Mug shots continue to have prominence in American newspapers. Singletary found 1,249 mug shots among 3,121 photographs in his study of 1,296 newspaper front pages. He found, however, that the amount of mug shots decreased from 59.4% of all photographs in the sample in 1936, to 42.2% in 1956, and to 29.9% in 1976. Singletary dismissed the mug shots as "relatively contentless and unimaginative." In later research, Lester found that nearly half of all front-page photographs in five major newspapers, including USA Today, were mug shots. Though perhaps not a pervasive as some fear, the USA Today-style has influenced other papers, especially smaller, group-owned papers. Unlike Singletary's assertion that mug shots are contentless, more recent research, including Lain and Harwood's experimental study, indicates that the content of mug shots, such as the facial expression, does affect reader reaction to stories, sources, and subjects of the photographs.

We know that the nature of a picture can influence reader attitudes toward the photo's subject. Van Tubergen and Mahsman found that the flattering or unflattering nature of a photograph influences viewer judgment of the attributes of the photo's subject. They found that the nature of picture seemed to have greatest impact on opinions about personality when the pictures were of unknown people. Their study, however, dealt with photographs in isolation. We also know that captions and text accompanying a photo can influence the interpretation of the photo. However, many people do not read such material. Research indicates that pictures have much higher readership, up to 10 times higher, than does copy. Lain found that readers react to people in newspaper stories and the mug shots which accompanied those stories in similar ways. This body of research indicates that photographs, including mug shots, do affect reader attitudes and opinions, especially about the subjects of the photographs.

Based on his study, Lain suggested that editors may frequently select photographs which correspond with the attitudes they already hold toward the people in the stories. There is evidence that political news reporting is affected by management attitudes. Likewise, there is evidence that journalists go into stories harboring assumptions about their story subjects, a tendency known as the "confirmation bias." Wanta and Roark, in their examination of the use of wire service photographs, found editors consistently overselected photographs that contained only one individual. Although their other results were mixed, they identified editors as gatekeepers for photographs as well as news stories. Studies of gender portrayals in photographs provide evidence that editors base photograph selection according to stereotype or personal bias. Miller found in her analysis of the Los Angeles Times and Washington Post that photo coverage appears to conform to accepted journalistic formulas and concepts of newsworthiness, rather than the roles women occupy in their respective communities. Wanta and Leggett found that sports editors tend to overselect photos which depict female athletes in emotional or dominated positions, even though the pool of photographs supplied by Associated Press treated male and female players in the same manner. This body of research has begun to identify the editorial role in selection and publication of photographs.

Experiments in agenda-setting have found a causal relationship between media agenda and a subsequent public agenda. Iyengar, Peters and Kinder conducted an experiment to examine the agenda-setting effects of television newscasts by altering the newscasts to stress certain issues. Some have argued that newspapers could have a stronger agenda-setting effect than television. Wanta, in an agenda-setting experiment with newspaper photographs, found newspaper reporting can have an influence on readers. Wanta manipulated the size of photographs. He found that it was possible to raise readers' salience on certain issues over a short period of time by merely increasing the size of the photographs. This body of research recognizes the potential influence of media content, including photographs, on the public and the public political attitudes and behaviors.

What happens when these factors -- the power of the photograph to influence attitudes about the photograph's subject matter, the power of the editor to select photographs, and the power of the media to influence the public -- coincide? This study was motivated by just such an incident.

In April 1993, five candidates went head-to-head in a non-partisan election to determine the next mayor of the City of Alton, Illinois. Alton is a small city of approximately 34,000 on the banks of the Mississippi River, just upstream of St. Louis, Missouri. The city is a blue collar, union town with a declining industrial basis. The city lost over 20 percent of its population in 10 years.

The candidates (see attached photographs) included the incumbent mayor (#4), two aldermen (#1 and #3), a former chief of police (#2), and a young man just out of the Navy (#5). The incumbent mayor's father, also an alderman, was chairman of the city's old Democratic machine and promised to deliver the votes for his son. One of the aldermen (#3), owner of a local pharmacy, had inherited the political mantle from the mayor and was backed by the local political kingmaker, a Democratic state representative and majority leader of the Illinois General Assembly. The original alliance between the mayor and the state representative had collapsed, and the "smart money," including substantial contributions from Illinois' first riverboat gambling operation, backed the pharmacist. The other alderman (#1), although in a nominally non-partisan race, was an employee of the state's Republican lieutenant governor and, hence, the Republican candidate. The former chief of police had challenged the incumbent four years earlier and, after a last minute and poorly funded campaign, lost in a two-way race by some 1,000 votes. With the backing of city employees, local unions and assorted citizen groups, the former police chief was bent on challenging the dominant political establishment. Popular discontent with the incumbent mayor, the Democratic machine, the political power of the state representative, and the city council in general was demonstrated in a dramatic referendum vote the prior November to cut the city council in half. Running as a reform candidate, the former police chief was considered by all as the frontrunner, hence the target of attacks by both the mayor and the "smart money" candidate.

