Loading...
HomeMy WebLinkAbout8 - Article: Secondhand effects of student alcohol use reported by neighbors of colleges: the role o~,~ ~ fi ~~4 ~ ~ :'rl , ~'~l'~~+' ~ ~~ l~ ~ t } ~ SOCIAL SCIENCE -&- MEDICINE PGRGAMON Soci.~l Saence & Mcd~ane 55 (2002) 425-435 www eltievier cum/locate/tiocsumed Secondhand effects of student alcohol use reported by neighbors of colleges: the role of alcohol outlets Henry Wechslera°*, Jae Eun Leea, John Hall~, Alexander C. Wagenaar°, Hang Leed '`Hm unrd Schoal oJ Pub6c Henlth, Depqrtment oJ He~dth nnd Soc+n/ Behno~or, 677 Pfuntr~t~ton Avenue, lioaron, MA 02!/5, US~ ^Mudiemnncu Pa6cy Research, 600 Alexaiider Pak, Princemn, NJ 08543, USif °Depur[men! of Epidem~aloyy, Schon! of Pubhc Henlth, Uruue~sny of M~mreaotn, l300 SouUi 2nd Sneet, Suite 300, Mmnenpolm, MN 55454-1015 OSA `~CenterJo~ Vnccme Resea~ch & Depmfineiit of Pedmnics, UCLA SJioo! of Medicwe 1124 Weat Cn~snn Streel, ResemeG «nd E~Gmunon Rmldrn9 E-6, Torrnnce, C~l 10502, USA A6stract Th~s is a shidy of the secondh.~nd eflects of student alcohol use expenen~ed by residents of ne~ghbodioods near college campuses We exammed the relaGonsh~p of a college's level of 6mge drmkmg and Ihe mmiber of alcohol outlets m lhe immed~a[e area, to lowered quahty of ne~gliborliood hfe throiigh such secondh.md effects Adults from 4661 houseliolds m the Umted States were mternewed through a stratified hs[-nss~sted random d~grt d~ahng [elephone survey The mterview schedule mcluded queshons abont res~dents' expenences of secondhand effects of alcohol use such as no~se, vand2lism or pubi~c d~sriirbances Reports abou[ the qiiahty of neighborhood I~fe provided by responJents res~dmg near colleges were compared witli tliose of respondents who d~d no[ live near coileges, and reports of nc~ghbors of colleges w~th Iugh rates of bmge dnnkmg were compared with [hose of ne~ghbors of colleges wrth lowe~ ra[es The presence ot alcohol outlets m these ~reas wes also compared Res~den[s near colleges end partwularly near colleges wrth heavy ep~sod~c drmkmg reported the presence of mora alcohol outlets withm a mde Those ne~gh6orhoods we~e charactenzed by lower soc~oeconom~c s[atus Ne~ghbors I~vmg near college campuses were more hkely to report << lowered quahty of neighborhood life through sncli secondliand effects oC hcavy alcohol use as no~se and dtstnrbances, vandalism, drunkenness, vomitmg and urm~[ion A patli analysis mdicated thut the number of nearby alcohol outlets was an important fac[or med~ahng the relat~onship be[ween col(eges, espeaally those widi high rates of bmge drmkmg, and such secondhand effects The results suggest that ne~ghborhood d~snip[~ons around colleges due to heavy ~tlcohol use may be reduced by 6miUng the presence of alcohol outlets m those arcas, ~md the marketmg pracuces that this engenders Qc 2002 Elsev~er Saence Ltd AII n~hts reserved Keyiva~d~ Ne~ghborhood, College ltudents, Alcohol use, Environment, Alcohol-ielateJ d~sruption, Alcohol aude[ dens~ty, Socioeconomw status, USA Introduchon In 1993, the Harvard School of Pubhc Health College Alcoliol Shtdy (CAS) found that two ~u five US college shidents werc bmge dnnkers (Weclisler, Davenport, Dowd~ll, Moeykens, & Cnsl~llo, 1994) and this rate '"Co[t'espondmg rmthoc Tel k1-617-43?-1137, fax +I- 617-432-3223 GmuJnddiesr hwechsle(n)hsphhaivardedu(H Wechsler) remamed constlnt m[wo follow up surveys Q997 .md 1999) over ~i 6-year penod (Wcchsler, Dowdall, Maen- iter, Gledhill-Hoyt, 8r. Lee, O98, 2000a) Among the problems assoc~ated wrtli these high levels of alcoliol use are what we liave termed "secondhand" etiects Weclis- ler, Moeykens, Davenport, C~tstillo, and Hansen (19956) fo~md that non-bmge drmkmg s[udents res~dmg on campuses where more than half at students were bmge drmkers were twice as likely to expenence secondhand eCfects [han iion-bmge drmkers Irvmg on campuses wrtl~ 0277-9536/02/$-see front mattcr [Q 200? Elsevwr Saence Ltd All r~glits ~escrved PIT 50277-953b(O1)00259-3 426 H Wech~ler et al ~ Souu1 Smence & Medmrne 55 (2002J 425-435 fewer bmge dnnkers These secondhand effects mclude (i~vmg sleep or sCUdy m[erruptcd, havmg to take care of a drunken student, bemg msulted or assaulted, bemg the victim of unwanted sexual advances, or liavmg personal property v~~ndalized Heary alcohol consumphon by college students and othe~s may be encouraged by a"wet" environment, tlia[ ~s, an env~ronment m whicli alcohol ~s promment and easdy accessible (Edwards et al , 1995) Physicdl, social, and econom~c ava~lability of alcohol is assoc~ated wrth alcohol consumptron among the general population (Parker, Wolz, & Harford, 1978, Rush, Stemberg, & Brook, 1986, Abbey, Scott, Ohnsky, Qumn, & Andreski, 199Q Abbey, Scott, & Smrth, 1993, Gruenewald, Madden, & 7anes, 1992, Gruenewa(d, Milter, & Treno, 1993) and among youny+ adolescents and older teenagers (O'Mnlley & Wagenaar, 1991, Wagenaar, 1993, Wa~en- aar et al , 1996, 7ones-Webb et al , 1997) H~gh dens~ty of alcohol outlets has been found to be associated wrth h~gher rates of alcohol-related health and soc~al problems such as honnc~de (Scnbner, Cohen, Kaplan, & Allen, 1999), assaultrve molence (Alamz, Parker, Gallegos, & Cartmdl, 1996, Alaniz, Cartmdl, & Plrker, 1998, Gorman, Speer, Labouv~e, & Subarya, 1998a, Scnbner, MacKmnon, & Dweyer, (995, Speer, Labou- v~e, & Ontkttsh, 1995), domestic vtolence (Gorman, L~ibomie, Speer, & Suba~ya, 19986), [raffic safery outcomes (Rabow & Wa[ts, I982, Jewell & Brown, 1995, Scribner e[ al , 1994), and mortalrty, morb~dity and economro costs (Tatlow, Clapp, & Hohman, 2000, Mann, Smart, Ang(m, & Adlaf, 1991, Rabow & Watts, 1982, Scnbner, Cohen, & Farley, 1998, Gorsky, Schwar[z, & Denms, 1988, Smart, Mann, & Suurvali, 1998) Alcohol outlets and advert~smg appear to be over-concen[rated m ethme mmonty commumties (Ala- mz, 2000, Hackbartli, Sdvestr~, & Cosper, 1995, Altman, Schooler, & Basil, 1991, LaVeist and Wallace, 2000), ~mplying tha[ ~[ is necessary to understnnd the socio- demograph~c and economic background of a commumty m copmg with dnnkmg problems As Gruenewald and others (1995) have pom[ed out, most of d~ese shidies find