1. Aufsatz | Mergel – “Connecting to Congress”: The use of Twitter by Members of Congress
“Connecting to Congress”: The use of Twitter by Members
of Congress
Ines Mergel
Key message 2 Communicating with the public using new technologies
How do political elites, such as the Members of the U.S. Members of Congress in the U.S. government are known for
Congress, decide to use innovative forms of Information their conservative use of new information and communication
and Communication Technologies, such as social media technologies in general, and specifically using advanced Inter-
applications? Communication between elected officials is net technologies to represent themselves online or to commu-
guided by outdated rules and regulations that are focusing nicate with their constituents (Esterling, Lazer, & Neblo,
on paper mailings. The apparent lack of formal guidance 2012). They use their websites mostly as a static, non-inter-
and outdated rules are not reflecting the changing online active information tool to push information out and remove
landscape and the requirements on Members of Congress the information after a 10-day press release period to be able
to interact with their constituents where they prefer to to control the current message. Members of Congress are
receive their information. New forms of highly interactive therefore oftentimes criticized for their slow adoption and
online communication tools, such as the microblogging ser- time lag of perceived response time. Nevertheless, Members –
vice Twitter are challenging the existing information para- or their staff on behalf of the representative – have rules in
digm. First year of tweets posted by Members of Congress place to respond to inquiries within 24 hours. Most of the
in combination with qualitative interviews with congressio- times they are responding with a direct phone call to avoid
nal offices show that the Members are mainly using Twitter leaving a digital trace of a response.
to complement their existing push communication style and 2 CoWith the advent public us networking services the question
ommunicat ting with th of social new technologies
he sing
automatically distribute vetted content via Twitter, using is now why and how are Members adapting to highly inter-
Member of Congr
rs ress in the U govern
U.S. nment are kknown for ttheir conser
rvative
the Microblogging service as an additional communication activenew inf innovative forms of tec
use of andformation a commuand unication online communication, fically
chnologies in general, and speci and
,
channel for their individual appearances and issues. The specifically Internet te they to represen themselve online orsuch as the
usin advanced how do
ng d echnologies use fast-pace services r to commu
nt es unicate
awareness network among tweeting Members specifically with their const
h tituents (Est
terling, Lazer, & Neblo 2012). Th use thei websites m
o, hey
micro-blogging service Twitter to push infor out to tand remov the
ir
to reach rmation out and interact
mostly
as a static, non
n-interactive informati tool
ion ve
shows that the potential for interactive conversations are with their ter a 10-day press online.
information aftconstituents rele
y ease period to be able to control current meessage.
not harnessed. Finally, Twitter’s potential as an innovative Mem mbers of Co ongress are t
therefore offtentimes cr
riticized for their slow a
adoption and time
lag of perceived response time. Neve in Congress - or lagging on behalf ofthe
Technology adoption ertheless, M Members is their staff behind the
f
mode for future democratizing interactions is discussed. general use have technologyo in otherinquiries w organizations. the
resentative - of rules in place to respond to types within 24 h
repr s o of hours. Most of In
the mid-1990 s, Membersect phone c to avoid leavingstarted trac use
time they are responding with a dire
resp
es
ponse.
of Congress slowly a digital to of a
call d ce
emailWith additionofto paper-based vices the qu they sent and how
in the advent f social netw
e working serv letters that is now why out to
uestion
1 The role of Members of Congress in the U.S. American their district. Nevertheless, interactiv andthenovative f letteronline
are Members adapting to highly they kept ve innstandard forms of for-
commmunication and specifically how do they us fast-pace services su as the m
n, w se e uch micro-
System mat and emails still includeand intera with thei constituen online. letter
gging servic Twitter to reach out t the traditional paper-based
blog ce o to act ir nts
headsTechnollogy adoptio in Congr is laggiing behind tthe general u of techn
as the following graphic shows:
on ress use nology
in otther types o organizati
of ions. In the mid-1990s, Members o Congress slowly star to
, of s rted
The U.S. federal government is comprised of three branches:
Figure 1: Old-style letter paper-bas letters that communication with consti-
use email in addition to heads used in Email they sent out to their d
o sed district.
Besides the executive and judicial branch, the legislative tuents (Schreiber, 2012)standard l
Nevvertheless, th kept the
hey e letter forma and email still inclu the traditional
at ls ude
branch includes the House of Representatives and the Senate. papeer-based lett heads as the followi graphic shows:
ter s ing
Among other tasks Congress is responsible for rule and law
making to properly execute power.
