Abstrct U.S. Muslims fter 9/11: Poll Trends 2001-2007 by Clrk McCuley nd Jennifer Stellr The terrorist ttcks on September 11 th 2001 brought incresed ttention to Muslims living in the United Sttes. Results from four ntionl polls of Muslim Americns conducted between 2001 nd 2007 indicte tht Muslim Americns feel incresingly negtive bout the direction in which Americ is heding nd incresingly see the wr on terrorism s wr on Islm. Introduction Since the events of September 11 th, 2001, substntil reserch hs been devoted to the thoughts nd feelings of Americns towrds Muslims living in the United Sttes. A recent nlysis of polls conducted in the U.S. from 2000 to 2006 showed tht reltively tolernt tmosphere of Americn sentiment towrds Muslim Americns immeditely fter September 11 th gve wy to incresing feelings of concern nd distrust s time pssed [1]. In this report we nlyze polls of Muslims living in the United Sttes in order to determine wht chnges hve occurred since 9/11 in U.S. Muslims views of Americ nd Americns [For the methodology utilized, see Appendix]. Results We serched the four polls described in the Appendix to find where the sme or substntively similr questions were used in more thn one survey. The Tbles note where items or responses were not identiclly worded. We considered ten repeted items nd divided them into four content res. Politicl prticiption nd stisfction With the Globl Wr on Terrorism coming to the forefront in domestic politics s well s foreign policy, Muslim voters re bringing their voices into the public ren. The Pew Center estimtes the popultion of U.S. Muslims of voting ge t 1.4 million. In res of high concentrtion, Muslims hve the potentil to hve significnt impct on elections. Polls in 2001, 2004, nd 2007 indicte tht the mjority of Muslims re registered to vote (79%, 82%, 67%). (W!!!!!!!!!!! -'./'&0'1!2334
Tble 1. Are you registered to vote? (percentges) 2001 2004 Pew 2007 Yes 79 82 67 No 21 17 30 NS, DK, Refused 1 1 3 It ppers tht Republicn registrtions my hve fllen (17%, 18%, 9%, 7%) s more Muslim Americns identified themselves s independents or in lignment with minor prties such s the Libertrins (21%, no dt, 24%, 35%). Tble 2. Wht is your politicl prty? (percentges) 2001 2002 N=531 2004 Pew 2007 Democrt 40 36 40 37 Republicn 17 18 9 7 Independent/Minor Prty 21 -- 24 35 NS, DK, Refused 33 46 28 21 reports Independent, Other, nd Refused s 35%, No Response s 11%. A disturbing trend is the growing Muslim disstisfction with the wy things re going in Americn society. Polls in 2001, 2004 nd 2007 show tht disstisfction grew shrply (38%, 61%, 55%). Tble 3. How stisfied re you overll with the wy things re going in Americn society tody? (percentges) 2001 2004 Pew2007 Stisfied 52 35 37 Unstisfied 38 61 55 NS, DK, Refused 10 4 8 (V!!!!!!!!!!! -'./'&0'1!2334
Overll, re you stisfied or disstisfied with the wy things re going in this country tody? Importnce of Islm Polls in 2001, 2002, 2004 nd 2007 show tht the importnce of Islm in the lives of Muslim Americns hs remined consistently high for the gret mjority of Muslims (79%, 81%, 82%, nd 72% sy very importnt ). Tble 4. Would you sy the role of Islm in your life is very importnt, somewht importnt or not very importnt? (percentges) 2001 2002 N=531 2004 Pew2007 b Very importnt 79 81 82 72 Somewht importnt 16 13 14 18 Not importnt 5 5 4 9 NS, DK, Refused 1 1 1 1 Originl scle from 1= not importnt to 10 = very importnt. Responses converted: 1-3 = Not importnt; 4-6 = Somewht importnt; 7-10 = Very importnt. b How importnt is religion in your life? Responses converted: not too importnt nd not t ll importnt = Not importnt. The sme polls show tht bout hlf of Muslims (54%, 51%, 54%, 40%) ttend mosque t lest once week. Note the stbility of results (54-54%) in contrst to the lower report of weekly mosque ttendnce from Pew2007 (40%); the lower percentge in the Pew poll is consistent with our suggestion in the Methods section tht the Pew poll my hve better represented Muslims who re less religious. (T!!!!!!!!!!! -'./'&0'1!2334
