Kull Ramsay Lewis 2003

Misperceptions, the Media, and the Iraq War Author(s): Steven Kull, Clay Ramsay, Evan Lewis Reviewed work(s): Source: Political Science Quarterly, Vol. 118, No. 4 (Winter, 2003/2004), pp. 569-598 Published by: The Academy of Political Science Stable URL: http://www.jstor.org/stable/30035697 . Accessed: 27/02/2012 11:39 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The Academy of Political Science is collaborating with JSTOR to digitize, preserve and extend access to Political Science Quarterly.

http://www.jstor.org

Misperceptions,the Media, and the IraqWar

STEVEN KULL CLAY RAMSAY EVAN LEWIS The Iraq war and its aftermathhave raised compellingquestions about the capacityof the executive branchto elicit public consent for the use of militaryforce and about the role the media plays in this process.From the outset, the Bush administrationwas faced with unique challengesin its effort to legitimateits decisionto go to war.Becausethe warwas not promptedby an overt act againstthe United Statesor its interests,andwas not approvedby the UN SecurityCouncil,the Bush administrationarguedthat the war was necessaryon the basis of a potentialthreat.Because the evidence for this threatwas not fully manifest,the Bush administrationled the public to believe that Iraq was developingweaponsof massdestruction(WMD) andprovidingsubstantial supportto the al Qaeda terroristgroup.The challengefor the administration was later intensifiedwhen the United States occupiedIraq and was unable to find the expected corroboratingevidence. Fromthe outset the publicwas sympatheticto the idea of removingSaddam Hussein, though only a small minorityof Americanswas ready to go to war with IraqwithoutUN SecurityCouncilapproval.'The majoritywas inclinedto believe that Iraqhad a WMD programandwas supportingal Qaeda.However, I Asked in a Chicago Council on Foreign Relations poll in June 2002 about their position on invading Iraq,65 percent said the United States "shouldonly invade Iraqwith UN approvaland the support of its allies";20 percent said "the US should invade Iraq even if we have to go it alone";and 13 percent said "the US should not invade Iraq." STEVEN KULL is the director of the Programon InternationalPolicy Attitudes (PIPA), a joint programof the Center on Policy Attitudes and Center for Internationaland SecurityStudies at Maryland of the School of Public Affairs, University of Maryland. CLAY RAMSAY is the director of research at PIPA. EVAN LEWIS is a research associate at PIPA. Political Science Quarterly Volume 118

Number 4

2003-04

569

570 1 POLITICALSCIENCEQUARTERLY

most were not persuadedthat the case was strongenough to justify taking action unilaterally.The majoritypreferredto continuelooking for more decisive evidence throughthe UN inspectionprocess and to continue seeking the support of the UN SecurityCouncil.2 Nevertheless,when the Presidentdecided to go to war,the majorityof the public expressedsupport.More significantly,when the United States failed to find the expected evidence that would corroboratethe administration'sassumptionsthat promptedthe war, the majoritycontinuedto supportthe decision to go to war.3 This polling data raisesthe questionof why the publichas been so accommodating.Did they simplychangetheir views about the war despite their earlier reservations?Or did they in some way come to have certain false beliefs or misperceptionsthat would make going to war appearmore legitimate,consistent with pre-existingbeliefs? A variety of possible misperceptionscould justify going to war with Iraq. If Americansbelieved that the United States had found WMD in Iraq or had found evidence that Iraq was providingsupportto al Qaeda, then they may have seen the war as justified as an act of self-defense even without UN approval. If Americansbelieved that world public opinion backed the United States going to war with Iraq, then they may have seen the war as legitimate even if some membersof the UN SecurityCouncilobstructedapproval. Of course,people do not develop misperceptionsin a vacuum.The administrationdisseminatesinformationdirectlyandby implication.The presstransmits this informationand, at least in theory, provides criticalanalysis.One's source of news or how closely one pays attention to the news may influence whetheror how misperceptionsmay develop. To find out more about the possible role of misperceptionsin public supportfor the Iraqwar,andthe role of the mediain thisprocessbefore andduring the war,the Programon InternationalPolicyAttitudes(PIPA) conducteda series of polls with the polling firm KnowledgeNetworks (KN). From January throughMay 2003, a more limitedset of questionswas asked in four different polls. Later, KnowledgeNetworks developed a more systematicset of questions that was includedin a series of three polls, conductedfrom June through 2 In August 2002, 55 percent thought Iraq "currentlyhas weapons of mass destruction,"and 39 percent thought Iraqis tryingto develop these weapons but does not currentlyhave them (CNN/USA Today). On al Qaeda, Newsweek asked in September 2002, "Fromwhat you've seen or heard in the news ... do you believe that Saddam Hussein's regime in Iraq is harboringal Qaeda terrorists and helping them to develop chemical weapons, or not?" Seventy-five percent said yes. Yet, in a 24-25 February2003 CBS News poll, only 31 percent agreed that "Iraqpresentssuch a clear dangerto American interests that the United States needs to act now"; 64 percent agreed that "the US needs to wait for approval of the United Nations before taking action against Iraq,"and 62 percent said that "the United States should wait and give the United Nations inspectors more time." 3 From May through November 2003, the Program on InternationalPolicy Attitudes/Knowledge Networks (PIPA/KN) has found a decliningmajorityof 68 percent to 57 percent saying "the US made the right decision ... in going to war with Iraq."

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR 1

571

September,with a total of 3,334respondents.These resultswere combinedwith the findingsfrom four other polls, conductedfromJanuarythroughMay, for a total dataset of 8,634respondents.In addition,relevantpollingdatafromother organizationswere analyzed,includingpolls that asked questionsaboutpossible misperceptions. The polls were fieldedby KnowledgeNetworksusingits nationwidepanel. Panel membersare recruitedthroughstandardtelephone interviewswith random digit dialing (RDD) samples of the entire adult population and subsequently providedinternet access. Questionnairesare then administeredover the Internetto a randomlyselected sample of the panel.4 This article first explores the degree of pervasivenessof misperceptions, particularlythe followingthree:that since the warU.S. forces have found Iraqi WMD in Iraq;that clear evidence has been found that SaddamHussein was workingclosely with al Qaeda;and that world public opinion was in favor of the United States going to war with Iraq. Second, it analyzesthe relationship between the holding of these misperceptionsand supportfor the Iraq war by using multivariateregression analysisto compare the strengthof this factor with a range of other factors.Third,it analyzesthe relationshipbetween the holding of misperceptionsand the respondent'sprimarynews source.Fourth, it evaluatesthe relationshipbetween attentionto news and the level of misperceptions. Fifth, it analyzesmisperceptionsas a function of political attitudes, includingintentionto vote for the Presidentand partyidentification.A binary logistic regression analysis includingmisperceptionsand eight other factors provides a rankingof factorsby power. The articleconcludeswith an analysis of the various factors that could explain the phenomenonof misperceptions, includingadministrationstatementsand media reporting. MISPERCEPTIONS RELATED TO THE IRAQ WAR

In the run-upto the war with Iraq and in the postwarperiod, a significantportion of the Americanpublic has held a numberof misperceptions5relevantto the rationalesfor going to war with Iraq.While in most cases only a minority has had any particularmisperception,a strong majorityhas had at least one key misperception.

Close Links betweenIraq and al Qaeda Both before and after the war, a substantial portion of Americans have be-

lieved that evidence of a link between Iraq and al Qaeda existed. Before the 4 For more informationabout this methodology,see the Appendix or go to www.knowledgenetworks. com/ganp. 5 Herein the term "misperceptions"is not used to refer to controversial beliefs about what U.S. intelligence has been able to infer, such as the belief that Saddam Hussein was directly involved in September 11. The term is limited to noncontroversialperceptionssuch as whether actual weapons or actual evidence have in fact been found. The misperceptionrelated to world public opinion is established based on polling data discussed later.