The House of Representatives consists of 435 voting mem-
bers. Each Member of Congress represents a congressional
district and serves a 2-year term. In addition two senators, for
a total of 100 senators, represent each state in the House. The
representatives are voting on behalf of the citizens in their local
district, work on committees to prepare legislation and stay in
close contact with their constituents. The communication
needs focus on interactions with professional associations, Figu 1: Old
ure d-style lette heads u
er used in Em
mail comm
munication w
with consti
ituents
Official 2012)
hreiber, websites as a means of online representation for indi-
(Sch
meeting with visitors from the district on Capitol Hill and
vidual Members of Congress did not start until the early
generally responding to inquiries from citizens.
2000 s. Generally, the degree of innovativeness, such as the
number of interactive elements to reach the member or for the
public to interact with the website, are widely differing among
Members. Some Members are highly innovative, as the exam-
108 ZPB 3/2012

2. “Connecting to Congress”: The use of Twitter by Members of Congress – Mergel | Aufsatz
ple of Mike Honda shows who represents the 15th congres- would go on the envelope and can then be send for free to the
sional district, California, who crowdsourced the design of his constituents (U.S. Senate Select Committee on Ethics, 2012).
website online and asked his constituents to decide about the However, the existing rules are regulating paper mail only,
type and position of the interactive elements (see: http:// and were not extended to include the changing technology
honda.house.gov/). Remarkably, Honda’s district includes Sil- landscape, such as email or Internet use and therefore made it
icon Valley, the core of America’s high-tech industry. Other difficult for staff members to decide how to apply the rules to
websites are clearly “hand-made” by non-professional staffers newer forms of electronic communication by Members of
and still provide very rudimentary services, mainly plain text, Congress (Lazer et al., 2011).
no interactive elements, or 80s-style animated comic gifs. It was not until 2008, when the Senate updated the Frank-
Diffusion of innovative practices of online representation ing rules to allow Members of Congress to post content on
and interaction is clearly driven by outside vendors, but not and from third party websites on their own website (Miller,
necessarily by one specific thought leader among the offices 2008). At this point, offices were allowed to use their own
(Lazer, Mergel, Ziniel, Esterling, & Neblo, 2011): Over time, YouTube videos or videos of third parties, such as TV stations,
a convergence of online practices is observable. Members are posted online to integrate into their own websites (Yehle,
using a handful of IT vendors on the Hill who are reusing 2008). It took another year, until summer 2009, when Mem-
online templates over and over again, changing only color, bers of Congress started to use the micro-blogging service
content and sometimes the positions of interactive elements. Twitter.com to reach out to the public in an even more inno-
Previous research has shown that Members of Congress vative way with very little prior experience or comparable ini-
are not making optimal use of the potential of new forms of tiatives in other corners of the federal government.
digital technologies. As Esterling et al. show, incumbents show
considerable path dependence in their Web site technology 2.2 Microblogging using Twitter.com
adoptions, while the sites of the freshmen are largely indepen- Twitter is a form of microblogging that allows users to write
dent of the Web designs of their corresponding predecessors short online text updates. The service is used for 140-charac-
(Esterling, Lazer, & Neblo, 2011; Esterling et al., 2012). Rep- ter-long updates that can point a user to other rich media con-
resentatives are learning political practices from each other, tent on a government organization’s website. The service is
but are not learning and enhancing their own website practices often used to interact with the public. Public-sector applica-
to adhere to the changing technology landscape. tions include, for example, the active distribution of mission-
The use of advanced and interactive social networking ser- relevant information, information searches, emergency alerts,
vices, such as blogs, YouTube channels, and Facebook pages and public diplomacy efforts. A more indirect, almost passive,
to communicate with constituents started in the late 2000 s way to use Twitter includes the participation or observing of
(2009/2010) with a 5-10 year time lag in comparison to other ongoing issue conversations for government organizations to
types of organizations. understand how public policy issues are currently being dis-
The potential effects of using highly interactive social net- cussed online (Mergel, 2012).
working elements to supplement the campaign brochure style Similar to users on Facebook and other social media plat-
of their current websites can provide Members of Congress forms, Twitter users set up personal accounts and follow the
with the following opportunities: updates of other users. The asymmetric follower model means
(1) The tools can enhance the ability of Members of that contact requests do not have to be confirmed; the result
Congress to fulfill their representational duties by providing is that users may have a high number of followers compared
greater opportunities for communication between the Mem- to the numbers of accounts they themselves are following.
