Impacts of Tax and Flavor Tobacco Policies on San Francisco Bay Area Tobacco Prices

Abstract

California Proposition 56 increased the state tobacco tax by $2 per cigarette pack effective April 1, 2017. Between 2015-2020 San Francisco (SF) and some cities in Alameda County enacted local flavored tobacco sales restrictions. SF also increased its Cigarette Litter Abatement Fee, from $0.20/pack in 2015 to $1.00 in 2020. Compare the change in tobacco prices before (2015) and after (2019/20) the implementation of a $2 increase in tobacco excise tax and local flavored tobacco policies in SF and Alameda Counties. Descriptive study of the pre-to-post policy analysis design. We drew a proportional random sample of retailers (N=463) in SF and Alameda Counties, by city. Using multivariable, single- and multiple-level linear regressions, we compared inflation-adjusted average tobacco prices in 2015 vs. 2019/20 by county and by flavor policy, accounting for socio-demographics. Change in inflation-adjusted average tobacco prices in 2015 vs. 2019/20 by county and flavor policy, accounting for socio-demographics. Between 2015-2019/20, the increase in cigarette prices was higher than the $2 tax increase, and higher in SF than Alameda County (+$4.6 vs +$2.5). SF retailers stopped selling Newport menthol cigarettes and Blu brand menthol e-cigarettes in 2019/20. Adjusted average cigarette prices increased significantly more in SF and Alameda County cities with comprehensive or partial flavor policies versus cities without flavor policies (by $3.23 and $2.11). Local flavor policies affected menthol product availability and may have had positive spillover effects and indirectly increased pack prices.

Increasing tobacco prices via taxation or other pricing policies can decrease tobacco use (U.S. Department of Health and Human Services, 2012) by reducing initiation, especially among youth (U.S. Department of Health and Human Services, 2012) who are more price-sensitive than adults (Corrigan et al., 2021; Ding, 2003), and by prompting smoking cessation (U.S. Department of Health and Human Services, 2020), especially among people of lower socio-economic status (SES) (Chaloupka et al., 2012; Hill et al., 2014). Price-related tobacco policies therefore have strong potential to mitigate disparities in smoking (Chaloupka et al., 2012; Hill et al., 2014). In California, Proposition 56 increased the state tobacco excise tax by $2 per pack of 20 cigarettes, representing an increase from $0.87 to $2.87 effective 1 April 2017 (California Proposition 56, tobacco tax increase, 2016). Simultaneously, Proposition 56 imposed a new tax on electronic cigarettes (e-cigarettes) and liquids with nicotine by adding them to the definition of “other tobacco products” (California Revenue & Taxation Code Article 2, 2016; California Revenue and Taxation Code Article 2.5, 2016; California Proposition 56, tobacco tax increase, 2016). According to the new tax policy, excise taxes for e-cigarettes and e-cigarette liquids with nicotine are now levied from distributors based on the percentage of wholesale costs from all distributed non-cigarette tobacco products (Tax guide for cigarettes and tobacco products, 2022). In the absence of federal taxation of e-cigarettes, California’s statewide tax including e-cigarettes, some of which can deliver approximately the same amount of nicotine as cigarettes (Prochaska et al., 2021) and appeal to youth (Cohn et al., 2019) and young adults (Rostron et al., 2020), was an important tobacco control measure.

Tobacco companies have exploited loopholes and used various strategies to undermine tobacco tax and pricing policies (Ross et al., 2017). For example, retailers regularly offer industry-sponsored promotions (i.e., coupons and discounts) that keep prices low (Apollonio & Glantz, 2020; Pierce et al., 2005). Tobacco companies also engage in differential shifting of tobacco taxes to consumer prices (Apollonio & Glantz, 2020), which may reduce the public health impact of increases in tobacco taxation. Both “under-shifting” and “over-shifting” of manufacturer pricing with tax increases may have advantages for the industry. “Under-shifting” taxes by reducing wholesale prices results in consumers paying less than the amount of a tax increase, which allows sellers to maintain demand among price-sensitive purchasers (Hiscock et al., 2017). “Over-shifting” taxes by increasing wholesale prices means consumers will pay more than the amount of tax increase, which maintains profitability (Ross et al., 2017). These promotional strategies have also targeted specific groups, such as racial and ethnic minorities (Henriksen et al., 2019), young adults (Choi & Forster, 2014), and people with lower SES (Osman et al., 2019), which may exacerbate tobacco use disparities (Ross et al., 2017). For example, a prior study that assessed the impact of California Proposition 56 on pricing revealed cigarette tax under-shifting in Black and Hispanic neighborhoods across the state (Henriksen et al., 2019).

