A Content Analysis of Campaign Literature

With all the attention given to the Internet and social media and TV it is easy to overlook the role that printed materials play in a campaign.  Literature ranging from brochures and palm cards to detailed plans and position papers to slick mailings helps spread the campaign's message.
Governor and U.S. Senate
Presidential Campaigns:
Primaries  |  General Election.
Primaries Dem. / Rep.  |  General Election.
Primaries and Iowa mail. |  General Election.
2008: Primaries Dem.Rep.
Revised Apr. 13, 2022.  An early edition of this page is at: http://www.p2012.org/comms/campaignlittext.html
See also: Campaign Managers.

A Content Analysis of Campaign Literature
Eric M. Appleman
This is adapted from an old paper I wrote looking at literature in U.S. gubernatorial, Senate and House campaigns.

Campaigns produce a wide assortment of printed campaign materials.  This article focuses on the introductory brochures, palm cards and flyers that summarize the image and appeal that a candidate is trying to make.  In a sense these are the "introductory handshake" of campaign literature.  Carefully selected pictures, choice adjectives, biography and career information, positions, and sometimes a "personal" note to the reader are combined to present the candidate in the best possible light.  For all the attention a campaign may devote to designing a brochure, anecdotal evidence suggests voters typically just take look at this material, scanning through it and glancing at the photos and headers. This paper presents thoughts on the analysis of the images and text in introductory campaign literature and on possible research questions based on my observations of this literature over three decades. 

Campaign literature has evolved over the years.  With the advent of the Internet, introductory literature has become less important.  The tri-fold brochure of yore has given way to the palm card which typically presents three to five bullet-pointed issues each supported by one or two carefully worded sentences. 
Campaigns are quick to refer voters to their websites; it is almost obligatory for a lit piece to include the campaign's web site address; Facebook, Twitter and YouTube and Instagram icons and links are frequently included; and recently more campaigns have been including QR codes which the voter can scan to connect with more information.  A few campaigns even pride themselves for doing without printed literature.

Graphic design looks and trends have changed over time.  For example, magazine ads from the 1950s look very different from magazine ads of today.  The same applies to campaign literature.  Photoshop allows designers to easily do cut-outs, layering, and other effects which are far advanced from the basic square or rectangular head shot of the candidate.  Designers have access to huge font libraries and are free to experiment with bold typographical treatment.  The palette of colors and typefaces used in the campaign's printed literature may also be used across the campaign's website, social media and advertising.

The substance of campaign literature differs from cycle to cycle.  Campaign literature from the 2002 cycle—the first cycle after 9/11—focused on a different set of issues than in the 2010 cycle, when the sluggish economy was the dominant concern, which in turn differed from the 2020 cycle when the COVID pandemic was a major preoccupation.  Differences apply not only to text, but to images.  Thus one would expect 2002 literature has more defense and law enforcement-oriented images, while the 2010 material might show more people at work.  Some of the 2020 literature shows the candidate or other people wearing masks. 

Partisan differences provide an obvious area for study.  Republican candidates emphasize different sets of issues than Democrats.  For example, crime, tax cuts and pro-life views have long been Republican themes, while Democrats tend to focus on themes of justice, protecting social programs and choice. 
Discussing the pandemic in the 2020 campaign, Republican and Democratic candidates used different language; Republican candidates were more likely to emphasize keeping the economy open while Democratic candidates stressed following the science.  A big picture review of 2020 Senate literature found health/healthcare was among the top themes for Democratic candidates, while support for the president/President Trump was a top theme for Republican candidates.  It is important to note, however, that the issues highlighted may vary according to the office a candidate is running for.  One expects that the issue sets for U.S. Senate and gubernatorial candidates would differ.
One could also consider regional differences.  Loss of manufacturing jobs has been a major issue in states of the industrial Midwest, while water is a top issue in the West.  There are also design elements specific to particular states.  Literature from New Mexico often has yellows and reds and browns, while Vermont literature often has a green palette.  Colorado literature has photos of the Rocky Mountains and oftentimes the campaign logo may include a representation of mountains.  Wyoming literature often has the graphic of the cowboy on a bucking horse.

