Community effects of the opening of the Niagara casinofaculty.buffalostate.edu/macleamg/440/Room 1999 Add.pdf · on horse races. Lotteries, bingo and parimutuel wagering are available

Addiction (1999) 94(10), 1449-1466

RESEARCH REPORT

Community effects of the opening of theNiagara casino

ROBIN ROOM/ NIGEL E. TURNER^ & ANC A IALOMITEANU'

^Centre for Social Research on Alcohol and Drugs, Stockholm University, Stockholm, Sweden& ^Addiction Research Foundation Division, Centre for Addiction and Mental Health,Toronto, Ontario, Canada

AbstractAims. The impacts on the community ofthe opening ofa casino in Niagara Palls are studied. Design. Thestudy uses a pre/post design for the community data, with pre/post data from Ontario as a whole as acomparison. Setting. The study site is the city of Niagara Falls, Ontario, where a casino opened in earlyDecember, 1996. Participants. Using random-digit dialing, telephone interviews were conducted with adultresidents of Niagara Falls in 1996 and 1997, and with adult residents of Ontario in 1995 and 1997.Measurements. Aside from demographic variables, measures included general attitudes to gambling,expectations about (1996) and experiences with (1997) the casino's opening, extent of participation in 11types of gambling and 18 items on problems with gambling: five key items from a standard gambling problemsscore (SOGS), five life-area problems items, and items on pressures from others concerning the respondent'sgambling and on gambling problems among family and friends. Findings. Attitudes to gambling remainedstable in Niagara Falls, while there was some evidence of decline in approval in Ontario as a whole. Whilestrong majorities of 1996 respondents had expected many positive and negative effects on The Community ofthe Casino's opening. Significantly fewer respondents in 1997 reported actually experiencing most of theseeffects. While a small increase in employment was found, it fell far below projections, a result probablyreflecting displacement effects. The rate and level of casino gambling increased in Ontario, but increased evenmore in Niagara Falls, with little displacement of other gambling. Reported gambling problems increasedsignificantly in Niagara Falls for two of 10 gambling problem items and for the short SOCS score, while rateswere generally stable or declining in the province. Pressure from others about gambling rose significantly inNiagara Falls (in contrast to the province), and reported rates of family members or friends with gamblingproblems also rose substantially. There was an increasing trend in Niagara Falls for all 18 problem indicators.Conclusions. The casino's opening brought more gambling by local residents, and an increase in reportedgambling problems; yet support for the casino, already strong, if anything grew. At least in the short term,problems from the increased availability of gambling manifested themselves not in the public arena but ratherin the arena of private life.

Introduction of Niagara Falls, Ontario, in late 1996. NiagaraThis is a first repon of the impacts on the Falls, located on the Niagara Penisula of Southcommunity ofthe opening ofa casino in the city Central Ontario, is a city of more than 75 000

Correspondence to: Robin Room, Centre for Social Research on Alcohol and Drugs, Stockholm University,S-106 91 Stockholm, Sweden.

ISSN 0965-2140 print/ISSN 1360-0443 online/99/101449-18 © Society for the Study of Addiction to Alcohol and Other Drugs

Carfax Publishing, Taylor & Francis Limited

1450 Robin Room et al.

which is within a 2-hour drive of two large urbancentres (Toronto and Buffalo) and has been awell-established tourist destination for more thana century. This report draws primarily on surveysof the adult population of Niagara Falls carriedout just before the opening of the casino, andagain about a year later. Data from surveys of theadult population of Ontario as a whole, carriedout in early 1995 and late 1997, are used forcomparison. This paper repons on expectationsof community members about the effects of thecasino, in comparison with their experience ayear later of what had actually happened. Inaddition, trends in the level of involvement ingambling in general, in-casino gambling and inthe rates of gambling-related problems in Nia-gara Falls are compared to trends in the provinceas a whole. Taken together, these comparisonsare the most comprehensive view yet of theeffects of opening a casino, as seen from theperspective of the quantified experiences, posi-tive and negative, of members of the host com-munity.

BackgroundAs the National Council of Welfare (1996) putit, "in less than a generation, gambling has be-come a multi-billion dollar industry in Canada",as in North America more generally. Before1970, Ladouceur noted,

legal gambling in Canada was restricted tooccasional charity bingos and raffles, midwaygames of chance, parimutuel betting on horseraces and friendly bets between individuals ...By 1993, legal gambling in Canada had ex-panded to include slot machines and videogaming devices, casinos, large-scale bingo op-erations, spons wagering and off-track bettingon horse races. Lotteries, bingo andparimutuel wagering are available in everyCanadian province^ casinos now operate inmore than half of the provinces (Ladouceur,1996).

The increase in the availability of gambling hasbeen fuelled to a large extent by the search byvarious levels of government for new sources ofrevenue. Gambling has indeed become a majorsource of governmental revenue, with the resultthat often "there is no impartial representative ofthe public's welfare among the various groupsthat influence the formation of public policy on

commercial gambling—not even government"(McGurrin & Abt, 1992).

In the wake of the increase in the availabilityand diversity of types of gambling has come agrowing awareness that there are problems aswell as benefits from gambling. With varyingdegrees of alacrity or reluctance, Canadianprovinces have moved to provide resources todeal with gambling problems (Kinross, 1995;Addiction Research Foundation, 1996), but rela-tively little attention has been paid in Canada topreventing gambling problems. There has alsobeen little explicit recognition that the pathologi-cal gambler is not the only possible negativeoutcome of gambling—that there may be prob-lems also at community and other aggregatelevels.

In the continuing expansion of gambling avail-ability, Ontario has been somewhere in the mid-dle of the pack among Canadian provinces. Inrecent years, a number of new forms of gamblinghave become available in Ontario, including hos-pital lotteries, charity casinos and pull-tab tick-ets. Full-scale casinos came to Ontario in 1994(Windsor), with three more (Windsor boat;Rama; Niagara) opened in the following years.In early 1998, the government phased outfloating "charity casinos", which moved aroundbetween locations, operating for two or threenights at a time, with the intention of replacingthem with 44 "neighborhood gaming clubs",with some revenues earmarked for charities,scattered throughout the province. In compari-son to the rest of Canada, Ontario is also some-where in the middle in making provision forreferral and treatment of those with gamblingproblems—well behind the prairie provinces inallocations and timing, but ahead of the mari-time provinces and British Columbia.

In the wake of the increase in gambling avail-ability has come increasing public unease aboutthe problems it is perceived as causing. To aconsiderable extent, this public unease has fo-cused on video lottery terminals (VLTs). Facingrising resistance to the introduction of VLTs, theOntario government decided instead in spring,1998 to introduce slot machines in the province.Community resistance to the introduction ofsmall-scale casinos—the "neighborhood gamingclubs"—also proved strong, so the number wasreduced to only four new permanent casinos andinstallation of slot machines at 18 racetracksaround the province.

Community effects of casino 1451

These major social developments—the expan-sion of gambling availability, and increasing pub-lic resistance to further expansion—havehappened largely in the dark, in terms of theavailability of a cumulative tradition of knowl-edge and experience. Communities have foundthemselves with a lack of coherent and balancedinformation for making decisions about whetherand how to introduce new forms of gambling,and how to deal with and minimize the new orincreased problems the new gambling may bring.Economic models are available for estimating thebenefits to the community from the new gam-bling—for instance, the Ernst & Young (1994b)model which was applied to estimate the"impact" of opening the casino in Windsor.Such models have been substantially criticizedon grounds of basic principles in economics,since they assume that none of the share of thelocal economy attributable to the casino wouldotherwise have been devoted to other purposes(Persky, 1995). The Ernst & Young model alsodoes not factor in the extra expenditures whichmay have to be made in the community becauseof the casino. In the case of Windsor, for in-stance, the Ernst & Young report mentions that25 new police were hired in Windsor and 41additional customs officers assigned to Windsorborder traffic duty in preparing for the casino(Ernst & Young, 1994a, pp. 2-4, 2-5), but theseextra costs for the taxpayer do not figure in theeconomic impact model.