Tble 5. On verge, how often do you ttend the mosque for Slh nd Jum'h Pryer? (percentges) 2001 2002 N=531 2004 Pew 2007 b More thn once week 31 22 29 17 Once week for Jum h pryer 24 29 25 23 Once or twice month 10 11 10 8 A few times yer, especilly for Slh 14 16 16 18 Seldom 9 11 9 16 Never 11 11 10 18 NS, DK, Refused 1 1 1 0 Approximtely how often do you ttend mosque for pryer? b On verge, how often do you ttend the mosque or Islmic center for Slh nd Jum h pryer? It is lso worth noting tht the religiosity of Muslim Americns is only slightly higher thn tht reported in ntionl surveys of ll Americns. Polls tken between 1996 nd 2004 show 59% - 64% of Americns sying tht religion is very importnt in their lives, nd 38% - 43% sying they ttend religious services t lest once week. [2]. Perception of Bis nd Discrimintion The belief tht Islm nd Muslims re unfirly portryed in the medi shows some indiction of decline: 77% believed this in 2001, 76% in 2004, nd 57% in 2007. Explntion of chnge towrd more positive view of the medi between 2004 nd 2007 is not obvious perhps due to growing medi criticism of the wr in Irq s the presidentil election pproched -- but the importnt result is tht consistently over the yers more thn hlf of U.S. Muslims feel the medi re bised ginst them. (S!!!!!!!!!!! -'./'&0'1!2334
Tble 6. Do you think the medi is fir in its portryl of Muslims nd Islm? (percentges) 2001 2004 Pew2007 Yes fir 13 17 26 No bised ginst 77 76 57 Bised in fvour/depends -- -- 6 NS, DK, Refused 10 7 11 Do you think tht coverge of Islm nd Muslims by Americn news orgniztions is generlly fir or unfir? Response converted: Unfir = Bised ginst. In contrst, report of personl experience of discrimintion shows some indiction of pek in 2004: 26% reported discrimintion in 2002, 40% in 2004 nd 27% in 2007. Agin, explntion of this vrition is not obvious. Increse from 2002 to 2004 my be function of the phrsing of the items, which sked bout discrimintion experienced since 9/11. Tble 7. Aside from restrictions on religious expression t work, hve you yourself suffered nti- Muslim discrimintion, hrssment, verbl buse, or physicl ttck since Sept. 11? 2002 N=531 2004 Pew 2007 b Yes 26 40 27 No 74 59 71 NS, DK, Refused 0 1 1 Hve you personlly experienced discrimintion since September 11th? b And thinking more generlly - NOT just bout the pst 12 months - hve you ever been the victim of discrimintion s Muslim living in the United Sttes? Any experiences of discrimintion between 2002 nd 2004 would contribute to n incresed reporting of discrimintion. However, this interprettion contributes nothing to understnding the drop from 40 to 27 percent reporting discrimintion in 2007; indeed the Pew 2007 item sks bout ever being victim of discrimintion in the U.S., suggesting tht, if nything, the Pew2007 percentge should be higher thn the 2004 percentge. Despite this uncertinty bout vrition over time, there is n importnt consistency in finding tht t lest qurter of U.S. Muslims report personl experience of nti-muslim discrimintion. (4!!!!!!!!!!! -'./'&0'1!2334
It is worth noting tht the percentges of Muslims reporting experience of discrimintion (26%-40%) re below the 46% of Africn-Americns who reported in 2007 tht they were victims of discrimintion [3]. Nevertheless, Muslims in the U.S. report significnt levels of discrimintion. Wr on Terrorism Polls in 2001, 2002, 2004, nd 2007 show tht Muslim Americn for U.S. militry ction in Afghnistn decresed (52%, 53%, 35%, 35%). Tble 8. Do you strongly support, somewht support, somewht oppose or strongly oppose the U.S. militry ction ginst Afghnistn? (percentges) 2001 2002 b N=531 2004 Pew 2007 c Support 52 53 35 35 Oppose 43 42 53 48 NS, DK, Refused 6 6 11 17 Responses converted: support nd somewht support = Support; somewht oppose nd oppose = Oppose. b U.S. militry ction in Afghnistn fter Sept 11 ws justified under the circumstnces. Do you strongly gree, somewht gree, somewht disgree, or strongly disgree? Responses converted: strongly gree nd somewht gree = Support; somewht disgree nd strongly disgree = Oppose. c Do you think the U.S. mde the right decision or the wrong decision in using militry force in