There are many other differences that could be considered.  Might there be differences among incumbents and challengers?  Male and female candidates?  Ultimately one would like to understand what makes for an effective lit. piece.  To get at this, one could


Different forms of campaign literature are described briefly here and outlined in the table below.  The detailed plan in booklet form conveys a sense that the candidate has thought through the issues and developed an overall strategy for addressing them. Position papers address isolated issues.  An issue brief summarizes positions on a number of key issues.  A biography presents the candidate's upbringing and career, often on a single page.  A contrast piece compares positions of the candidate with those of his/her opponent.  The brochure, often but not always glossy, is aimed at a general audience.  Tabloid-style newspapers are similar to brochures, but may be cheaper to produce and offer certain advantages with their larger format.  The palm card (or hand card or push card) is a single panel piece that is now the prefered choice of many campaigns.  Low-budget campaigns may opt for a simple xerox flyer, either full or half page.  For other low-budget and long shot candidates a business card may be all the literature there is.  Campaigns may produce different variations of brochures or palm cards targeted at different audiences, for example a version in Spanish, a version targeted at African-Americans, and so forth. 

Door hangers are frequently distributed near Election Day with the campaigns' get out the vote message.  Newsletters report on the campaign's progress and upcoming events; these may be directed to supporters and contributors rather than the general public.  Direct mail, including persuasion mail and get-out-the-vote (GOTV) pieces, is frequently used to target specific segments of the electorate such as seniors or voters in inefficient media markets.  A miscellaneous category includes such items as football schedules, Mrs. Candidate's favorite recipes, etc. emblazoned with the candidate's name.  Finally, campaigns sometimes send out copies of news articles in response to requests for information.

Some candidates do not produce literature.  Incumbents in safe seats may wage minimal campaigns, including no literature.

Not considered here are specialized communications such as faxes directed to the media, press releases and fundraising letters. Note that in some cases literature may not emanate from the campaign but state party or the coordinated campaign.

Table 1. A Classification of Campaign Literature.
 -detailed plan in booklet form
  -position papers
  -issue brief
  -contrast piece
  -glossy brochure
 -tabloid-style newspaper 

  -palm card (hand card or push card)
  -business card
  -direct mail pieces
  -miscellaneous (for example, football schedules, recipes, etc.)
  -news articles

It is important to acknowledge at the outset that today's campaigns are largely fought on television, on the Internet, and by direct mail.  Even back in the pre-Internet days of 1994 one campaign worker stated, "People don't read them."  The images, slogans and issues presented there may be completely different from those in the general candidate brochure.  However, one of the hallmarks of a good campaign is that the various efforts to communicate the message all reinforce and build upon each other.  TV ads and direct mail often focus on negative aspects of an opponent's record or character.  Candidate brochures and web sites provide a more convenient means for studying how the candidates are presenting themselves.  This is an analysis of literature.  It will concentrate on images, slogans, and highlighted issues, assuming that a glance at the photos and a quick look inside may be all the attention a voter commits to a brochure.

Literature Review

A review of the literature reveals a number of studies that can guide a visually-based content analysis of campaign brochures.

Kaid and Davidson (1984) examine political advertising and develop the notion of video style.  They describe television advertising as "pseudo-interpersonal communication."  While these terms do not translate exactly to printed communications, the brochure, which often contains a "personal" note from the candidate, can also be thought of as a pseudo-interpersonal 
communication.  The brochure serves to introduce the candidate. It is the "introductory handshake" of campaign literature, and the presentation and layout of information it contains influence our impressions of the candidate.

Shyles (1984), in his study of images and issues in candidates' TV advertisements, points out that "image" can refer to "the candidate's perceived or projected cluster of personality traits" or to the visual appearance of the candidate.  This study considers both.  Photographs, the second kind of image, help shape our views of the candidate.  Slogans attempt to capture in a few words those traits and image that the campaign wants us to associate with the candidate.