Furthermore, such models are not balancedwith any equivalent accounting of the problemsfor the community from gambling. Citing a USreport, the British Columbia Gaming Policy Re-view noted that "there is little reliable data aboutthe costs of some impacts such as problem gam-bling, crime and infrastructure expenditures, andeffects on other businesses. Most of the analysisdone respecting specific ventures is done by theindustry itself and tends to ignore or understatecosts" (British Columbia, 1994, p. 14). TheErnst & Young study, for instance, apparentlydoes not consider problems such as crime, em-ployee absenteeism or pathological gambling tohave an economic dimension at all, since theyare discussed under the heading "Other Non-Economic Impacts" (p. 5-1). Estimates of thecosts of these problems would normally be in-cluded, for example, in economic cost studies inthe cost-of-illness tradition (Single et al., 1996).

A balanced analysis at the community level

also needs to consider impacts of gambling,negative as well as positive, at the collective level.These are often intangible, but may neverthelessbe experienced as substantial (Thompson et al.,1993).

Studies of the full range of impacts of gam-bling at a community level have been few and farbetween. A series of studies in three Minnesotacommunities (Aasved & Schaefer, 1992; Aasved& Laudergan, 1993; Aasved et al, 1995) exam-ined the situation in the community after anexpansion in gambling, but had no data frombefore the expansion. Evidence on adverse socialimpacts was confined to opinions, rumours andclinical anecdotes.

A longitudinal study of a state-wide youthsample examined the change in gambling behav-iour of 15-18-year-olds with the introduction ofa state lottery. The 18-month follow-up studydid find an increase in frequency of legal gam-bling, but this increase was nearly offset by adecline in the frequency of informal, extralegalgambling (Winters, Shinchfield & Kim, 1995).Given the weaknesses in the study design (anon-probability sample, no control group andwith the "before" measurement shortly after thestate lottery first functioned), the study's resultsmust be regarded with caution. Another studycompared adult gambling frequencies among lo-cal residents before and a year after the openingof the first Ontario casino, in Windsor; a prelim-inary report found no net change (Govoni &Frisch, 1996). A 4 March 1999 press releasefrom Frisch and coworkers reports no significantchange in problem or pathological gambling in anew survey after 4 years (http://www.uwindsor.ca/pgrg/fyear.htm). However, if a one-tailed test isused, the reported rise in pathological gamblingscore from 0.8% to 1.4% is significant at the0.05 level. Other studies have compared theperceptions of community residents before andafter the opening of new casinos in small com-munities. In general, these studies have foundthat residents' attitudes were positive before thecasino opened, but were less enthusiastic a cou-ple of years later (Caneday & Zeiger, 1991;Perdue et al, 1995; Stokowski, 1996).

MethodsUsing random-digit dialing techniques, a proba-bility sample of the adult population of the Nia-gara region was interviewed before or around the


time of the opening of the Niagara casino in late1996. A new sample of adult residents of Nia-gara Falls were interviewed 1 year later, while atthe same time the 1996 respondents werereinterviewed. The present analysis compares re-sponses of residents of Niagara Falls in 1996with responses of residents in 1997.

In addition, two province-wide surveys wereused to compare trends in the province as awhole to those found in the city. These province-wide surveys were repeated cross-sectional sam-ples. The analysis thus uses answers from a totalof five datasets.

Samples(1) A probability sample of 1002 adults (18years and older) residing in the Niagara regionwere interviewed by telephone: 70% in the cityof Niagara Falls, 30% in the surrounding region.The interviews were conducted in English, andaveraged about 24 minutes in length. The surveywas conducted for the Addiction Research Foun-dation by the Institute for Social Research atYork University between 18 November and 19December 1996. \XTiile 93% ofthe respondentswere interviewed prior to the casino's opening,the last 7% of the sample was interviewed in the10 days following the opening date. The surveyasked questions about approval of the casino,expectations regarding the impact of the casinoon the community, attitudes toward gambling,gambling behavior in the past 12 months anddemographic and background information. An-swers from the 677 residents of Niagara Falls inthis sample are used in the present analyses. Theresponse rate was 60% of estimated eligiblehouseholds.

(2) In November and December 1997, abouta year after the casino opened, 662 of the re-spondents from the whole Niagara region samplewere reinterviewed. This represented a reinter-view completion rate of 66% of first-wave re-spondents. The interview schedule repeatedmost of the questions asked in the first interview,and included some questions specifically aboutexperiences with casino gambling, employmentin the casino, questions regarding economic di-version of expenditures and some questions ongambling strategies and knowledge ofthe odds ofwinning. The 1997 answers from the 468 re-spondents resident in Niagara Falls are com-bined with (3) below as the "after" dataset on

Niagara Falls. Longitudinal analysis of the Nia-gara data will be reported in a later paper. Itshould be noted that the two 1997 Niagarasamples were similar in terms of the effects re-ported here.

(3) A new probability sample of 608 adultsresident in the city of Niagara Falls was alsointerviewed in November/December, 1997. Thequestionnaire was primarily composed of itemsasked in component (2) above. The responserate was 60% of estimated eligible households.

(4) A province-wide probability sample of1030 adults, interviewed about gambling behav-ior and problems and attitudes towards gamblingand gambling policies in January-February 1995(Ferris, Stirpe & Ialomiteanu, 1996) is used as acomparison control for the baseline communitysample. Funding exigencies precluded the bestchoice for a "before" control sample, whichwould have been a new sample collected in late1996. However, the 1995 sample is an adequatesubstitute, particularly to the extent that prov-ince-wide patterns are stable in the 1995/1997comparisons. The response rate was 65% ofestimated eligible households.

(5) A new province-wide probability survey of1005 adults was interviewed in November andDecember 1997. The questionnaire was com-posed primarily of items about gambling andgambling-related problems as well as some ques-tions about experiences with casino gambling,gambling strategies and knowledge ofthe odds ofwinning. The primary purpose of the data fromthe new province-wide sample was as an "after"control for the Niagara comparisons. Trends inthe provincial data from 1995 to 1997 will thusbe used as a control, in comparison with trendsfrom 1996 to 1997 in the Niagara samples, toestimate the specific effects of the casino openingin the intervention community. The responserate was 62% of estimated eligible households.

Niagara Falls is, of course, a part of Ontario,and thus the intervention site is included withinthe control site in this study. Comparisons of thetwo will thus tend to underestimate very slightlydivergences in status and in trend. However,since the Niagara Falls population is about 0.7%of the population of the province as a whole, theeffect of this on the study results is negligible.

All samples were defined in terms of a ran-domized choice of one adult per household. Toattain representative samples of the adult popu-lation, each respondent was weighted by the


number of adults in the household and by thereciprocal of the number of telephones, and theresulting weights were proportionately reducedso that the weighted dataset matched the un-weighted sample size. All analyses in this reportare based on these weighted datasets.

MeasuresExpectations and perceptions of the casino's effects.Niagara region respondents were asked 28 itemsin 1996 concerning their expectations about theeffects of the casino's opening on the com-munity. Factor analyses (Turner, Ialomiteanu &Room, 1999) found that responses on most ofthese items clustered on three factors: expecta-tions about social problems and disruption,about environmental problems and about econ-omic and amenity benefits. In 1997 25 of theitems were repeated, but now asked in terms ofthe respondent's perception of what had actuallyhappened in the community as a result of thecasino's opening.