In addition to the state-level tax increase, the variety of local tobacco control measures concentrated in a limited geographic region of the San Francisco Bay Area (Andersen-Rodgers et al., 2021) provided a unique setting for research on tobacco price variability in association with local tobacco regulations, including policies that are not necessarily tax- or price-related (Golden et al., 2016). Local tobacco ordinances were enacted in several cities between 2015 and 2020 (included an increased Cigarette Litter Abatement fee in San Francisco (SF), flavored tobacco sales restrictions, and other policies (Holmes et al., 2022). The present study compared changes in retail prices of tobacco products stratified by two San Francisco Bay Area counties before and after the implementation of the state excise tax increase, accounting for socio-demographic characteristics by city. Past studies focused on Proposition 56 impacts on pricing were limited to cigarettes only (Boettiger & White, 2020; Gunadi et al., 2021; Henriksen et al., 2019). We expand on previous research by also analyzing price changes for non-cigarette products, including flavored e-cigarettes. Knowledge obtained from this study is relevant for policy research in other related industries (e.g., alcohol, sugary drinks) that use similar strategies to undermine public health gains from increased taxation by using price promotions or engaging in differential tax shifting (Nelson & Moran, 2020; Rojas & Wang, 2021).

Furthermore, we found no prior studies assessing the impacts of local policies restricting sales of flavored tobacco products (i.e., flavor policies) on tobacco pricing in California; however, one recent study showed an association between a comprehensive policy restricting sales of menthol-flavored tobacco and decreased cigarette sales in Massachusetts. The same study revealed higher average prices for both menthol and non-flavored cigarettes when compared to other states without such flavor policies (Asare et al., 2022). Therefore, we hypothesized that reduced availability and sales of flavored tobacco products in SF (Vyas et al., 2021; Yang et al., 2020) and some Alameda County cities (Kurti et al., 2020), due to comprehensive or partial flavor policies, also affected tobacco pricing. To evaluate this, we compared changes in prices for the cheapest cigarettes in cities with comprehensive (i.e., restrictions apply to all retailers, products, and flavors except for tobacco flavor) and partial (i.e., exceptions for buffer zones, adult-only tobacco retailers, menthol-flavored products) flavor policies versus cities without any flavor policies as described in previous research (Holmes et al., 2022; Pravosud et al., 2022). Findings of our study will provide insights for future research on tobacco pricing in localities that have enacted similar flavored sales restrictions. Tobacco companies continue circumventing flavor policies (Hiscock et al., 2020) by introducing new products, which are replacements for prohibited items, as has been seen with marketing of new non-menthol cigarettes that taste like menthol (likely due to unrestricted synthetic cooling agents) to bypass comprehensive flavor policies enacted in U.S. jurisdictions and states like California (“California Health and Safety Code: Sec. 104559.5” 2022; Jewett & Baumgaertner, 2023). Most tobacco products, especially combustible cigarettes, are price elastic: the demand responds to price changes of the same product—own-price elasticities, or of other related products—cross-price elasticities (Cotti et al., 2022; Huang et al., 2018; Jawad et al., 2018; U.S. National Cancer Institute and World Health Organization, 2016; Zheng et al., 2017). Although evidence is scarce, positive cross-price elasticities of eliminated flavored products and their alternatives (Denlinger-Apte et al., 2021) suggest that flavored sales restrictions may have indirect effects on pricing of tobacco products that remain on the market.

