Cigarette Smoking Before and After an Excise Tax Increase and an Antismoking Campaign--Massachusetts, 1990-1996

In November 1992, residents of Massachusetts approved a ballot petition(Question 1) that increased the tax on each pack of cigarettes from 26¢to 51¢ beginning January 1, 1993, and requested that the legislaturespend the proceeds on tobacco control and health education. The MassachusettsTobacco Control Program (MTCP), administered by the Massachusetts Departmentof Public Health (MDPH), was established in response to the approval ofthe petition. In October 1993, the MTCP initiated a statewide mass-mediaantismoking campaign. In early 1994, the program began funding local boardsof health and school health and other youth programs to promote policiesto reduce public exposure to environmental tobacco smoke and to restrictyouth access to cigarettes. Efforts also included support to health educationprograms, primary-care providers, and other services to help smokers quit.Through June 1996, MTCP spent $116 million, including $43 million for themass-media campaign.

Assessing the Effects of the Tax

To assess the effects of the excise tax increase and the antismokingcampaign on cigarette smoking in Massachusetts, the CDC and MDPH analyzeddata on the number of packs of cigarettes taxed per capita and the prevalenceof cigarette smoking during the period preceding (1990 to 1992) and following(1993 to 1996) implementation of the ballot petition. This report summarizesthe findings of the assessment and compares trends in cigarette consumption(ie, purchases) in Massachusetts, in California (where a voter-mandatedcigarette tax increase in January 1989 funded a statewide antismoking campaignthat began in April 1990), and in the 48 remaining states and the Districtof Columbia combined. The findings suggest that the number of packs ofcigarettes taxed per capita declined substantially in Massachusetts afterimplementation of the ballot petition.

For each full calendar year from 1990 through 1995, taxable cigaretteconsumption for Massachusetts, California, and the other states and theDistrict of Columbia combined was derived from monthly reports from theTobacco Institute on tax receipts for wholesale cigarette deliveries. Taxableconsumption for 1996 was estimated as twice the cumulative values for January-June.Per capita rates (in packs/year) were based on the resident populationage 18 or more years in each state.

Data on the average retail price of a pack of cigarettes in Massachusettsat 4-week intervals during 1990 to 1995 were based on bar-code scanningdata provided by Information Resources, Inc. Data were obtained for a seven-countyregion (including the Boston and Worcester metropolitan areas) that represented83% of residents of Massachusetts based on 1990 census estimates. The observedretail prices of cigarettes were adjusted for inflation by using the consumerprice index for urban workers in the Boston metropolitan area.

Data from the Behavioral Risk Factor Surveillance System (BRFSS) for1990 through 1995 (the most recent year for which data were available)were used to estimate the annual prevalence of cigarette smoking amongadults in Massachusetts, California, and the remaining participating statescombined. The BRFSS is a population-based, random-digit-dialed telephonesurvey of the noninstitutionalized US population age 18 or more years.The District of Columbia and seven states (Alaska, Arizona, Kansas, Nevada,New Jersey, Rhode Island, and Wyoming) were excluded because they did notparticipate in BRFSS one or more years during 1990 to 1995. Because samplingerrors for annual BRFSS estimates precluded precise year-to-year comparisons,3-year average prevalences were estimated for 1990 to 1992 and 1993 to1995.

Defining a Smoker

A current smoker was defined as any respondent who answered "yes"to the following two questions: "Have you smoked at least 100 cigarettesin your entire life?" and "Do you smoke cigarettes now?"Estimates were weighted based on the number of telephones per householdand the age, sex, and racial/ethnic composition of the residents of theindividual states. The prevalence of smoking for the remaining participatingstates combined was computed as a population-weighted average of the prevalencesestimated for the 41 states that participated in BRFSS every year during1990 to 1995. SESUDAAN was used to calculate 95% confidence intervals (CIs).

During 1990 to 1992, taxable per capita consumption of cigarettes byadults declined 6.4% in Massachusetts, 11.0% in California, and 5.8% inthe 48 remaining states and the District of Columbia combined (Table1). In Massachusetts, from 1992 (the year before implementation ofthe petition) to 1996, taxable per capita consumption declined by 19.7%(from 117 to 94 packs) (Table 1); inCalifornia and the remaining states, per capita consumption declined by15.8% and 6.1%, respectively.