Afghnistn? Responses converted: right decision = Support; wrong decision = Oppose. Events between 2002 nd 2004 my explin the shift in sentiment. The entrnce of the opposition forces into Kndhr nd the end of the Tlibn uthority in tht region in December of 2001 ppered to open n opportunity for democrcy nd stbility in Afghnistn. However, Tlibn resurgence in Afghnistn nd stgntion in the wr in Irq my hve contributed to incresing doubts bout U.S. militry opertions in Afghnistn. In Gllup polls conducted in 2001, 2002, 2004 nd 2007, ntionl smples of Americns were sked whether the U.S. hd mde mistke in sending militry forces to Afghnistn. The percentge sying it ws mistke incresed (9%, 6%, 25%, 25%) [4]. Thus opposition to the wr in Afghnistn is consistently t lest 25 percentge points higher for Muslim Americns thn for other Americns. R3!!!!!!!!!!! -'./'&0'1!2334
Tble 9. Do you strongly support, somewht support, somewht oppose, or strongly oppose the wr in Irq? (percentges) Pew 2004 2007 b Support 13 12 Oppose 88 75 NS, DK, Refused 6 13 Responses converted: strongly support nd somewht support = Support; somewht oppose nd strongly oppose = Oppose. b Do you think the U.S. mde the right decision or the wrong decision in using militry force ginst Irq? Responses converted: right decision = Support; wrong decision = Oppose. Only 2004 nd Pew2007 inquired bout Muslim opinions towrd the wr in Irq, but the gret mjority of respondents (88% nd 75%) opposed this wr nd few supported it (13%, 12%). In contrst, USA Tody/Gllup polls (http://www.gllup.com/poll/106783/opposition-irqwr-reches-new-high.spx) show tht the percentge of Americns sying tht the U.S. mde mistke in sending troops to Irq rnged between 40% nd 50% in 2004, rising to bout 60% in 2007 [5]. As for the wr in Afghnistn, Muslim opposition to the wr in Irq is consistently higher thn for other Americns, t lest 15 percentge points higher. A centrl question for U.S. Muslims is whether the US is fighting wr ginst terrorism, or whether the Globl Wr on Terrorism is wr on Islm. Muslim Americns show striking decrese in belief tht the US wr is ginst terrorism (from 67% in 2001 to 26% in 2007), nd prllel increse in the belief tht the wr is ctully bout Islm (from 18% in 2001 to 55% in 2007). R,!!!!!!!!!!! -'./'&0'1!2334
Tble 10. In the ftermth of the September 11 ttcks, do you feel the U.S. is fighting wr on terrorism or wr ginst Islm? (percentges) 2001 2002 N=531 2004 b Pew 2007 c Terrorism 67 41 33 26 Islm 18 31 38 55 Neither/Both -- 20 -- 2 NS, DK, Refused 16 7 29 17 Some describe the U.S. worldwide response to the Sept. 11 ttcks s wr on terrorism. Others sy it is wr on Islm. Which do you think is more ccurte? b Do you feel the U.S. is fighting wr on terrorism or wr ginst Islm? c Do you think the U.S.-led wr on terrorism is sincere effort to reduce interntionl terrorism or don t you believe tht? Responses converted: sincere effort = Terrorism; don t believe tht = Islm. Conclusion Polls of minority groups re mde difficult nd expensive by the need to screen out mjority members from rndom smpling of respondents. Muslims mke up bout one hlf of one percent of the U.S. popultion, very smll minority for which ordinry smpling methods re prohibitively expensive. Nevertheless we found four ntionl polls of U.S. Muslims conducted between 2001 nd 2007, nd looked for trends over time for items tht were repeted or substntilly the sme cross polls. The four polls show both stbility nd chnge in U.S. Muslims. Results stble over time include lrge mjorities registered to vote, lrge mjorities sying tht Islm is very importnt in their lives, lrge mjorities opposing the wr in Irq, nd simple mjority reporting ttendnce t mosque t lest once week. Indictions of smll chnges included switching from Republicn to Independent prties, possible pek in personl experience of discrimintion round 2004, nd possible decrese in perceived medi bis ginst Muslims between 2004 nd 2007. Lrge nd consistent chnges over time too lrge in our judgment to be ttributed to ny kind of smple vritions were found for three issues. Stisfction with the wy things re going in the U.S. declined from 52% in 2001 to 35% in 2004 to 12% in 2006. Support for U.S. militry ction in Afghnistn declined from 52% - 53% in 2001 nd 2002 to 35% in 2004 nd 2007. Perception of the wr on terrorism s wr on Islm incresed from 18% in 2001, to 31% in 2002, to 38% in 2004, nd to 55% in 2007. R2!!!!!!!!!!! -'./'&0'1!2334