Alton is a single medium town, dominated by the regional newspaper, the Telegraph, with a circulation of over 36,000. The Telegraph made media history by losing a notorious libel suit in 1980. There is no local television station and only one local radio station, an AM talk/news station. As a result, the Telegraph is in a dominant position to inform and influence. Based on reader feedback after the last municipal election, the Telegraph editors decided not to endorse any municipal candidate. However, as was its tradition, the newspaper asked each candidate to respond to a questionnaire and published the responses along with mug shots of each of the candidates. The full page article appeared just two weeks before the municipal election, which occurred April 20, 1993.

To the reform candidate's dismay, his picture was dramatically out of character from the other four mug shots. The newspaper had used photographs of the other four candidates, dressed in business suits, posed in professional studios, with appropriate studio smiles, utilizing flattering lighting, and sufficient blurriness to hide blemishes. His photograph, on the other hand, was detailed and grainy, with windbreaker, and squinting against the sunlight. This was in spite of the newspaper having published a studio-style, business suit photograph of the policeman only two weeks earlier. The photographs, when published in the newspaper, were the identical size. The newspaper apologized, claiming that the staff had been unable to find the earlier photograph and was forced to use what was at hand due to the deadline. The publication of the contrasting photograph reinforced the candidate's belief that the newspaper was taking sides, though he could not identify who the newspaper was supporting.

To the surprise of many, the reform candidate (#2) lost the five-way mayoral election to the Republican (#1) by 50 votes, less than one percent of those voting. Both the incumbent mayor (#4) and the smart-money candidate (#3) were strongly rejected by the voters, neither getting more than 20 percent. The mayor's father was also defeated in his bid to remain a city alderman, as were the majority of aldermen running for reelection to the city council.

There are many explanations for the defeat, including aggressive attacks against the reform candidate by the incumbent and the smart-money candidate, under the assumption that the reform candidate was the frontrunner. Anecdotal evidence suggests that voters were driven away from the reform candidate rather than attracted to the attackers, thereby benefitting the Republican alderman. The campaign staff, in debriefings, attempted to identify the turning points in the campaign. They identified the publication of the mug shot as a potentially seminal event. The publication of the photo reinforced their belief that the newspaper had secretly endorsed another candidate, although the staff was never clear as to which candidate was so favored. This, of course, raised the question of how much impact the photograph may have had in such a close election.

This experimental study was undertaken within a week of the publication of the original set of mug shots and a week prior to the municipal election. This methodology was chosen in the absence of any polling data, either before or after the publication of the mug shots.

Two sets of mug shots were clipped from the newspaper and mounted on poster board. (see attached photographs) The first set consisted of the five original mug shots, including the controversial photograph. The second set consisted of the five mug shots, with the an earlier, studio photograph in the same style as the other four replacing the controversial photograph. Each set was shown to two different groups of college students and parents at Principia College. In order to account for order bias in the alignment of the photographs, the mug shots were randomly reordered for each group. A total of 88 students, from American government classes, and parents, visiting for a parents' visitation weekend, responded. From two separate groups, 47 people examined the original set and 41 people examined the replacement set. Principia College, located 15 miles upriver from Alton, draws its students from across the country. None of the students or parents were familiar with Alton politics or any of the five candidates. Each person rated each of the five candidates, based on their impression of the candidates from the mug shots, on a five-point Likert scale on five characteristics: competence, experience, friendliness, honesty, and leadership. Means were calculated for each characteristic of each candidate, separating the original mug shots from the replacement mug shots. Paired-sample T-Tests were run on each candidate's scores between the two sets of mug shots at a 95% confidence interval.

In addition, respondents were invited to write down their reactions to the mug shots, both individual photographs and as a set.

The methodology does not account for the influence of the accompanying captions or stories, which were not shown to the respondents, nor does it account for the reaction of Alton voters, who were not included as respondents.