572

I POLITICAL SCIENCE QUARTERLY TABLE1

Evidenceof LinkbetweenIraqandal Qaeda (percentages) Is ityourimpressionthatthe US has orhas notfoundclearevidenceinIraqthat SaddamHusseinwas workingcloselywiththe al Qaedaterroristorganization? 8-9/03 7/03 6/03 (6/03-9/03) US has US has not (Noanswer)

49 45 6

45 49 6

52 43 5

48 46 6

Source:Programon International Networks. PolicyAttitudes/Knowledge

war, in the JanuaryPIPA/KN poll, 68 percent expressed the belief that Iraq played an importantrole in September11, with 13 percenteven expressingthe belief that "conclusiveevidence"of Iraq'sinvolvementhad been found.Asked in June,July,andAugust-September(Table 1), largepercentages(45 to 52 percent) said they believed that the United States had "foundclear evidence in Iraq that SaddamHussein was workingclosely with the al-Qaeda [sic] terrorist organization." HarrisInteractivein June and August asked, "Do you believe clear evidence that Iraq was supportingal Qaeda has been found in Iraq or not?" In June, 48 percentsaid that clear evidence had been found, with just 33 percent sayingthat it had not and 19 percentsayingthey were not sure. Despite intensive discussionof the issue in the press,in August the numberswere essentially the same: 50 percent believed evidence had been found, 35 percent believed that it had not been, and 14 percentwere unsure. Weapons of Mass Destruction Before the war, overwhelming majorities believed that Iraq had WMD. Though it now appearslikely that this belief was incorrect,it does not seem appropriateto call this a misperceptionbecause it was so widespreadat the time, even withinthe intelligencecommunity. However,a strikingmisperceptionoccurredafterthe war,when the United Statesfailed to find any WMD or even any solid evidence of a WMD program. PIPA/KN first asked in May whether respondents thought that the United States has or has not "foundIraqiweapons of mass destruction"in Iraq, and 34 percent said the United States had (another 7 percent did not know). In June, HarrisInteractivesubsequentlyasked, "Do you believe clear evidence of weaponsof mass destructionhas been found in Iraqor not?"and 35 percent said that it had. PIPA/KN asked againin late June--during a period with much discussion in the press about the absence of WMD- and found that the percentagehold-

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 573

TABLE2

Existenceof Weaponsof MassDestruction in Iraq (percentages) Since the warwithIraqended, is ityourimpressionthatthe US has or has not foundIraqiweaponsof mass destruction? 9/03 7/03 6/03 3/03 (6/03-9/03) US has US has not (No answer)

24 73 3

21 76 3

23 73 4

34 59 7

22 75 3

Source:Programon International Networks. PolicyAttitudes/Knowledge

ing this belief had droppedto 23 percent.This numberthen stayedroughlythe same in July and early September.In late July,NBC/WallStreetJournalasked whetherthe United Stateshas been successfulin "findingevidence of weapons of massdestruction,"and22 percentsaidthatit had.Harrisaskedagainin midAugust and found 27 percent saying that evidence of WMD had been found (Table 2). Americanshave also incorrectlybelieved that Iraq actuallyused WMD in the recent war with the United States. PIPA/KN asked respondentswhether "Iraqdid or did not use chemicalor biologicalweaponsin the warthathadjust ended."In May,22 percentof respondentssaidthat it had. In mid-June,ABC/ WashingtonPost presenteda slightlyadaptedversionof the questionandfound 24 percent said that that they thoughtit had. When asked by PIPA/KN again in August-September,the percentagesayingthat Iraqhad used such weapons slipped only slightlyto 20 percent. World Public Opinion A key factor in Americanpublic supportfor going to war with Iraq has been its internationallegitimacy.Rightup to the periodimmediatelybefore the war, a majorityfavoredtakingmore time to buildinternationalsupport.A key question, then, is how the public perceivedworld public opinion on going to war with Iraq. PIPA/KNpolls have shown that Americanshave misperceivedworldpublic opinion on the U.S. decision to go to war and on the way that the United States is generallydealing with the problem of terrorism.This has been true duringand afterthe warand appliesto perceptionsaboutworldpublicopinion as a whole, Europeanpublicopinion,and publicopinion in the Muslimworld. In March2003, shortlyafter the war started,PIPA/KN asked respondents "howall of the people in the worldfeel about the US going to war with Iraq." Respondentsperceivedgreatersupportfor the war than existed at the time or has existed since.6Only 35 percent perceived correctlythat the majorityof 6 Gallup Internationalconducted two internationalpolls (in Januaryand April-May2003) and Pew Research Center conducted one (in April-May2003), which included poll questions that directlymeasured support or opposition to the Iraq war. In the three polls taken together, fifty-sixcountrieswere

574 1POLITICALSCIENCEQUARTERLY

TABLE3 WorldOpinionabout the U.S. Decision to Go to War (percentages) abouthowallthe peoplein the worldfeel aboutthe US going/havinggone to warwith Thinking 3/03 6/03 9/03 7/03 (6/03-9/03) Iraq,do you think of peoplefavorit The majority of peopleoppose it The majority Viewsare evenlybalanced (Noanswer)

27 38 33 2

24 42 30 4

25 41 32 2

31 35 31 3

25 41 31 3

Network. Source:Programon International PolicyAttitudes/Knowledge

people opposedthe decision.Thirty-onepercentexpressedthe mistakenassumption that views were evenly balancedon the issue, and another31 percentexpressedthe egregiousmisperceptionthatthe majorityfavoredit. Asked againin June,July,and August-September,these views changedvery little (Table3). Perceptionshave been a bit more accuratewhen it comes to perceivingEuropeanpublicopinion,but still there are widespreadmisperceptions.Asked in June and August-September,nearlyhalf (48 to 49 percent) correctlysaid that the "majorityof people oppose the United States havinggone to war."But 29 to 30 percent believed incorrectlythat views are evenly balanced,and 18 percent believed that the majorityeven favors it.7 A substantialnumber of Americans also misperceiveattitudes in the IslamicworldtowardU.S. effortsto fightterrorismand its policies in the Middle East. Respondentswere asked in August-Septemberwhetherthey thought "a majorityof people in the Islamicworldfavoror oppose U.S.-led effortsto fight terrorism."A pluralityof 48 percentincorrectlyassumedthat a majorityof Islamicpeople favorsU.S.-ledeffortsto fightterrorism,while46 percentassumed that they do not. When askedwhetherrespondentsthought"amajorityof people in the IslamicworldthinkU.S. policies in the MiddleEast make the region more or less stable,"35 percentincorrectlyassumedthat the majorityof people surveyed. The JanuaryGallup Internationalpoll asked, "Are you in favor of military action against Iraq:under no circumstances;only if sanctioned by the United Nations; unilaterallyby America and its allies?" Of the thirty-eightcountriespolled (includingtwenty European countries), not a single one showed majoritysupport for unilateralaction, and in nearly every case the percentage was very low. When asked, "Ifmilitaryaction goes ahead againstIraq,do you think [surveycountry]should or should not supportthis action?"in thirty-fourof the thirty-eightcountriespolled (seventeen out of twenty in Europe), a majorityopposed having their country support this action. In April-May, the Pew Global Attitudes Surveyasked respondentsin eighteen countrieshow they felt about their country'sdecision to participateor not participatein "us[ing]militaryforce against Iraq."Among the thirteen countries that had not participated,in every case, a large to overwhelmingmajorityapproved of the decision. For the three countries that contributedtroops, in the United Kingdom and Australia, a majorityapproved; in Spain, a majoritywas opposed. For the two countries that had allowed the United States to use bases, in Kuwait,the majorityapproved;in Turkey, the majoritywas opposed. For full results, see www.gallup-international.comand www.people-press.org. 7Ibid.

THE MEDIA,AND THE IRAQWAR MISPERCEPTIONS,

I 575

FIGURE1 Percentage HavingKey Misperceptions: Evidenceof Linksto al Qaeda, WMDFound,and WorldPublic OpinionFavorable No Misperceptions GlobalAttitudesare "EvenlyBalanced" 1 Misperception

30% 40% 10% 32%

2 Misperceptions 3 Misperceptions 0

60%1 ormore

20% 8% 20

40

60

80

100

Percent Compositeof Polls ConductedJune-September2003 Source: Programon International PolicyAttitudes/KnowledgeNetworks,October2003.

in the Islamicworld feel that U.S. policies make the region more stable, while 60 percentperceivedattitudescorrectly.8 Combined Analysis Most specificmisperceptionsare held by a minorityof respondents.However, this does not tell us if these misperceptionsare held by the same minorityor if large percentageshave at least one misperception.To find out, we repeated three key perceptionquestions over three polls, conductedin June, July, and August-Septemberwith 3,334 respondents. The three key perceptionquestionsused were the ones thatfound the most egregiousmisperceptions,and to qualifyas a misperceptionthe most extreme form of the misperceptionwas used. These were the beliefs: * Clear evidence that SaddamHussein was workingclosely with al Qaeda has been found. * Weapons of mass destructionhave been found in Iraq. * Worldpublic opinionfavored the United States going to war with Iraq. To determinethe pervasivenessof misperceptions,we focused on the 1,362respondentswho heard all three of the perceptionquestions. Misperceptionswere not limitedto a smallminoritythat had repeatedmisperceptions.A majorityof 60 percenthad at least one of these three unambiguous misperceptions,and only 30 percenthad no misperceptions(Figure1). AnPew Global Attitudes survey in summer 2002 and May 2003 asked in seven countries with primarilyMuslim populations (Turkey,Indonesia, Pakistan,Lebanon, Jordan,Kuwait, and Morocco, plus the Palestinian Authority): "Whichof the following phrases comes closer to your view? I favor 8 The

576

I POLITICALSCIENCEQUARTERLY

other 10 percenthad the more modest misperceptionthatworldpublicopinion was evenly balancedbetween supportand oppositionto the Iraqwar. MISPERCEPTIONSAND SUPPORTFORWAR

The misperceptionsaboutthe warappearto be highlyrelatedto attitudesabout the decision to go to war, both before and after the war. In every case, those who have the misperceptionhave been moresupportiveof the war.As the combined analysisof the three key misperceptionswill show, those with none of the key misperceptionshave opposed the decision while the presence of each additionalmisperceptionhas gone togetherwith sharplyhigher support.