ber and individual constituents; The core of the service is the news feed that automatically
(2) They have the potential to support the fundamental displays updates from those whom an account holder actively
democratic role of spreading information about public policy chooses to follow, as the screenshot in figure 1 shows. Many
and government operations; and users combine Twitter updates with other social media
(3) They provide Members with the ability to easily com- accounts and automatically post updates to their news feed
municate information traditionally sent to the district or state from Facebook, blogs, or other content-sharing sites, such as
only, also to non-constituents and thereby broadening the Flickr or YouTube. The rich-media sharing function allows
reach beyond the local district. Twitter updates to extend the character limit of 140 words;
and pictures, links to websites, or videos can be embedded in
2.1 Restrictions and challenges for the use of innovative ICTs a Twitter update, so that readers are directed to longer ver-
All types of communication between Members of Congress sions and texts outside of Twitter, for example on a govern-
and the public are regulated by the “Franking rules”. The rules ment agency’s website.
were designed in 1789 to provide guidance on how physical
2.3 Drivers for the use of Twitter
mailings have to be designed, printed, and mailed at the tax-
payers’ expenses. As an example, the signature or stamp with Traditionally, formal press releases, memos, or other news are
the Member’s name is placed at the position where a stamp added by a web manager or public affairs director to a Mem-
ZPB 3/2012 109

3. Aufsatz | Mergel – “Connecting to Congress”: The use of Twitter by Members of Congress
ber’s website. While there are a few mechanisms, such as mail- help users to access their data, given that Twitter only displays
ing lists to direct the messages to the right audiences, most of the last two weeks of updates on its website.
the online traffic is anonymous. Beyond hits on the website, Twitter has the potential for “fast-and-furious” bi-direc-
there is relatively little insight available into the web audience. tional exchanges with individual constituents in real time.
Social media channels like Twitter, on the other hand, Moreover, the publicness of the tool allows access to the con-
allow for the reuse of messages, for redistribution of official versations because all exchanges are publicly observable even
content in a snowballing mechanism through each user’s net- by non-constituents and collapse many different online audi-
work and through a diverse set of social media platforms. ences into one news stream (Grudz, Wellman, & Takehteyey,
Once a message is posted, it can potentially reach unlimited 2011; Marwick & boyd, 2011; The Economist, 2010). Online
numbers of citizens. Messages can automatically be reposted exchanges allow Members of Congress to gauge the “temper-
to other social media channels, such as Facebook. Moreover, ature” among their constituents and to understand their sen-
interactions can easily occur in a bi-directional, reciprocated timents towards a specific policy issue (Thelwall, Buckley, &
manner. Twitter is therefore opening possibilities for interac- Paltoglou, 2011).
tive exchanges that traditional websites currently do not Why should a Member of Congress jump on the Twitter
allow. bandwagon? Who is the audience? It is generally not the
Among the overall top trending topics of 2010 were two “American public;” instead, each Member has very specific
in which the U.S. government was heavily involved: the BP constituents in his or her local district and choosing the right
Gulf oil spill and the Haiti earthquake. In 2011, the top 10 tool should therefore follow the preferences of the local audi-
trending topics worldwide included the Japan earthquake and ence.
tsunami, the Libyan conflict, Egyptian protests, or Bin Laden
updates. 3 Methodology
In the past three years, Twitter has grown significantly to
The focus of this article is on Members of Congress and their
over 500 million registered accounts in early 2012. News
innovative use of a new information and communication tech-
organizations, corporations, and more recently, government
nology. Twitter.com challenges governments existing infor-
agencies adopted this trend. Most Members of Congress now
mation paradigm. The traditional press release paradigm
maintain at least one Twitter account — some even manage
includes a meticulous process through which public affairs
multiple accounts, based on their operational needs and their
officers, press secretaries, or web managers have to go before
diverse audiences.
the final approved and vetted short updates are officially
The use of social media platforms and specifically Twitter
released to the press and the Members’ websites (Lazer et al.,
has expanded significantly in the last two years (Blanchard,
2011). For most Members of Congress adopting a third party
2012; Congressional Management Foundation, 2012; Lux
platform such as Twitter or any other social media application
Wigand, 2011). Goldbeck et al. showed in their study of the
hosted outside of the congressional ICT infrastructure there-
200 most recent tweets collected in February 2009, that Mem-
fore constitutes a departure from the existing routines of com-
bers of Congress are using Twitter primarily to disperse infor-
munication with the public, professional groups, or the press.
mation, especially links to news articles about themselves and
Previous waves of online interactions, such as emails and
to their own blog posts or reports of their daily activities
congressional websites have shown that Congress is an espe-
(2010). The Congressional Research Service studied a 61 days
cially slow adopter of innovative technologies. Each office
time period of the 111th Congress and found that the fre-
constitutes and individual organizational unit with its own
quency of tweets while Congress is in session is higher than
mission and audience. Even though all offices are combined
during recess (2010). Gulati and Williams determined in their
within the larger organization of Congress, they are divided
study that party affiliation and available campaign resources
by party lines, and their focus is inherently on the local polit-
determine early adoption of Twitter: Republican Members of
ical landscape in their own congressional district within a spe-
Congress are more likely to adopt Twitter as a new form to
cific state (Salisbury & Shepsle, 1981).
communicate with constituents than Democrats (2010).