relat~onshtps be[ween outlets, demograph~es, and drmkmg patterns, but most do not provide a[heoretical basis for understandmg such mterrelat~ons One such theoreGCal approach receiving mereased attentiou recently is Yhe "routme actrvrt~es" theory (Fox & Sobol, 2000) Most commonly appl~ed to cnme victmuzat~on, routme act~vity Uieons[s find that more frequent °~omg ouP' mcre~ses one's nsk of vict~m~zat~on (Mustame & Tewksbury, 1998) In the contex[ of college drmkmg, one nuyht ~rgue [ha[ h~gh rates of heavy dnnkmg and alcohol-reldted probleme among college students are "s~mply° the result of tlie~r frequent and routme activ~ty of gomg ou[, part~cularly to bars and ntghtcfubs Tltus, ~us[ as time spent walkmg tlie stree[ mcrenses exposure Co nsk of (one type o~ assaull, tmie spenf m bars mereases exposure [o the nsk of experiencmg sccondl~and etTects of heavy drmkmg The pomt rema~ne, liowever, tha[ a high densiCy of bars and clubs around campuses may encourage hcavier dnnkmg among s[udents Alcohol usc rates and related problems have been reduced by strateg~es to restr~et alcohol availab~liCy Coate and Grossman (1988) reported that as alcol~ol excise taxes mereased, youth drmkmg [ates and deaths resultmg from motor veh~cle acc~dents s~gmficantly decreased O'Malley and Wagenaar Q991) found that as sta[es mcreased nvmmum dnnktn~ age laws, alcohol use and problems associated with ~t s~gmficantty decreased Chiu, Perez, and Parker (1997) reported that an alcohol ban, rts hf[mg, and its re-~mpos~hon had stat~stical(y st~+mficant effects on the number of alcohol- related outpatient visits m a geograph~cally isolated eommumty Reshrouve alcoho( control pohcies sigmfi- cantly affected m~ury death rates m a popu(at~on w~th extremely high m~ury mortahty (Berman, Hull, & May, 2000) Colleges with large numbers of bmge drmkers are ~ charactenzed by gre~[er v~s~b~hty and ava~lab~hty of ; elcohol m the~r environment College sh~dents' bmge I drmktng ~s assoc~ated w~[h the degree of ease of access to i alcohol (Wechsle~, Kuo, Lee, & Dowdall, 20006), I Iocahon of a bar wrthm a mdc from campus (Wechsicr et aI , 1994), pr~ce (Chaloupk~i, Grossman, & Saffer, 1998, Wechsler et al , 20006), and state alcohol control pohcies (Chalottpka et al , 1998) Cleady, drmkmg levels and ra[es of alcohol-related problems are assoc~ated with state and local pohc~es as well as alcohol avadabdity, pnce, and marketmg prachces For many dimensions of the poLcy and markeUng env~ronment (eg, nlcohol taxes, drmkmg age), we know that the causal mHuence runs &om pohcy to drmkm~ For others (e ~, outlet densrty), the causal ~ mfluences mr~y be reciprocal, w~th the environment encouragm~ drmking, and heavy drmkmg encouragm~ detenorahon of the commumty envuonn~ent Tlie curren[ study exammes tl~e mterrelahonsh~ps be[ween a commumty envvonment tl~at encourages drmkm~ and a concentrat~on of heavy drmkerz (on college c~mpuses) [I~at shape the commumty env~ronment Specifically, we used surveys of commumty residents aronnd colleges, along w~th surveys of student behavior on those campuses to nnswer the followmg quesUOns Are there more alcohol outlets m ne~ghborhoods near colleges than m smt~lar neighborhoods wli~ch are not nenr colleges~ Do residcn[s I~vm~ m commum[~es nea~ a college expenence more secondhand e(fects of alcohol use than residents of s~m~la~ areas no[ near a college~ Are the mcreased secondhand effects rela[ed to more alcoliol oude[s near ~i college~ H Wedislcr el nl / Sacml Scmnce & MerLune 55 (2002J 425-435 • Do res~dents of areas near colleges widi h~gh levels of bmge dnnking expenence more secondhand e(i'ects Ih2n residents of areas near colleges wrth low levels of bmge dunkmg~ Methods Shady procedure We conducted a telephone survey oF ndult residents of tlie cont~guous Umted Stntes plus the Distnct of Columb~a usmg a strat~fied Lst-nssisted random d~git dtalmg (2DD) sample purchased from Genesys Sam- pl~ng Systems ~ The hsbass~sted method used covers an estimated 96 5% of all households wrth telephones (Brick, Waksberg, & Starer, 1995) Ach~al coverage may be higher because [he sample wus selected at multiple points m timc, so some households excluded carly m[he survey could have been ~ncluded latcr on Bnck et al (1995) concluded that hst-ass~sred RDD samplmg ~s "effic~ent and no[ suU~ect to unportant coverage bias" The survey was conduc[ed by M1[hemet~ca Policy Rescarch of Pnnceton, NJ The mterv~ew schedule mcluded qneshons about residents' expenences of secondhand effects of heavy alcoliol use siich as noise, vandal~sm or public d~sturbances Questwns were p~.tterned nfter [4ose mclnded m[he Hatvard School of Publro Healtli College Alcohol Study s[udent ques- t~onnave (Wedisler, Dowdall, Davenport, & Castillo, 1995a, Wechsler, Kelly, Wertzman, G~ovanm, & Seibr- mg, 2000a) Respondents were nlso asked about thev views on alcohol control pohcies, as well as personal backgronnd charactensncs. The schedule was pre-tested on a small sample of respondents res~dmg near colleges diat were not part of tlie samplmg frame Mmor revis~ons were done as a result of the pretest Survey m[erviews were conducted between Marcli and August 1999 Up to IS callz were attempted to obtam a completed mterview for each sampled [elephone num- ber Enghsh-speakm~ adults (age 18 and above) Lvmg m a household setfmg who were not full-time college students were eLgible for the sncvey In honseholde wrth more than one eli~~ble adult, one was randonily selected for the mterview Intervtews were condncted by tramed mterv~ewecs uemg convent~ottal Camputer Ass~sted Telephone [nterv~ewmg (CATI) me[hods A total of 9248 telephone numbera were called, w~th 4661 households ident~Ged Of these, 2621 were study elig~ble and 2300 were successfully mterviewed, yieldmg ~i completion rate of 86% and 2n es[im<<ted over2ll response rate of 50% nsmg methods recommended by ~ L~s[ ass~sted RDD snmp6ng me[imds aro descnbcd m Lepkowsk~ (1988) 427 the Counci] of Amer~can Survey Research orgamzations (CASRO, Frankel, 1983) Desp~ta tlie level of response, a comparison of selected demograph~c char~ic[eristics of the respondents w~th US census