Immediately after the Massachusetts petition became effective on January1, 1993, the real price of cigarettes increased sharply but subsequentlydeclined (Figure 1). In response to increasingsales of discount brands, in April 1993 one US cigarette manufacturer announceda nationwide, 40¢-per-pack price discount on its major premium brand,and in May, another manufacturer matched the discount on its major premiumbrands.

In August, all manufacturers announced a permanent wholesale price reductionof 39¢ per pack on all premium-brand cigarettes. As a result of thesenationwide price reductions, by the end of October the real price of cigarettesin Massachusetts had declined to the 1992 level (Figure1).

The prevalence of current smoking among adults in Massachusetts was23.5% (95% CI, ±1.4%) during the 3 years before implementation ofthe petition (1990 to 1992) and 21.3% (95% CI, ±1.2%) during the3 years after implementation (1993 to 1995). In comparison, the prevalenceof adult smoking declined 2.7% in California (from 20.1% [95% CI, ±0.9%]during 1990 to 1992 to 17.4% [95% CI, ±0.9%] during 1993 to 1995)and 0.8% in the 41 other BRFSS participating states combined (from 24.1%[95% CI, ±0.3%] during 1990 to 1992 to 23.4% [95% CI, ±0.2%]during 1993 to 1995).

Editorial Note from the CDC

The findings in this report indicate that, in the state of Massachusetts,the number of packs of cigarettes taxed per capita decreased significantlyduring 1992 to 1996, following implementation of a ballot petition to increasethe excise tax on cigarettes and initiate an antismoking campaign. Thischange was similar to decreases in California, the only other state toinitiate an extensive statewide antismoking campaign in conjunction withan increase in cigarette taxes. However, complexities related to the accuratemeasurement of changes in smoking prevalence among adults in Massachusettsrequire further study to determine the combined impact of the excise taxincrease and antismoking campaign on adult smoking prevalence in the state.

Although some smokers in states that implement increased cigarette excisetaxes may attempt to avoid higher prices by purchasing cigarettes in neighboringstates with lower prices, the 19.7% decline in per capita consumption ofcigarettes in Massachusetts during 1992 to 1996 probably reflects the effectsof the tax increase and antismoking campaign rather than increased cross-borderpurchases by Massachusetts smokers. During 1993 to 1994, cigarette excisetaxes in Connecticut and Rhode Island were increased to levels comparablewith those in Massachusetts; however, in New Hampshire, the real priceof cigarettes declined during 1992 to 1993, and taxable cigarette consumptionincreased by 17 million packs. Increased taxable consumption in New Hampshiremay reflect either a real upward trend in smoking by state residents orincreased cross-border purchases by Massachusetts smokers. However, evenif the 17 million-pack increase were attributed entirely to cross-borderpurchases by Massachusetts smokers, the decline in per capita consumptionin Massachusetts during 1992 to 1996 would still have been 17.0%, a significantreduction.

The findings in this report are subject to at least two limitations.First, the estimates of per capita consumption were based on tax receiptsat the wholesale level and not the actual number of cigarettes consumed.Distributors may delay or advance cigarette shipments in anticipation ofannounced wholesale price changes or excise tax increases. Such shiftingof wholesale deliveries may produce year-to-year changes in tax receiptsthat do not reflect actual changes in per capita consumption. However,temporal trends in taxable consumption over a period of several years probablyreflect actual consumption more accurately. Second, a decline in the numberof cigarettes taxed in a single state may result in an overestimation ofthe actual decline in consumption if resident smokers increase their out-of-statepurchases. However, the data on taxable per capita cigarette consumptionin Massachusetts and three adjacent states suggest that the increased purchaseof cigarettes by Massachusetts smokers in neighboring New Hampshire wasnot a major source of the reported decline in per capita consumption inMassachusetts.

Increases in the price of cigarettes can reduce per capita consumptionand the prevalence of smoking. In Massachusetts, however, the tax-inducedincrease in cigarette price was soon offset by coincidental national, industry-wideprice reductions that began during the spring of 1993. While real cigaretteprices returned to pre-1993 levels, per capita consumption in Massachusettscontinued to decline. This finding suggests that a tax increase combinedwith an antismoking campaign can be more effective in reducing per capitaconsumption than a tax increase alone. The MTCP plans additional evaluationsof this preliminary finding, including changes in smoking prevalence amongadults and further comparisons with findings from California and otherstates.

Adapted from Morbidity and Mortality Weekly Report 45(44):966-970,1996.