Innovations in these independent subunits are therefore
On April 14, 2010, the Library of Congress announced
driven by local and not necessarily joint national priorities and
that it had acquired the entire Twitter archive—a step forward
budgets. Local dependencies vary by district and rural districts
in reducing some of the hesitation social media directors, espe-
show different levels of broadband diffusion, internet access
cially in the federal government, were facing (Library of
or social media literacy.
Congress, 2010). Up to that point, it was unclear how to keep
public records of Twitter messages—or any messages created Qualitative and quantitative data collection and analysis
on social networking services. The Library of Congress’ col- In collaboration with the Congressional Management Foun-
laboration with the microblogging site Twitter.com now cre- dation, a representative sample of 25 Members of Congress
ates a lifetime archive of all Twitter updates ever sent, but it was drawn to include social media innovators, slow adopters,
does not necessarily relieve government agencies of the and laggards who will most likely never use Twitter. Among
responsibility to archive their own records. It does, however, the 25, 20 can be characterized as early adopters of social
110 ZPB 3/2012

4. “Connecting to Congress”: The use of Twitter by Members of Congress – Mergel | Aufsatz
media application in Congress, two offices just started to use specific analysis to focus only Member-to-Member interac-
Twitter, and three Members did have a website, but had not tions. The Member-by-Member matrix is used as an indicator
adopted social media applications. The Congressional Man- to analyze if and how Members are using Twitter as an inter-
agement Foundation (CMF.org), a small nonprofit organiza- active tool to discuss policy issues. The network representa-
tion helping Members of Congress to manage their offices tion was analyzed and visualized using the social network
effectively, provided initial support in the recruitment of inter- analytical software Ucinet and Netdraw (Borgatti, Everett, &
view partners. While the sample selection clearly oversamples Freeman, 2002). Each node was color coded using a Member’s
on the side of innovators and early adopters, non-adopters party affiliation.
provided important insights to understand the reasons why The main research questions guiding the multiple data
congressional offices might not adopt Twitter. analyses steps are to understand why and how are Members
Semi-structured interviews were conducted with 25 com- of Congress using the microblogging service Twitter. What are
munication or new media directors of the preselected congres- the perceived benefits for early adopters? What are the main
sional offices. The interviews lasted about 45-90 minutes, challenges they are experiencing and how are they adminis-
were recorded with the permission of the interview partners, tering the accounts and updates? For those offices actively
and transcribed verbatim. Each interview was hand-coded using Twitter, what is the main content Members are sharing
line-by-line in an iterative processes going back and forth bet- and do they actively engage in online interactions?
ween the data and the existing literature using the qualitative
data analysis software QSR NVivo9 (2011). Using a 4 Findings
grounded-theory approach main themes were extracted from
the data (Glaser & Strauss, 1967). Each website of the participating interview partners was coded
to understand the extent to which they promote the use of
Parallel to the qualitative data collection efforts, each
different forms of online media and products. In addition, the
respondent’s website was coded for the type and quantity of
interview partners reported the use of ICT and interaction
traditional and interactive online components. The data was
practices that are not publicly observable directly on their
complemented with findings from the interviews to gain an
website, such as Tele-Townhalls or individual phone calls.
understanding of the number and quality of channels Mem-
bers of Congress use to interact with their constituents. The interviews show that Members of Congress use a wide
variety of tools and mechanisms to publish their content,
The quantitative data collection included multiple sources:
including traditional media mechanisms and a variety of social
All tweets Members of Congress have sent in their first year
media platforms. The mix and match of ICTs reported by the
of Twitter use were downloaded. The final database included
interview partners is shown in the following table 1:
16,397 tweets sent by all existing 144 congressional Twitter
handles between their first day of use as early as late 2008 until
the December 31, 2009. Each tweet was then hand-coded and Table 1: Mix and match of online communication tools used in
Congress
sorted into emerging categories. The author coded tweets with
the support of two graduate students and compared the extent
Traditional media tools Social media tools
of overlaps in a shared set of tweets. Differences in coding were Website Website incl. social media tools
discussed and in a second round of coding confidence in the Press releases – Photosharing (Photobucket, Flickr,
Mailings etc.)
codes increased, so that intercoder reliability was increased Emails – Twitter
(Kurasaki, 2000). E-Newsletters – Facebook
TV (CSPAN, national news networks,
Blogs & RSS feeds
The major categories that emerged include professional vs. district/local channels)
Radio YouTube Channel with official TV
private content of tweets, individual issues Members are pro- Phone coverage, unofficial footage and other
moting, and Members’ public appearances. Each account was – Individual phone calls types of online videosharing services,
– (In person) town halls such as Vimeo
coded based on the number of followers and number of – Tele-town halls Podcasts
accounts followed by the Member, and the number of tweets (Group) visits in Washington Internet town-hall meetings
sent.