Gambling behaviors. There was some variationbetween samples in the lists of types of gamblingasked about, reflecting the advent of additionaltypes of gambling in the province between 1995and 1997 and greater or lesser subdivision oftypes. For the purposes of this analysis, 11 typesof gambling were used: instant lottery tickets,sports lottery tickets, 6-49 lottery tickets, hospi-tal lottery, bingo, betting on horse races, charitycasino, casino (non-charity), betting on card-playing, video lottery and betting on sportsevents. The first three varieties of lottery are runby a provincial corporation, while hospital lotter-ies are run intermittently for the benefit of aparticular hospital. For each type of gambling,respondents were asked if they had engaged inthat type in the previous 12 months and, if so,how often they did so, and what their averagemonthly expenditure was for that type of gam-bling. Outliers were a problem with the expendi-ture data; but because the outliers are in fact thegamblers that we are most interested in, deletingor modifying the outliers would distort the data.In order to assess the statistical significance ofdifferences but minimize the effect of the out-liers, the data were analyzed using the natural logof each person's expenditure.

Frequency of gambling was recorded on aseven-point scale ranging from once a day tonever. To compute an aggregate frequency across

different types of gambling, these categories werefirst converted to the interval midpoints (e.g. 2-3times a month = 30 days a year), aggregated andthen normalized using a log transformation.

Respondents were classified as having gambledin a non-charity casino if they had done so oneor more times in the past year. Respondentswere counted as having participated in gamblingactivities if they had engaged in one or more ofthe 11 types of gambling asked about on at leastone occasion in the past year.

Gambling problems. Several different measureswere used to gather converging evidence on ratesof problem gambling in the community.

Short SOGS. The South Oaks GamblingScreen (SOGS) has been used widely in gam-bling problems research (Lesieur & Blume,1993), but is lengthy for use in a general-population telephone survey. The version ofSOGS used in this study limits the questions toevents in the past 12 months. From an itemanalysis of the full list of SOGS items in the1995 provincial survey, five items were selectedfor use in the Niagara study: in the past 12months.

Was there ever a time when you gambled morethan you intended to?Have people criticized your gambling?Have money arguments centred on your gam-bling?Have you felt guilty about the way you gambleor what happens when you gamble?Have you claimed to be winning money gam-bling when you were not?

A score based on these items was correlated 0.87with a respondent's full SOGS score in the 1995provincial survey and also maximized correla-tions with life area problems, gambling partici-pation, gambling expenditures and DSM-IVgambling symptoms. All respondents that scoredin the pathological range (5 + on full SOGS)scored 2 and above on the Short SOGS andnearly all the people that scored 2 and above onthe Short SOGS were in the problem area (3 +)on the full SOGS. Thus, a score of 2 on theShort SOGS is taken here as indicating a prob-lem or pathological gambler; however, no dis-tinction can be made in this study betweenproblem and pathological gamblers.


Gambling life-area problems. Respondents wereasked whether they felt that their gambling hadharmed each of five areas of their life in the past12 months: friendships and social life, physicalhealth, home life, work and studies and financialposition. These items, adapted from questionslong used to measure alcohol problems (Caha-lan, 1970), were first used concerning gamblingin a 1994 Ontario population survey (Paglia,1995; Smart & Ferris, 1996). Positive responsesto two or more of these problems have com-monly been taken as a indication of fairly seriousproblems.

Others' responses. Respondents were askedwhether a family member and whether a friendor acquaintance had said anything about therespondent's gambling in their life-time and inthe past 12 months. Positive responses weretaken as an indirect indication of problematicgambling by the respondent. Current and life-time scores for pressures from others were cre-ated by assigning one point for each "yes"response.

Others' problems. Respondents were askedwhether a family member or relative and whethera friend had had a problem with gambling intheir life-time and in the past 12 months. Re-sponses on these items were taken as indicatoryof the level of gambling-related problems in thecommunity. Current and life-time scores for oth-ers' gambling problems were created by assign-ing one point for each "yes" response.

Demographics. A full range of demographicvariables were assessed. In the present analysisgender, summary age category, marital status,education and income category are used.

ResultsLevels of gambling involvement prior to casino: Nia-gara Falls vs. OntarioAbout 86% of adults in both the 1996 Niagaraand the 1995 Ontario samples reported havinggambled during the last year. The percentagewas slightly higher for males (87%) than forfemales (83%), although the difference wassmall. The percentage of people gambling wasfairly stable across ages except for the oldestgroup (71%), which gambled less than the others(87%). Gambling was also somewhat lower in

the lowest income group (80%) compared to therest of the sample (87%). Overall, the demo-graphic distribution of gambling was very similarin Niagara Falls and in the province.

The percentages of people participating inmost of the 11 types of gambling were verysimilar for the Niagara Falls and Ontario sam-ples. The amounts of money spent on each ofthe 11 types of gambling were also similar. Onaverage, the residents of Niagara spend some-what less (S36) than the residents of Ontario asa whole ($47) [r(l,1705)= - 1.22, NS]. How-ever, the Niagara Falls residents reported spend-ing more than the provincial average onsports-lottery tickets [t(l,1705) = 2.09, p<0.05]and horse-race betting, [r(l,1705) = 2.29,p<0.05]

Most respondents who gamble at all reportspending an amount which could be consideredwithin reasonable limits, with 93% reportingspending less than SI00 per month, and 94%reporting spending less than 5% of their incomeon gambling. However, expenditures range fromzero to more than $1000 in the past 30 days. Aminority of respondents are spending a greatdeal on gambling.

Changes in reported gambling behaviourChanges in the percentage of people particpatingin non-charity casino gambling are portrayed inFig. 1. The prevalence of engaging in otherforms of gambling besides the casino hardlychanged at all between 1996 and 1997; but theproportion of the population which has gambledin a non-charity casino increased significantly[X^(l, «= 1753) = 191.7, p<0.01].

The provincial comparison between 1995 and1997 shows results similar to those for NiagaraFalls, but the change in non-charity casino gam-bling is considerably smaller [x^(l, « = 2035)= 38.7,/)<0.01].

Figure 2 shows the amount of money thatrespondents reported spending on non-charitycasino gambling and total gambling in the past30 days for the four samples. VChen the differenttypes of gambling were entered as predictors in alogistic regression with study year as the crite-rion, only non-charity casino gambling [b = 0.44,Wald = 47.8, p<0.01] and horse racing[fe= -0.19, Wald = 5.3, /X0.05] significantlydifferentiated between the pre- and post-casinoopening samples.

90

80

70

60

50

I 40I 30^ 20

10

0Casino

GamblingNiagara

ni996• 1997


CasinoGamblingOntario

TotalGamblingNiagara

TotalGamblingOntario

Figure 1. Percentage participating in casino gambling and gambling in general: Niagara vs. Ontario and 1996 vs. 1997.

The average amount spent on gambling byNiagara Falls residents rose by $9 per month(25%) between 1996 and 1997. Almost all thisrise was accounted for by the increase in expen-diture on casino gambling, which increased from$2.30 to 11.10 in Niagara (rounded to thenearest 10 cents). Non-charity gambling alsoincreased in the province from 5.90 to 14.50.In 1997, casino gambling accounted for one-quarter of all gambling expenditures reported byNiagara Falls residents.

The trends in reported spending in NiagaraFalls are matched mainly by those in the prov-ince as a whole. Reported expenditures on bingoand non-charity casinos went up. Expenditureson most other forms of gambling stayed aboutthe same, except for an increase in horse-racingexpenditures in the province, which reflect re-sponses by a few outliers.