Deutsch (1992) provides an invaluable theoretical perspective for understanding a politician's image.  He identifies three ingredients essential to a politician' success: familiarity, appeasement and power.  The candidate (or officeholder) must present himself as being like the voter, caring for him or her as an individual, and, at the same time being more powerful than the voter and thus able to solve his or her problems.  Campaigns, and the consultants they hire to design brochures, do not have these criteria in mind when they select which images they will use, but they have learned from experience what works (Cabarga, 19--).  Photos of the candidate interacting with ordinary citizens or pictures of citizens without the candidate communicate the idea that the candidate is a regular guy who cares about people (familiarity and appeasement).  Family shots also show the candidate is "like me."  There should be many of these kinds of images. Images of the candidate in an office setting or on the phone imply problem-solving ability and correspond loosely to the power component.

Various studies provide insights into coding print and televised images.  Cutler and Javalgi (1992) compared visuals used in magazine advertisements in the United States and Europe.  Among other factors, they looked at size of the images, use of black and white or color, use of photographs or illustrations, and frequency with which minorities, elderly and children were 

Moriarty and Popovich (1993, 1991) and Moriarty and Garramone (1986) analyzed photographs of the presidential candidates appearing in fall 1992, 1988 and 1984 issues of Time, Newsweek, and U.S. News & World Report.  In their effort to determine whether one or the other of the candidates received more favorable coverage, Moriarty and Popovich considered 15 attributes in three broad categories: behavior (posture, expression ... ), context (props, setting, dress) and perspective (camera angle, position on page).  Whereas the focus of these studies is on possible bias, the candidate images in brochures should be positive, obviating the need for subjective judgements about favorableness or unfavorableness. (An exception is occasional unflattering portrayals of the opponent, but the focus here is self-presentation).

A coding scheme from the Center for Media and Public Affairs to evaluate TV coverage of presidential candidates in 1988 includes two categories for analyzing images—"candidate in context" and "uses of props."  Candidate in context lists 12 elements ranging from "factory/construction/work situation" to "conducting official business."  Eleven items are in the use of props list including hard hats, other hats and the American flag.

A coding instrument devised by the Center for Media and Public Affairs to evaluate TV coverage of the 1996 elections considered more than 300 policy issues grouped in broad areas such as foreign policy and economic policy.  Coding also considered "political issues."  Further the coding considered whether an issue was the "main focus of the story," "extensively discussed," or "just mentioned."  Such considerations also apply to campaign literature where a campaign may just list things the candidate is for or against or may go more in-depth, devoting one or two paragraphs to a particular subject. 

General Considerations

The first consideration is what to include in the sample.  Soliciting literature from campaigns can be a difficult process; some campaigns are helpful and others are nonresponsive.  One does not get contextual information about when the pieces were produced or how widely they were circulated.  As noted at the top, a campaign may be circulating many different types of lit. pieces.  A piece from the primary campaign, used when the campaign is trying to connect with base voters, can have very different content than literature produced for the general election, when the campaign is trying to reach a wider audience.  Yet it is also true that some campaigns produce a piece before the primary and stick with it through to November, particularly in cases of noncompetitive races or limited resources.  In some states the general election campaign lasts for six or seven months and there may be early and final versions of lit. pieces circulating.  The 2020 cycle was a difficult cycle for literature.  Because of COVID, some campaigns eschewed canvassing and did not produce basic lit. pieces.  In these cases I tried to obtain a representative mail piece.  In some cases all that was available was a primary piece, in some cases produced before the outbreak of the pandemic.  The question arises, if one considers these various lit. pieces together, is one mixing apples and oranges? 

A number of independent variables are expected to effect findings of a content analysis.  Most important among these are the office the candidate is seeking, the candidate's party (Democratic, Republican, Independent/ Other), their status as an incumbent or a challenger, and the part of the country they are campaigning in.
The office a candidate is seeking will certainly affect the content of his or her brochure.  Governors are executives and Senators are members of a legislative body; different offices have different constituencies.  The most significant issues in Senate races and gubernatorial contests may well be different, with topics such as welfare and taxes possibly assuming more importance in state contests.  Local issues such as transportation may be prominent in House races.  Governors' races offer another peculiarity which must taken into account.  In a number of states, the candidates for governor and lieutenant governor run as a team, so the brochure may show, or at least mention, both.  Party identification is another variable to be considered; candidates may highlight party identification in their primary literature and then drop it for general election when they are seeking to appeal to the broader public.