data mdtcated no signdicant d~fferences, prov~dmg no s[rong evidencc of selection b~as on [he basis of these var~ables Sampltn~ desi~n We defined 7 strata for sampla selechon Strata 1-4 mcluded areas near h~gli and low binge sdiools A h~gh bin~e schoo( is one of tlie 30 schools wrth the lughest prevalencc of bmge dnnkmg amon~ the 116 colleges par[ic~patmg m the 1997 H~.rva~d School of Pubhc Healtli College Alcohol Study (CAS) Similarty, s low bmge school ~s one of the 30 schools that were lowest m [he prevalence of bmge drmking (Wechsler et al , 1998) The h~gh bmge areas ~nchide strata 1(pubhshed numbers) and 2(unpubhshed) The Iow bmge areas are covered by strata 3(pubhshed) and 4(Linpnbhshed) More prec~sely, strata 1 and 2 mcluded [elepltonc numbers assoc~ated w~th census tracts [hat were eshmated to be wrthm a 1 m~le rad~us of colleges that had been clnss~fied as h~gh bmge drinking schools Strata 3 nnd 4 were s~milarly near colleges tl~at had been class~fied as low bmge drmkmg sclzools Pubhshed and unpubhshed refer to whether a household's telephone number appeared m the telephone directory Pubhshed nnmbers were ass~gned to stratum 1 or 3 based on tlie~r street addresses Unpubl~shed mimbers were assigned to stratum 2 or 4 if they belonged to a telephone exchange where at least 30% of the publ~slied numbers were assigned to stratnm 1 or 3, respecGvely Strata 5 und 6 included households m count~es [liat have colleges on ti~e sample frante used m selecnng the sample for [he earlier s[udent snrvey Stratum 7 is [he balance of the US More spec~fically, stratum S mcluded otlier counhes wrth colleges prov~ded the coimty had ~ large enougli population to be selected with eertam[y when usmg proba6~Lty proportional to size (PPS) metliods Strahim 6 mcluded any other counties wrth one or more colleges on t6e sample frame Stratum 7 compnsed court~es wrth no college on frame Telephone numbers m strata l-4 also could have been sampled m crther stratum S or 6 Tliese multiple chances of selectwn were lccounted for ~n sample we~gh[mg Tlie sample ~s a nmlt~s[age design W~thm strat~ l, 2, 3, and 4 the pnmary samplmg unrt (PSU) ~s [lie college and the surroundmg area For stratum 6 the PSU is the county In each case [he secondary sampl~ng umt ~s the house- hold The s~tmples of households tn sV.it~i 5 1nd 7 are not I~kely to be clustered Data were weiglited to reflect d~fferences m prob- ~ib~hty of selecuon and response rates across s[rn[a Otlier components of the we~glits mcluded ad~us[men[s for mulhplc telephone Imes and for mterniptions m 428 K Wechsler et u! / Somu! Science & Med~cme 55 (2002J 425-435 telephone service,z and posbstrat~fication ad~ustments to n~it~onal esGm~~tes of the populaUOn distr~bution by sex, ~ige, race and home ownersh~p 3 All analyses were conducted usmg weighted data Measures Almost all of [lie measures m the stndy were obt~med from responses to the comple[ed mterv~ews Respon- dents we~e asked ~f they I~ave seen or witnessed negative consequences of others' drinkmg (li[ter, noise or d~shirbance, vandal~sm, people who are drunk, figlitmg or assault to otliers, vonut or unuatton, and antomob~le acc~dents) one or more t~mes m their ne~gliborhood m the p~ist year (secondtiand effects) The number of neighbors exper~encm~ four or more of these second- hand effects was exammed The secondhand effects were broken down mto the mc~dents attnbu[ed to c,ollege students by askmg ~f the college students were pr~mar~ly responsible for the mcident Commimrty problems were measured by askmg respondents ~f tliey thought ne~ghborhood concerns and ~ssues were a ma~or problem or a problem ~n [lie~r ne~ghborhood Neigltborhood concerns <<nd issues m- cluded homelessness, cnme, pubhc drunkenness, drug use, vandahsm, drunk drrvm~, underage drmkmg, anJ loitermg Respondents were asked [o esGmate the d~stance of the nearest college from the~r home "How many miles f~om your home is the c(osest college or umversity (Plcase exelude commumty college m your answer)~" They were also asked to esdma[e how many alwhol outlets (on-prem~se and off-prem~se, sep~ra[ely) were located withm 1 mile of Cheir home (n addit~on to survey data, we also nsed some var~ables from census data Of the commumty back- ground vanables, eshmates of mcome, racial composo- [ion, home ownersh~p and a~e distribution were estimates a[ the [elephone exchange level provided by Genesys Samplmg Systems (Marketmg Systems Group) or the US Census Bureau zAd~ustments for mterrupnon m telephone service allow the survey to compens<<te for Ihe omission ot non-telephone households Very few honseholds wrthout telephone serv~ce on a given d.~y never have [elephone servwe Mos[ fall mto wha[ Keeter (1995) calls Ihe "trans~en[" oategory-havmg serv~ce tome 4mes and bemg wrthout at othen By usmg n we~ght acl~ustment factor thn[ ~s proporhonal to Ilte number of monthti w~fhout telephone servwe, Ihe tranment telephone household populn[~on can be appropn~~[ely ~epresented m sample es[~- m.i[es ~AdJtts(ments for Itome ownersh~p were b,tsed on eshmates prov~ded by die s~tmple vendor, Genesys SAmp6ng Systems (Markenng Systems G~oup) Ad~usunents for age, race ~md sex were baaed on Census Bureau pro~ectiona (Bure.iu of the Census, ?000) Analysrs SUDAAN V 7 5(Shah, Barnwell, & B~eler, I997) was used for all Chi-square tests and multtple log~stic regression analyses, SUDAAN empioys a Taylor senes Imear~zahon to approximate corxecf standard errors for sample eshmates grven [he mult~stage samphng dcsign of the survey and the effects of sample we~glitmg 2 x Z Cho- square tests were used to examme the difference m socioecononve backgrounds between commumt~es de- fined by the d~stancc (w~tlim and Farther dian 1 m~le) from t0e closest college Multtple log~stic regress~ons were conducted to test tf repor[ed secondhand effects of dnnkmg attnbutable to college students diftered among residents of h~gh and fow heavy-ep~sod~c dnnkmg school areas, and whether secondhand eftects vaned dependmg on [he d~stance from the closest colle~e The logist~c regress~ons controlled for soc~oeconom~c background variables est~mated at the