Figure 2 gives the average dollar values forspending on total gambling and casino gambling.These data are skewed, with many people spend-

6 0 , -

50 -

g 40Q .CO

I 30Eora 20

I10

ni996E31997

CasinoGamblingNiagara

CasinoGamblingOntario

TotalGamblingNiagara

TotalGamblingOntario

Figure 2. Average money spent on casino gambling and gambling in general: Niagara vs. Ontario and 1996 vs. 1997.


ing very little and a small number spending alarge amount. To counteract the skewness andoutlier problems, the statistical tests conductedon the expenditure data were first subjected to anatural log transformation [xln = ln(x-l-1)].Overall, compared to Ontario residents, Niagararesidents reported spending less money on casi-nos, cards and gambling in general, and moremoney on instant lottery and sports lotteries. Inaddition, gambling expenditure decreased overthe study period on instant lotteries, lotteries andsports betting, but increased on casino gambling.Since the primary purpose of this paper is toexamine differences in trends between the Prov-ince and Niagara Falls, we will only report theinteractions. The analyses of variance showedsignificant interactions for three types of gam-bling. Reponed spending on 6-49 lottery ticketswas down in the province between 1995 and1997 ($10.34 vs. $7.88) but not in Niagara Fallsbetween 1996 and 1997 (S9.45 vs. $9.50)[F(l,3783) = 5.3, p<0.05]; reported spendingon horse race betting fell in Niagara Falls ($4.78vs. $1.03) but rose in the province ($1.89 vs.$10.15) [F(l,3783) = 14.59, p<0.01]; and re-ported spending in non-charity casinos rose pro-portionally more in Niagara Falls ($2.26 vs.$11.12) than in the province ($6.00 vs. $14.45)[F(l,3783) = 38.76, p<0.01]. It should benoted that although the main effects and interac-tion are significant with analyses of either logmeans and raw means, the interaction for non-charity casinos is more clearly visible with the logmeans: the mean natural logs were 0.13 (SD0.69) for Niagara 1996, 0.53 (1.35) for Niagara1997, 0.11 (0.74) for Ontario 1995 and 0.34(1.22) for Ontario 1997.

A similar pattern of results was found for thefrequency of gambling. Instant lotteries, cardsplaying, VLTs outside of casinos and sportsgambling decreased from 1995/6 to 1997 andnon-charity casino gambling increased[F( 1,3783) = 249.3, /XO.Ol]. Three types ofgambling showed significant interaction effects.The frequency of sports betting fell more in theprovince than in the Niagara Falls[F(l,3783) = 27.3, /)< 0.01], horse-track bettingfell in Niagara Falls, but rose in the province[F(l,3783) = 4.0,/)< 0.01] and spending at non-charity casinos rose more in Niagara Falls thanin the province [^(1,3783) = 72.3, /XO.Ol].Overall, Niagara Falls residents gamble morefrequently than respondents in the province as a

whole [F(l,3783) = 12.6, p<0.01]j however,this is due largely to their more frequent lotterypurchases.

There was also a dramatic increase in thenumber of people going to casinos on a regularbasis Ix^ (18, Ar= 3790) = 402.34, p<0.01]. InNiagara Falls, 3.6% of the population reponedgoing to a casino two or more times a month,whereas less than 1% of the population in theother three samples reponed attending casinostwo or more times a month. Although there wasan increase in the number of people attendingnon-charity casinos in the province from 1995 to1997, the entire increase was found for the oncea month (0.7% in 1995 vs. 2.3% in 1997) or lessthan once a month (10.1% vs. 18.6%) categories.

An examination of the demographics for gam-bling in general indicated that there was verylittle change in either Ontario or Niagara in thepercent panicipating in gambling. However, asindicated in Table 1, there was a large change inthe percentage panicipating in casino gambling.In the province as a whole, the proportion ofadults gambling in non-charity casinos almostdoubled, the increase is widely spread across alldemographic categories, with relatively little vari-ation by demographics. The increase has beeneven more dramatic in Niagara Falls: the pro-ponion of the population gambling in a casino in1997 (43%) is almost four times what it was in1996 (11%). The increase is panicularly strongamong the young, the non-married and the mid-dle and lower income groups. For example, in1997 the number of low-income people panici-pating in casino gambling increase 4.6 times(from 6% to 34%), whereas the number of high-income people panicipating in casino gamblingincreased only 2.5 times (from 20% to 49%).Prior to the opening of the local casino, casinogambling was more or less limited to the rela-tively well-off in Niagara Falls, but losing moneyat the table has become more of an equal-opponunity proposition.

Changes in general attitudes to gamblingWe turn now to the question of change or stabil-ity in attitudes in the wake of the coming of theNiagara casino. Table 2 shows that general atti-tudes to gambling stayed stable in Niagara Fallsbetween 1996 and 1997. The casino's adventseems to have had no effect in this time periodon the general disposition of the communitytowards gambling. In both years, a majority of


Table 1. Percentage participating in non-charity casino gambling during the past 12 months, bydemographic characteristics, adults (18 + ) , Niagara Falls, 1996-97 and Ontario, 1995-97

Total sample ( ± %y

GenderMaleFemale

Age18-2930-3940-4950-6465 +

Marital statusNever marriedMarriedPreviously married

EducationLess than high schoolCompleted high schoolSome college/universityUniversity degree

Income< 30 000$30 000-49 000$50 000-79 000$80 000 +Not stated

Niagara Falls1996

(n = 677)

11(±2.4)

139

101412109

10138

98

1710

68

132019

Niagara Falls1997

(n=1076)

43(±3.0)

4243

5246463721

514237

35404945

3442544933

Ontario1995

(M=1030)

12(±2.0)

1211

129

16117

121113

119

1214

512131511

Ontario1997

(n=1005)

22(±2.6)

2519

2324212314

202223

17212520

1221252720

' 95% confidence interval.

respondents clearly recognized positive aspectsof gambling, but two-thirds also saw a downside:that it is "usually addictive". A stable one-quar-ter of the population rejected gambling morethoroughly, seeing it as immoral.

In contrast to the Niagara Falls results, there issome evidence in Table 2 that in Ontario as awhole attitudes turned somewhat less approvingof gambling. A somewhat smaller proportion ofrespondents in 1997 than in 1995 gave positiveresponses to the three items concerning goodthings about gambling. Positive responses in-creased for one of the negative items ("is usuallyaddictive"). The end result in 1997 was thatprovincial attitudes, which had been more favor-able to gambling than attitudes in Niagara Falls,came closer to matching the attitudes in NiagaraFalls.

The right-hand side of Table 2 shows theresults of tests of the significance of differences,using a 2 X 2 factorial analysis of variance on thefull distribution of each item. Ontario residents

were more likely to feel that gambling providesan opportunity for friendly socializing[F(l,3736) = 58.1, / )< 0.01], and that gamblingis a good way for a person to take a little risk[F(l,3736) = 35.8, p<0.01]. From 1995/96 to1997 scores significantly decreased on Friendlysocializing [F(l,3736) = 15.1, p < 0 . 0 1 ] , a littlerisk [F(l,3736) = 7.98, />< 0.01] and gambling isimmoral [F(l,3736) = 3.9, / )<0.05] . In 1997more people endorsed gambling is usually addic-tive in both Niagara and Ontario samples[F(l,3736) = 3.0, p < 0 . 0 5 ] . For all three posi-tive items—socializing [F(l,3736) = 4.7,/><0.05], exciting [F(l,3736) = 3.9, p<0.01]and little risk [F(l,3736) = 3.8, p = 0.05]—therewere significant interactions, reflecting thatagreement had dropped more in the Ontariosamples than in the Niagara Falls samples. How-ever, despite the trend, the provincial means forsocializing [r(2049) = 4.1, p<0 .01] and little risk[r(2027) = 2.9, p<0.0\] are still significantlyhigher in 1997 than the Niagara Falls means. In


+00

=* 2O aJ

8^

_ ino o

*C t o

2

O i^'C u on ov „c2 IIO >

n ON ^^

CO

«

s

c _a-3:5

i e, s - = , s |

nilo oo

S •2 " i2 V

t; ii o -s."ii X o .9 "i.