Close Links to al Qaeda Before the war,those who believed that Iraqwas directlyinvolvedin September 11 showed greatersupportfor going to war even without multilateralapproval.In the JanuaryPIPA/KNpoll, amongthose who wronglybelieved that they had "seenconclusiveevidence"that "Iraqplayedan importantrole in September 11 attacks,"56 percent said they would agree with a decision by the Presidentto proceedto go to warwith Iraqif the UN SecurityCouncilrefused to endorse such an action. Among those who said they had not seen such evidence but still believed that Iraqwas involvedin September11, 42 percentsaid they would supportsuch a decision.Among those who said they had not seen such evidence and were not convincedthat it was true, only 9 percentsaid they would agree with such a decision. In the FebruaryPIPA/KN poll, supportfor going to war was high among those who believed that SaddamHussein was directlyinvolved in September 11 but was progressivelylower as the perceivedlink between Iraqand al Qaeda became more tenuous. Among those who believed that Iraq was directlyinvolved in September11, 58 percent said they would agree with the President decidingto go to war with Iraqeven withoutUN approval.Among those who believed that Iraqhad given al Qaeda substantialsupportbut was not involved in September11, supportdroppedto 37 percent. Among those who believed that a few al Qaeda individualshad contactwith Iraqiofficials,32 percentwere the US-led efforts to fight terrorism,or I oppose the US-led efforts to fight terrorism."In six of the eight cases, strong majorities ranging from 56 to 85 percent in summer 2002, and rising to 67 to 97 percent in May 2003, said they opposed "US-led efforts to fight terrorism."In only one case-Kuwait in May 2003-did a majoritysay they favored U.S. efforts. In the case of Pakistan, a plurality of 45 percent opposed U.S. efforts in the summer of 2003, rising to 74 percent in May 2003. In May 2003, respondents were asked: "Do you think US policies in the Middle East make the region more stable or less stable?" In six of the eight cases, majoritiessaid that U.S. policies in the Middle East make the region less stable. These majorities ranged from 56 percent in Lebanon to 91 percent in Jordan. In Pakistan,43 percent said U.S. policies make the Middle East less stable, but another 43 percent said U.S. policies either "made no difference" (12 percent) or that they did not know (31 percent). In Kuwait, a 48 percent pluralitysaid U.S. policies made the Middle East more stable.

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 577

FIGURE2 for War and Support Misperceptionof Evidenceof IraqiLinksto al Qaeda

The U.S. Has Found Evidence

67%

Among Believed: War forWho

The U.S. Has NOT FoundEvidence

29%

Those Support

0

20

40

60

80

100

Percent Compositeof Polls ConductedJune-September2003 Source: Programon International PolicyAttitudes/KnowledgeNetworks,October2003. Note: The question also offered respondentsthe option of saying that they did not know if going to war was the best thing to do, but that they nonetheless supportedthe President.Here and in comparisons discussed later,we have limitedouranalysisto those who took an unequivocalpositioninfavoror against the decision to go to war.

supportive,whilejust 25 percentexpressedsupportamongthose who believed that there was no connection. During the war, Americanswho supportedthe war also said that the supposed linkwas a majorreasonfor supportingthe decisionto go to war.An April poll for Investor'sBusinessDaily and the ChristianScienceMonitor askedthe 72 percentwho saidthey supportedthe warto rate the importanceof a number of reasonsfor theirsupport."Iraq'sconnectionwithgroupslike Al-Qaeda"was rated as a majorreason by 80 percent. After the war,nearlyhalf of the respondentsmistakenlybelieved thatclear evidence that SaddamHussein was workingclosely with al Qaeda had been found. PIPA/KN found a strongrelationshipbetween the belief that evidence of suchlinkshas been foundand supportfor the decisionto go to war.Combining data from June throughSeptember,among those with the misperception, 67 percentheld the view that going to war was the best thing to do, while only 29 percentexpressedsupportamongthose who did not have the misperception (Figure2). Among those withoutthe misperception,52 percentsaid it was the wrong decision. Just as before the war, in the postwarperiod there was also a strongrelationshipbetween beliefs about the natureof the connectionbetween al Qaeda andIraqandsupportfor the war.Among those who believedthatSaddamHussein was directlyinvolved in September11, 69 percent said going to war was the best thing to do. Among those who believed that Iraqhad given al Qaeda substantialsupportbut was not involved in September11, approvaldropped

578 I POLITICAL SCIENCE QUARTERLY

FIGURE3 Supportfor the Warand Misperceptionthat IraqiWMDFound

The U.S. Has Among FoundWMD Believed: War forWho

73%

The U.S.HasNOT

Those Support

41%

FoundWMD 0

20

40

60

80

100

Percent Compositeof Polls ConductedMay-September2003 Source: Programon International PolicyAttitudes/KnowledgeNetworks,October2003.

to 54 percent.Among those who believed that a few al Qaeda individualshad contact with Iraqi officials,39 percent were supportive,while just 11 percent expressed support among those who believed that there was no connection. Among those who believed that there was no connection,73 percent thought that going to war was the wrong decision. Weapons of Mass Destruction The mistakenbeliefs that WMD have been found in Iraq, or that Iraq used WMD in the war,have been highlyrelatedto supportfor the decisionto go to war. Consolidatingall respondentsasked by PIPA/KNin four polls conducted fromMaythroughSeptember,amongthose who believedthatWMD have been found, 73 percent thoughtthat going to war was the best decision (Figure3). Amongthosewho didnot havethismisperception,only41 percentheldthisview. Similarly,consolidatingtwo polls conductedin May and August-September, amongthose who believedthatIraqhadused chemicalandbiologicalweapons in the war, 64 percentsaid they thoughtgoing to war was the best thing to do. Among those who did not have this belief, only 48 percent thought it was the best thing to do. World Public Opinion Perceptionsof world public opinion on going to war with Iraq have been significantlyrelated to supportfor the war. This has been true duringand after the war. In the PIPA/KNpoll conductedin late March,shortlyafterthe onset of the war, amongthose who wronglybelieved that the majorityof the people in the

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 579

FIGURE4 Viewsof WorldPublic Opinionand Supportfor WarDuringand Afterthe War Who

I81%

Favors the U.S.

Opinion: Those Going to War World Among

77% 58% 52%

Is Evenly Balanced

War Majority for Opposes the U.S.

28% 28%

Going to War

Believed Support

0

20

40

60

80

100

Percent of Polls Conducted Composite March-September2003 U March2003 0 June-September2003

Source: Programon International PolicyAttitudes/KnowledgeNetworks,October2003.

world favored the United States going to war with Iraq, an overwhelming81 percent said they agreed with the President'sdecision to go to war with Iraq, despite his failure to garner UN Security Council approval. Among those who- also incorrectly--believedthatviewswere evenly balancedon this question, 58 percentsaid they agreed.Among those who correctlybelieved that the majorityof people opposed it, only 28 percentsaid they agreedwith the President's decision.When polled after the war (May-September)the patternwas basicallythe same, though a differentquestion was used to measure support for the war (Figure4). Combined Analysis To determinethe cumulativestrengthof the relationshipbetween variousmisperceptionsand supportfor the war, we analyzedthose who had been asked all of the three key misperceptionquestions--whether evidence of links between Iraqand al Qaeda have been found,whetherWMD have been found in Iraq,and whetherworldpublicopinionfavoredthe United States going to war with Iraq--in threepolls conductedfromJune throughSeptember.These polls revealed a strongcumulativerelationship(Figure5). Multivariate Analysis To determinehow stronga factormisperceptionsare in predictingsupportfor the war as comparedto other factors,a binarylogistic regressionanalysiswas

580 I POLITICALSCIENCEQUARTERLY

FIGURE5 CumulativeEffectof HavingKey Misperceptionson Supportfor the War No Misperceptions

23%

Among Have:1 Misperception War Who for 2 Misperceptions

53%

78%

Those 3 Support Misperceptions 0

86%

20

40

60

80

100

Percent Compositeof Polls ConductedJune-September2003 October2003. Source:Program on International Networks, PolicyAttitudes/Knowledge Note:Misperceptions includedwerethatclearevidenceof Iraq-alQaedalinkshavebeenfound,WMD havebeenfound,andworldpublicopinionfavoredthe Iraqwar.

performedtogether with eight other factors.Four of the factors were demographic:gender,age, householdincome, and education.Two other categorical factorswere partyidentificationand intentionto vote for the Presidentin the next election as opposed to an unnamed Democratic nominee. In addition, there were the factors of how closely people follow events in Iraq and their primarynews source.The odds ratiostatisticwas used to determinethe relative likelihood that respondentswould supportthe war. Supportfor the war was defined as the respondentsayingthat he or she thoughtthe war was the right decision and the best thing to do, not that he or she was just supportingthe President.For this analysis,the numberof respondentswas 1,219. When all respondentswith one or more of the three key misperceptions were put into one categoryand comparedto those with none of these misperceptions, the presence of misperceptionswas the most powerful predictorof supportfor the war, with those misperceivingbeing 4.3 times more likely to supportthe warthanthose who did not misperceive.The secondmost powerful predictorwas the intention to vote for the President,with those intendingto vote for the Presidentbeing 3 times more likely to supportthe war than those who plannedto vote for the Democraticnominee.Those who intendedto vote for the Democraticnominee were 1.8 times less likely to supportthe war. All other factors were far less influential.Those who followed the news on Iraq very or somewhatclosely were 1.2 times more likely to supportthe war than those who followed it "not very closely" or "not at all." Men were 1.5 times more likely to supportthe war than women. Those with higherincomes were very slightlymorelikely to supportthe war.All other factorswere insignificant,