telephone exchange level (racial composi[ion, % meome 0-lOk, °/a owner occnp~ed, °/a age 18-24, and rural/urban) We used structnral equation models to examme how the d~stance From the closest college or the college bmge dnnktng rate is rela[ed to numbers of Hlcohol outlets and Yhe number of secondhand efTec[s (controllmg commum[y's socioeconom~c characterishcs) We created an mdex of soc~oeconom~c status reflechng race, mcome, home ownerslup, and popnlat~on age distribnt~on to s~mphfy the model and avo~d potential multroollmeanty When we conducted the path analys~s, we assumed a nmdirechonal causal relat~onslup between alcohol out- lets and the environment even though there was the pose~bd~Cy of a bi-dvechonal relaGOnsh~p between the two Smce our ma~or concem through the path model was to determme the med~atmg role of alcohol outlets between college bmge drmkmg and secondhand el~'ec[s, we used a recursrve rather [han non-recursrve model The imtial path model was based on our hypotheses F~t of the model was eval~iated by comparahve fit mdex (CFI), Bender and BonetPs (1980) non-normed fit mdex (NNFI), Bentler and Bonett's (1980) normed fit mdex (NFI), and tlle Chi-square goodness of fit The statlstical vi~tb~hty of Ihe restnchons m[he model wns determmed by i.agrange MulGpher test The SAS CALIS procedure was used for structuial equation modelmg (Hatcher, 1994) Resulte Communtty Background Income was sigmficantty lower antong ~espondents hvmg witlun a nule than those hmng mare than 1 nule from ~i college (Table 1) More A&ioan Amencans, fewer whites, and, ~is expected, more young people aged 18-24 H Weclnle~ e! u! ~ Soa~ul Suertce & Mvdame 55 (2002) 425-435 Table l Socioeconom~c charac[erm[~cs of commum[y by dtstance from college Prevalence m (%) To[al More [hun Withm a m~le` I mde° (n = 1692) (n = 526) 429 Chrsquare p-value More Ihan 10% of households huve annual mcome less [h,~n $1Q000 Yes 56 2 No 43 8 More than 12% of md~vidualy are African Amerman (non-Hispamc) Yes 31 0 No 69 0 More than 1l% of md~viduels arc H~spemc Yes 28 4 No 71 G More than 71"/0 of mdrviduals nce White (non-Htspamc) Yes 61 8 No 38 2 More than 50% of housmg unrts arc awnei occupied Yea 84 9 No 15 I Mo~e than 10'% of mdrnduals are age 18-24 Yes 20 9 No 79 I 53 8 72 7 0 0005 46 2 27 3 28 8 46 6 0 0028 71 2 53 4 27 7 33 1 0 3095 72 3 66 9 64 3 43 9 0 0009 35 7 56 1 87 7 65 4 0 0001 I23 346 17 4 38 8 <0 0001 826 61? ' Respottdent's est~m~[e of chstunce oC home &om college n = vnhd snmple s~ze hved w~thm a mile from the college Are2s wrthm a mile of a college had a lower prevalence of homeowners On-premise (bars/mghtclubs) and off-prem~se (Itqttor stores) alcohol outle[s were more often loca[ed withm a mile from a college Nmety-Iwo percent of res~den[s Irvmg w~tltm a mile from the closes[ college reported one or more alcoliol ouUets w~thm a mile from the¢ house compared to 75% of those who hved more than 1 m~le away After controlLng for mcome, race, urbamsm, and home ownersh~p, respondents who I~ved w~thtn a mile from the nearest college were sigmficantly more likely to report the presence of alcol~ol outlets nearby (nd~us[ed OR= 2 83, 95% CI I 47-5 47, p<0 001, Tablc 2) Commumty problems repor[ed by respondents are presented m Tab~e 3 Commumty problems reported most &equently were underage dnnkmg (60 8%), cnme (55 6%), vandalism (52 3'%), and drunk dnvmg (47.9%) Ne~ghbors who Irved wrthin a mde from a college more otten repor[ed liontelessness, crime, pubhc drunkenness, dn~g use, undernge dnnkmg, and lo~teung than those hvmg one or more m~les from a colle~c (Table 3) d~cturbances, v~inda6sm, dnmkenness, and vom~t and unnnhon than those lrvmg tnore than a mile from the school They ware s~gmficantly more hkely to report fottr or more sucli e11'ects (Table 4) College shidents were not v~ewed as primardy respons~ble for most of these secondhand effects Only about one-fourteenth of tl~e respondents v~ewed coilege students to be responsible for vom~t/unnation (7 8%), no~se/d~s[urbance (6 9%), fi~hhng/assault (6 3%), and litter (6 1%) Coilege students were more often v~ewed to be responsible for I~tter, noise/d~s[urbance, vandal- ism, and dnmkenness by respondents hvmg wrthm 1 m~le from a college, than by tliose hvmg more than a mile from the school One m five (19 5°/n) respondents who Irved w~tltm a mile from a college mewed college students to be respons~ble for 2t leas[ one such effect, wh~le one m twelve (8 3%) Irvmg more than a m~le away d~d Those who lived wrthm a mile were sigmficantly more hkely to report at le.ist one of these effects Secondhnnd eJJects m/mv and /tt~h bm~e dr ~n/c~n~ colle~e s~tes D~stance /rom colleye nnd secondhnnd efJ'ects Respondents who Irved w~thm I mile from a college were sigmficantly more likcly to report no~se and Wh~le more respondents m Ivgh bmge drmkmg school areav than w low bmge drmkmg ~ireati ~eported the p~esence of alcohol outlets wrthm a mile of the~r homes 430 H Wechs[er e! al / Sonul Suence & Med~une SS (2002) q25-435 Table 2 Presenca of alcohol ontle[s by d~s[ance of cesponAent's home from college Mo~e [han I m~le" (n = I692) W~[hm a mile' (n = 526) ~"~o) (%) Ad~us[ed ORs (95%CI)n PreSence of 6ar/mgh[clttb` 49 8 73 9 2 17 (1 32-3 57)*** Presence of hquor s[are 52 4 77 3 2 33 (I 32-4 17)*** Presence oF other store that sells ,~Icohol 63 9 74 2 I 20 (0 75-I 92) Presene,e of any one of above alwhol outlets 74 9 92 I 2 83 (I 47-5 47)*** "Respondent's eshmate oF d~sWnce of home &om closest college ~ORs ~~e ad~usted for % mcome, % rnce, rural/urban and % owner occupied OR = odds rat~o 95% CI = 95% confidence mterval `One or more self-repor[ed alcohol ou[leta wrthm 1 mile Cmm house °**p <0 001, n =vaLd sample s¢e Table 3 Reported commumty problems by dts[ance ot responden['s home from college More thun I mde` (n = 1692) W~thin n m~le~ (n = 526) ~ ~o) (%) Ad~usted ORs (95%CI)~ Commumty problemt` Homelessness 19 I 35 1 1 82 (I I I-3 03)* Cnme 53 7 68 4 t 75 (1 12-2 78)** Publ~c Dru~~kenness 30 3 43 1 1 61 Q OI-? 