S » l lIlli=J| i s "g ^ s ,2 ^

c c -s-S =•2 - g 'S ^O H O .. 1>

£i . 2 cn -U rt <u

3 c 2 t « 2

•S fi .S c s Si


contrast, the average for excitement was higher inthe province in 1995/96 [t(1256) = 2.1,p<0.05], but was higher in Niagara in 1997 [t(1985) = -1 .9 , p<0.05].

Approval of the casinoAs is indicated in Table 3, two questions askedNiagara Falls respondents if they approved of thecasino. On both, more than 70% of the respon-dents approved of the casino. The 1997 dataindicate a non-significant increase in approval. Itshould be noted, however, that although morepeople approved of the casino overall, the num-ber of people that strongly approved of thecasino dropped. For example, on the first ap-proval question, "strongly approve" droppedfrom 34.9% in 1996 to 29.9% in 1997, while"strongly disapprove" decreased from 16.8% to13.0%. In comparing 1996 and 1997 data, theincrease for "somewhat approve" compared to"strongly approve" was significant[Wald(l) = 8.9, p<0.01].

In the provincial surveys, one question asked ifthe respondents were in favor of having a casinoin their community. In contrast to the NiagaraFalls data, only slightly more than half the re-spondents agreed or strongly agreed with thisstatement, and this figure decreased from 53.1%in 1995 to 50.3% in 1997 [t(1592) = - 9.0,p<O.Ol].

Expectations and experiences on the effects of theNiagara casinoThe remainder of Table 3 focuses on what Nia-gara Falls residents see as the effects of thecasino on the community. Looking first at thedata for 1997, we see that there is a wide percep-tion (90%) of an increase in gambling addictionin the community (item 8). On the other hand,relatively small percentages see some potentialproblems as having occurred—only 16% thinkpeople have moved away because of the casino(item 3), and only 18% perceive an increase inthe proportion on welfare or other public assist-ance (item 5; this result may also reflect generalprovincial reductions in the welfare rolls in thisperiod).

There is wide agreement (88%) on a percep-tion that the number of jobs in the communityhas increased as a result ofthe casino (item 13);about half the sample perceives that the average

personal income of residents has increased (item14); but almost two-thirds ofthe sample perceivethat most of the profit from the casino will go tooutsiders, rather than stay in the community(item 20). Only a third of the respondents seethere having been an increase in the variety ofentertainment in the community (item 10).

Traffic congestion is the most pervasive en-vironmental problem perceived by Niagara Fallsresidents (item 15); in concert with this almosteveryone perceives the number of American visi-tors to have risen (item 27).

Comparing expectations in 1996 with theseperceptions of what actually happened, the gen-eral picture is that reality turned out to be lessdramatic, both for good and for bad, than ex-pected. The majority of the differences betweenrates for 1996 expectations and for 1997 experi-ences are statistically significant. While 77% hadexpected the number of serious crimes to in-crease, now only 44% perceive this as havinghappened (item 4). The reality also turned out tobe less frightening than the expectations foryoung people getting into trouble with the law,for family break-ups, and for people movingaway (items 7,6,3). Except for traffic congestion,expectations of environmental problems turnedout to have been overblown: litter on the streetswas much less often a problem in the event thanhad been expected (item 16), and fewer hadexperienced higher noise levels and drunkendisturbances in their neighborhoods than hadexpected to do so (items 18, 19).

On the other hand, the reality was also oftenless dramatic than expected on the positive side.Compared to what they expected, fewer peoplethought that shopping, entertainment opportuni-ties, property values or business income hadincreased (items 9-12).

Only two items showed a significant increasefrom 1996 to 1997. More people believe thatmost of the profits are going to outsiders[r(1487) = 2.97, p < 0.01], and more people be-lieve that the number of people addicted togambling has increased [r(l 548) = 1.95,p = 0.05)].

Change in reported employment among NiagaraFalls respondentsRespondents were also asked about their em-ployment status. The survey data indicate asmall increase in the proportion of adults in


Table 3. Opinions about the effect of the opening of the casion, Niagara Falls 1996 (expectations) and 1997 (experiences)

Questions

Niagara Falls1996 (%)(N=677)

Niagara Falls1997 (%)

(N= 1076)

Approval of the casino1. The casino to be opened soon in Niagara Falls (approve)^2. A casino is a good idea for the Niagara region (agree)''

Community impacts: expected/perceived social problems and disruption3. People who live in Niagara Falls will move away because of

the casino (agree)4. The number of serious crimes (increase)""5. The number of people on welfare or other social assistence

(increase)6. The number of marriages and families breaking up

(increase)7. The number of young people who will be in trouble with the

law (increase)8. The number of people who become addicted to gambling

(increase)Community impacts: expected/perceived economic and amenity benefits

9. The variety of stores and services (increase)10. The variety of entenainment (increase)11. Property values (increase)12. The amount of money going to stores and local businesses

(increase)13. The number of jobs (increase)14. The average personal income of residents (increase)

Community impacts: expected/perceived environmental problems15. The level of traffic congestion (increase)16. The amount of litter on the streets and lawns (increase)17. The size of crowds in public places (increase)18. The noise levels (increase)19. Disturbances caused by people who have been drinking

(increase)Community impacts: other expectations/experiences

20. Most of the profits of the casino will go to outsiders (agree)21. The more opportunities people have to gamble, the more

they will gamble (agree)22. Because of the casino, the kinds of tourists visiting the

Niagara region will change (yes)'23. A job in the gambling industry is as good as any other job24. The cost of goods and services (increase)25. The availability of parking (decrease)*^26. The amount of time people spend volunteering in the

community (decrease)27. The number of Americans who visit (increase)28. Household income will change because of the opening of

the casino (increase)

7375

28

7724

50

63

87

85876684

9356

9859907172

5071

65

78634819

9818

7677

16*

44*18

42*

35*

90*

64*39*46*65*

88*49*

88*17*64*32*31*

62*73

58*

7930*3822

9711*

Missing data (i.e. "don't know" responses and refusals) were excluded from analyses. *diflference significant by t-teston means of full distributions, p<O.Ol. "Percentage responding "somewhat approve" or "strongly approve" asopposed to "somewhat disapprove" or "strongly disapprove". ""Percentage responding "somewhat agree" or "stronglyagree" as opposed to "somewhat disagree" or "strongly disagree". 'Percentage responding "yes" versus "no".Percentage responding "a large increase" or "a small increase" as opposed to "no change", "a small decrease" or

"a large decrease". "Percentage responding "a large decrease" or "a small decrease" as opposed to "no change", "asmall increase" or "a large increase".

Niagara Falls working full-time (47.6% in 1996vs. 49.7% in 1997) and a reduction in tfie pro-portion of people working part-time (13.7% vs.13.3%) or unemployed (7.5% vs. 6.2%). In con-

trast, in the provincial sample there was verylittle change in the rates of full-time (53.6% in1995 vs. 52.7% in 1997), part-time (11.0% vs.9.7%) or unemployed people (4.3% vs. 4.2%).