Interprettion of these chnges cnnot be mde with confidence, but tenttively we suggest tht ll three of the lrge chnges cn be linked with the Globl Wr on Terrorism. GWOT ws the most slient U.S. government policy between the ttcks of September 11, 2001, nd the Pew poll in 2007. Incresing disstisfction with the wy things re going in the U.S. ws likely linked with incresing opposition to the wr on terrorism. In turn, mjor element of GWOT in these yers ws the continuing U.S. militry ction in Afghnistn; Muslim Americns moved towrd incresing opposition to this intervention. In the context of very high nd stble Muslim opposition to the wr in Irq, the mjor shifts in the opinions of Muslim Americns reported here my signl reduced willingness of Muslim Americns to cooperte in the wr ginst terrorism. We begn by noting tht polls ssessing Americn opinions bout Muslims since 9/11 indicte lck of understnding coupled with growing distrust of Muslim Americns (1). The polls we hve reviewed indicte tht t lest hlf of Muslim Americns feel the medi re bised ginst Islm nd t lest qurter feel victimized by nti-muslim discrimintion. Those seeing the wr on terrorism s wr on Islm hve incresed nd included bout hlf of U.S. Muslims by 2007. Tken together, the polls of Americn opinions bout Muslims nd the polls of Muslim Americns suggest tht there my be difficult times hed for Muslims in Americ. Perhps President Obm, himself the son of n Africn Muslim, cn hlt or reverse these unhppy trends. Interprettion of these chnges cnnot be mde with confidence, but tenttively we suggest tht ll three of the lrge chnges documented here cn be linked with the Globl Wr on Terrorism. GWOT ws the most slient U.S. government policy between the ttcks of September 11, 2001, nd the Pew poll in 2007, nd incresing disstisfction with the wy things re going in the U.S. ws likely linked with incresing opposition to the wr on terrorism. Specificlly, the most slient elements of GWOT in these yers were the continuing U.S. militry ction in Afghnistn tht begn in 2001, nd the invsion of Irq in 2003. In ddition to high nd stble opposition to U.S. militry forces in Irq, U.S. Muslims show incresing opposition to U.S. forces in Afghnistn nd incresing perception of the wr on terrorism s wr on Islm. Thus it seems likely tht incresing disstisfction with the wy things re going in the country nd incresing doubts bout the wr on terrorism re both reflecting opposition to U.S. forces in Muslim countries. Consistent with this interprettion, nlyses not reported here show significnt correltions between opposition to the Irq wr nd seeing the wr on terrorism s wr on Islm. In other words, individuls opposed to U.S. forces in Irq re more likely thn other U.S. Muslims to see wr on Islm. Finlly, we note tht incresing doubts bout the wr on terrorism cnnot esily be ttributed to the ctions of Al Qed. In our results, it is incresing disgreement with U.S. ctions in Afghnistn tht most obviously prllels incresing percentges seeing the wr on terrorism s R(!!!!!!!!!!! -'./'&0'1!2334
wr on Islm. But there is little doubt tht frming the wr on terrorism s wr on Islm is useful for AQ, s this frming represents sympthy for AQ s professed gols of defending Islm even if sympthizers do not gree with AQ s ttcks on civilins nd mrtyrdom missions. A wr on Islm is the simple populist messge identified by Brynjr Li s n AQ selling point for Muslims in mny countries [6]. From this perspective, the ppel of AQ is in prt function of Muslim perceptions of U.S. ctions. This is dynmic perspective in which AQ s brnd or imge depends s much on U.S. ctions s on AQ s ctions. APPENDIX on Methodology Utilized We reviewed four ntionl polls of U.S. Muslims conducted between 2001 nd 2007, including brief description of their smpling