Methods

Study Design and Sample

There were two waves (2015 and 2019/20) of retail data collection, described in more detail elsewhere (Holmes et al., 2022), which were part of a larger study including a household panel survey of young adults in SF and Alameda Counties (Holmes et al., 2019). Tobacco retail samples were drawn from address lists of all licensed tobacco retailers in the two counties, provided by the Alameda and San Francisco County Departments of Health. The sampling method differed slightly between 2015 and 2019 due to changes in the nature of the parent research project. For the 2015 data, we sampled 100% of retailers located in the Census block groups where randomly selected participants in our corresponding household survey resided and added a five percent random sample of remaining retailers in the two counties (N = 253 of 2,353, or 11% of stores). For the 2019/20 data, we revisited stores that had been audited in 2015 and were still in business, and we further drew a proportional random sample by city of remaining tobacco retailers. We ultimately audited 369 tobacco retail stores in this wave (of 1707 total or 22%). The study’s analytic sample size is described below.

Measures

Outcome

The outcome variables were tobacco prices for the following four tobacco products: (1) the cheapest cigarette pack in the store (N = 463), (2) a pack of Newport menthol cigarettes (N = 287), (3) a Blu brand menthol e-cigarette (N = 96), and (4) the cheapest vape pen in the store (N = 58). Not all retailers sold every tobacco product; therefore, we used available case analysis and the number of retail stores varied by tobacco product. One outlier with an extremely low recorded price of Newport cigarettes (0.70/pack) was excluded from the analysis for this product. To compare final retail prices, we adjusted all prices for state excise and local sales taxes. For comparison purposes, 2015 prices were also adjusted for inflation (2015 final price^X$1.08). We also described the number of reported brands for the cheapest cigarettes in the store by year of data collection and the most frequently reported brand(s) by county.

Socio-Demographic Characteristics

Studies have shown associations between greater tobacco retail density (Kong et al., 2021; Mills et al., 2022) and lower tobacco prices (Lee et al., 2015), particularly in neighborhoods with larger proportions of Black, Hispanic/Latino, and lower-income residents. Thus, to control for the socio-demographic composition of cities, we linked retail data to 2015–2019 American Community Survey (U.S. Census Bureau, 2021a) estimates at the place level to obtain 5-year period estimates by city for median income (in U.S. dollars) and proportions (multiplied by 100%) of non-Hispanic Black and Latino populations.

Statistical Analyses

We assessed both average and median price values to account for non-normality in the distribution of prices of the four tobacco products (p < 0.05 for Kolmogorov–Smirnov and Shapiro–Wilk tests). SAS software, version 9.4 (SAS Institute; Cary, NC, USA) was used to conduct all analyses. We deemed results with two-sided p-values <0.05 to be statistically significant.

Stratification by County

We compared 2015 vs 2019/20 median prices using unadjusted quantile regression models (PROC QUANREG), stratified by SF and Alameda Counties. Few stores (N = 16) sold vape pens in 2019/20, thus, this product was not included in further adjusted analyses. For the other three products (i.e., cheapest cigarette pack, Newport menthol pack, and Blu brand menthol e-cigarettes) with larger sample sizes and approximation to normality in data based on the distribution shapes, we assessed increases over time in the average prices for these tobacco products, stratified by county and accounting for socio-demographic composition of cities.

Because SF is a consolidated city-county being composed of only one city, we carried out an unadjusted, single-level linear regression (PROC GLM) to assess the increase in SF average prices for tobacco products from 2015 to 2019/20. In contrast, to determine change over time in the average tobacco prices for 16 (out of 20) Alameda County cities, we used multivariable multi-level linear regression models, adjusted for each city’s demographic profile, with city at level 2 to account for possible clustering (PROC MIXED with empirical “sandwich” estimator for robust standard errors).

