You are on page 1of 9

Quantifying Debate Deception: Tracking Media Fact Checking Trends in the October Debates Introduction This report uses

a quantitative framework to compare and contrast the fact checking efforts of CNN (CNN), The Huffington Post (HP), and FactCheck.Org (FC) in week brackets surrounding the four, October of 2012 Presidential and Vice Presidential debates. The former two sources are often perceived to be liberal leaning, and are therefore referenced against the self-proclaimed non-partisan reporting of the latter source. Proceeding the tracking of trends in these sources fact checking is a brief, correlative analysis of said trends against an aggregate poll index. This analysis is performed in an effort to assess the relationship, if any, between the frequency of biased fact-checking in relation to the relative strength of the Obama and Romney campaigns, testing the research question: When Obama was lagging in the polls, did liberal media sources lend more face time to fact checking efforts that developed Obamas strengths and highlighted the validity of his claims? Articles and video review information were pulled from CNN.Com, HuffingtonPost.Com and FactCheck.Org, and a complete document with rough quantitative findings can be found in the Appendix. October 3rd Debate Domestic Policy Domestic policy matters were the focus of the first Presidential debate. As defined by frequency of mention, the most salient topics across the profiled media sources dealt, predominantly, with health care, financial sector, macro economycentric, and federal deficit policy concerns in both camps. A total of 34 fact checking topics were addressed on sum, encapsulated in 10 individual articles or video 1

reviews in the three sources of interest CNN (CNN), The Huffington Post (HP), and FactCheck.Org (FC). CNN, FC and HP covered 15, 10, and 9 topics, respectively, over the week period bracketing the 3rd of October. Obamas assertions were checked for validity 4, 5, and 4 times, respectively, while Romneys were assessed 11, 5, and 5 times, respectively. Fact checking is an arguably objective art, however there was still obvious subjectivity in the quantity of fact checking for each candidate. Relative to HP and FC, CNN showed the most support for Obamas arguments, presenting in favor of his policy points 100% of the time across 4 topic mentions, and only 57% for Romneys contentions in 11 articles. Comparatively, no support was explicit in FCs fact checking, as to be expected from a presumably un-biased source, while HP reaffirmed 25% and 20% of the Democratic and Republican positions, respectively. Across all sources, the first debate was proximate to the average in terms of total stories reported. October 11th Debate Vice Presidents on Domestic and Foreign Policy The Vice Presidential debate afforded time to elaborate on positions of both domestic and foreign policy. The points of supreme interest, as defined by those topics that were highlighted at the highest frequency by the three sources, were the administrations response in Libya, women in the labor force, educational subsidies and the auto industry and macro-economic conditions. A total of 32 fact checking topics were addressed on sum, representing a 6% decrease in topic reporting when compared to the debate on October 3rd. CNN, FC and HP covered 9, 8, and 15 topics, respectively, over the week period bracketing the 11th of October. CNN checked

Obamas and Romneys arguments 7 and 6 times respectively, while FC and HP checked 1 and 7 times, and 6 and 6 times, respectively. In the second debate, FC, surprisingly, showed the most implicit support for Obama relative to HP and CNN. FC reaffirmed the incumbents policy points 1 out of 2 topic mentions, and only 1 out of 6 times for Romneys contentions. In contrast to the first debate, CNN showed no support for Obamas points, and only validated 33% of Romneys arguments. While HP reaffirmed 25% and 20% of the Democratic and Republican positions, respectively, following the October 3rd elections, they allotted most of their assessment space to disproving each candidates positions proceeding the second debate confirming only 17% and 14% for Obama and Romney, respectively. October 16th Debate Town Meeting Format on Domestic and Foreign Policy The Presidential nominees reconvened for the third debate, which centered on both and foreign policy in a town hall like forum, on October the 16th. 37 factchecking topics were addressed by CNN, HP and FC proceeding this debate, equating to a 14% increase over the second debate and an 8% increase over the first debate. CNN, FC and HP covered 17, 1, and 6 topics, respectively, over the week period bracketing the debate. Obamas assertions were checked for validity 4, 5, and 4 times, respectively, while Romneys were assessed 11, 5, and 5 times, respectively. As defined by frequency of mention, the most salient topics across the profiled media sources surrounded health care, financial sector, macro economy-centric, and federal deficit policy concerns in both camps.

Relative to HP and FC, and similarly to the first debate, CNN showed the most support for Obamas arguments, presenting in favor of his policy points 100% of the time across 7 topic mentions, and only 25% for Romneys contentions in 10 articles. Comparatively, no support was explicit in FCs fact checking of Obama, however they did reaffirm Romneys points in 33% of their 7 mentions. HPs assessment showed impartial favor for both Obama and Romney; in both cases, HP upheld 33% of the candidates assertions. October 22nd Debate Foreign Policy Foreign policy was the topic of the fourth and final third and final between the Presidential candidates debate in October. 57 topics were covered by the three media sources following the October 22nd bout, representing 68%, 78% and 54% increases over reporting in the first, second and third debates, respectively. The main topics considered by CNN, HP and FC were the Iranian nuclear program, the strength of Al-Qaeda and other terrorist groups, the US trade imbalance, and issues concerning the wars in Iraq and Afghanistan. CNN, FC and HP covered 35, 12, and 10 topics, respectively, over the week period bracketing the 22nd of October. CNN checked Obamas and Romneys arguments 19 and 16 times, respectively, while FC and HP checked candidate arguments 5 times. As aforementioned, the final debate realized a tremendous increase in the frequency of fact checking journalism, however this increase did not correspond with a relative increase in the space allotted to supporting one candidate over the other. Fact checking following the final debate was the least partial of the four. CNN, FC and HP reaffirmed Obamas and Romneys points, 86% and 75% (in 19 and 16

