Ethnicity and Episodic Framing in the Wake of Hurricane Katrina

Download data and study materials from OSF

Principal investigators:

Eran N. Ben-Porath

Social Science Research Solutions


Lee K. Shaker

Portland State University



Sample size: 504

Field period: 8/3/2006-8/9/2006


Employing an experimental design, this study looks at the effects of photo images and race on attribution of responsibility for the consequences of Hurricane Katrina. Subjects drawn from a national representative sample, oversampling for African Americans, were asked who they thought was responsible for the disaster. Before answering the questionnaire, subjects read a story that included pictures of either white Katrina victims, black victims or no images at all. We find attribution of blame to the federal government significantly higher when no pictures of victims (of any race) were included in the news. Furthermore, without the pictures there were no significant differences between black and white subjects as far as blaming the government was concerned. However, when pictures were included, black subjects attributed significantly higher levels of responsibility to the government than whites. There were no statistically significant effects driven by the race of victims in the photos.


H1: News images emphasizing the circumstances of individuals (episodic framing) would lead people to attribute less responsibility to the government and more to the victims.
H2: The race of victims appearing in images of victims will affect estimates about the racial distribution of victims overall.
H3: White respondents exposed to images of black victims will be less inclined to hold government responsible for the plight of Katrina victims than those exposed to white victims.
H4: Blacks will be more inclined to attribute responsibility to the government for the fallout following Katrina than whites.
H5: The effects of responsibility framing will be more pronounced for whites than for blacks.

Experimental Manipulations

There were five experimental conditions: In all conditions, subjects read the same news story about Katrina (which was devoid of any discussion of responsibility). In one condition the story included an image of one black male victim; in the second, a comparable white male; in third, a group of blacks walking on the highway; in the fourth, a group of whites in the same setting; and the fifth condition included no images.


-Attribution of responsibility to the federal government (measured by a seven-item scale and S´s assessment of percentage of blame ascribed to the federal government)
-Attribution of responsibility to the residents of New Orleans (measured by a seven-item scale and S´s assessment of percentage of blame ascribed to the residents)
-Attribution of responsibility to local authorities (based on percentage of responsibility ascribed).
* Attribution was also assessed through an open-ended question.

Summary of Results

We find that images affect attribution. When pictures of victims were included in the news story, people were likely to attribute less responsibility to the federal government than when there were no pictures. Looking within racial groups, we find that this effect was far weaker for blacks than for whites, meaning white respondents were much more likely to be affected (and blame the government less) as a result of exposure to images of victims. Contrary to expectations, this effect took shape irrespective of the victims´ race or whether they saw just one victim or a group. Images did not seem to impact attribution of blame for the residents of New Orleans. We do find, however, that blaming of local authorities in Louisiana increases when the images are included.


Our findings suggest that news images can act as episodic framing mechanisms, reducing the accountability attributed to situational factors, namely the actions of the government. Second, this consequence is not inevitable, particularly if the force of framing is contradicted by another force, in this instance, race. When readers of our fabricated news account were not provided with a concrete image of victims they were inclined to hold the government highly responsible for what had transpired in New Orleans after the storm. This was true for blacks and whites alike. Yet the inclusion of victims in the story, through the subtle cue of images and an entirely non-committed caption, led to a lowered sense of government culpability expressed by white respondents. This attests both to the power of the images and the strength of beliefs held by blacks. The effect of the victims´ race did not materialize most likely because of a pre-existing high base-rate assumption about the victims´ ethnicity.

Additional Information

-The story used was the Wikipedia entry for Hurricane Katrina. No references to responsibility for the events following the storm were included. Images used in the stories were of real Katrina victims, photoshopped into the same highway setting. The study took place in the summer of 2006, about 10 months after the storm. This means the events of Katrina may have still been salient to the subjects but were no longer a pressing issue for most or highly visible in the press. People at this point may have likely had some preset notion of what had happened and who was to blame.


Ben-Porath, E.N., & L. K. Shaker. 2007. "Ethnicity and Episodic Framing in the Wake of Hurricane Katrina." Paper Presented at the Annual Meeting of the Midwest Political Science Association Political Communication Division. Chicago, IL

Ben-Porath, E. N. and L. K. Shaker. 2010. News Images, Race, and Attribution in the Wake of Hurricane Katrina. Journal of Communication, 60: 466–490. doi: 10.1111/j.1460-2466.2010.01493.x