The difference in trends between the provinceand Niagara Falls failed to reached significance[Wald(l) = 0.54, NS], nor was there any overalltrend according to year; however, the data sug-gested that the unemployrrient rate was higher inthe Niagara area than in the province as a whole[Wald(l) = 9.97, ;.<0.01].

Although non-significant, the trend in NiagaraFalls was, as expected, an increase in the numberof people working full-time. There are approxi-mately 60 000 adults residing in Niagara Falls. A2.1%) increase in the proportion of adults em-ployed (49.7-47.6%) would thus translate intoapproximately 1260 new jobs.

In 1997 respondents were also asked if theyhad worked at the casino in the past year. Alto-gether 39 (3.6% of the sample) responded thatthey had. Among 60 000 adults, that wouldtranslate into 2160 people. The actual employ-ment at the casino is 3500 people, but this wouldinclude employees residing outside NiagaraFalls. To the extent that the responses to thesurvey give a reliable estimate, they suggest thatonly 62% of the casino employees reside in Nia-gara Falls. Tfie increase in employment reportedby the respondents in the Niagara Falls surveys isthus less than the reponed direct employment inthe casino. However, those employed in the1996 survey would include some who werealready employed by the casino prior to itsopening.

In the survey data, the increase in numbers ofjobs in the city (1260) is lower than might beexpected, given that the casino has also beencredited with creating "more than 6,000 jobs" inthe Niagara region beyond the direct jobs in thecasino (Government of Ontario, 1998). Thesample sizes of the present study have a powergreater than 0.99 to detect the impact of theprojected 9500 additional jobs (direct + indirect)on the proportion of respondents that are em-ployed. Assuming that only 61% of those jobsare taken by Niagara Falls residents and theremainder filled by commuters, the sample stillhas a power of 0.98 to detect an increase. Eventaking into account only 61% of the 6000 pro-jected indirect jobs, the sample still has a powerof 0.81 to detect an increase. Thus the power ofthe study is sufficient to suggest with a highdegree of confidence that a net increase in jobsdid not occur at the levels which had beenprojected.

Data from Statistics Canada (Statistics

Canada, 1998) also do not indicate any increasein employment in the Niagara-St Catherines re-gion. According to Statistics Canada data, theseasonally adjusted unemployment rate in theregion in November 1996 was 9.0%, and inNovember 1997 was 9.5%. For the 6 monthsprior to the Casino opening, the unemploymentrate in the Niagara region had been very similarto the provincial average; however, for the first 6months after the casino opened the unemploy-ment rate rose, reaching its highest level since1994 in May, 1997. After May, the rate fell,returning to the falling trend in the province as awhole (Statistics Canada, 1998).

On the 1997 Niagara Falls survey, respon-dents were asked to estimate how much of themoney spent at Casino Niagara they would havespent on entertainment, another form of gam-bling or necessities. A total of 432 Niagara resi-dents provided answers to these questions.Responses that added up to more than 100%were excluded. Altogether, 299 respondents pro-vided usable displacement data. On average,these respondents reported that 80% of themoney spent at Casino Niagara was divenedfrom some other type of expenditure: 62% fromentertainment, 11% from other forms of gam-bling and 8% from necessities of life (food, rent,etc.). Another 88 respondents from the provincealso provided information for this question andtheir displacement figures were very similar tothose in the Niagara Falls sample.

Changes in reported rates of problems from gamblingAs noted earlier, we used several sets of variablesas indicators of rates of gambling problems in thecommunity, both in terms of respondents' owngambling and concerning the gambling of thosearound them. Comparison of 1997 Niagara datawith 1996 data and with the Province were con-ducted using a priori r-tests.

Short SOCS. Results for the five items of theShon SOGS measure are shown in Table 4. Inthe Niagara Falls samples, the proportion ofpositive responses increased on all five of theseitems between 1996 and 1997. The proportionof the sample with a score of 2 or more on theseitems rose from 2.5% to 4.4%, and those with ascore of 3 or more on these items rose from0.7% to 2.3%. The rate of those reporting two ormore problems thus increased by about three-


Table 4. Mean and standard deviations for gambling problem indicators for the past 12 months: Niagara Falls,1996-97 and Ontario, 1997

Niagara Falls1996

(N=677)

Niagara Falls1997

(N= 1076)

Ontario1997

(N= 1005)

Average Short SOGS. Score out of 5 (SD)

Average number of life area problems. Score out of 5(SD)

Number reporting being criticized about gamblingfrom family or friend. Score out of 2 (SD)

Number reporting a friend or family member with agambling problem. Score out of 2 (SD)

0.131*(0.461)0.060

(0.347)0.022*

(0.177)0.191*

(0.451)

0.198(0.640)0.079

(0.484)0.044

(0.256)0.278

(0.515)

0.140*(0.469)0.078

(0.365)0.023"

(0.176)0.173''

(0.413)

* Significantly lower than Niagara Falls 1997, p < 0.05 (t-test)

quarters. The average number of positiveresponses was 0.131 in 1996, and 0.198 in1997—a significant increase of about one-half [t (1719) = 2.54, p<0.05]. As can be seenin Fig. 3, the increase in problematic gamblingbehaviors is spread across the entire distribution.This indicates that the differences found werenot the net result of a small number of outliers,but reflect a broadly-based upward shift in thedistribution.

The average number of problems in the prov-ince as a whole in 1997 was 0.139, which is verysimilar to the 1996 Niagara data andsignificantly smaller than the 1997 Niagara data[£(2104)= -2.47, p<0.05]. This suggests thatthe increase found in the city of Niagara wasisolated to that city and not found across the restof the province.

Unfortunately, data for problems in the pastyear for the reduced SOGS were not availablefrom the 1995 provincial survey. However, life-time data for these questions were available fromthe 1995 data. There was no significant differ-ence between the life-time rates in the Provincein 1995 and in 1997 [j(2033) = 0.38, NS].These results from the life-time data stronglysuggest that problem gambling measured by theShort SOGS did not increase in the province asa whole, while it increased significantly in Nia-gara Falls.

Life-area problems items. Reported rates ofproblems in five life-areas due to gambling alsoincreased uniformly in Niagara Falls between1996 and 1997; however, these changes were

small and non-significant [r(1705) = 1.0, NS; seeTable 4]. The provincial data show a non-significant decrease in the average number of lifearea problems reported [t(2033) = 0.41, NS].The interaction was also non-significant.

Pressure from others on gambling. Table 4 alsoshows the mean scores (range 0-2) for respon-dents reporting that a friend or family memberhad said something to them about their gamblingin the past 12 months. There was a significantincrease from 1996 to 1997 in respondents' re-ports of having family members or friends talk tothem about their gambling [r(1734)= -2.19,p<0.05]. The mean for Niagara Falls in 1997was also significantly higher than for Ontario,[F(1962) = 2.13, p<0.05]. A comparison ofdata for the province in 1995 and 1997 did notindicate any trend towards an increase in theprovince as a whole [t(2033) = 0.67, NS].

The life-time data for this variable also showthe same pattern of results. Interestingly, theentire increase in these indicators in NiagaraFalls on a life-time basis can be attributed tochange in the past year. While both life-time andpast year measures increased, the difference be-tween the two—the rate of past gambling prob-lems (life-time but not current)—stayed stable.

