Depends on the Scandal. New
York Times, p. http://www.nytimes.com/2005/05/23/technology/23blog.html?_r=0.