In the models for the cheapest cigarette pack and the Blu brand menthol e-cigarettes, our analysis showed no variation between Alameda County cities (ICC: interclass correlation coefficient = 0), thus, we report results from adjusted, single-level linear regression models for these two outcomes. We observed a potential collinearity between the percentage of Black population and median income variables (proportions of variation >0.92 and VIF: variance inflation factor >12, condition index of <29), but no significant interaction between these two factors. Therefore, we included all three socio-demographic variables in the adjusted models: percentage of Latino (1) and non-Hispanic, Black (2) population concentrations, as well as the median income (3). In the stratified analysis, we categorized the continuous income variable by tertiles due to skewness in the distribution.

Price Comparison by Flavored Tobacco Sales Restrictions

We categorized retailers into three policy groups based on the comprehensiveness of restrictions for sales of flavored tobacco products effective before we terminated data collection on 9 January 2020 (Holmes et al., 2022): 241 retailers were located in five cities with comprehensive flavor policies; 172 retailers in four Alameda County cities with partial flavor policies; and 50 retailers in eight Alameda County cities without restrictions on sales of flavored tobacco products.

Using two-way ANOVA (PROC GLM with LSMEANS statement and Tukey–Kramer adjustment for multiple comparisons) with the interaction term for the year and location by policy group (pre- and post-policy analysis), we estimated inflation-adjusted average prices for the cheapest cigarettes in 2015 vs 2019/20 for the three comparison groups and assessed differences in pre- to post-policy changes in cigarette prices for retailers in cities with comprehensive and partial flavor policies relative to cities without flavor sales restrictions. We report unadjusted and adjusted for socio-demographic variables (i.e., median income, and percentages of non-Hispanic Black and Latino populations) results.

Results

Median Price Change by Tobacco Products in SF and Alameda Counties

Table 1 shows median prices for the four tobacco products stratified by year and county. In 2015, there were no significant differences in the median prices for the cheapest cigarettes between SF and Alameda County. However, in 2019/20, median prices for the cheapest cigarettes were substantially higher in SF compared to other cities in Alameda County ($10.4 vs. $7.6, p < 0.0001). Based on available data, among 31 reported brands for the cheapest cigarettes in 2015, Fortuna was the most frequently observed in both SF (44.3%) and Alameda County (22.7%). Among 25 reported brands of the cheapest cigarettes in 2019/20, Fortuna was still the most common brand among SF retailers (35.1%), whereas in Alameda County, the Maverick (19.9%) and Pall Mall (19.1%) brands were more frequently reported as cheapest.

Table 1.

Median price change for tobacco products in San Francisco and Alameda County: 2014–2020 San Francisco Bay Area young adult health study retail data.

	San Francisco				Difference in medians	Alameda County				Difference in medians
Tobacco Product	2015		2019/20		2019/20 vs 2015	2015		2019/20		2019/20 vs 2015
	N	Median IQR	N	Median IQR	Estimate (SE) (95% CI) a	N	Median IQR	N	Median IQR	Estimate (SE) 95% CIa
Cheapest cigs	97	5.75 5.13, 7.02	105	10.36 8.50, 12.52	4.61 (0.43)*** 3.78–5.45	119	5.48 5.10, 6.07	142	7.58 6.98, 9.00	2.10 (0.14)*** 1.82–2.39
Newport menthol cigs	92	8.12 7.55, 8.80	0	–	–	113	7.12 6.87, 7.61	82	9.28 8.78, 10.30	2.16 (0.19)*** 1.79–2.53
Blu menthol e-cigs	26	11.33 10.79, 11.88	0	–	–	53	10.79 10.79, 11.87	17	12.48 11.98, 13.98	1.69 (0.53)** 0.63–2.75
Cheapest vape pen	18	17.27 10.79, 21.60	9	31.38 19.98, 40.00	13.03 (7.90) −3.24–29.30	25	19.44 10.79, 21.60	6	15.70b 15.00, 31.38–16.57–9.09	−3.74 (6.27)b

Notes: Data shown are summary statistics (mean, std, min, max, median, IQR: 25^th and 75^th percentiles) for tobacco products in U.S. dollars. Prices for 2015 were adjusted for inflation.

^***p < 0.001

^**p < 0.01

^*p < 0.05.

^aEstimates obtained from unadjusted quantile regression models using PROC QUANREG in SAS 9.4.