Gambling problems in the family and amongfriends. The study also included information onthe respondent's experience with others' gam-bling problems (see row 4 on Table 4). Theproportions reporting this in the past 12 monthsin Niagara rose from 5.0% to 7.5% for family


co

o 4o>c 3

D Niagara 1996• Niagara 1997

1 2 3 4 5Number of problem behaviors

Figure 3. Problem gambling behavior: Niagara 1997compared to Niagara 1996.

gambling-related problems in the Niagara re-gion. In allj there are 18 separate items behindthe data given in Table 4 measuring the preva-lence of a gambling-related problem. There wasa statistically significant increase in the NiagaraFalls rates for five of these items. However, forall 18 of them the rate in the 1997 Niagara Fallssample is higher than in the 1996 Niagara Fallssample. If these were fully independent observa-tions, the chances that 18 comparisons would allfall the same way on a random basis are one in262 144. Since the same respondents areanswering each of the 18 questions, this calcu-lation overestimates the unlikelihood of this re-sult, but it underlines the finding that there isstrong evidence in the data from this study thatthe rates of gambling-related problems rose inNiagara Falls in the year after the opening of thecasino, and that this rise appears to be greaterthan any rise in problem rates in the province asa whole.

members, and from 14.0% to 20.5% for friends.The means for the family or friends problemsscore (range 0-2) given in Table 4 follow a verysimilar pattem to the data on pressure fromothers, although at considerably higher rates.The number of people in Niagara Falls thatreport having family members or friends with agambling problem increased significantly from1996 to 1997 [r(1750)= -3.99, p<O.Ol]. Themean for Niagara Falls in 1997 was alsosignificantly higher than the mean for Ontario inthe same year [r(2079) = -4.94, p < 0.01]. Onlylife-time data were available in the 1995 provin-cial sample; a comparison of this with life-timedata from 1997 did not indicate any trend to-wards an increase in the province as a whole.

It should be noted that responses to thesequestions cannot be used to estimate prevalencerates of gambling problems in the community,since there may be many family members orfriends aware of and concerned about a singlepathological gambler. However, changes in ratesof respondents with family or friendship links tothose with gambling problems are a useful indi-cator of how the prevalence of gambling prob-lems is changing in the community.

Problem indicators: summaryIn summary, there were several findings whichprovided converging evidence for an increase in

DiscussionThe Niagara Falls casino was built on the borderwith the United States, with the explicit goal ofattracting gamblers from the United States. Inthe jargon of the gaming industry, it is a"destination" casino, designed to attract cus-tomers from elsewhere, and not simply depen-dent on local trade. The observations ofrespondents in our study support data fromother sources in concluding that the Niagaracasino has been successful in this goal: 97% ofthe sample considered that "the number ofAmericans who visit" Niagara Falls had in-creased because of the casino.

In evaluating costs and benefits, the balancesheet for a community which hosts a"destination casino" will be different from thatfor a community with a casino serving primarilylocal customers. For a community with a desti-nation casino, there is the hope of attractingincreased commerce, revenue and employment,offset only by the minor inconvenience and en-vironmental troubles related to crowding, trafficcongestion and parking problems. This would bethe situation if all the gamblers in the casinowere from elsewhere. As events in North Amer-ica are showing, this policy of making money byattracting tourists to a casino tends to work onlyin the short term. Neighboring cities tend tobuild counter-attractions to win the commerce.


revenue and employment back to their side ofthe river. As an example, the casinos in Windsor,Ontario have precipitated plans to build threecasinos across the river in Detroit.

Apart from that, the data from the presentstudy show clearly that the casino's customersare not limited to those from out of town. Theproportion of Niagara Falls residents who hadengaged in casino gambling during 1996 was11%; 1 year later the proportion was 43%. Theaverage amount of money that the Niagara Fallssample reported spending in non-charity casinosin that year quadrupled, from $2.30 in a monthto 511.10, and 3.6% of the people in Niagarareport going to the casino 2 or more times amonth. Some of this extra spending came at theexpense of other forms of gambling, notablyhorse racing. However, most of it appears tohave come as a diversion from other expendi-tures, such as on entertainment, that would havebeen made in the community.

In terms of reported employment in the com-munity, data from our Niagara Falls respondentsdo show a non-significant gain. There are jobs inthe community directly attributable to buildingand operating the casino; but our data suggestthat gains in employment and commerce areprimarily at this direct level. There is little evi-dence on a net basis of the secondary, pass-alonggains that are projected in conventional modelsof the economic impact of opening a casino. Alarge number of factors affect the employmentrate. Perhaps, as the mayor of Niagara has sug-gested, the unemployment rate would have beenmuch higher if the casino had not been opened(Gray, 1997). Another possible explanation forthe missing indirect jobs is that direct and indi-rect employment due to the casino is, to a con-siderable extent, diverted from other economicactivities, rather than simply added onto thelocal economy. Any new business or industrythat is brought into a community contributes tothe local economy, but some of the new businessrepresents displacement (Persky, 1995) fromother economic activities: a person that mighthave spent $100 on a restaurant dinner mayinstead spend that $100 gambling. This displace-ment hypothesis is supported by the data. Ac-cording to the respondents, most of the moneythey spent at the casino was diverted from someother form of entertainment. Gambling losses atthe casino by local residents quietly subtractfrom other parts of the local economy, in terms

of other expenditures forgone, and this effectmore or less cancels out the added employmentand commerce from visitors' meals, stays atmotels, etc.

When questioned just before the casino'sopening, Niagara Falls respondents had vividexpectations about its effects. Strong majoritiesexpected not only increases in the number ofjobs, but also improvements in the variety ofentertainment, stores and services in the com-munity. On the other side of the balance sheet,strong majorities expected an increase in seriouscrimes, in public disturbances and in noise levelsand crowding in the community, as well as anincrease in the level of traffic congestion. After ayear of actual experience, the picture is more inshades of gray. The expectations of more jobsand more traffic congestion were borne out, ac-cording to the residents' perceptions, but muchsmaller proportions reported each of the othereffects, good or bad.

The effects mentioned so far are all fairlypublicly visible. A casino also has potential ef-fects more in the area of private life, in the familyor in individual pleasure or despair. A strongmajority of respondents had expected an increasein the number of people who become addicted togambling in the wake of the casino. A year latereven more respondents (90%) saw this as areality. Respondents' attitudes to gambling ingeneral showed no net change, but they werevery clearly aware of this downside to the casino.

Our data on changes in actual gambling be-havior and the experience of problems in generalbear out the respondents' perceptions. Theamount Niagara Falls respondents reportedspending on gambling did increase, with most ofthe increase from spending in the type of gam-bling (non-charity casinos) which includes theNiagara casino. There were increases in rates oneach of the items used to measure problem gam-bling. Given the restricted sample sizes, onlysome of these results are significant at the 5%level, but the cumulative effect is highlysignificant. We can state with considerableconfidence that rates of self-reported gambling-related problems rose among Niagara Falls resi-dents in the wake of the opening of the casino.

These results are also supported by data onthe proportion of respondents reporting that theyhave family members or friends with problemswith gambling in the last 12 months. This rise inthe proportions reporting gambling problems in


the family or among friends was highlysignificant.

The casino has thus clearly brought problemsto the Niagara Falls community; it has particu-larly brought problems into the arena of privatelife, behind curtains rather than out on thestreet. Despite this, Niagara Falls residents re-main firm in their support for the casino. In fact,the proportions supporting the casino rose alittle. Afrer a year of experience, three-quartersof the residents approved of the Niagara casinobeing there.