56)* Drug use 44 7 58 8 1 67 (1 04-2 70)* Va~daham 51 4 58 5 I 33 (0 87-2 04) Drunk dr~vmg 48 0 47 2 1 09 (0 72-I 64) Underage d~mkmg 59 5 69 9 I 64 (I OS-2 50)• Lo~term6 346 541 192 (123-294)^"* Four or more pro6lems reported 44 0 59 7 1 89 (I 22-2 94)*""" "Respondent's es[~ma[e of dislance of liome from closes[ college ~'ORs are ad~usted fo~ '% mcome, % race, "/o age 18-24, rural/urban, .ind % owner occup~ed OR°odds ra[~o 95% CI=95% wnfidence ~nterval °% repor[mg lh~s ns a problem *p <0 05, **p <0 01, *"*p <0 001, n= vahd sample size (903% vs 82 1%, ad~usted OR=233, 95% CI 1 45- 370, p<0001), no s~~ntficant d~fference m soc~ceco- nomic status was found between the two school areas ~ Respondents who hved m h~gli-bm~e school areas j more often reported htter and no~se/disturbance by ~ college students than [hose m low-bmge dnnkmg school are~s (Table 5) One m five (18 6%) respondents m high- bmge drmkmg school srtes reported at least one such secondhand e1~eU, compared [o only one m ten respondents m low-bmge school areas Rale of a(coho! outlets as medtahny fnctor We conducted a path analys~s to explore the degree to wh~ch alcohol outlets med~a[e the relat~onship between wllege factors (distance from college and college bmge dnnkmg levels) and the secondhand effects (F~~ 1) Tlie Chi-square statishc was not sigmfichnt and tlie CFI, NNFI, and NFI nll exceeded 0 98, md~ca[mg the model fits the observed data well All path coefficients shown were sigmficant at p<005 Distance from [he closest college and college bmge drmkmg level had an md~rec[ effect on rates of sewndhand problems through the number of alcohol outle[s m the area No direct effect of distance from a college on secondhand problems was found Soc~oeconom~c s[atus had both dtrect ~md mdirect effects on secondhnnd problems The md~rect or med~ated effects of college, student dnnkmg, and soc~oeconom~c steh~s on secondhand problems is ~~ stronger tlian duect effects, mdicatmg thaY the presence of atcoliol ontlets appears to be essenCial for colleges ~Lnd theu bmge drmkmg students to have a s~gmficant effect on neighborhood d~srupt~on Discussion A survey of a nat~onal sample of households revealed s~gmficant correla[wns be[ween tlie dwtance from the uearest coilege and such secondhand efTects of heavy H GVechsle~ e~ ul ~ Somal Sc~ance & Medreme 55 (2002) 425-435 Table 4 Repor[ed sewndliand efTee.ts of dlcoliol by d~stance af respondenPs hame from college 431 More thon 1 m~le° (n = 1692) Wrthm a mde° (n = 526) (%)0 ('%) ORs (95%CI)e Secondhnnd effects` Lrt[er 72 8 79 3 I 22 (0 76-2 00) No~se or d~shvbance 53 4 70 8 l 72 (1 10-2 70)* VandaLsm 31 7 48 7 2 00 Q 27-3 23)*"'" People who ~~re d~ unk 35 6 58 5 2 00 (1 22-3 33)*'" Figh[mg or nssnult m others 17 8 28 5 I 41 (0 85-2 38) Vom~t or unna[~on 10 5 32 2 2 70 (1 54-4 76)*** Automobde acrndent o~ others 40 2 46 1 1 19 (0 78-1 79) ('our or mare problems observed 30 8 53 2 2 00 (1 25-3 23)*** College student-at[nbuCed secondhand effects`~ Littcr 5 2 11 9 2 27 (1 09-4 76)* Noine or d~sturbnnae 6 0 11 S 2 63 (1 20-5 88)* V~mdalism I 7 8 9 5 00 (1 39-16 fi7)** People who are d~ unk 4 3 12 5 2 94 Q 19-7 14)• F~ghtmg or nstiault to others 4 9 12 2 3 45 (0 81-14 29) Vomit or urma[ion 5 I 13 8 3 23 (075-14 29) Automobile .tcctden[ or o[hers 3 2 5 4 2 27 (0 54-10 00) Any one of above problems 8 3 19 5 2 78 (l 54-50 00)*'"" °RespondenPti est~m~tte of d~stunce of home fmm collebe nORs are ~d~us[ed for % mcome, % raae, and % owner occup~ed OR=odds rat~o 9$% CI= 95 % oonfidence m[erval `% repordny observmg even[ one or more dmes °'% who abserved event nnd .ittnbu[ed rt to college students 'p <0 O5, *'"p <0 Ol, ***p <0 001, n= val~d sample s~ze Ta61e 5 Secondhend efi'ects a[tnbu[ed to colle~;e s[udents by respondents near h~gh nnd low bmge level co lleges Low bmge dnnkmg achool s~[e (n = 817) H~gh bmge dnnking school s~te (n = 490) (%) (%) Ad~us[ad ORs (95%aC])" Colle6e student-nt[r~bu[ed secondliand efiectse Lrtter 4 7 15 8 3 36 (1 77-6 40)*"* Noise or d~sturbance 8 3 13 9 1 97 (1 12-3 44)* Vandahsm 2 S 7 4 2 70 (0 76-9 68) People who are drmik 7 9 I S 8 ? 3? (0 98-5 83) ~igh4ng or ~ssault [o o[hers 4 0 5 8 I 60 (0 46-6 34) Vom~[ or urm~~don I l 8 7 3 93 (0 85-18 10) Automobde ace~dent or others 2 5 2 4 1 OS (0 32-3 44) Any one of above problems 10 3 IS 6 2 I I (1 21-3 68)'"* "ORs nre ~~d7ttsted for'% mcome, "/o r.ice, and "/o owner occup~ed OR=odds mho 95'% CI= 95% ~onfidence mterval °i% who observed event and attn6u[ed ~[ [o caliege titudents *p<005, **p<001, *'"'p<OOOI, n=vnl~d sample size alcohol use as noise, ht[er, and vandahsm 2espondents residmg near a college werc a[ h~gher i~sk of cxpcnen- c~ng such secondhand effects They were ~ilso more hkely to liave alcohol oL~tlets Iocated near them Path analys~s md~clted that residing near a college does not appear to be suffic~ent for experiencmg high r~t[es of secondhand problems Tlie colleges' contnbuLOn to ne~gliborhood pioblems appears Co operatc [hrough the presence of alcohol outlets Our findings suggeat tliat alcohol outlets are more often located m areas near colleges, pardctt- larly those w~[h h~gh rates of bmge dnnkmg Commu- mty iesidents m these areas aie I~kely to expe~ience lugher iates of neigh6orhood disniphon Such an mte~pretaLOn ~s cons~stent wrth the hteratu~e on alcohol ~ 432 H Wechale~ ct af / Sounl Scrence & Med~cme 55 (2002J 425-435 DISTANCE from Colle e Cam us - ll+*• R'=29 R '= I8 /~ 1 • ALWIIOL SECONDHAND OUTLETS Number EPFECTS BINGEDRINKING 08*** ofalcoholoutlets 33*** Numberof 33 *• LEVEL at collegc w~thm a mile from -~ secondhand etfects +~• responAent abserved _ qq SCS Commumty's SomoeCOnomw Stafus _ ~ q+++ *** p<0001 %Z (2) =4 966I, p=0 O8, CFI=O 9964, NNFI= 0 9822, and NFI= 0,9941 Valid samples=2,206 Fig 1 Reduced path model of secondhand elfec[s oudet density m general, which finds that h~gher outlet dens~hes mcrease percerved availabihty of alcohol, lower reta~l price through mcreased compehhon, lower total eoet to tlie dnnker (mcludmg C~avel t~me), mareases consumptton of alcohol, aud mcreases violence and other cnme and disrupt~on assoctated with drm_ k~n~ (Abbey ct at , l99Q Alamz et al , 1998, Scrtbner, Cohen, & F~sher, 200Q Berman et al, 2000) Gyvep t cro s- swhonel des~~n of the current study, we cannot answer the queshon of which came first Does [he presence of a college, espec~ally w~th a li~gh ra[e of heavy drmkmg, encourage more alcohol outlets7 Or does the presence of meny compepng alcohol outlets encourage h~gh rates of heavy drmkmg by the students of the nearby college~ Our resul[s sugges[ however, [ltat ra[es of nei~hborhood disruption around colleges may be si~mficantty reduced by ImnGng the presence of alcohol outlets m those areae - tier ac[ors contnbute to the presence of elcoio outlets .