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 581

includingeducationand age. Partyidentificationby itself would be predictive, but when intention to vote for the Presidentis included,party identification also becomes insignificant. To determinethe cumulativestrengthof misperceptionsas a predictorof war support,the smallersample that received all of the three key misperceptions questionswas analyzed.Respondentswere divided into four categories of no misperceptions,exactly one misperception,exactly two misperceptions, and all threemisperceptions.Those withjust one misperceptionwere 2.9 times more likely to supportthe war,risingto 8.1 times morelikely amongthose with exactly two misperceptionsand to 9.8 times more likely among those with all three misperceptions.In this sample, all other factors remain essentiallyunchanged,withthose intendingto vote for the Presidentbeing2.8 timesandmen 1.5 timesmore likelyto supportthe war.Those intendingto vote for the Democraticnominee were 1.6 times less likely to supportthe war.Attention to news coded as a binaryform,however,becameinsignificantwhile remainingsignificant as a continuousvariable. When the three key misperceptionsare treatedas separatefactors,there is wide variationin theirpower to predictsupportfor the war.By far, the strongest is the perceptionof worldpublicopinion,withthose who perceivethe world public opinion as approvingof the war being 3.3 times more likely to support the war themselves. Those with the perception that evidence of links to al Qaeda have been found were 2.5 times more likely to support the war, and those who perceived that evidence of WMD have been found were 2.0 times more likely. MISPERCEPTIONS AS A FUNCTION OF SOURCE OF NEWS

The widespreadpresence of misperceptionsnaturallyraises the question of whetherthey are to some extent a functionof an individual'ssourceof news.To findout, in three differentPIPA/KNpolls conductedin June,July,andAugustSeptember,an aggregatesample of 3,334 respondentswas asked, "Wheredo you tend to get most of your news?"and offered the options of "newspapers and magazines"or "TVandradio."Overall,19 percentsaid theirprimarynews source was print media, while 80 percent said it was electronic.Respondents were then asked, "If one of the networksbelow is your primarysourceof news please select it. If you get news fromtwo or more networksabout equally,just go on to the next question."The networks offered were ABC, CBS, NBC, CNN, Fox News, PBS, andNPR. Because the PBS andNPR viewerswere such a small percentage,we combinedthem into one categoryof public networks. In the case of ABC, CBS, and NBC, we do not knowhow manypeople primarily got theirnews fromlocal affiliatesand how manyfromnationalnews shows. Likewise,we do not know if all of those who said that they got theirnews from

582 I POLITICALSCIENCEQUARTERLY

TABLE4

Frequencyof Misperceptions per Respondent:WMDFound,Evidenceof al QaedaLink,and WorldMajority SupportforWar (percentages) Numberof misperceptions per respondent Noneof the three One or more misperception

Fox

CBS

ABC

CNN

NBC

PrintMedia

NPRIPBS

20

30

39

45

45

53

77

80

71

61

55

55

47

23

Networks. Source:Programon International PolicyAttitudes/Knowledge

Fox News primarilygot their news from the nationalcable news networkand how many from local Fox affiliates.9 The same respondentswere also asked about their perceptions,with 1,362 respondentsreceivingall threekey perceptionquestionsand3,334respondents receivingat least one of them-that is, whetherevidenceof close linksbetween Iraq and al Qaeda has been found, whether WMD have been found in Iraq, and whetherworldpublicopinion approvedof the United States going to war with Iraq. COMBINED ANALYSIS

Because it provides the best overview of the relationship between media sources,this articlefirst analyzesthe relationshipbetween media sources and the presenceof multiplemisperceptionsto explorethe variationin the level of misperceptionsaccordingto the respondents'news source. Afterward,it analyzes the variancefor specificmisperceptions. An analysisof those who were asked all of the key three perceptionquestions does reveal a remarkablelevel of variationin the presenceof misperceptions accordingto news source. Standingout in the analysisare Fox and NPR/ PBS, but for oppositereasons.Fox was the news sourcewhose viewershad the most misperceptions.NPR/PBSare notablebecausetheirviewersand listeners consistentlyheld fewer misperceptionsthan respondentswho obtained their informationfrom other news sources.Table 4 shows this clearly.Listed are the breakoutsof the sampleaccordingto the frequencyof the threekey misperceptions (that is, the beliefs that evidence of links between Iraq and al Qaeda has been found,that WMD have been found in Iraq,and that worldpublicopinion approvedof the United States going to war with Iraq) and their primarynews source.In the audiencefor NPR/PBS,therewas an overwhelmingmajoritywho did not have any of the three misperceptions,and hardlyany had all three. 9 Numbersfor those naminga network as their primarynews source were as follows:Fox, 520; CBS, 258; CNN, 466;ABC, 315;NBC, 420;NPR/PBS, 91. All findingsin this section were statisticallysignificant at the p<0.05 level, except where noted.

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 583

TABLE5

RatesamongViewersand Averageof ThreeMisperception Listeners:WMDFound,Evidenceof al QaedaLink, and WorldMajority SupportforWar (percentages) News Source

AverageRateperMisperception

Fox CBS CNN ABC NBC PrintMedia NPR/PBS

45 36 31 30 30 25 11

Source: Program on InternationalPolicy Attitudes/Knowledge Networks.

To check these strikingfindings,the data were analyzeda differentway by using the larger sample of 3,334 who had answeredat least one of the three questions just mentioned. For each misperception,it was determined how widespreadit was in each media audience,and then for each media audience this frequencywas averagedfor the three misperceptions.Table 5 shows the averagesfrom lowest to highest.Again, the Fox audienceshowed the highest average rate of misperceptions(45 percent) while the NPR/PBS audience showed the lowest (11 percent). Close Links to al Qaeda The same patternin the distributionof misperceptionsamongthe news sources was obtainedin the cases of each specificmisperception.When askedwhether the United States has found "clearevidence in Iraqthat SaddamHusseinwas workingcloselywiththe al-Qaedaterroristorganization,"amongthe combined samplefor the three-monthperiod,49 percentsaidthatsuchevidencehad been found (Table 6). This misperceptionwas substantiallyhigheramongthose who get their news primarilyfrom Fox, 67 percent.Once againthe NPR/PBS audience was the lowest at 16 percent. TABLE6

Viewers'Beliefson Whetherthe UnitedStatesHasFound Evidenceof an al Qaeda-lraq Link (percentages) ClearEvidenceof al QaedaLink

NBC

CBS

ABC

Fox

CNN

NPR/PBS

PrintMedia

US has found US has not found

49 45

56 41

45 49

67 29

48 47

16 85

40 58

Source: Program on InternationalPolicy Attitudes/Knowledge Networks.

584 I POLITICAL SCIENCE QUARTERLY

TABLE7

Perceptionthatthe UnitedStatesHasor HasNotFoundWMD (percentages) Weaponsof Mass Destruction

NBC

CBS

ABC

Fox

CNN

NPRIPBS

PrintMedia

US has found US has not found

20 79

23 75

19 79

33 64

20 79

11 89

17 82

Networks. Source:Programon International PolicyAttitudes/Knowledge

Variationswere muchmore modeston the perceptionthatIraqwas directly involvedin September11.As discussed,the view thatIraqwas directlyinvolved in September11 is not a demonstrablemisperception,but it is widely regarded as fallaciousby the intelligencecommunity.In this case, the highestlevel of misperceptionswas in the CBS audience (33 percent) followed by Fox (24 percent), ABC (23 percent), NBC (22 percent), and CNN (21 percent). Respondents who got their news primarilyfrom printmedia (14 percent) and NPR or PBS (10 percent) were less likely to choose this description. Combiningthe above groupwith those who had the less egregiousbut still unprovenbelief that Iraqgave substantialsupportto al Qaeda,the patternwas similar.Among CBS viewers,68 percent had one of these perceptions,as did 66 percentof Fox viewers,59 percentof NBC viewers,55 percentof CNN viewers, and 53 percent of ABC viewers. Print readerswere nearly as high at 51 percent,while NPR/PBS audienceswere significantlylower at 28 percent. Weapons of Mass Destruction When respondentswere asked whether the United States has "found Iraqi weaponsof massdestruction"since the warhad ended,22 percentof all respondents over June through Septembermistakenlythought this had happened. Once again,Fox viewerswere the highestwith 33 percenthavingthis belief. A lower 19 to 23 percent of viewerswho watch ABC, NBC, CBS, and CNN had the perceptionthat the United States has found WMD. Seventeen percent of those who primarilyget their news from print sources had the misperception, while only 11 percentwho watch PBS or listen to NPR had it (Table 7). World Public Opinion Respondentswere also asked to give their impressionof how they think "people in the world feel about the US having gone to war with Iraq."Over the three-monthperiod,25 percentof all respondentssaid,incorrectly,"themajority of people favor the US having gone to war" (Table 8). Of Fox watchers, 35 percentsaid this. Only 5 percent of those who watch PBS or listen to NPR misperceivedworldopinionin this way. As usual,those who primarilyget their

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 585

TABLE8 WorldPublic Opinionon the UnitedStates Going to War (percentages) of people Majority in world...