^bBecause of even number of observations (n = 6), the regression model used the 3rd rank value of 2019 prices as the reference value and the difference in medians over time shown is between $19.44 and $15.70. If defining median as the value that is between the 3rd and 4th rank price values, then 2019 median price = $17.84.

In 2019, median prices for Newport menthol cigarettes and Blu menthol e-cigarettes increased significantly as well (by +$2.16 and +$1.69, respectively), and these products were sold only by retailers in Alameda County. None of the audited stores in SF sold menthol cigarettes or e-cigarettes in 2019 due to comprehensive flavor policies enacted in 2018 (Holmes et al., 2022).

The number of stores selling vape pens reduced from 43 in 2015 to 15 in 2019/20, with data on the cheapest vape pen in 2019/20 limited to stores mostly in SF (n = 9) and only four cities in Alameda County, including Hayward and Oakland (two in each) as well as Castro Valley and Newark (one in each). The increase in median prices of the cheapest vape pen in both counties was statistically insignificant.

Average Price Variability by County and Socio-Demographic Characteristics

Table 2 shows significant increases in average prices of the cheapest cigarette pack from 2015 to 2019/20 by +$4.61 in SF and by +$2.53 in Alameda County, adjusted for socio-demographic characteristics. In Alameda County, adjusted average prices for Newport menthol cigarettes and Blu menthol e-cigarettes increased significantly by $2.88 and $2.82, respectively. Our analysis for Alameda County revealed no significant associations of socio-demographic variables with increases from 2015 to 2019/20 in average prices for the three tobacco products assessed.

Table 2.

Results from single-level unadjusted and multi-level adjusted linear regressions for increase from 2015 to 2019/20 average prices of tobacco products.

	San Francisco	Alameda Countya
	Cheapest Cigs	Cheapest Cigs	Newport Menthol cigsb	Blu Menthol e-Cigs
Variable	n = 202	n = 261	n = 195 (ICC = 0.15)	n = 70
Increase in 2019/20 compared to 2015 Race/Ethnicity	4.617 (0.288)***	2.532 (0.171)***	2.881 (0.560)***	2.822 (0.751)***
% of Black, non-Hispanic population	—	−0.069 (0.059)	−0.003 (0.052)	−0.056 (0.163)
% of Hispanic population	—	−0.008 (0.012)	−0.003 (0.012)	−0.003 (0.039)
Median household income (in U.S. dollars)	—
1st tertile (n = 128, $73,692)	—	1.944 (1.148)	0.609 (1.051)	3.269 (3.255)
2nd tertile (n = 68, $78,003–95,400)	—	0.483 (0.416)	0.240 (0.361)	0.660 (1.310)
3rd tertile (n = 99, $97,286–156,400)	—	Ref	Ref	Ref

Notes: Results shown are unadjusted (for San Francisco) and adjusted (for Alameda County) b-coefficients and their corresponding standard errors.

^***p < 0.001

^**p < 0.01

^*p < 0.05.

^aModels assessed increase in the average prices from 2015 to 2019/20 adjusting for percentages of non-Hispanic, Black and Latino populations, and median household income for cities of Alameda County.

^bA multi-level model (using PROC MIXED with empirical “sandwich” estimator for robust standard errors) included random effect for city to account for possible clustering by cities in Alameda County (i.e., Alameda, Albany, Berkeley, Castro Valley, Dublin, Emeryville, Fremont, Hayward, Livermore, Newark, Oakland, Pleasanton, San Leandro, San Lorenzo, Sunol, and Union City).

Change in Cheapest Cigarette Pack Average Prices by Flavored Tobacco Sales Restrictions

Table 3 shows that 2015 average prices were similar before the enactment of flavor sales restrictions, and in 2019/20, average prices in cities with comprehensive and partial flavor policies were significantly higher compared to cities without flavor policies ($10.21 and $8.47 vs $7.21, p < 0.0001 and p = 0.012, respectively). We observed significant increases from 2015 to 2019/20 in average prices per the cheapest cigarette pack in cities with comprehensive and partial restrictions on sales of flavored tobacco, compared to cities without flavor policies (respectively, by +$2.94 and +$1.81, p < 0.0001 and p = 0.002, respectively). Adjustments for socio-demographic variables provided similar results regarding differential increments in pricing by flavored tobacco sales restrictions.