methods, then discussed some of the issues rised in compring results cross these polls. Three nrrower polls of U.S. Muslims were not included in our nlyses: Interntionl s 2003 Detroit Arb Americn Survey (http:// www.icpsr.umich.edu/cocoon/icpsr/study/04413.xml), s 2006 Four-Stte Arb- Americn Survey (http://www.truthout.org/rticle/how-rb-mericns-will-vote-2006 ), nd s 2007 Arb-Americn Identity Survey (http://www.ius.org/pge/-/polls/ AAIIdentity2007Finl%20Report.pdf ). These three polls did not im for ntionl surveys of U.S. Muslims. Four Polls of U.S. Muslims 2001-2007 2001: The Americn Muslim Poll (n=1781) ws conducted in November nd December 2001 by the Muslims in Americn Public Squre (MAPS), supported by the Pew Reserch Center (http://pewreserch.org/bout/ ) in collbortion with Interntionl (http:// www.zogby.com/ ). In this nd other polls considered here, ll respondents were 18 yers of ge or older. A telephone list ws creted by mtching the zip codes for 300 Islmic centers ntionwide ginst their respective locl telephone exchnges; listings of common Muslim surnmes were then identified from the locl telephone exchnges nd rndom smple of nmes were clled using Rndom Digit Diling (RDD). This pproch under-represents Africn- Americn Muslims, mny of whom to do not hve Muslim nmes. Thus, n dditionl smple of Africn Americn Muslims ws obtined in fce-to-fce interviews conducted 7-9 December 2001 t loctions in New York, Wshington, D.C., Atlnt, GA, nd Detroit, MI. The percentge of Africn-Americn respondents ws weighted to reflect 20% of the Americn Muslim popultion. 2002: The Hmilton College Muslim Americn Poll (n=521) ws designed by Sociology Professor Dennis Gilbert nd tem of Hmilton students nd supported by the Arthur Levitt Public Affirs Center. It ws conducted in April 2002 in collbortion with Interntionl. A ntionl cll list ws creted by softwre tht identifies common Muslim nmes in telephone RR!!!!!!!!!!! -'./'&0'1!2334
listings, n pproch tht, s lredy noted, probbly underweights Africn-Americn Muslims (http://www.hmilton.edu/news/muslimameric/muslimameric.pdf ). Published results were gender-weighted to correct for 60/40 representtion of men nd women, but results presented in this pper re unweighted. 2004: The Americn Muslim Poll (n=1846) ws conducted in August nd September 2004 by Muslims in the Americn Public Squre (MAPS) supported by the Pew Chritble Trusts in conjunction with Interntionl. Telephone interviews were crried out with ntionwide smple of Americn Muslims using the sme methods s for 2001 - except tht no dditionl fce-to-fce smple ws included. Pew2007: The Muslim Americns Survey (n=1050) ws conducted by Schulmn, Ronc, nd Bucuvls Incorported (SRBI) between Jnury nd April of 2007, ccording to the specifictions of the Pew Reserch Center. Telephone interviews were conducted in English (83%), Arbic (11%), Urdu (3%), nd Frsi (3%). The smpling frme hd three prts: two RDD smples nd one re-contct smple. The first smple cme from smple frme tht rnked U.S. counties by percentge Muslims (using dt from government surveys, commercil lists of Muslim nmes, nd self-identified Muslims in previous Pew RDD ntionl polls). RDD within the top three qurtiles of counties produced 354 completed interviews. The second smple begn with list of 450,000 households thought to contin t lest one Muslim, purchsed from Experin, commercil credit nd mrket reserch firm. Anlysis of nmes by Ethnic Technologies, nother commercil firm, produced smpling frme from which RDD produced 533 completed interviews. The third smple re-contcted ll Muslims identified in previous Pew ntionl surveys 2000-2006, producing 163 completed interviews. The complexities nd dvntges of this multiple-strt polling re detiled in pges 57-71 of the Pew Reserch Center report, Muslim Americns: Middle clss nd mostly minstrem. (2) Comprbility of Surveys Although there hve been other polls of U.S. Muslims since 9/11 (see for instnce Council of Americn-Islmic Reltions poll, http://www.cir.com/portls/0/pdf/ Americn_Muslim_Voter_Survey_2006.pdf), we believe tht the four polls reviewed here provide the best