~round many colleges Onr resnlts md~cate neighborhoods near colleges are more hkely to be lower soc~oeconomic arcas These cond~twns might tncrease Lhe ease of obtammg alcoho] hcenses, and produce a lugher presence of outlets Others have reported panc~ularly h~gh ra[es of alcohol outlet density m poor urban lreas (Goiman & Speer, 1997, LaVeisr & Wallace, 2000), and res~dents of tliese ne~ghborhoods are more likely to report a range of social problems such as homelessness, crime, pubhc drunkenness <md lorter- ~ng Current attemp[s ro change studenr behavior througli educddon and bnef motrvat~onal techmques are ~imong the mam mtervenhons colleges are usmg to reduce he.~vy drtnkmg Results of Ph~s stndy suggest that deahng w~Cl1 tlre h~gh density of alcohol outlets and [he marketing~ prachces th~s engenders ~n ne~ghborhoods ~mmediately surroundmg campuses may also be an ~mportant strategy Str~cGy Imvnng hcenses for new outlets and pliasing out Iicenses of establishments tl~at ~epeaCedly violate servin~ and marketmg regnlahons are means to ieducmg aleohol ouCle[s In many comnmmGes, half of all alcohol outlets regularty v~olate laws aglmst sellmg or servmg alcohol [o those under tlie legal dnnkmg age (Forster, Murray, Wolfson, & Wagenaar, t995), and a recent study revealed three-quarters of outlets v~olate laws proh~bitmg sales [o patrons who already show s~gns of obv~ous mtoxicatton (Toomey et al ,(999) Aetive enforcement of these laws is needed throu~h regular comphance checks of all alcohol outlets, espec~ally m colle~e areas wliere sales to nunors and sales to mtoxicated mfrachons may be panc~ularly pievalent Sucli enforcement has ~mmediate benefits m reducmg risky sales pract~ces (Jeffs & Saunders, 1983, Preusser, Wilhams, & Wemstem, 1994), and may have furtlier benefits via the revocaC~on of Che hcenaees of particulaely problem-prone ontlets, and a gradual reduchon m alcohol outlets m college ne~ghborhoods Res~dents who suffer the secondhand etCects of lteavy drmkmg can be enhsted m tli~s eftbrt, nsmg a type of `ne~ghbor- (iood watch' operation Raismg licensmg fees and alcohol [~~xes to pay for the preventmn and cle~nup ot ne~ghborhood disntp[~on should be constdered, espe- cially smce substanual ma~onties of the US general populat~on support such poi~cies (Wagenaar, Harwood, Toomey, Denk, & Z~inder, 2000) Another noteworthy findmg suggests that (ower soc~oeconomic conditions around college campuses H GVech~ler el a! / Socra/ Scrence & Med~mne 55 (2002) 425-435 may contribute to the presenec of aleohol outlets D~sadvan[aged neighbors may be less able to preven[ [he grantmg of hcenses to sell alcohol Th~s may be part of a v~c~ous c~rcle Iowe~ soc~oecononuc s[atus near colleges may result m more alcohol outlets, more alwhol outlets may lead to more secondliand effects, and more secondhand cffects may contrtbu[e to decreased real estate values and st~ll lower SES Eff'orts shonld be focused on how to d~sconnect tlie v~cious c~rcle A few cau[~ons are important to cons~der when mterpretmg data from th~s s[udy Tlie results are based on a[elephone stirvey, and are sub~ect to the I~m~tations ~nherent m such metliods Persons w~thout telephones canno[ be pazt of [lie sumple However, m a large sc21e general population survey w~th adequale coverage <<nd response rate the results for those who have phones werc found to no[ ddfer sigmf~cantly from those of [he popul.itwn ~is .i whole (Aday, 1989) Sample atlnhon also oecurs hecduse of fadure to obtam and oomplete mterv~ews with the setected telephone numbers Tlie response r~tte of 50% may have mlrodnced s~impLng b~ns However, a comparison of selected demograpliic char~ctens[ics of the responden[s w~th US census data mdicated no s~gmficant dilTerences Whde other sources of 6~as may ex~st, the sample of responden[s mdtches the charactenshcs of the general populahon In addition to poss~ble samplmg b~as, selt-reports m1y mtroduce a whole set oF measurement error components (Det Boca 3c Noll, 2000) However, such errors are Iikely to be random, and should not alter the nahire of [lie relaLOnslups Smce we exammed relat~onships at the aggregate or ne~ghborhood level, est~mates of 2 college's heuvy dnnkmg rate or a neighborhood's alcol~ol out(e[s and level of alcohol-related disruption represent an avera~e for overall responden[s at that s~te, by which the potenti~~l measurement errors may be averaged out In onr study, d~stance from [he nearest college, and number of alcohol outlets w~thm a ntile of home were based on respondents' estimates r2ther than phys~cal measures, and may not exactly re(lect real d~stances and actu<<I mmiber of outlets However, usmg an admims- tr~tor survcy developed to obtam mformaGOn on campus alcoliol poliaes from denns of students or otlier adm~mstrators, Wechsler, Lee, Kuo and Lee (2000c) also found a stanstwally sigmficant associat~on of campus dnnkmg levels w~th adm~mstrators' report of ~tlcohol outle[s located wrthm a mile of their college These consistent results usmg reports of d~sYance from nearest alcoliol ouUet obtamed trom [wo d~fferent types of respondents seive to valida[e tlie measure Fur[hermoie, wh~le not re8echng actual miles, respondents may be repoctmg the number of elcohol outlets wrthm the area [hat [liey perceived as °Iheir ne~gli6orhood" One poss~ble source of error tha[ muy no[ be random, relates to the dnnktng 6eltavror ot respondents It is poss~ble that respondents who drank more frequently 433 were more aware of the outlets m the~r envvonmenl, and conld provide morc accurate, and prob~tbly fuller counts of them Allhongh we mchided quest~ons