Table 3.

Increase in the average price per the cheapest cigarette pack from 2015 to 2019/20 by flavored tobacco sales restriction policy: 2014–2020 San Francisco Bay Area young adult health study retail data.

Policy Category for retailers	2015 Mean (SD)	2019/20 Mean (SD)a	Increase from 2015 to 2019/20	Estimate (SE)b 95% CI	Adjusted Estimate (SE)b,c 95% CI
Comprehensive (n = 241)	6.03 (1.36)	10.21 (2.59)	+4.18***	2.94 (0.56)*** 1.84–4.03	3.23 (0.56)*** 2.13–4.32
Partial (n = 143)	5.41 (0.85)	8.47 (1.81)	+3.06***	1.81 (0.58)** 0.68–2.95	2.11 (0.58)*** 0.97–3.25
No restrictions (n = 79)	5.96 (1.05)	7.21 (0.85)	+1.25	Ref	Ref

Notes: Cities with comprehensive policies effective as of 9 January 2020, include Alameda, Albany, Fremont, Livermore, and San Francisco. Cities with partial policies include Berkeley, Hayward, Oakland, and San Leandro. Cities with no policies include Dublin, Newark, Emeryville, Castro Valley, Pleasanton, San Lorenzo, Sunol, and Union City.

All 2015 prices were adjusted for inflation.

^***p < 0.001

^**p < 0.01

^*p < 0.05.

^aIndicates significantly different prices among all policy groups during 2019/20 retail audits.

^bResults shown are estimates for b-coefficients for the interaction term of the two variables indicating a three-level policy group and a two-level year predictor. Results obtained from a two-way ANOVA (PROC GLM) with adjustment for multiple comparisons using Tukey–Kramer test.

^cEstimates adjusted for the percentages of non-Hispanic, Black and Latino populations, and median household income (in U.S. dollars).

Discussion

Our study of retail stores in SF and Alameda Counties revealed significant increases in tobacco prices following the implementation of Proposition 56 and other local tobacco ordinances between 2015–2019/20. The increase in the average prices for the cheapest cigarettes in both SF and Alameda Counties exceeded the $2 increase in the statewide tax. We also found complete elimination of menthol cigarette and e-cigarette products in SF and significant increases in prices for Newport menthol cigarettes and Blu brand menthol e-cigarettes in Alameda County cities that exceeded the raise in the statewide tax of $2 dollars for cigarettes and the expected $1.25 for Blu menthol e-cigarettes with 1 mL of fluid (Pesko, Courtemanche, and Catherine Maclean 2020). These results might indicate over-shifting of tobacco taxes to consumers, consistent with the study by Henriksen et al. (2019) which found over-shifting of the $2 tax for four popular cigarette brands, including Marlboro, Natural American Spirit, Newport, and Pall Mall, based on data from a random sample of licensed tobacco retailers in California.

Other studies of price changes following the tax increase in California found significant, although less than $2 increases in cigarette prices (i.e., by $1.84 (Gunadi et al., 2021) and by $1.89 (Boettiger & White, 2020)), indicating under-shifting the tax. Together with these findings, our study, which had a longer follow-up to observe the effect of the tax policy over time, can indicate a common tobacco industry strategy of absorbing a part of the tax increase through reduced profits initially, but gradually moving to price increases that would eventually exceed the increase in taxation (Ross et al., 2012). On the other hand, these variations in the magnitude of price change and the direction of tax shifting may be related to differences in the definition of the outcome variable (e.g., price paid for the last pack of cigarettes in Gunadi et al. (2021) vs. a focus on popular brands, rather than discount brands, in the study by Henriksen et al. (2019), and the cheapest cigarette pack in our study) and/or the nature of data sources. Similar to Henriksen et al. (2019), we obtained information on tobacco prices through direct retail audits and found over-shifting, whereas other studies that found under-shifting used self-reported cigarette pricing data from a population survey (Gunadi et al., 2021) or calculated cigarette prices as a retail revenue-to-sales ratio (Boettiger & White, 2020). Retail audits eliminate consumer recall bias; consumers’ self-reported price might also include use of price promotions that undermine (Apollonio & Glantz, 2020; Pierce et al., 2005) the effects of tax increases. Indeed, Henriksen et al. (2019) reported that many tobacco retailers were more likely to offer and advertise discounts following the increase in California tobacco excise tax and revealed Proposition 56 under-shifting in Black neighborhoods.