Based on the original set of mug shots, the respondents evaluated the reform candidate as lowest in combined characteristics. The respondents viewing the replacement set gave the reform candidate a much higher total score, propelling him from last position to second place. Likewise, they gave the other four candidates lower scores, though the four kept their positions relative to each other. The difference between the two sets of photographs was significant (p < .05) for all, including the minor candidate (p < .10). Based on the scores, the studio photograph put the reform candidate in a much better competitive position. (see attached table)

There was a high and significant correlation between the reaction to the replacement mug shot scores and the final vote in the mayoral election (actual vote: .9580, p=.010; percentage of vote: .9585, p=.010), whereas little correlation was found between the reaction to the original mug shots and the final vote (actual vote: .3328, p=.584; percentage of vote: .3117, p=.610). The replacement mug shot scores appeared to predict the outcome of the election. If voting can be taken as any indication of the public attitude toward a candidate, the replacement mug shots, as evidenced though the scores, are a better reflection of public sentiment than the original mug shots.

The respondents wrote a variety of comments about the pictures. Several respondents were repulsed by the posed or studio nature of the photographs. They said: "They all look fake," or "They all looked pretty 'slimy' to me." Several of the respondents volunteered a dislike for facial hair, since two of the candidates wore moustaches: "Facial hair makes me question integrity," "Facial hair makes me feel they are dishonest," or "I don't particularly like mustaches. I think they make men look shifty." One respondent summed up the power of the photograph:

Respondents had generally similar of comments about the photograph of the Republican candidate and ultimate mayoral winner. They noted experience, the look of a professional, and his caring. His photograph scored the highest in both sets. The respondents said:

In their comments, the respondents noted an array of negative reactions to the original photograph of the reform candidate. A number of respondents referred to him as a "crook," while others thought that his choice of attire indicated a lack of interest in the position. The respondents said:

The photograph probably hurt the reform candidate, especially among the uninformed voters or those who did not know him personally. While it was impossible to convince the losing candidate, it is not possible to demonstrate that the publication of the mug shot lost the election for him. Causality requires temporality, which existed, correlation, which appears to exist, and the absence of "third variables," of which there were many, including an extensive negative campaign waged against the reform candidate by two of the other losing candidates and siphoning of Democratic votes through the five-way race. Nevertheless, the candidate lost by only 50 votes, one half of one percent of those voting, and he appears to have a legitimate complaint against the newspaper, at least on ethical grounds.

The results of this study cannot be generalized, either to the citizens of Alton or to readers in general. Alton citizens were purposefully avoided in this study, under the expectation that the readers would be familiar with the various candidates through prior news coverage, campaign literature and advertising, or personal contact. As a result, this is not an indication of how the mug shots influenced the Alton voters, the people in whom we should be most interested. Future experiments might be conducted with subjects having demographic characteristics more similar to those of the actual voters. However, this study reinforces the growing body of research that has found that portrayal of a subject in a mug shot does influence the attitude readers have about the subject and the attributes readers assign to the subject.

This study continues, at least to some critics, the limited focus of studying photographs in isolation. However, as Lain and Harwood and Van Tubergen and Mahsman found, mug shots can have a differential effect on the meaning newspapers readers attribute to individuals who are subjects of accompanying news stories. The next step is to have respondents evaluate the actual and replacement mug shots within the context of the captions and the full story. As the respondents indicated, other characteristics of the mug shots could be relevant, including facial expression or their physical features, such as age, weight, facial hair, or baldness.

One of the assumptions that readers held, based on the content and tone of their remarks, is that the subjects of the photographs had some control over the publishing of the photographs. This is a logical assumption, given that the candidates were asked to submit a "candidate photo." However, as this incident demonstrates, it is the editor, not the candidate, who controls the publication decision. Photographs were not picked by the subjects, but by the editors. Hence, the attributes that readers gave to candidates were based on faulty assumptions about decisions the readers assumed the candidates made. This incident should serve as a warning to readers.

Newspaper editors, if they are concerned with objectivity, should be aware of the impact of mug shots have on reader attitudes and exercise care in their selection of photographs. This should be especially true when editors have alternative mug shots available and when the stories pertain to political candidates, in which image is a key element in the campaign. While the impact of their decision can never be truly assessed, the editors of The Telegraph still bear responsibility for their selection of the controversial mug shot.