about respondents' dnnkmg behavioi, we could no[ wntrol for th~s factor because of the large nnmber of no answers [o tlus question (45%) Smce most analyses werc condncted wrth d~chotomous vanables (no outlet vs some outlets), the potent~al confoundmg effect of this fnctor may be mm~mized, thougli not fully discounted Another I~m~tat~on m mterpretmg tlie resnlts of the study ~s tlie cross-secUonal des~gn Whde complex and expens~ve, future studies arc needed to examme thc role of alcohol outlets m heavy drmkmg on college c~mpuses wli~ch track changes over tmze m botlt drmkmg rntes and [he density and prachces of alcohol outle[s The best opportnmt~es for such stud~es are mos[ Itkcly s~hiahons m wlucli there aic ma~or changes m law, regulation or economic condrtions [ltat result m substanhal changes m alcoliol onUe[s over a rel<ttively short penod of t~me Contiolled [~me-ser~es shidies (B~glan, Ary, & W2gen- aaq 2000) of such na[ttral expenments m selec[ wllege commumties may help furthei our understandmg of the apparen[ly important role alcohol outlets play m encouragmg heavy drmkm~; on college c~impuses Acknowledgements We gratefully acknowledge the assistance of Ehssa R We~[~m2n m[he development of [he survey qnestwn- naire Th~s shidy was supported by the Rober[ Wood Johnson Founda[ion References Abbey, A, Sco[t, R O, OLnsky, D M, Qumn, B, & Andreski, P M Q990) Sub7ecdve, sooial, and physic~~l <<v~ulabdity ^ Thev simuitnneous ef7ects on aicohol consumphon /nteiimnonul Jou~n~d of ~ddapott, 25(9), 1011-1023 Abbey, A, Scott, R O, & Smrth, M J(1993) Plrysmdl, sub)ec[rve, and soaal ~rva~lab~l~ty 'Lhea reladonship to alcohol consumpt~on m rurnl and urban areas ~IddiU~on, 88(4), 489-499 Adny, L A Q989) Des~ynmg n~id conducnng hau!!h aurueys San I'ranc~sco, CA Jossey-IIass Iuc Al~ma, M L(2000) Commumtyadentifled alcohol nsues m Ihe Mexican Amenam communrty Resenrch des~gn and utthzatlon Subs(nnce Use qnd Mnuse, 35(I-?), I57-169 Al.tmz, L A, C.irtmill, R S, & Parkcr, R N(1998) Immigcan[ and violence The unportance of neighhorhood contex[ Hi spnntc Journal oJ IJehuwortd Scrences, 20(2), 155- 174 Alan¢, M L, P.irker, R N, Gallegos, A,& CnrMt~ll, R S Q996) Alcohol mitlet densrry rtnd Merrcan Ame~¢nri youlh uiole~~ce Final Progress Repoct, In[ervUnrvem~ty Program for L,iuno Research, The Fo~d ~oundahon 434 H Wechsler et n! / Socrn[ Suence & Med~cine 55 (2002J 425-435 Altm.~n, D G, Schooler, C, & Basil, M D Q991) Alcohol and mgnrette adve~hsmg on billboacde Kenlth &lucalmn 2esearch Theory and Pracnce, 6(4), 487-490 Ben[ler, P M, & Bonett, D G(1980) SigmFlcance tes[ nnd goodness-of-ft[ m[he analys~s of covananee structwes P~yciiolo9~col Bidlenn, 8$ 588-606 Berman, M, Hull, T,& May, P(2000) Alcohol control and ~n~ury de.~[h m Alaska nat~ve commumheti We[, damp nnd dry nndec Alaskh s looal opt~on Inw Journn[ oJ Student A[mhoGem, 6](2), 31I-319 Diglan, T, Ary, D, & Wagenaar, A C(2000) The vailue oF mterrupted t~me senes expenments for commumty inter- venROn and pohoy resenrch Preuenpon Scrence, /(l), 31-49 Br~ek, 7 M, Waksberg, J, Culp, D,& Starer, A Q995) 6~as m hsh.tss~sted [elephone samples Publ~c Opmron Qumteily, 59, 218-235 Bureau of tha Cena~s (2000) Pro~ecdous of The voting aye pop~dn[mn for NouemGcr, 1998 Repor[ P25, I132, Govern- ment Pnn[mg Office Chnloupka, F], Grossman, M,& S.iffeq H(1998) The effects of pnce on the wnsequences of .tlcohol use and ~tbuse 2ecer~! Developmenta m ~Ilwholnm, 14, 331-346 Chm, A Y, Perez, P E, & Pndcer, R N Q997) Impact of b.imm~g alcohol on outpatient v~s~ts m Barrow, Alaska Journa! of ~he dmer rcan Me~hcul Aosomnnon, 278(21), 1775- 1777 Coate, D,& Grossmnn, M(1988) Effects of .ilcoliol beverage pnces nnd le~al dmkmg ages on you[h alcoliol use Jnurn~d of S~uden~ ~IlcolroGem, 48, I67-175 Del Boca, F K, & Noll, J A(2000) Truth or consequences The vt~Ldrty of selL~epor[ data m health aerv~cea iesearch on nddiotions dddret~on, 95(SUppl 3), 5347-S3b0 Cdwards, G, Anderson, P, 6abor, T P, C~~sswell, S, I'errence, R,& G~esbrecht, N, et al (1995) ~llcoho! podcy unJ the pubhc yood (pp 125-152) New Yark Oxford Umvers~ty Press Fors[eq 7 L, Murray, D M, Walfson, M,& Wagenauq A C Q 995) Commerc~al nvu~lab~Lty of alcohol to young people 2esults of alcohol purchase at[emp[s Prevendue Med~eme, 29, 342-347 P'ox, J G, & Sobol, J 1(2000) Dnnkmg patterns, social mteraction, and b~irroom behnvior A rouGne ecuvmes approach Deurnnt Behuwor, 21($), 429-450 I~rankel, L R p983) Cha 2aport of tha CASRO Task Farce on Reaponse Rates In ~ Wiseman (Ed ), fmpromny dntn 9utdrry ~n rt aumple suruev Cambndge, MA Marketing Science Inshtu[e Gorman, D M, Labouvie, E W, Speer, P W,& Su6u~ya, A P(1998a) Alcohol avadability and domeshc violence Ame~~enn Jo~~rnnl oJ Dru9 ~Ilcoha! ~lbuse, 24(4), 6fiI-673 Gorman, D M,& Speer, P W Q997) Concentranon of Lquor ou[lets m an economically d~sadvantaged aty m the Northeaste~n Unrted States Substnncc (lae nnd Miane, 32(14), 2033-2046 Gorman, D M, Speeq P W, Lahouvie, G W,& Subnry.i, A P(19986) R~sk of nussalhve violence nnd alcohol nvad- ab~l~ty m New Sersey Ameraun Journul o/ Publm Heuf(h, 88, 97-100 Gorsky, R D, Schwar[z, E, & Denms, D Q988) The morta~hty, morb~di[y, ~~nd econom~c ws[s of .ilcohol ebuse m New Hampsh~xe Preuennue MeJicine, /7(6), 736-745 Gruenewnld, P 7, Madden, P,& 7anes, K(1992) Alaohol avadabdity and formal power and resources of sta[e atcohol beverage control agenc~es A[cohoGsm C[miud & Exper~- mentul Resemdi, 16(3), 591-597 Gruenewald, P 1, Miller, A B,& Treno, A J Q993) Alcohol avaJ26~hty and [he ecology of dnnkmg behawor Alcohol Health nnd Resenrch World, 17(1), 39-45 Gruenewald, P J, Treno, A J, Nephew, T M,& Ponicki, W R(1995) Rou[me nctrv~tles and aloohol use-constramts an outlet uGl~zation ~I[cohoham Clnuca! & Expe~~mentn! Reseurdi, 19(l), 44-53 Hnekbarth, D P, S~Ives[n, B,& Cosper, W(1995) Tobaceo and nlcohol b~llboards m 50 Chwago ne~ghborhoods Market segmentahon to sell dangerous products to the poor Journul of Pub6c Henlih Polacy, 16(2), 213-230 Ha[cher, L(1994) ~I slep-by-slep npproach ta u~my Uie sus system for Jncto~ annlyers and structurul equnnon modeGn~ Cary, NC SAS Researeh