mentions), 20% and 0% (in 5 and 5 mentions), and 50% and 50% (in 6 and 6 mentions), respectively. Debates on Sum Tracking General Trends and Their Relationship with Polls The total count for fact checking topics across CNN, The Huffington Post and FactCheck.Org over the course of four October debate frames (7-day periods bracketing the 3rd, 11th, 16th and 22nd of October) was 160. Of these 160 counts, the Republican camp was checked for validity 91 times, or 57% of the time, while the Democratic camp was assessed 69 times, or 43% of the time. On a source-by-source basis, CNN, HP and FC represented 76, 48 and 36 or 48%, 30% and 23% of the total counted checks. CNN checked the Republican camp 57% of the time while HP and FC allotted 52% and 64% to the Romney/Ryan ticket. Most interestingly, however, was the discrepancy between the sources in regards to support as defined by the frequency at which the source reaffirmed contentions of the candidates. 71% of CNNs checks were supportive of the Democratic ticket, while 66% were in unsupportive of the Republican candidates. In contrast, support for both candidates hovered well below 50% for HP and FC at 31% and 29%, and 18% and 13%, respectively. These finding point to the style of journalism that each of these publications employ. In terms of the supportive spectrum, CNN was reaffirming while FC was both speculative and disavowing. To place these perceived media biases into context, it is prudent to frame these trends against the recorded trends in polls in the month of October. Figure 1 exhibits aggregated polling values, which are functions of synthesized polling results from 30 national polls, for three-day periods surrounding each debate date.

As illustrated, the significant Obama polling lead that existed during the first debate cycle, which was around 300 basis points, receded substantially by the conclusion of the third debate cycle, where Romney was instead leading in the polls by around a 100 basis points. Furthermore, it is crucial to note the competitive relative stances of each candidate during the 3-day period surrounding the second and third debates, occurring between October 10th and 12th and between 15th and 17th of October. Figure 1

Aggregate Polls in October


50.00 49.00 Percent Favor 48.00 47.00 46.00 45.00 44.00 2 3 4 10 11 12 15 16 Date in October 17 21 22 23 Obama Romney

Original Table, Data Sourced From the Wall Street Journal Tables 2 and 3 (shown below) illustrate the movements within each source in terms of both fact checking frequency and supportive checks for each camp. As explicated above, CNN performed the most fact checks on sum, however they also performed the most during and following the final debate on the 22nd of October. Even more interestingly is the fact that this larger frequency of checks did not translate into more implicit support for Obama; in the first debate, there was an incredible discrepancy between the quantity of supportive checks for Obama on

the part of CNN and the quantity of supportive checks for Romney, yet there was effectively equal supportive checks for both Romney and Obama following the final debate. Throughout all debate cycles, the percentage of supportive checks from The Huffington Post and FactCheck.Org were relatively constant. Figure 2

Fact Checking Frequency


20.00 18.00 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 2 3 4 10 11 12 15 16 17 21 22 23 Date in October

Frequency

(D) Checks - CNN (R) Checks - CNN (D) Checks - HP (R) Checks - HP (D) Checks - FC (R) Checks - FC

Figure 3

Checks in Support of Candidates


1.20 1.00 % Support 0.80 0.60 0.40 0.20 0.00 2 3 4 10 11 12 15 16 Date in October 17 21 22 23 (D) Support - CNN (R) Support - CNN (D) Support - HP (R) Support - HP (D) Support - FC (R) Support - FC

An interesting phenomenon can be observed when comparing the trends in support for the Democrats with that of the polling results for the same three-day segments in October. In specific regards to CNN, there is an observable relationship between the levels of implicit support and how Obama is fairing in the national polls. Statistically speaking, there is a correlation of .67 (out of 1) between the level of support and polling results, pointing, with a statistically significant value, to the fact that there is a relationship between positive press and points at which Obama was polling well. Contrastingly, there is only a .13 relationship between Romneys polls and CNNs reporting, and statistically irrelevant relationship values for HP and FC for both candidates. Conclusions The initial research question was as follows: When Obama was lagging in the polls, did liberal media sources lend more face time to fact checking efforts that developed Obamas strengths and highlighted the validity of his claims? CNNs track in terms of frequency and support was the most compelling in assessing this question. To revisit, their fact checking frequencies were comparable for both camps across all the October debates, but there were considerable fluctuations in their support of each candidates debate arguments hovering at or close to 100% for the Democrats in most of October, and only reaching a peak of around 80% for the Republicans in late October. A limited number of conclusions can truly gleaned from this data set in regards to the true biases of the liberal media sources of CNN and HP, however there is an obvious pop in the frequency of reporting following the final debate. This

could speak to an effort on the part of media to inform the public as they entered into the vote on November 6th, or one could perhaps develop this fact further and hypothesize that with Obama lagging in the polls in late October, the liberal media with specific attention being paid to CNN began to both cover and support the Democratic camp at a larger rate (as illustrated by the discrepancy in frequency and support between the Democrats and Republicans in Figures 2 and 3). And yet these observations must be taken with a grain of salt. Only twelve value points were considered when interpreting the relationship between the polls and the levels of support attributed to each source. Furthermore, there were assumptions employed regarding this presumed support; just because these sources reaffirmed either of the camps debate contentions does not necessarily translate into explicit bias. Statistically speaking, there was indeed a relationship between levels of support and polling results, but in no way were the correlations strong enough to prove that CNN was consciously favoring the Democratic camp in their reporting and little evidence to even suggest the possibility of explicit bias at the Huffington Post and FactCheck.Org.

You might also like