When asked to say which paper “reports fairly on the LDS Church,” 12% said The Salt Lake Tribune, 44% said The Deseret News. Only 14% said neither paper reported fairly.
James Madison eloquently argued for a free press in a constitutional republic. The press, he reasoned, helped to hold elected officials accountable by publishing the information citizens would need to make decisions about the power wielded by their elected officials.
The State of Utah finds itself in the enviable position of having two daily statewide newspapers to provide such information. Citizens can go to either one to learn about civic affairs. Each has its own distinct history and editorial stance. Overtime it appears that the histories and stances have created different loyalties among the state’s voters.
As reported in both The Deseret News and The Salt Lake Tribune, 81% of respondents to the October 2014 Utah Voter Poll “strongly agreed” or “agreed” that Utah needs two daily newspapers.
However, the overall agreement masks important subgroup differences in the ways in which individuals read the two papers and view their coverage. Thirty-eight percent of those who say they are not LDS say they read The Salt Lake Tribune regularly. Only 7% of those who are LDS say they read the paper regularly. The proportion almost flips for reading The Deseret News. Interestingly, and certainly in line with national trends, a large proportion of both LDS and non-LDS say they do not read either paper regularly.
When asked to say which paper “reports fairly on the LDS Church,” 12% said The Salt Lake Tribune, 44% said The Deseret News. Only 14% said neither paper reported fairly.
But as might be expected, these proportions change when controlling for whether or not one actually reads a newspaper. When we restrict the analysis to those who say they read one or both of the papers regularly, perceptions of fairness change. Forty-three percent of those who read a paper regularly indicate that The Deseret News reports fairly about the LDS Church. Only 15% of those who read a paper regularly say The Salt Lake Tribune does, and 34% say both papers report fairly about the LDS Church.
When we ask respondents (analysis limited only to those who read a paper regularly) to state which paper “performs the press’ role as a watchdog,” 36% say The Salt Lake Tribune and 6% say The Deseret News. What seems to emerge in the data is that readers value different things about the two papers. In the case of the Tribune, respondents to the poll who regularly consume the news appear to value the Tribune more than the News as fulfilling that traditional function of a free press.
The differences become even more evident when controlling for both regular readership and religion. Those individuals who read a paper regularly and are LDS are much more likely to say The Deseret News “reports fairly” than those who read a paper regularly and are not LDS. The same difference exists when assessing the fairness of The Salt Lake Tribune. Only 4% of individuals who read a paper regularly and are LDS believe The Salt Lake Tribune reports fairly while 38% of those who read a paper regularly and are not LDS conclude that The Tribune reports fairly.
Such differences raise interesting questions about what “fairness” means. Some individuals probably think “fairness” involves being tough on an institution while others certainly think “fairness” involves understanding and respecting that institution’s mission. What particular form of “fairness” individuals use to evaluate the reporting of the two major newspapers will have to wait for another poll.
The analysis reported above isolates only two important variables. A better method controls for many more factors in order to infer that the differences between the categories persist. Where the respondents have a choice between different categories but those categories are not ranked in any meaningful way, an appropriate method is multinomial logistic regression. This method permits a researcher to test what independent variables predict the probabilities that a person will choose a specific category of a dependent variable that is not ordered.
The model controls for religion, party identification, gender, age, religiosity, income, and education. The dependent variable is selecting The Salt Lake Tribune, The Deseret News, or Both when responding to the statement, “Reports fairly about the LDS Church.” Then it is possible to convert the results into probabilities for specific categories of the independent variables. Once again, only those respondents who “read a paper regularly” are included in the analysis.
As the figure below shows, a person’s probability of saying The Salt Lake Tribune “reports fairly on the LDS Church” drops 13 percentage points—holding all of the other variables constant—when the category changes from not LDS to LDS.
The change in the probability of selecting The Deseret News is even larger when comparing LDS to non-LDS respondents.
The other two figures essentially show that there is not much difference between LDS and non-LDS respondents in their probabilities of selecting the categories of “both” or “neither.”
What all of the data in this blog post seems to indicate is that religion matters for the ways in which respondents view the different newspapers and their roles. If Madison was correct, then the press matters. However it certainly behooves those who publish newspapers to always bear in mind that the decisions they make do indeed shape public opinion about their newspapers.