The use of promotional strategies or differential shifting is not unique for tobacco companies. Price promotions have been used to cover up tax increases that could modify purchasing behavior of gasoline (Black, 2015), and the relationship of policy and price shifting have also been raised in studies of food, alcohol, and sugar-sweetened beverages (Falbe et al., 2020; Nelson & Moran, 2020; Rojas & Wang, 2021). The study of the 2010 Washington excise tax on carbonated drinks (i.e., “soda tax”) revealed rapid price increases of most products assessed and an overall tax over-shifting (Rojas & Wang, 2021). The 2015 Berkeley taxation on sugar-sweetened beverages (Falbe et al., 2016) indicated mixed results showing either nonsignificant price increases, or very low pass-through rates, or a significant tax under-shifting to consumer prices for selected products (Rojas & Wang, 2021). After the 2016 Oakland and San Francisco sugary beverage tax, not only the prices for sugar-sweetened beverages have been raised by almost the same amount of the tax increase (“fully-shifted”), but the tax increases might have indirectly contributed to price increases for some unaffected products, such as unsweetened carbonated drinks, or “diet sodas” (Falbe et al., 2020). As for the alcohol industry, longitudinal data from 27 countries have suggested that excise taxes tend to be over-shifted, especially for more expensive products, and can be “fully” or under-shifted to lower-priced products (Shang et al., 2020). These findings indicate relevance of our research to studies of other industries outside tobacco and show that our results can be informative for a broader policy assessment.

Our study adds to prior research by providing a more detailed analysis of specific Bay Area Counties with a variety of tobacco control regulations—a unique context for examining local policy and pricing. We found the increase in price for the cheapest cigarette pack was notably higher in SF compared to Alameda County. In addition to over-shifting the tax, this might be attributable to the numerous other tobacco ordinances implemented in SF (San Francisco Department of Public Health, 2018). Particularly relevant to pricing, unlike Alameda County cities, SF has a Cigarette Litter AbatementQuarterly Paper Return, 2021 which increased from $0.20 per cigarette pack in 2015 to $0.85 in 2019 (“Cigarette Litter AbatementQuarterly Paper Return, 2021: Quarterly Paper Return 2019″), and to $1.00 in 2020 (“Cigarette Litter Abatement Fee, 2019: Quarterly Paper Return 2020”; “Imposition of Cigarette Litter Abatement Fee, Chapter 105: Sec. 105.3.”).

Another novel finding of our study is that average prices for the cheapest cigarettes in cities with either comprehensive or partial flavor policies increased significantly and were higher than in cities without restrictions on sales of flavored tobacco products. This is consistent with findings of Asare et al. (2022), whose study reported higher cigarette prices in Massachusetts, which had enacted a comprehensive menthol flavor ban, compared to states without restrictions on sales of flavored tobacco; unlike our study, Asare et al. (2022) did not assess differential increases between the two policy groups. In 2018, SF passed comprehensive flavored tobacco sales restrictions (“San Francisco Health Code, 2021”) that were enforced in 2019 with extensive educational efforts by the Flavored tobacco sales ban: Frequently Asked Questions, 2019. Consistent with prior research finding high compliance with this policy (Vyas et al., 2021), we found in 2019 none of the SF retailers sold Newport menthol cigarettes or Blu menthol e-cigarettes. It is possible that complete elimination of flavored tobacco, including menthol, could have affected SF retail prices for unflavored tobacco that remained on the market.