There should also be some correlation between a candidate's party and the issues he or she chooses to emphasize. Republicans are stereotyped as being tough on crime while Democrats champion social concerns such as education.  Cynics would argue that there is no difference between the two major parties and that candidates from both parties are driven by polling.

Incumbents will likely highlight their experience and accomplishments in office, while challengers point out the incumbent's shortcomings and promise change.

Finally, regional variations may be evident.  In the West, immigration may be a big issue and imagery can include cowboy hats the Rocky Mountains and such.

A candidate's organization may be experimenting with several pieces in the course of the 
campaign.  In some cases one or two panels or photos in a brochure may be varied in a printing 
run or the brochure may be revised in a later printing.  A campaign may also deploy a range 
of brochures designed for different parts of the state, or for women, seniors, Hispanics or 
African-Americans, or addressing specific issues.


The candidate may be shown with his/her family, interacting with citizens or officials or in a portrait or cut-out image.  There may be other images of the candidate as a youth or growing up.  Some images may not include the candidate, but show children, people at work, scenery, etc.  Just as one can speak of video style to describe a candidate's television advertising, so too different styles of presentation may be discerned in their brochures.  Some lit. pieces include lots of images, others relatively few; in addition to considering each of the images, one may want to consider the proportion of space devoted to images as compared to text or to the entire area of the piece.

Many different coding schemes are possible.  To keep coding decisions simple, I fit the images into four main categories.  In a portrait, the candidate is the only individual in the picture (there may be somebody's elbow or shoulder at the edge, but the candidate is dominant).  Portraits encompass a variety of images ranging from cut-outs, in which information about the candidate's surroundings has been obliterated, to scenes with a dramatic background.  Drawings of the candidate are included in this category. 
 Governor-Lt. Governor images (or Presidential/Vice Presidential candidate) were coded as portraits.  Images showing the candidate as a youth, growing up and in his or her pre-political life are a bit tricky; these are somewhat akin to family snapshots, and could possibly be coded under family.  Interaction shots 
show the candidate interacting with citizens. (The brochure designer could create a portrait by 
cropping out other people in the image). The family shot shows the candidate and his or her 
family, and sometimes even the family dog.  Family shots might include pictures of the candidate as
a youth.  Other images include stock photos of kids, seniors, people at work, a crime scene, the 
Capitol Dome and so forth, as well as photos of the candidate's opponent.  The candidate is not 
present.  Included in the "other" category are wallpaper images, encompassing photos and 
drawings of subjects such as a field of wheat or the Grand Canyon used to fill the background. 

In addition to the main categories above, images were coded for symbolic content.  The context 
or settings, props, and kinds of people shown in the brochure all provide cues about the 
candidate.  An image of the candidate talking with law enforcement officials implies he or she 
is "tough on crime;" a classroom scene conveys concern about education; a scene with workers
denotes jobs and the economy.  One question to consider here is whether or not to code images which do not contain the candidate.  This seems logical; a generic classroom scene "says" 
education even if the candidate is not in the picture.  A picture of a wilderness expresses environment.

We coded for an official context (i.e. an office setting, on the phone, shuffling papers, on the 
steps of the Capitol, near the official seal, etc.), a workplace setting (with hardhats, on the 
factory floor), a classroom or instructional setting, hospital/doctor situations, an agricultural 
setting (with farmers), an environmental setting (fishing, hunting, in the wilderness), the presence of 
patriotic symbols such as an American flag or a American Legion banner, and the presence of 
regional symbols (cowboy hat, state flags). 

For the kinds of people shown in the interaction and 
other images we identified kids (embodying youth, the future), seniors (our heritage), members 
of minority groups, law enforcement officials, and VIPs.  Coding did not distinguish between 
candidates and citizens (sometimes the candidate themselves are doctors, African-Americans ... ).