NBC

CBS

ABC

Fox

CNN

NPRIPBS

PrintMedia

FavorUS goingto warin Iraq

20

28

27

35

24

5

17

Source:Programon International Networks. PolicyAttitudes/Knowledge

news fromprintmediawere the second lowest,with 17 percenthavingthismisperception. Numerousrespondentsalso chose the option of sayingthat in worldpublic opinion, views are evenly balanced between favoringand opposing going to war-a misperception,thoughless egregious.Combiningthose who said views were evenly balancedwith those who assumedthat the majorityfavored the Iraqwar--a more inclusivedefinitionof misperception--thesame patternobtained. Fox viewers had the highest level of misperceiving(69 percent) and NPR/PBS the lowest (26 percent).The others also formed a familiarpattern: CBS viewersat 63 percent,ABC at 58 percent,NBC at 56 percent,CNN at 54 percent,and printmedia at 45 percent. The same questionwas asked about Europeanopinion.Perceptionsof Europeanviews are more accurateamongthe U.S. public:only 17 percentthought there had been majoritysupportamongEuropeansfor the war.Over the three months,CBS viewersmost frequentlymisperceivedEuropeanopinion (24 percent);Fox viewerswere second (20 percent).The NPR/PBSaudienceandthose relyingon printedmedia were lowest, both at 13 percent. If one adds togetherthose who thoughtthere was Europeanmajoritysupport with those who thoughtviews in Europewere evenly balanced,47 percent misperceivedEuropeanopinion;CBS viewerswere highestat 56 percent,NBC and Fox viewerswere next at 52 percentand 51 percentrespectively,while the NPR/PBS audience was lowest at 29 percent. ABC viewers and those using print sourceswere tied for second lowest at 41 percent. The Effect of Variations in Audiences The questionthusarisesof whetherthe variationin misperceptionsis a function of variationsin the demographicsor political attitudesof the audience.Some audiencesvaried accordingto education,partyidentification,and supportfor the President.However, as is evident in the regressionanalysis,when all of these factors are analyzedtogether, the respondent'sprimarysource of news is still a strong and significantfactor;indeed, it was one of the most powerful factorspredictingmisperceptions.

586 I POLITICALSCIENCEQUARTERLY

MISPERCEPTIONS AS A FUNCTION OF LEVEL OF ATTENTION TO NEWS

It would seem reasonableto assumethat misperceptionsare due to a failureto pay attentionto news and that those who have greaterexposureto news would have fewer misperceptions.All respondentswere asked, "Howclosely are you followingthe news aboutthe situationin Iraqnow?"For the summeras a whole (June,July,August-September),13 percentsaid they were followingthe news very closely, 43 percent somewhatclosely, 29 percentnot very closely, and 14 percentnot closely at all. Strikingly,overall,there was no relationbetween the reportedlevel of attention to news and the frequencyof misperceptions.In the case of those who primarilywatchedFox, greaterattentionto news modestlyincreasedthe likelihood of misperceptions.Only in the case of those who primarilygot their news from print did misperceptionsdecreasewith lower levels of attention,though in some cases this occurredfor CNN viewers as well. The most robust effects were found among those who primarilygot their news from Fox. Among those who did not follow the news at all, 42 percent had the misperceptionthat evidence of close links to al Qaeda has been found, risingprogressivelyat higherlevels of attentionto 80 percentamongthose who followedthe newsveryclosely.For the perceptionthatWMD have been found, those who watchedvery closely had the highestrate of misperceptionat 44 percent, while the other levels of attentionwere lower, thoughthey did not form a clearpattern(not at all, 34 percent;not very,24 percent;somewhat,32 percent). Among those who did not follow the news at all, 22 percentbelieved that world public opinion favored the war,jumpingto 34 percent and 32 percent among those who followed the news not very and somewhatclosely, respectively,and then jumpingeven higher to 48 percent among those who followed the news very closely. With increasingattention, those who got their news from print were less likely to have all three misperceptions.Of those not followingthe news closely, 49 percenthad the misperceptionthat evidence of close links has been found, decliningto 32 percentamongthose who followed the news very closely.Those who did not follow the news at all were far more likely to misperceive(35 percent) that WMD had been found than the other levels (not very, 14 percent; somewhat,18 percent;very, 13 percent).Twenty-fivepercentof those who did not follow the news at all had the misperceptionthat world public opinion favored the war, droppingto 16 percentfor all other categories. CNN viewersshowed slightly,but significantly,lower levels of misperception on findingWMD and world public opinion at higher levels of attention, thoughnot on evidence of links to al Qaeda. MISPERCEPTIONS AS A FUNCTION OF POLITICAL ATTITUDES

Not surprisingly,politicalattitudesdid play a role in the frequencyof misperceptions. The intention to vote for the Presidentwas highly influential.Party

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR 1

587

FIGURE6 Supportfor Presidentand Frequencyof Misperceptions

Among for: VoteGeorge W. Bush to

45%

Misperceptions of Planning Democratic Nominee

17%

Those Frequency

0

20

40

60

80

100

Percent Compositeof Polls ConductedJune-September2003 Source: Programon International PolicyAttitudes/KnowledgeNetworks,October2003.

identificationwas also influential;however, this effect disappearedafter controllingfor intention to vote for the President.

Intentionto Votefor the President The polls of June, July, and August-Septemberall includeda question,placed nearthe end, askingwhetherthe respondentsthoughtthey wouldvote for Bush or for the Democraticnominee in the presidentialelection (Figure 6). In all cases, the responseswere very similarto those in numerousother polls at the same time--and showed either a slight edge for Bush or a statisticaltie. Only 10 percentdid not answerthe question.When Bush supportersand supporters of a Democraticnominee are compared,it is clear that supportersof the President are more likely to have misperceptionsthan are those who oppose him. Multivariateanalysisindicates that intention to vote for the Presidentis the single most powerfulpredictorof misperceptions. Takingthe averagesof the percentagethat had each of the three key misperceptions- evidenceof al Qaedalinksfound,WMD found,andworldpublic opinion favorswar-those who said they wouldvote for the Presidentwere far more likelyto misperceive.On average,those who wouldvote for the President held misperceptions45 percentof the time,whilethosewho say theywillvote for a Democratheld misperceptions,on average,17 percentof the time (Figure6). Lookingat the specificcases,in responseto the question"Hasthe US found clear evidence SaddamHussein was workingclosely with al-Qaeda?"a strikingly large 68 percent of Bush supportersbelieved that the United States has

588 I POLITICALSCIENCE QUARTERLY

found such evidence. On the other side, an equally striking66 percent of supportersof a Democraticnominee knew that such evidencehas not been found. When asked to characterizethe relationshipbetween the previous Iraqi governmentand al Qaeda given four choices, 29 percentof Bush supporterssaid, "Iraqwas directlyinvolvedin the 9/11 attacks."Only 15 percentof Democratic supporterschose this description. Only minoritiesof either Bush supportersor supportersof a Democratic nominee believe that the United States has found evidence of WMD in Iraq. However, three times as many Bush supportersas Democrat supportershold this misperception.Thirty-onepercent of Bush supportersthink the United States has found such evidence,while only 10 percentof Democrat supporters think this. When asked,"Howdo you thinkthe people of the worldfeel about the US having gone to war with Iraq?"Bush supporterswere more than three times more likely than supportersof a Democraticnominee to believe that "the majority of people favor the US havinggone to war."Thirty-sixpercent of Bush supportershad this misperception,while only 11 percent of Democratic supportersdid. The PIPA/KNpolls asked the same questionabout Europe, on whichmisperceptionsare less widespreadamongAmericans.Twenty-sixpercentof Bush supportersmistakenlythoughtthat a majorityof Europeansfavored the war, while only 7 percentof supportersof a Democraticnominee believed this. Party Identification Republicansare also more likely thanDemocratsor independentsto have misperceptions.However,when the analysiscontrolsfor supportfor the President, this partydifferencelargelydisappears.For example,amongBush supporters, Republicans,Democrats,andindependentswere similarlylikelyto believe that the United States has found clear evidence that SaddamHusseinwas working closely with al Qaeda (pro-Bush Republicans,68 percent;pro-Bush Democrats,77 percent;pro-Bushindependents,67 percent). On whetherthe United States has found evidence of WMD, the same patternamongBush supporters was present (31 percent of pro-BushRepublicansbelieving such evidence has been found, 29 percent of pro-BushDemocratsbelieving this, and 29 percent of pro-Bush independentsbelieving this). The same pattern appeared in all cases tested. RELATIVE STRENGTH OF VARIOUS FACTORS RELATED TO LEVEL OF MISPERCEPTION