In addition to the quantitative coding of each tweet, the @- 4.1 Use of Twitter
mentions in each tweet were captured. This means that every
Members of Congress have set up 144 Twitter handles at the
time a tweet included another Twitter handle, the name was
time of the data collection. This constitutes ~ 32% of the total
extracted and a 2-mode network diagram was created
number of 441 Members.1 During the first year of Twitter use
(Wasserman & Faust, 1994). The resulting mxn matrix was
in Congress, Members or their press secretaries have created
then converted into a symmetric nxn matrix which includes
the Twitter handles from Members of Congress and all their
1 This number has significantly increased especially in 2012, but those
corresponding communication partners. Twitter mentions Twitter handles were not included in the data collection. Data collec-
were not symmetrized in order to conserve the directions of tion has been restricted by Twitter and automatic data collection is
restricted by the Twitter API. It does not allow researchers to easily
the conversations. In a second step, all those Twitter handles download data anymore and requires an application process with the
from outside of Congress were omitted for the purposes of this company that is rarely accepted.
ZPB 3/2012 111

5. Aufsatz | Mergel – “Connecting to Congress”: The use of Twitter by Members of Congress
16,397 tweets. The MOCs number of Twitter followers varies mated and pushed out through social networking sites as an
widely: On average, MOCs have ~ 2.3 total followers. In the add-on. Important mechanisms that are making Twitter an
sample collected for the first year of tweets, 97 Members of interactive tool to reach specific subgroups and participate in
the Republican Party and only 45 Democrats in Congress have quick online conversations are not utilized – leaving Twitter
set up Twitter accounts to interact with the public. as just one of the push channels to educate the public.
Non-adopters, those offices included in this study who
Reasons for using Twitter
have not started to use Twitter, are highly reluctant and skep-
The use of Twitter is generally seen as challenging. The
tical of the usefulness of online conversations via Twitter.
absence of clear guidance for the appropriate use of social
Some offices are thinking about the future use of Twitter: “We
networking services has left many offices in the dark, as the
haven’t dove into the Twitter world yet. […] It’s something
following statement shows: “[There is] No policy that tells me
we have talked about and just haven’t gone forward with
I should (not) do this.” As a result not all offices have used
yet.” Other offices are making a conscious decision and are
their Twitter account – even though the majority has set up
arguing that according to their research, their own district is
the @handle to secure the name for future purposes.
not tech-savvy enough to use Twitter or their constituents are
Those offices that have set up accounts use Twitter for a
not interested in participating in online conversations, as the
variety of reasons. The most prominent response points to the
following statement shows: “Like Twitter for example, was
necessity to be able to contact constituents on all available
something that we’ve decided, doesn’t have particular useful-
online channels and be present where constituents are inter-
ness in our district. […] Twitter is useful in a certain context,
acting with each other. As one office puts it, Members of
[…] for official purposes, it doesn’t seem to be something that
Congress need to “Reach people where they are”. While not
would be utilized by our member.“
all constituents are reachable on Twitter, it became clear in
the interviews that the majority of Members who are using
4.3 Content of tweets
Twitter, understand that a part of their constituents are reach-
able through this medium who are otherwise not involved in Members of Congress use their Twitter accounts mainly for
democratic processes. The following statement highlights the professional purposes (~ 41% of all tweets were coded as pro-
necessity to reach those voters who are otherwise not reach- fessional), and only 3% of the updates Members tweeted
able or are not willing to interact with their representative focused on personal issues, such as family members, personal
using traditional means: “It’s a nice way to get your message taste in a specific sports team, etc. Most of the professional
out to a different crowd. […] and Twitter is a nice way to reach tweets focus on issues a Member is passionate about, chairs
out, so we wanted to add that link onto our website there.” or participates in committees. A prominent topic in all tweets
– independent from other policy issues Members of Congress’
4.2 Administrative routines in the Washington offices of are usually tweeting about – is the health care reform with
Members of Congress ~15% of all tweets.