Not coded are details such as facial expression, whether the candidate is waving his or her arms 
around (Moriarty and Popovich's behavior category) or camera angle, image size and so forth 
(perspective category).  These points are generally considered in studies looking for evidence 
of positive or negative portrayals of individuals.  Candidates' dress (formal/informal), a frequent component of image coding schemes, was not deemed worth analyzing as a separate category.


Many campaign logos are done in the traditional red, white and blue; often there is a star or stars, and sometimes there are wavy stripes.  Various typefaces are used--serif and sans serif--and they are done in different styles, for example, bold or regular, all caps or with lower case letters.  Some logos highlight or include only the candidate's last name, some include both the first and the last name, and others emphasize or include just the first name.  Most include the office the candidate is running for but a few do not.  As noted above, some graphic elements or color palettes are particular to specific states.  A strong logo can help with branding the campaign.  The Obama campaign's 2008 blue O with with the red and white stripes at an angle is instantly recognizable.  Designer Sol Sender states the logo means "the sun rising over the horizon--the dawning of a new day in American politics."  The rising sun logo was incorporated in the 2012 re-elect campaign's logo.  (see 2008 Campaign LogosAs noted above, a lit. piece can include other logos such as for social media as well as for endorsing organizations.

Text: Slogans

Very often a lit. piece will have a slogan across the front, typically from three to seven words in large font designed to summarize the appeal of the candidate.  An analysis could break out just the slogans or look at them as part of the overall text.  Slogans typically include action words (fighting/working/standing 
up ... doing, making), objects including change, jobs, people/working people/families and maybe a more specific 
identification (America, Americans; or in the case of state and local race for Maryland or for Marylanders or for the people of the # District), and qualities such as independence, leadership, experience, common sense integrity and courage.  In 2008 Obama had the memorable slogan "Change you can believe in" (not simply change).  Trump, borrowing from Reagan, used "Make America Great Again" (or simply MAGA) incessantly.

Text: Copy

A piece of campaign literature may cover the candidate's biography and accomplishments; issues and positions; and other details such as how to get involved in the campaign.  It may also consider an opponent's or opponents' record; indeed many of the slick direct mail pieces do only that.  How much depth a piece goes into on particular issues depends on its purpose and intended audience, but generally a palm card or brochure handed out at a campaign rally is not going to have much detail. 

The most basic question one can consider is the number of words on a lit. piece.  Applying wordcounter.net to the texts of 2020 U.S. Senate introductory pieces, I found the Democratic lit. pieces had an average of 271 words and Republican lit. pieces an average of 275 words, but there are a number of caveats. 

In considering the text, I chose not to include campaign logos, which, in addition to the candidate's name and often the office they are running for, can include key words or a slogan.  In some instances a campaign provided a couple of different introductory pieces and I considered both.  However, one campaign provided both a palm card and a widely distributed tabloid newspaper.  In this instance, I only considered the palm card, viewing the tabloid as a outlier which would have skewed the word average.  

I included disclaimers in the word counts, but one could argue that it makes sense to exclude them.  Almost every lit. piece has the "Paid for by..." disclaimer, and there are other disclaimers, most notably for use of photos of a candidate in uniform or warnings about signing up for text messages.  These last two disclaimers can result in word inflation.  For example, the palm card from John James' 2020 Senate race in Michigan had 150 words of which 27 were the disclaimer ("Use of John James' military rank, job titles & photographs in uniform does not imply endorsement by the Department of the Army or the Department of Defense.")  A typical text message disclaimer has about 34 words ("By participating, you agree to the terms and privacy policy for recurring autodialed marketing messages from us to the phone number you provide. No consent required to buy. Message and data rates may apply.") 

In sum, even a most basic content analysis such as tallying the number of words on a lit. piece presents coding challenges.