Institu[e Inc Jefl's, B W,& Saunde~s, W M(1983) Mm~m¢mg alcohoi rele[ed atTences by enforcement of the ewsfmg hcensmg Iegmlapon Bnt~sh Jown~d of ~Iddicnon, 7R, 67-77 Jewell, R T, & Brown, R W (1995) Alcohol availebtlity and aleohol-ralatetl motor vehicle accidenW Apphed Gconomies, 27, 759-765 Jones-Webb, R, Toomey, T L, Short, B, Murray, D M, Wagenaaq A, & Wolfson, M(1997) Relahonship among alcohoi nvn~lab~l~ty, drmkmg locanoq alcohol consump- hon, end dnnkmg piablema m.tdolescen[s Subelonce (lee and Miwne, 32Q0), 1261-1285 Keeteq S Q995) Est~mnhng noncoverage b~as from a phone survey PubGc Op~man Quurlerly, 59, 196-216 LaVe~s[, T A, & Wallace Jr, J M(2000) Heulth nsk and meyuitable d~stnbution of hquo~ s[orea tn Afrtcan Amer- ~can ne~ghborhood Soual Scrence und Med~une, SI(4), 613- 617 Lepkowsk~, J Q988) Telephone samphng me[hods ~n the Um[ed SW[es In Robert M Groves, et al (Ed ), Telephorre suruey mediodola~y New York W~ley Mann, R E, Smart, R G, Anghn, L,& Adlaf, E M(1991) Reduchons m c~rchasis deaths m Ihe Umted States Assoc~ations wrth per cuprta consumptwn and AA memberv ship /ouruu( of Studen~ Alcoholrsm, 52(4), 361-00365 Mustame, C E, & Tewksbury, R Q998) Pred~chng nsks of I.~rceny IUeR victim¢nhon-a xouUne acnv~Ty ~inalyvs usmg refined lifestyle measures Crimmolo~y, 36(4), 829-857 O'Malley, P, & Wagenaar, A C Q991) BRects of mm~mum drmkmg .ige Iaws on alcohol use, related behavmrs, and [ratfic crnsh mvolvement among Amencan youth 1976- 1987 Jmnnal oJ Student Alcnhohsm, 52, 478-491 Puker, D A, Wolz, M W,& Ha~focd, T C(1978) The prevent~on of alcohohsm An emp~ncai report on [he eflects of outlets .tvaila6~ltty Aleohohsm Cbmcul 8c L'~Perimenful Reseurch, 2(4), 339-343 Preusser, D P, Wdhams, A P, & Wemstem, H 6 Q994) Policmg underuge <Acohol aales Journu! of S~~ery Resenrcli, 25, 127-133 Rnbow, J, & W.~tts, R K(1962) Alcahol av~ulab~hty, alcohoLc beverage sales and ~Qcohoi-reia[ed pro6lems Journp! oj Student AlcoHohsm, 43(7), 767-801 Rush, -, Stemherg, M,& Brook, R Q986) The relationsli~ps among alcoliol ~~vn~Inb~L[y, alcohol consumpdon and H Wech~[e~ et n! / Socio! Smenca & Med~cu~e SS (2002J 425-435 alcohol-~elated damage m tl~e provmce oC OnCa~lo and the st~~te of M~ch~gan 1955-I982 Adunnces m~Iltoho! SuGetnnce Abuse, 5(4), 33~{5 Scnbneq R A, Cohen, D A, & I'arlcy, T A(1998) A geographic relnhon between alcohol availnbiLty anci gonor- rhen rates Sex Trnnamiaeion D~aeusee, 25Q0), 54¢548 S~nbner, R A, Cohen, D A,& Fishe~, W(2000) 8vidence of n stcuctural efleet for .tlcohol outlet density A mulhlevel analysis Alwhol~am Gm~cul & E~penmemal Resemch, 24(2), 188-195 Scnbner, R, Cohen, D, Kaplan, S,& Allen, S H(1999) Alcohol ava~labd~ty and homic~de m New Odeans Con- ceptuai cons~dera[~ons for sm~~ll area analys~s of [he effec[ of .tiwhol ou[let densrty Jomnn! of Student ~IlcohoGsm, 60(3), 310-316 Scnbne~, R A, M.tcKmnon, D P, & Dweye~, J H Q994) Alcohol outle[ density and mo[or vehu,le ornshes m Los Angeles cottnty c~Ges Jotnnnf oJ Student Alcahol~cm, 55, 447-453 Scr~bner, R A, MaclGnnon, D P, & Dweyeq J H(1995) Tlie nsk of assnuihve v~olence and alcohol ava~labJity m Los Angeles County A~nencun Jmnnnl of PubGe Neu[dr, 85(3), 335-340 Sh.ih, R V, Barnwell, B G, & B~eler, G S(1997) Sudoan user's mnnunl, re[ense 75 Reseurch "R~angle P.vk, NC Recearch Tnmigle Inshnrte S~nart, R G, Mann, R C,& Suurvali, H(1998) Ch.ingcs m Lver c~rrhosw death rn[es m d~fferent coun[nes m relation to per caprta alcohol consump[~on ,md alcohohcs hnonymous membenh~p Journa! of Student ~llcoboLsm, 59(3), 245-249 Speer, P W, Gorman, D M, Lnbouvie, E W,& On[kush, M J Q998) V~olent cnme .md alcohol avail~~bil~ly Rel.~hon- ships m an urban commumty Journul of ft~b6c Heaftli Pohcy, 19(3), 303-3I8 Tatlow, J R, Clapp, J D,& Hohmnn, M M(2000) The ~elationship between the geographw dens~ty of alcohol ou[lets and alcohol-related hospital admiss~one m San Diego County Journ~d o/ Canmunity Heultli, 25(I ), 79-88 Toomey, T L, Wagenaar, A C, KiLan, G, ~rtch, O, Ro[hstem, C, &['leroher, L(1999) Aloohol sales to pseudo-mtoxicated bar pa[rons Public Henllh 2epor/s, 774(4), 337-342 435 Wagenaar, A C Q993) Minrmtrm drmkin~ uye und rdcoho! nunduhdtry to youth les~ies nnd ieaenrch neede Reyearch Monograph 2S, US Departme~tt of Health and Hum~tn Serv~ces, N,~ponal [nsh[ute on Alcohoi Abuae and Alco- hohsm, Rockville, MD Wagen~~lr, A C, H.~rwood, E M, Toomey, T L, Denk, C E, & Zander, K M(2000) Pubhc op~mon on alcohol pohues m Ihe Um[ed States Retiul[s from a nahonal survey Journnl oJ PubGc Nenltd Po[tcy, 2!(3), 303-327 W~~gen.inq A C, Toomey, T L, Murr~ry, D M, Short, 6 J, Wolfson, M,& Jones-Webb, R(l996) Sources of alcohol for underage dnnkers Journul of Student AlcohaGsm, 57, 325-333 Wechsler, H, Davenport, A, Dowdnll, G, Moeykens, A,& CasGllo, 5 Q994) Health nnd behevior~l connequences of bmge dnnkmg m collebe A nadonal survey of studen[s .~[ l40 campuses Jmonp( of !he Amencun Medicnl ~l ~sauunai, 272, 1672-Ib77 Weohsler, H, Dowdall, G W, Davenpor[, A, & Cas[~Ilo, S (1995n) Co[relates of oollege atudent bmge dnnking dmer~crm Jonrnnl of PuG6c Henlth, 85, 921-926 Wechaler, H, Dowdall, G W, Maenneq G, Gledh~ll-14oyt, J, & Lee, H Q998) Changea ~n bmge dnnkmg .md related problems among Amencan college studenta between 1993 and 1997 Jouro~ul ~ J Nte Ame~ ~um Co(leqe of Hea[tli, 47, 57- 6S Wechsleq H, Kelly, K, Wertzmnn, E 2, Giovanm, J P,& Se~brmg, M(2000a) What colleges a~e domg nbout student bmge dnnkmg A survey of ~ollege admimstrn[ors Journn( of the Amencnn Collepc of Heald~, 48, 219-226 Wechsle~, H, Kuo, M, Lce, H,& Dowdall, G W(20006) Env~ronmen[a~l correlates of underage alcohol use and related problems of oollege s[uden[s Amencmt Journu! oJ Preuenn~e Medicme, /9(I), 24-29 Wechsleq H, Lee, J ~, Kuo, M,& Lee, H(2000c) College bmge d~mkmg m[he 1990s A conhnumg problem ~esults uf [he Harvard school of pubLc liealth 1999 coliege alcohol study Journul af !he Amerrcnn CoJle9e of Heulth, 48, 199- 2I0 Wechsleq H, Moeykem, B, Davenpor[, A, Cast~llo, S,& Hansen, J Q995b) The adversc ~mp~~c[ of heavy ep~sodm drmkers on ather college s[udants Joumul oJ Sn~dent Alcohodsn~, 56, 628-634