Besides policy issues tweets also reveal the target audience
Similar to other forms of interactions, Members of Congress of MoCs tweets: The second most tweeted content points
are most of the time not involved in the direct exchanges with Twitter followers to the Member’s appearance in his or dis-
their constituents on Twitter. Instead, a lot of the communi- trict, clearly targeting the local constituency and not Wash-
cation efforts are distributed among the staff members in the ington’s media audience. In more than 27% of all tweets
Washington office and in the district’s office. Especially when MoCs shared locations of public appearances, TV interviews,
it comes to routine tasks, such as updating the website with a and online media with coverage about the Member’s appear-
press release or a video of a TV appearance, the office staff ance in the district.
has set up routines to distribute the updates through all avail- Overall, Twitter is largely used to facilitate one-way trans-
able channels, including Twitter, as the following response mission of information from Members’ offices to the public.
shows: “It’s really a collaborative effort. So he’ll be out and MoC use Twitter to convey information about their official
about in the district, and he’ll say: ‘Have [staffer’s name] actions, media appearances, or policy positions, and in a one-
Twitter about this.’ So really, I’m the one that is in control of directional push tactic. Only 3.7% of all tweets counted and
the Twitter. He just tells me some different things that might analyzed were direct replies to others, indicating at least a
be interesting to put on there.” This implies that most Mem- response to a question or statement. The remaining 10% of
bers are not actively tweeting by themselves, instead they are tweets were not directly identifiable as either professional or
deferring the task to their staff members and as a result to the private and did not fit into the other issue codes.
question one staffer politely responds: “He’s “involved” with
it.”
4.4 Bidirectional conversations to create awareness for issues
Other offices do not create content tailored for the use of
Twitter updates at all: “We have automatic blog and Twitter A subset of messages were extracted that are mentioning other
updates.“ As a result, content curated and vetted for general MOCs Twitter handles. This subset was coded every time a
publication through all other ICT channels is simply auto- tweet included an @mentions in the 12-months of Twitter
112 ZPB 3/2012

6. “Connecting to Congress”: The use of Twitter by Members of Congress – Mergel | Aufsatz
messages posted by MOCs. Those are messages that are either early adopters see more constraints than opportunities in the
directly addressing other Members or passively attacking a use of Twitter. Instead of embracing the media richness poten-
Member from the opposite party. The result is an attention tial of Twitter as an interactive mode of communication the
network that reduces the total number of 144 congressional potential is largely ignored and Members are mostly using the
Twitter handles to only 39 MOCs who are willing to veer tool as a “me-machine”, pushing out information about them-
away from their rather static push tactic that only focused on selves (Naaman, Boase, & Lai, 2010).
their own promotion, to a potentially conversational Neither a significant observable number of bi-directional
approach with other Members. interactions or real-time exchanges occur. Twitter is used as a
Of the 39 Members included in the resulting attention net- push mechanism to advertise personal appearances or
work, only 16 are directly included in @-replies. Most promi- announce voting preferences and standpoints on political
nently, the GOP leader on the republican side receives issues, such as the healthcare reform. The potential to create
responses to his tweets. Remarkably, there are however no issue networks and online communities around specific policy
loops in the network. This means that even though Members issues is not used so far and Members are also not involved in
are responding or retweeting messages from each other, they so-called fast-and-furious exchanges with their constituents or
do not actively keep a conversation going or are willing to other Members of Congress.
publicly start longer back-and-forth conversations. Future research needs to explore online behavior beyond
The following network diagram shows the attention net- the publication process of MOCs themselves. Instead it is
work among Members of Congress who are responding to imperative to understand to what extent new ICTs have a
each other’s tweets. Nodes colored in blue represent Members democratizing impact. How can Members of Congress include
of the democratic party, and nodes colored in red represent those online audiences that are otherwise excluded from the
republican membership. The arrows indicate the direction of political deliberation processes beyond using Twitter as a mere
the tweet: as an example @timryan mentioned @RepSteveIs- broadcasting mechanism (Marlin-Bennett, 2011)? To what
rael, but do not activelly keep a cdid notnrespond. This is to publicly start extent do constituents feel that their representatives are truly
othe they @RepSteveIsrealconversation going or are willing the practice
er, y
for almostd-forthtweets collected in this attention network and
long back-and all conv
ger versations. more efficient and effective in communicating with them?
The foll
lowing netw
work diagram shows th attention network am
m he mong Memb of
Twitterwho arethereforetoused other’s twcollectedcolored in blue reptime-
bers Does engagement on Twitter result in higher trust in govern-
Conngress was respond ding eac in theweets. Nodes dataset and present
ch
frame onlyhe democra party, and nodesone-way conversations.