There is, of course, no substitute for actualexperience in evaluating the effects of an initia-tive such as the opening of the Niagara casino,but the results we have reported still represent arelatively early stage in the community's re-sponse to the Niagara casino. It is possible thatthere will be some differences in the effects in thelonger term. There will certainly be some remis-sions among those with gambling problems. Onthe other hand, a pathological gambling trajec-tory often takes several years to build up in anindividual's life, so that the coming years maybring a further rise in rates of gambling-relatedproblems. Competing casinos across the borderand elsewhere in Ontario are likely to cut intothe tourist trade of the Niagara casino, which isthe source of the most unambiguous gains to thecommunity. On the other hand, the casino isnow a social reality in the community, and it willgamer some support from those who tend tosupport the status quo. It would be worthwhile toobserve how trends in community benefits, prob-lems and perceptions unfold in the longer term.

There are limitations to the findings of thisstudy. The reliability of the findings—and partic-ularly the findings on employment and onamounts spent—is limited by the restricted sam-ple sizes, and by the sample attrition reflected inthe response rates. The measures of gamblingproblems used here rely on the reports of respon-dents, and are not equivalent to a full clinicalassessment of pathological gambling. A 1-yearfollow-up study cannot determine whether theincrease in problem gambling behavior will besustained, increase or disappear in subsequentyears.

Perhaps the most important implication ofthese data is the indication that more needs to bedone to prevent the development of gamblingproblems. While the majority of people thatgamble do so without experiencing gambling

problems, the increased availability of casinogambling does appear to result in an increase inproblems. Our recent studies have found thatgambling problems are related to not keeping toa spending limit, the use of systems, playing untilclosing time, not quitting when tired and to thebelief that random events are self-correcting.There is thus some potential to limit the increasein gambling problems by public education ongambling (see also Turner, 1999) and by en-vironmental modifications of gambling venues.

AcknowledgementsThis study was supported by the Donner Cana-dian Foundation and by the Ontario Ministry ofHealth. The views expressed are those of theauthors and do not necessarily reflect those oforganizations with which they are affiliated or ofthe funding agencies. The study was carried outwhile the first author was at the Addiction Re-search Foundation. We acknowledge the contri-butions of Josh Paul, Jurgen Rehm, Phil Lange,Susan Bondy, Ed Adlaf and Ron Roizen to thestudy, and helpfril comments from CarolynNutter.

ReferencesAASVED, M . J. & SCHAEFER, J. M. (1992) Who Needs

Las Vegas? Gambling and its impacts in a southwesternMinnesota agricultural community (St Paul, Min-nesota, Department of Health Services, MentalHealth Division).

AASVED, M . J. & LAUDERGAN, C . J. (1993) Gamblingand its impacts in a northeastern Minnesota com-munity: an exploratory study. Journal of GamblingStudies 9, 301-319.

AASVED, M . J., SCHAEFER, J. M., et al. (1995) Legalizedgambling and its impacts in a central Minnesotavacation community: a case study. Journal of Gam-bling Studies, 11, 137-163.

ADDICTION RESEARCH FOUNDATION (1996) Building aStrategic Response to Gambling Problems in Ontario(Toronto, ARF).

BRITISH COLUMBIA, MINISTRY OF GOVERNMENT SER-VICES (1994) Report of the Gaming Polity Review,(Victoria, Queen's Printer).

CAHALAN, D . (1970) Problem Drinkers (San Francisco,Jossey-Bass).

CANEDAY, L . & ZEIGER, J. (1991) The social, econ-omic, and environmental costs of tourism to a gam-ing community as perceived by its residents. Journalof Travel Research 30, 45-48.

ERNST & YOUNG (1994a) Assessment of Casino Windsor:final repon, November 1994 (Toronto, Emst andYoung).

ERNST & YOLING (1994b) The Econometric ResearchLimited economic impact model, appendix 2 to:


Assessment of Casino Windsor: final report, November1994, (Toronto, Ernst and Young).

FERRIS, J., STIRPE T . & IALOMITEANU, A. (1996) Gam-

bling in Ontario: a report from a general populationsurvey on gambling-related problems and opinions, re-vised edn, research document series no. 130,(Toronto, Ontario, Addiction Research Foun-dation).

GOVERNMENT OF ONTARIO (1998) Falls ManagementCompany chosen as preferred proponent for the NiagaraFalls Casino/Gatetvay Project (Toronto, Govemmentof Ontario Press Releases).

GovoNi, R. & FRISCH, R . G . (1996) TTie Impact of theWindsor Casino on Adult Gambling in the City ofWindsor, preliminary repon, (Windsor, Ontario,Problem Gambling Research Group, Department ofPsychology, University of Windsor).

GRAY, J. (1997) Niagara Falls to take its biggest gam-ble yet: when a honeymoon haven goes after highrollers, it wants to build Las Vegas North instead ofwax museums. Globe and Mail (Toronto), 5 Dec.

KINROSS, I. (1995) Problem gambling: a special report.The Joumai (ARF, Toronto), 24(6), 5.

LADOUCEUR, R. (1996) The prevalence of pathologicalgambling in Canada, Journal of Gambling Studies 12,129-142.

LESIHUR, H . R . & BLUME, S. B . (1993) Revising theSouth Oaks Gambling Screen in different settings,Joumai of Gambling Studies, 9, 213-219.

MCGURRIN, M . C . & ABT, V. (1992) Overview ofpublic policy and commercial gambling, Joumai ofGambling Studies, 8, 325-329.

NATIONAL COUNCIL ON WELFARE (1996) Gambling in

Canada, Cat. no. H68-40/1996E (Ottawa, Ministerof Supply and Services Canada).

PAGUA, A . (1995) Alcohol, Tobacco, and Drugs: depen-dence, problems and consequences of use: a report of the

1994 Ontario Alcohol and Other Drug Opinion Survey,research document no. 121 (Toronto, AddictionResearch Foundation).

PERDUE, R . R . , LONG, P. T., et al. (1995) Residentsupport for gambling as a tourism developmentstrategy, Joumai of Travel Research, 33, 3-10.

PERSKY, J. (1995) Impact studies, cost-benefit analysisand casions, Joumai of Gambling Studies, 11, 349-360.

SINGLE, E . , EASTON, B . , COLLINS, D . , HARWOOD, H . ,LAPSLEY, H . & MAYNARD, A. (1996) IntemationalGuidelines for Estimating the Costs of Substance Abuse,(Ottawa, Canadian Centre on Substance Abuse).

SMART, R . G . & FERRIS, J. (1996) Alcohol, drugs andgambling in the Ontario adult population, 1994,Canadian Joumai of Psychiatry, 41, 36-45.

STATISTICS CANADA (1998) Employment statistics for theNiagara-St. Catherines region, located at http://www.statcan.ca/cgi-bin/Cansim/modify.cgi?format= KW, on the world wide web.

STOKOWSKI, P . A . (1996) Riches and Regrets: betting ongambling in two Colorado mountain tovins, (Niwot,Colorado, University Press of Colorado).

THOMPSON, W . N . , SCHWHR, R . K . , HOYT, R . & BROS-NAN, D. (1993) Not in my backyard: Las Vegasresidents protest casinos, Joumai of Gambling Studies,9, 47-62.

TURNER, N . E . (1999) Doubling vs. constant bets asstrategies for gambling, Joumai of Gambling Studies,in press.

TURNER, N . E . , IALOMITEANU, A. & ROOM, R . (1999)Checkered expectations: predictors of approval ofopening a casino in the Niagara community, Joumaiof Gambling Studies, in press.

WINTERS, K . C , STINCHFIELD, R . D . & KIM, L . G .(1995) Monitoring adolescent gambling in Min-nesota, Joumai of Gambling Studies, 11, 165-183.

Documents

Community effects of the opening of the Niagara casinofaculty.buffalostate.edu/macleamg/440/Room 1999 Add.pdf · on horse races. Lotteries, bingo and parimutuel wagering are available