vilble foundtion for comprisons over time. The idel of comprison over time would be repeted rndom smples from the sme smpling frme, with the sme poll items used by the sme polling orgniztion for the sme sponsor. Frme nd orgniztion re linked becuse the polling orgniztion produces the smpling frme, nd two orgniztions will seldom produce the sme frme, even for the sme trget popultion of interest. Similrly, sponsorship is importnt becuse the preferences of those supporting survey cn mke difference in the smpling frme nd the items used. Thus it is importnt to note tht the four surveys exmined here ll involved either Interntionl or the Pew Reserch Center, who RW!!!!!!!!!!! -'./'&0'1!2334
together produced the first survey of interest, 2001. The two polls tht come closest to n idel comprison over time re 2001 nd 2004. Both were conducted by telephone by Interntionl in ssocition with Muslims in the Americn Public Squre (MAPS) nd the Pew Foundtion. Both imed t ntionl smple nd used smple frme bsed on Muslim nmes. The only difference ws the ddition of fce-to-fce Africn-Americn interviews to 2001 but not 2004. Similrly, 2002 ws ntionl Muslim survey using nme identifiction to crete the smple frme, lthough it ws sponsored by Hmilton College rther thn the Pew Center. Finlly, Pew2007 ws ntionl poll tht differed from ll the polls in joining three different smple frmes, interviewing in four lnguges, nd developing pinstking pproch to estimting the popultion nd distribution of U.S. Muslims (1.4 million Muslims 18 yers old or older). This gold stndrd of Muslim surveys cost over one million dollrs. It my hve reched less religious Muslims thn the surveys, which begn in 2001, s lredy noted, with the zip codes of 300 Islmic Centers. Sttisticl nd Substntive Significnce of Differences over Time Sttisticlly, the stndrd error for percentge in smple of 1000 is bout two percentge points, nd responses to the sme item in two independent rndom smples of n=1000 re significntly different (p<.05) if the percentges differ by more thn bout five percentge points. The complex smpling frmes used in Muslim polls suggest lrger stndrd errors for percentges; for instnce the Pew Reserch Center (2007, p. 57) estimtes stndrd error of 2.5 percentge points for Pew2007 percentges. Using this estimte, percentges from two smples of 1000 differ significntly if the difference exceeds bout 7 percentge points. Using more conservtive criterion of substntive significnce, we ignored in this rticle differences between surveys of less thn ten percentge points, nd focus especilly on consistent trends over time tht mount to chnge of 15 percentge points or more. The uthors thnk Dennis Gilbert for generously shring the dt of the Hmilton College Muslim Americn poll; they lso thnk Gry LFree nd Susn Brndon for suggestions towrds improving the report. This reserch ws supported by the United Sttes Deprtment of Homelnd Security through the Ntionl Consortium for the Study of Terrorism nd Responses to Terrorism (START), grnt number N00140510629. However, ny opinions, findings, nd conclusions or recommendtions in this document re those of the uthors nd do not necessrily reflect views of the U.S. Deprtment of Homelnd Security. About the Authors: Clrk McCuley is Rchel C. Hle Professor of Mthemtics nd the Sciences t the Psychology Deprtment, Bryn Mwr College; he is lso Editor of Dynmics of Asymmetric Conflict ; Jennifer Stellr is t the University of Cliforni t Berkeley where she is specilizing in pro-socil emotions nd cross-culturl foundtions of morlity. Notes RV!!!!!!!!!!! -'./'&0'1!2334
[1] C. Pngopoulos, Public Opinion Qurterly 70, 608 (2006). [2] Pew Reserch Center, (2007; http://pewreserch.org/ssets/pdf/muslim-mericns.pdf ). [3] Pew Reserch Center, (2005; http://pewforum.org/publictions/surveys/religion-politics-05.pdf). [4] Gllup.com, (http://www.gllup.com/poll/116233/fghnistn.spx ). [5] J. M. Jones, (2008; http://www.gllup.com/poll/106783/opposition-irq-wr-reches-new-high.spx),opposition to Irq Wr Reches New High, April 24. [6] B. Li. Perspectives on Terrorism, II (8), (2008). RT!!!!!!!!!!! -'./'&0'1!2334