To determine which factors had the most power to predict the likelihood of misperceiving, we performed a binary logistic regression analysis, together with eight other factors. Four of the factors were demographic: gender, age, house-

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 589

hold income, and education.Two other categoricalfactorswere partyidentification and intention to vote for the Presidentin the next election, as opposed to an unnamedDemocraticnominee. In addition,we includedthe factors of how closely people follow events in Iraq and what their primarynews source was. The odds ratio statisticwas used to determinethe likelihoodthat respondents would have misperceptions. In the regressionanalysis,the most powerful factor was the intention to vote for PresidentBush. As comparedto those who intended to vote for the Democraticnominee or were undecided,those who intended to vote for the Presidentwere 2.9 timesmore likelyto believe thatclose linksto al Qaedahave been found, 3.0 times more likely to believe that WMD had been found, and 2.6 times more likely to believe that worldpublic opinionwas favorableto the war.Overall,those who intendedto vote for the Presidentwere 3.7 times more likely to have at least one of these misperceptions. The second most powerful factor was one's primarysource of network news. Analysisshowsthe factorto be highlysignificant,but assessingeach network is difficult.Though several networks are significant,others are not. To determinethe relativeimportanceof each networkas a primarysourceof news, anotherregressionwas performed,treatingeach networkas a binaryvariable and comparingeach network'srespondentsto other respondents.When this analysisis performed,having Fox, CBS, or NPR/PBS as one's primarynews source emerges as the most significantpredictorof a particularmisperception and of misperceptionsin general. To determinethe overallimportanceof one factorto another,a comparison of statisticalmeasuresis necessary.10 Overall,Fox viewinghas the greatestand most consistentpredictivepower in the analysison a varietyof these statistical measures.Table 9 presentsthe results. Fox is the most consistentlysignificantpredictorof misperceptions.Those who primarilywatchedFox were 2.0 timesmorelikelyto believe thatclose links to al Qaeda have been found, 1.6 times more likely to believe that WMD had been found, 1.7 times more likely to believe that world public opinionwas favorableto the war,and2.1 timesmorelikely to have at least one misperception. Interestingly,when asked how the majorityof people in the world feel about the war,if the response"viewsare evenlybalanced"is includedas a misperception alongwith "favor,"only Fox is a significantpredictorof thatmisperception. Those who primarilywatched CBS were 1.8 times more likely to believe that close links to al Qaeda have been found, 1.9 times more likely to believe that world public opinion was favorableto the war, and 2.3 times more likely to have at least one misperception.However,they were not significantlydifferent on beliefs about the uncoveringof WMD. On the other hand, those who primarilywatchedPBS or listened to NPR were 3.5 times less likely to believe that close links to al Qaeda have been 'oPIPA compared two statisticalmeasures,the Wald statisticand the difference in the -2 log likelihood if the factor is removed from the analysis.

590 1 POLITICAL SCIENCE QUARTERLY

TABLE9

in Misperceptions NewsSource SignificantVariances by Primary MediaSource Primary

Odds/Ratio

N

Fox

1.6

361

CBS*

1.4

182

NPR/PBS*

-1.3

53

US has foundclearevidenceof Iraqilinkto al Qaeda (N = 2,202)

Fox CBS NPR/PBS

2.0 1.8 -3.5

366 188 59

of worldfavorsUS Majority havinggone to warwithIraq (N = 1,827)

Fox CBS NPR/PBS

1.7 1.9 -5.6

294 168 55

Fox

2.1

414

CBS

2.3

213

-3.8

66

US has foundWMD (N = 2,202)

At leastone misperception (N = 2,506)

NPR/PBS

Networks. Source:Programon International PolicyAttitudes/Knowledge * Notstatistically significantat the 0.05 level.Allothermediasourcesdidnotvarysignificantly. Fordata, please contactthe authors.

found, 5.6 times less likely to believe that world public opinion was favorable to the war, and 3.8 times less likely to have at least one misperception.However, they were not significantlydifferenton the issue of WMD. Level of attentionto news was not a significantfactor overall,with the exception of those who primarilygot their news from Fox. This is consistentwith the findingthat Fox viewerswere more likely to misperceivethe more closely they followedeventsin Iraq.Multiplicativevariableswere derivedfor each network by multiplyingattention to news by each network dummyvariable. A multivariateanalysiswas performedon misperceptionsin whicheach new combined network-attentionlevel variablewas added to the previousmodel. The resultsshow that Fox viewers are the only ones to be significantlymore likely to misperceivewith higherlevels of attentionto news. The third most powerfulfactor was intention to vote for the Democratic nominee. As comparedto those who intended to vote for PresidentBush or were undecided,those who intendedto vote for the Democraticnominee were 2.0 times less likely to believe that close linksto al Qaedahave been found and 1.8 times less likely to believe that world public opinion was favorableto the war.Overall,those who intendedto vote for the Democraticnominee were 1.8 times less likely to have at least one of these misperceptions,but did not quite achieve significanceon the WMD question. The fourthmost powerfulfactorwas education.Those who had no college, as comparedto those had at least some college, were 1.3 times more likely to believe that close links to al Qaeda have been found and 1.4 times more likely to have at least one misperception,but did not quite achievesignificanceon the other misperceptions.

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 591

Age was a very weak factor,with olderpeople beingvery slightlyless likely to misperceive.All other factors-gender, partyidentification(when intention to vote for the President was included), level of attention to news, and income-were not significant.In a separateanalysis,region of the countrywas includedand also not found to be significant. ANALYSIS

These data lead to the question of why so many Americanshave misperceptions that appearto be havinga significantimpacton attitudesabout the Iraq war and why these misperceptionsvaryaccordingto one's source of news and political attitudes.This analysisstartswith possible explanationsbased on exogenous factorsand then moves inward. The firstand most obviousreasonthat the publichad so manyof these misperceptionsis that the Bush administrationmade numerousstatementsthat could easilybe construedas assertingthese falsehoods.On numerousoccasions the administrationmade statementsstronglyimplyingthat it had intelligence substantiatingthat Iraq was closely involved with al Qaeda and was even directlyinvolvedin the September11 attacks.For example,in his 18 March2003 PresidentialLetter to Congress,PresidentBush explainedthat in going to war with Iraq he was taking "the necessaryactions againstinternationalterrorists and terroristorganizations,includingthose nations, organizations,or persons who planned, authorized,committed, or aided the terroristattacks that occurred on September 11, 2001.""When Secretaryof State Colin Powell addressed the UN Security Council on 5 February2003, he presented photographs that were identified as al Qaeda trainingcamps inside Iraq, leaving unclearthe fact that the campin questionwas in the northernpartof Iraq,not underthe controlof the centralIraqigovernment.12 Administrationfigurescontinued to refer to the purportedmeeting between MohammedAtta and an Iraqi official in Prague even after U.S. intelligence agencies establishedthat Atta was in fact in the United Statesat the time.13Morerecently,on 14 September 2003,Vice PresidentRichardCheneymadethe followingambiguousstatement: "Ifwe're successfulin Iraq ... so that it's not a safe haven for terrorists, now we will have strucka majorblow rightat the heartof the base, if you will, the geographicbase of the terroristswho have had us under assaultnow for Sometimesthe associationhas been manyyears,but most especiallyon 9/11."O4 established by inserting a reference to September 11 that is a non sequitur and 11

President George W. Bush, "Presidential Letter," 18 March 2003, available at http://www. whitehouse.gov/news/releases/2003/03/20030319-1.html. 12 Secretary Colin L. Powell, "Remarksto the United Nations," New York City, 5 February2003, available at http://www.state.gov/secretary/rm/2003/17300.htm, 12 October 2003. 13Dana Priestand Glenn Kessler,"Iraq,9/11 Still LinkedBy Cheney,"WashingtonPost, 29 September 2003. 14Vice President Richard Cheney, "Meet the Press," 14 September 2003.