Memmbers of th for one-directional ors colored i red repr
atic in resent repu
ublican ment operations and information constituents receive from
mem
mbership. T arrows indicate the direction of the twee as an ex
The e et: xample @timryan their representatives? Tools such as Twitter or Facebook
Figure 2: Attention network@R
menntioned @R
RepSteveIsra but RepSteveIsre did not respond. Th is the pr
ael, eal
among Members of Congress his ractice
therefore still need to show to what extent they can be labeled
for a
almost all tw
weets collec in this a
cted attention ne
etwork and TTwitter was therefore u
s used in
the c
collected daataset and tim
meframe on for one-d
nly directional o one-way conversatio
or ons. as politically transformative ICTs (Hong & Nadler, 2011).
Moreover, research similar to the new Twitter index (see
election.twitter.com) is needed to understand the sentiments
among voters for each district or specific policy areas. Seman-
tic text analysis can help to understand better how con-
stituents perceive their representative’s online interactions
with the public.
Implications for political consulting
Citizens increasingly use innovative and highly interactive
ICTs such as Twitter.com. The tool has become a platform for
the discussion of policy issues and hot button issues rise to the
top. Political appointees and elected officials are oftentimes
criticized for their slow adoption of social media and as shown
in this article for their relatively reluctant use of all the func-
tionalities. While Twitter was designed to be a public conver-
Figu 1: Attent
ure tion networ among M
rk Members of C
Congress
sation platform, behavior of users is quickly changing and
5 Discussion memes as well as behavioral patterns evolve and disappear
iscussion
5 Di
after a while. Ultimately, Twitter should be used in a way an
Why Why an how do p political elites Member as CongreMembers of
and nd how do political elite such as th such rs of the ess use inno
es he ovative elected official feels comfortable with. Changing his or her
Congress cuse innovativeucratic and that are challenging ? their
ICTs that are challenging their bureau ICTs learned com mmunicatio routines? In an
on online personalities based on the expectations of a perceived
appaarent “vacuuum” of rule and conf
es fronted with rising pres
h ssure from their consti
ituents
bureaucratic and ress have sttarted to ex
man Members of Congr learned communication Twitte to extend their
ny s xperiment w with routines? In an
er d public will result in “fake” interactions that don’t reflect the
apparent “vacuum” withrules and confronted withpart of popres-
existing reach and intera of au
act udiences tha are othe
at erwise not rising olitical actual preferences of a politician or public manager.
ractions. Th findings have shown that even those high innovativ Member who
inter he n n hly ve rs
sure from their constituents many Members of Congress have Nevertheless, innovative forms of transferring already
started to experiment with Twitter to extend their existing accepted interactions into the online sphere will help members
reach and interact with audiences that are otherwise not part to reach those audiences that they are otherwise not able to
of political interactions. The findings have shown that even reach and are only receiving their news and updates on social
those highly innovative Members who are characterized as media platforms such as Twitter or Facebook. As an example,
ZPB 3/2012 113

7. Aufsatz | Mergel – “Connecting to Congress”: The use of Twitter by Members of Congress
Barack Obama has started to use the platform to host Twitter Lux Wigand, F. D. (2011). Tweets and retweets: Twitter takes wing in
government. Information Polity: The International Journal of Govern-
Townhall meetings. Questions from the public were collected ment & Democracy in the Information Age, 16(3), 215-224.
for a period of time using the hashtag #AskObama. The Pres-
Marlin-Bennett, R. (2011). I hear America Tweeting and Other Themes
ident’s team then picked questions the President responded to for a Virtual Polis: Rethinking Democracy in the Global InfoTech Age.
live and ignored other questions. Journal of Information Technology & Politics, 8(2), 129-145.
Marwick, A. E., & boyd, d. (2011). I tweet honestly, I tweet passionately:
It is also important for political consultants to understand
Twitter users, context collapse, and the imagined audience. New Media
that not all districts or all parts of a Member’s audience are & Society, 13(1), 114-133.
reachable on Twitter. Advice should therefore be focused on Mergel, I. (2012). Working the Nework: A Manager's Guide for Using
a detailed analysis of the demographics in each individual dis- Twitter in the Public Sector. In I. C. f. t. B. o. Government (Ed.), Using
trict, broadband diffusion, age groups, etc. before an elected Technology. Washington, D.C.
official starts to use a platform that might end up not having Miller, E. (2008). Congress can Tweet, Following them with Capitol
Tweets Widget. Retrieved from http://sunlightfoundation.com/blog/
an audience and will in turn be criticized. 2008/10/06/congress-can-tweet-follow-them-with-capitol-tweets-widget/
Naaman, M., Boase, J., & Lai, C.-H. (2010). Is it really about me?: mes-
References sage content in social awareness streams. Paper presented at the CSCW
'10 Proceedings of the 2010 ACM conference on Computer supported
Blanchard, B. (2012). Congressional use of social media platforms cooperative work Savannah, Georgia, USA.
expands. The Daily Texan. Retrieved from The Daily Texan website on NVivo 9. (2011). NVivo qualitative data analysis software (Version 9).