Ideally one would like to know not just how many words there are, but what they are about.  To do this properly would require a detailed coding instrument such as the one used by the Center for Media and Public Affairs, cited above, as well as rigorous quality assurance to insure that coding is consistent.  One interesting aspect of examining content for issue is how to differentiate between piece which just lists a bunch of things the candidate is for (or against) and one which goes into more detail on several specific issues.  With sufficient resources, one could determine the weight given to the various issue by dividing the words devoted to a specific issue by the total word count.

Considering word frequencies is a poor substitute, but can yield some insights.  Word processing software as well as websites such as wordart.com allow one to determine the most frequently used words in a block of text.  For example, applying wordart.com to the 2020 U.S. Senate sample showed somewhat surprisingly that "protect" was one of the top words across both parties.  Language about protecting Social Security and Medicare and protecting the Second Amendment is to be expected, but looking at the texts one sees many references to "protecting lives," referring to the pandemic.  There are many direct ways to refer to the pandemic—COVID-19, coronavirus, the virus, the crisis—but none of these individual words made the top 15.  In addition, to these specific words, there are also many indirect references ("Governor Parson has worked hard to protect the health and well-being of all Missourians."  From Sununu (NH): "his insistence on being guided by data and the advice of public health officials."  From Forest (NC): "open and rebuild our state's economy").  Simply looking at word frequencies has severe limitations, but in combination with a review of the texts can be of some use.

Big Picture

Beyond analyzing and understanding the content of campaign literature, it is worth considering what makes for an effective lit. piece.  Most people are not going to study in great detail a brochure handed to them at the farmer's market or outside the Metro on their way to work in the morning, but they may at least glance at it.  A good lit. piece, presenting a well-designed combination of words and images, can convey at least some positive information about the candidate.  The limited space afforded by a piece of paper forces the campaign to highight the candidate's priorities, although some of the more slick pieces contain remarkably little about what the candidate stands for (+).  Indeed, a well thought-out black and white photocopied flyer can be more effective than a slick glossy brochure.  This is a digital era, but a campaign that goes all digital and ignores the literature is neglecting a tool that can help it.


Cabarga, Ted. 19--. "On the Art and Science of Designing Political Brochures." San Francisco: Winning Directions.

Center for Media and Public Affairs.  1996.  "The 1996 Elections: Primary Season Codebook." [Unpublished coding scheme for television news coverage of the candidates]. Washington, DC.

Center for Media and Public Affairs. 1988. "Visuals Another Interesting Election Variable." [Unpublished coding scheme for television news coverage of the candidates]. Washington, DC.

Cutler, Bob D., and Rajshekhar G. Javalgi. 1992. "A Cross-Cultural Analysis of the Visual Components of Print Advertising: The United States and the European Community." Journal of Advertising Research 32:71-80.

Deutsch, Robert D. 1992. "Dissecting the TV Image." Nieman Reports 46:59-64.

Kaid, Lynda L. and Dorothy K. Davidson. "Videostyle: Candidate Presentation of Self Through Television Advertising, " paper presented to the International Communication Association. San Francisco, May 1984. (From footnote 8 in Moriarty and Popovich).

King, Karen N. 2002. "The art of impression management: self-presentation in local-level campaign literature."  The Social Science Journal 39:31-41.

Moriarty, Sandra E., and Mark N. Popovich. 1993. "News Magazine Visuals and the 1992 Presidential Election." Draft paper presented to the Association for Education in Journalism and Mass Communication, April 1, 1993.

Moriarty, Sandra E., and Mark N. Popovich. 1991. "Newsmagazine Visuals and the 1988 Presidential Campaign." Journalism Quarterly 68:371-80.

Moriarty, Sandra E., and Gina M. Garramone. 1986. "A Study of Newsmagazine Photographs of the 1984 Presidential Campaign." Journalism Quarterly 63:728-34.

Graber, Doris A. 1987. "Kind Pictures and Harsh Words: How Television Presents the Candidates." In Elections in America, ed. K. Schlozman. Winchester, Mass: Allen & Unwin.

Shyles, Leonard. 1984. "The Relationships of images, issues and presentational methods in televised spot advertisements for 1980's presidential primaries." Journal of Broadcasting 28:405-21.