592

I POLITICAL SCIENCE QUARTERLY

then simplymoving on, or implyingthat the connectionis so self-evidentthat it does not require explanation.For example, PresidentBush's own remarks at his press conferenceof 28 October2003 could appearto reinforcemultiple misperceptions: The intelligencethatsaidhe [SaddamHussein]hada weaponsystemwasintelligencethathadbeenusedby a multinational agency,the U.N.,to passresolutions. It's been usedby my predecessorto conductbombingraids.It was intelligence gatheredfroma varietyof sourcesthatclearlysaidSaddamHusseinwasa threat. And giventhe attacksof Septemberthe 11th-it was-you know,we neededto enforceU.N. resolution(sic)for the securityof the world,andwe did.We took actionbasedupongood,solidintelligence.It wasthe rightthingto do to make Americamoresecureandthe worldmorepeaceful."5 Here the listener could mistakenlyinterpretthe President'scomments as meaning that the same intelligencethat determinedthe United States' policy on warhad been acceptedas correctby the UN SecurityCouncilin its deliberations and that the September11 attacks,a UN SecurityCouncilresolution,and the choice to invade Iraqall followed a logical progression("giventhe attacks of Septemberthe 11th-it was-you know"). In anycase, it is quite clearthatthe publicperceivedthat the administration was assertinga stronglinkbetweenIraqand al Qaeda,even to the point of Iraqi directinvolvementin September11. When PIPA/KN asked in June, "Do you thinkthe Bush administration did or did not implythatIraqunderSaddamHussein was involvedin the September11thattacks?"71 percentsaid that it had. The administrationalso made statementsthat came extremelyclose to assertingthatWMD were foundin postwarIraq.On 30 May2003,PresidentBush madethe statement,". .. for those who say we haven'tfoundthe bannedmanufacturingdevices or bannedweapons, they'rewrong.We found them."16 Another possible explanationfor why the public had such misperceptions is the way that the media reportedthe news. The large variationin the level of misperceptionsdoes suggest that some media sources may have been making greaterefforts than others to disabusetheir audiencesof misperceptionsthey may have had so as to avoid feeling conflictaboutgoing or havinggone to war. Of course,the presenceor absenceof misperceptionsin viewersdoes not necessarilyprove that they were caused by the presence or absence of reliable reportingby a news source. Variationsin the level of misperceptionsaccording to news source may be related to variations in the political orientations of the audience. However, when political attitudes were controlled for the variations 15Full transcript:"BushDefends Foreign Policy,"WashingtonPost, online edition, 28 October2003, available at http://www.washingtonpost.com/ac2/wp-dyn?pagename=article&node=&contentld= A29127-20030ct28¬Found=true. 16 Mike Allen, "Bush: 'We Found' Banned Weapons; President Cites Trailers in Iraq as Proof," WashingtonPost, 31 May 2001.

THE MEDIA,AND THE IRAQWAR MISPERCEPTIONS,

I 593

between the networksand the same attitudesstill obtained,it suggeststhat differences in reportingby media sourceswere playinga role. There is also evidence that in the run-upto, during,and for a period after the war, many in the media appearedto feel that it was not their role to chalor thatit waseven appropriateto takean activepro-war lenge the administration News' Fox posture. programmingon the war included a flag in the left-hand corner and assumedthe Defense Department'sname for the war:"Operation Iraqi Freedom."When criticizedin a letter for taking a pro-warstance, Fox News' Neil Cavutoreplied,"So am I slantedand biased?You damnwell bet I am.... You say I wear my biases on my sleeve? Better that than pretendyou have none, but show them clearly in your work.""Interestingly,even CBS News, which tends to have a more liberal reputation,seemed to think along these lines. CBS anchorDan Rather commentedin a 14 April 2003 interview with LarryKing, "Look, I'm an American.I never tried to kid anybodythat I'm some internationalistor something.And when my countryis at war,I want my countryto win. .... Now, I can'tanddon't arguethatthatis coveragewithout a prejudice.About that I am prejudiced."'8 A study of the frequenciesof pro-warand anti-warcommentatorson the major networks found that pro-war views were overwhelminglymore frequent.'9In such an environment,it would not be surprisingthat the media would downplaythe lack of evidence of links between Iraq and al Qaeda, the fact thatWMD were not beingfound,andthatworldpublicopinionwas critical of the war.Furthermore,the fact shownin the presentstudythat the audiences of the variousnetworkshave variedso widely in the prevalenceof misperceptions lends credence to the idea that media outlets had the capacityto play a more criticalrole, but to varyingdegrees chose not to. Reluctantto challengethe administration,the media can simplybecome a means of transmissionfor the administration,ratherthan a criticalfilter.For example,when PresidentBush made the assertionthat WMD had been found, the 31 May 2003 edition of the WashingtonPost ran a frontpage headlinesaying, "Bush:'We Found' Banned Weapons.'"20 There is also strikingevidence that the readinessto challengethe administrationis a variablethat correspondsto levels of misperceptionamongviewers. The aforementionedstudy of the frequencyof commentarycriticalof the war found that the two networksnotablyleast likely to presentcriticalcommentary were Fox and CBS-the same two networksthatin the presentstudyhadview17 David Folkenflik,"FoxNews defends its patrioticcoverage:Channel'sobjectivityis questioned," BaltimoreSun, 2 April 2003. 18Dan Rather, during the 4 April 2003 Larry King Show. Quoted in Steve Rendell and Tara Broughel, "AmplifyingOfficials, SquelchingDissent: FAIR study finds democracy poorly served by war coverage," Extra! (May/June2003), Fairness and Accuracy in Reporting, available at www.fair. org/extra/0305/warstudy.html. 19 Steve Rendell and Tara Broughel, "AmplifyingOfficials,SquelchingDissent." 20Allen, "Bush:'We Found' Banned Weapons," 31 May 2001.

594

I POLITICAL SCIENCE QUARTERLY

ers most likely to have misperceptions.This is clarifiedby statisticsfrom Rendell and Broughel'scontent analysisof networkcoverage:"Thepercentageof U.S. sourcesthat were officialsvariedfrom networkto network,rangingfrom 75 percent at CBS to 60 percent at NBC.... Fox's Special Report with Brit Hume had fewer U.S. officialsthan CBS (70 percent) and more U.S. anti-war guests (3 percent)than PBS or CBS. Eighty-onepercentof Fox's sourceswere pro-war,however,the highestof any network.CBS was close on the Murdoch network'sheels with 77 percent.NBC featuredthe lowest proportionof prowar voices with 65 percent."21 Another contributingfactormayalso have been a dynamicin reportingthat is not unique to the Iraq war:the absence of somethingdoes not constitute a compellingstory, while even the prospect of the presence of somethingdoes. Thus, shortly after the end of the war, numerousheadlines trumpetedeven faintprospectsthat evidenceof WMD were aboutto be found.However,when these prospectsfailed to materialize,this did not constitutea compellingstory and, thus, reportingon it was given a far less prominentposition.The cumulative effect of repeatedlyhearingthe expectationthat weapons were about to be found, while hearinglittle or no disconfirmation,could well contributeto the impressionthat at least one of these leads was indeed fruitful. Other more subtle dynamicsmay also have been at work. The fact that world public opinion was so opposed to the United States going to war with Iraq may have been obscuredby givingsuch high visibilityto the U.S. conflict with Francein the SecurityCouncil.The key story became one of Frenchobstructionism,eclipsingthe fact that polls from aroundthe world, as well as the distributionof positionsin the UN SecurityCouncil,showed widespreadopposition to U.S. policy.22 One could well argue that this plethoraof exogenous factorsobviates the need for any explanationsbased on endogenousfactors.Indeed, the fact that no particularmisperceptionstudiedwas found in a clear majorityof the public and the fact that 40 percenthad none of the key misperceptionsbuttressconfidence in the capacityof the public to sort throughmisleadingstimuli.At the sametime, a majorityhad at least one majormisperception,raisingthe question of why so many people have been susceptible. The seeminglyobvious explanation--that the problemis that people just do not pay enough attention to the news--does not hold up. As discussed, 21 See footnote 19. Forthcomingstudies by Susan Moeller are likely to offer a much more comprehensive view of these dynamicsthan is available at the time of writing.A report on media coverage of WMD under the aegis of the Center for Internationaland Strategic Studies at the University of Marylandis in preparationfor release in early 2004. See also Susan Moeller, "A Moral Imagination: The Media's Response to the War on Terrorism"in Stuart Allen and Barbie Zelizer, eds., Reporting War (London: Routledge, forthcoming). On the issue of embedded reporters,see a content analysis by the Project for Excellence in Journalism,"Embedded Reporters:What Are Americans Getting?" at www.journalism.org/resources/research/reports/war/embed/default.asp. 22 See footnote 6.

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR

I 595

higherlevels of attentionto news did not reduce the likelihood of misperception, andin the case of those who primarilygot theirnews fromFox News, misperceptionsincreasedwithgreaterattention.Furthermore,the presenceof misperceptions was not just noise found randomly throughout an inattentive public-the presence of misperceptionsformed strongpatternshighlyrelated to respondents'primarysource of news. Perhaps the most promisingexplanationis that the misperceptionshave performedan essential psychologicalfunctionin mitigatingdoubts about the validityof the war.Polls have shown that Americansare quite resistantto the idea of using militaryforce except in self-defense or as part of a multilateral Even if a countryis developingnuclearweapons, operationwithUN approval.23 there is not a consensusin the public that the United States would have the right to use militaryforce to prevent it, though a very strongmajorityagrees that the UN SecurityCouncilwould have this prerogative.24 Thus, to legitimate the war without UN approval,the President had to make the case that the warwouldbe an act of self-defense.The waragainstthe Talibanhad been overwhelminglyapprovedas legitimatebecause the Taliban had providedsupportto al Qaeda and, thus, was a partyto the September11 attack on the United States. Americansshowed substantialreceptivityto the administration'sassertionthat Iraqalso had linksto al Qaeda and that the possibilitythat Iraqwas developingWMD that couldbe passedto al Qaeda,creating a substantialthreat to the United States. But the public also appearedto recognize that the evidence was circumstantial-and this was not a president who commandedso muchrespectin foreignpolicy realmsthat they could simply take his word for it. The public felt the need for UN approvalas an alternatenormativebasis for war. Early polls showed that a very strongmajoritywas ready to act with UN approval,but less than a thirdwere readyto act unilaterally,and even days before the war a majoritywas still sayingthat the United Statesneeded to wait for UN approval.25 But even monthsbefore the war,a clearmajorityof the pub23 Evidence from the