2012/07/08/: www.dailytexanonline.com/university/2012/07/08/con- Melbourne: QSR International Pty Ltd.
gressional-use-social-media-platforms-expands
Salisbury, R. H., & Shepsle, K. A. (1981). Congressional staff turnover
Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for Win- and the ties-that-bind. American Political Science Review, 75(2),
dows: Software for Social Network Analysis (Version 6.39). Harvard: 381-396.
Analytic Technologies.
Schreiber, P. (2012). Message from Representative Nancy Pelosi, from
Congressional Management Foundation. (2012). SocialCongress: Percep- http://paulschreiber.com/wp-content/uploads/2009/09/Screen-
tions and Use of Social Media on Capitol Hill. In C. M. Foundation (Ed.). shot-2009-08-27-at-12.06.47-AM.png
Washington, D.C.: CongerssFoundation.org.
The Economist. (2010). Twitter's transmitters. The Economist, A special
Esterling, K., Lazer, D., & Neblo, M. (2011). Representative Communi- report on social networking, 5.
cation: Web Site Interactivity and Distributional Path Dependence in the
U.S. Congress. Political Communication, 28(4), 409-439. Thelwall, M., Buckley, K., & Paltoglou, G. (2011). Sentiment in Twitter
Events. Journal of the American Society for Information Science and
Esterling, K., Lazer, D., & Neblo, M. (2012). Connecting to Constituents: Technology 62 (2), 406-418.
The Diffusion of Representation Practices among Congressional Web-
sites. Political Research Quarterly. (Forthcoming) U.S. Senate Select Committee on Ethics. (2012). Franking, from
www.ethics.senate.gov/public/index.cfm/franking
Glaser, B. G., & Strauss, A. L. (1967). The Discovery of Grounded
Theory: Strategies for Qualitative Research. New York: Aldine Publis- Wasserman, S., & Faust, K. (1994). Social network analysis: methods and
hing. applications. Cambridge: Cambridge University Press.
Glassman, M. E., Strauss, J. R., & Shogan, C. J. (2010). Social Networ- Yehle, E. (2008). Senate Updates Web Rules to Allow YouTube Videos.
king and Constituent Communications: Member Use of Twitter During a Roll Call – The Newspaer of Capitol Hill Since 1955, Sept. 24, 2008.
Two-Month Period in the 111th Congrss. In C. R. Service (Ed.), CRS
Report for Congress (Vol. 7-5700). Washington, D.C.: Congressional
Research Service.
Goldbeck, J., Grimes, J. M., & Rogers, A. (2010). Twitter Use by the U.S.
Congress. Journal of the American Society for Information Science and
Dr. oec. HSG Ines Mergel is an Assis-
Technology, 61(8), 1612–1621.
tant Professor of Public Administra-
Grudz, A., Wellman, B., & Takehteyey, Y. (2011). Imagining Twitter as
an Imagined Community. American Behavioral Scientist, 55(10), 1294– tion and International Affairs at
1318. Syracuse University’s Maxwell
Gulati, J., & Williams, C. B. (2010). Communicating with Constituents School of Citizenship and Public
in 140 Characters or Less: Twitter and the Diffusion of Technology Inno- Affairs. In her research she focuses
vation in the United States Congress. Retrieved from http://ssrn.com/
abstract=1628247.
on the adoption of social media app-
lications in the public sector. She is
Hong, S., & Nadler, D. (2011). Does the Early Bird Move the Polls? The
Use of the Social Media Tool 'Twitter' by U.S. Politicians and its Impact the author of the forthcoming
on Public Opinion. Paper presented at the Dg.o'11, College Park, MD, book: “Social Media in the Public
USA.
Sector: A Guide to Participation, Col-
Kurasaki, K. S. (2000). Intercoder Reliability for Validating Conclusions laboration, and Transparency in the
Drawn from Open-Ended Interview Data. Field Methods, 12(3), 179-194.
Networked World”. Her thoughts on
Lazer, D., Mergel, I., Ziniel, C., Esterling, K. M., & Neblo, M. A. (2011). the development of innovative ICTs
The multiple institutional logics of innovation. International Public
Management Journal, 14(3), 311-340. can be read on her blog: http://ines-
Library of Congress. (2010). Twitter Donates Entire Tweet Archive to
mergel.wordpress.com or on Twitter
Library of Congress Retrieved 11/16/2010, from www.loc.gov/today/pr/ @inesmergel. E-mail: iamer-
2010/10-081.html gel@maxwell.syr.edu
114 ZPB 3/2012