1990s is reviewed in Steven Kull and I.M. Destler, Misreadingthe Public: The a New Isolationism Myth of (Washington,DC: Brookings Institution Press, 1999), 42-57, 67-80, 94110. The public's views at the outset of the currentIraq experience are documented and analyzed in the report "IraqDebate 2002," available at www.americans-world.org. 24 In January2003, PIPA/KN asked a series of general questions about whether a right existed "to use militaryforce to prevent a countrythat does not have nuclearweapons from acquiringthem."Only 46 percent thoughtthat, without UN approval,a countryhad the rightto use militaryforce on another country in this situation;virtuallythe same number (48 percent) thought the United States had this right. Seventy-six percent thought the UN Security Council had the right to authorize militaryforce for this purpose. PIPA/Knowledge Networks Poll, "Americanson Iraq and the UN InspectionsI," 27 January2003, available at www.pipa.org/online_reports.html. 25 For early polls, see footnote 1. Just days before the war in a CBS News poll conducted4-5 March, only 36 percent agreed that "Iraqpresents such a clear danger to American interests that the United States needs to act now," while 59 percent agreed that "The US needs to wait for approval of the United Nations before taking any action against Iraq."

596

I POLITICAL SCIENCE QUARTERLY

lic said that if the Presidentwere to decide to go to war withoutUN approval they wouldsupporthim,26andwhen the time camethey did.Thiswas a standard effect, no doubt intensifiedby a felt imperativeto rally-round-the-president close ranksin the post-September11 environment.27 Such rally effects, though, are fairlysuperficial.Even duringand after the war, when asked whether they really approve of the decision to go to war as distinguishedfrom just supportingBush "becausehe is the president,"only about half or less have said they think that going to war was the best thing, while another15 to 22 percenthave said that their approvalof the warwasjust a way to supportthe president.28 Americanshad expected that once the United States went into Iraq, they would find evidence that Iraq was linked to al Qaeda and was developing WMD, thus vindicatingthe decision to go to war as an act of self-defense. Therefore,it is not surprisingthatmanyhave been receptivewhen the administrationhas stronglyimplied or even assertedthat the United States has found evidence that Iraq was working closely with al Qaeda and was developing WMD, andwhen media outlets- some more than others- have allowedthemselves to be passive transmittersof such messages. CONCLUSION

Fromthe perspectiveof democraticprocess,the findingsof this studyare cause for concern.They suggestthat if the publicis opposed to takingmilitaryaction 26 In December 2002, January2003, and February,PIPA/KN presented respondents the following scenario: "Imaginethat President Bush moves that the UN approve an invasion of Iraq to overthrow Saddam Hussein, but most of the other members of the UN SecurityCouncil want to continue to use threats and diplomaticpressure to get Iraq to comply, and the motion does not pass. President Bush then decides that the US will undertake an invasion of Iraq, even if the US has to do so on its own. Just based on this information,what do you thinkyour attitudewould be about this decision?"Respondents were offered the option of agreeingwith the President'schoice, disagreeing,or choosing "Iwould not agree with this decision, but I would still supportthe President."In all cases only a minorityof 33 to 43 percent said they would agree, but another 25 to 27 percent said they would support but not agree, thus creating a majorityready to support the President should he decide to proceed. 27 On the rallyeffect, see RichardA. Brody, "Crisis,War and Public Opinion:The Media and Public Support for the President"in W. Lance Bennett and David L. Paletz, Taken by Storm: The Media, Public Opinion,and U.S. ForeignPolicy in the Gulf War(Chicago:University of ChicagoPress, 1994); R.A. Brody and C.R. Shapiro, "A Reconsideration of the Rally Phenomenon in Public Opinion" in S. Long, ed., Political Behavior Annual, vol. 2 (Boulder, CO: Westview Press, 1989); and John E. Mueller, War,Presidentsand Public Opinion (New York: Wiley, 1973):208-213. 28Seven times in March and April 2003, Pew Research Center for the People and the Press asked, "Do you think the US made the right decision or the wrong decision in using military force against Iraq?"Those who said it was the right decision were asked whether they supported going to war because they think it was "the best thing for the US to do" or whether they were not sure if it was the best thing to do but they "supportBush's decision, because he is the president."During the war, 69 to 74 percent said the United States made the right decision, of which 48 to 54 percent thought it was the best thing to do, while 15 to 22 percent were unsure of this but supported the President. Each month from May through September, PIPA/KN repeated this question. Over this period, 45 to 53 percent thought that the war was the best thing to do, and 14 to 18 percent were unsure but supported the President.

MISPERCEPTIONS, THE MEDIA, AND THE IRAQ WAR 1

597

withoutUN approvaland the Presidentis determinedto do so, he has remarkable capacitiesto move the public to supporthis decision.This in itself is not worrisome-to the degree it is the productof persuasion,based on the merits of an argument.Whatis worrisomeis that it appearsthat the Presidenthas the capacityto lead membersof the publicto assumefalse beliefs in supportof his position. In the case of the Iraq war, this dynamicappearsto have played a criticalrole: among those who did not hold the key false beliefs, only a small minoritysupportedthe decisionto go to war.In a regressionanalysis,the presence of misperceptionswas the most powerfulfactorpredictingsupportfor the war,with intentionto vote for the Presidentclose behind.This does not prove that the misperceptionsalone causedsupportfor the war.It is more likely that it is one key factorthat interactedwith the desireto tally aroundthe President and the troops. However, it does appearthat it would have been significantly more difficultfor the Presidentto elicit and maintainsupportfor the decision to go to war if the public had not held such misperceptions. The President'sinfluenceis not limitless.He does not appearto be capable of gettingthe publicto go againsttheir more deeply held value orientations.If he did, then it would not be necessaryfor the public to develop false beliefs. But he is capable of promptingthe public to supporthim by developing the false beliefs necessaryto justify the administration'spolicies in a way that is consistentwith the public'sdeeper value orientations. It also appearsthat the mediacannotnecessarilybe countedon to play the criticalrole of doggedly challengingthe administration.The fact that viewers of some media outlets had far lower levels of misperceptionsthan did others (even when controllingfor politicalattitudes)suggeststhatnot all were making the maximaleffort to counterthe potentialfor misperception. To some extent, this period may be regardedas unique.We are still living in the aftermathof September11.Withthe persistingsense of threat,the public may be more prone to try to accommodatethe President,and the media may be more reluctantto challengethe Presidentor to impartnews that calls into question the validityof his decisions.And yet, it is also at times of threat that the most criticaldecisionsare likely to be made. It is likely that with time, publicmisperceptionswill tend to erode. For example, aftermedia coverageof David Kay'sinterimprogressreporton the activities of the Iraq Survey Group, the belief that WMD have been found droppedto 15 percent, althoughthe belief that evidence of links to al Qaeda has been found did not drop. At the same time, there was a significantrise in the percentagethat said they thoughtthat the Presidentat least stretchedthe truthwhen he made the case for war based on Iraqhavinga WMD program.29 However,when the mechanismsfor informingthe publicare in some way compromised,the process of the publicgraduallycatchingon is a slow one. In the meantime, the administration,by giving incorrectinformation,can gain sup29

See Steven Kull, "AmericansReevaluate Going to War with Iraq,"PIPA/KnowledgeNetworks Poll, 13 November 2003, available at www.pipa.org.

598

I POLITICAL SCIENCE QUARTERLY

port for policies that might not be consistent with the preferences held by the majority of Americans. APPENDIX

Methodology The poll was fielded by KnowledgeNetworks--a polling, social science, and market researchfirmin Menlo Park,California-with a randomlyselected sampleof its largescale nationwideresearchpanel.Thispanelis itselfrandomlyselectedfromthe national populationof householdshavingtelephonesandsubsequentlyprovidedinternetaccess for the completionof surveys(and, thus, is not limited to those who alreadyhave internet access).The distributionof the samplein the web-enabledpanel closely tracks the distributionof United States Censuscounts for the U.S. populationon such variables as age, race,Hispanicethnicity,geographicalregion,employmentstatus,income, and education.The panel is recruitedusing stratifiedrandomdigit-dial(RDD) telephonesampling.RDD providesa non-zeroprobabilityof selectionfor everyU.S. household havinga telephone.Householdsthatagreeto participatein the panel areprovided with free Web access and an Internetappliancethat uses a telephone line to connect to the Internetand uses the televisionas a monitor.In return,panel membersparticipate in surveysthreeto fourtimesa month.Surveyresponsesareconfidential,andidentifyinginformationis never revealedwithoutrespondentapproval.When a surveyis fielded to a panel member,he or she receives an e-mail indicatingthat the survey is availablefor completion.Surveysare self-administered.For more informationabout the methodology,please go to www.knowledgenetworks.com/ganp.