After the policy implementation in SF, some users of flavored tobacco sought other ways to obtain these products either out of SF or from online vendors, or switched to available alternatives (Yang et al., 2020). Given significant own-price elasticities of cigarettes (Cotti et al., 2022; Huang et al., 2018; Jawad et al., 2018; U.S. National Cancer Institute and World Health Organization, 2016; Zheng et al., 2017) and positive cross-price elasticities of flavored and unflavored tobacco or other alternatives (Cotti et al., 2022; Denlinger-Apte et al., 2021), the demand for alternative products and economic substitutes for prohibited flavored tobacco should have grown, which may have led to increased prices. This aligns with our findings showing significant price increases for the cheapest cigarettes that were driven largely by cities that had enacted flavor sales restrictions, especially by jurisdictions with comprehensive policies like in SF. Taken together, the large increases in cigarette prices in SF might reflect not only over-shifting but also other efforts to maintain profitability in response to increases in Cigarette Litter Abatement fee and comprehensive flavor tobacco sales restrictions.

Alameda County includes several cities, and so did not implement flavor policies evenly across all local jurisdictions (Holmes et al., 2022). When we terminated the data collection, only Alameda, Albany, Fremont, and Livermore had enacted comprehensive flavor policies; Berkeley, Hayward, Oakland, and San Leandro implemented partial flavor policies (i.e., exceptions for tobacco stores, buffer zones, or menthol flavor); other cities did not restrict sales of flavored tobacco. In addition to potential impact on tobacco pricing, equitable implementation of comprehensive policies to restrict sales of flavored tobacco in all nearby lower-price localities (Kurti et al., 2020; Schroth, Kurti, and Delnevo 2022) of the SF Bay area is an important step for addressing tobacco disparities among racial-ethnic groups. Latino and non-Hispanic Black young adults in SF Bay area are more likely to report menthol use compared to non-Hispanic Whites (Holmes et al., 2019), and in Alameda County stores menthol products increased in price, but remained available, compared to San Francisco where there was a comprehensive policy, and we found no menthol cigarette and e-cigarette product availability in our sample.

Despite substantial declines of overall smoking prevalence among U.S. adults in the past two decades (Jamal et al., 2018), racial-ethnic minority groups are still disproportionately affected by smoking and report higher rates of flavored tobacco use (Delnevo et al., 2020; Gardiner, 2004; Watkins et al., 2022). Sales restrictions that eliminate flavored tobacco, including menthol, and may indirectly lead to price increases for unflavored products can play an influential role in reducing the initiation of tobacco use among young people and encourage tobacco use cessation, particularly for people of color (Delnevo et al., 2020; Villanti et al., 2022; Yerger, 2022). African American communities have advocated for prohibition of mentholated tobacco products for reasons of equity as the disproportionate toll of menthol products among this group is a social injustice (Mills et al., 2021; Yerger, 2022). Other racial-ethnic groups with high prevalence of menthol tobacco consumption (Mukherjea et al., 2014; Watkins et al., 2022) can also benefit from flavored sales restrictions, and arguments about increased criminalization for possession or use of menthol tobacco products are unfounded (Yerger, 2022). However, to minimize any unintended consequences that can contribute to tobacco-related racial-ethnic inequities (Breland et al., 2022) and to increase the reach to all subgroups, flavor sales restrictions should be comprehensive (e.g., applied to all products, flavors, and point of sale) and paired with strong enforcement and equitable implementation (Rose et al. 2020, 2022). In our study, we accounted for a potential impact that race-ethnicity could have had on pricing in the San Francisco and Alameda counties that feature substantial variation in racial-ethnic minority groups (U.S. Census Bureau, 2021). Our adjusted analyses revealed no differences in tobacco price changes over time for localities with greater proportions of Black and Latino populations, likely due to the use of aggregated city-level demographic data. The nonsignificant findings prevent us from making definite conclusions but warrant further research using more granular neighborhood-level data to examine tobacco-related racial disparities and evaluate the impact of local tobacco control policies on racial equity.

Limitations

First, we used two cross-sectional samples, and the sample size of stores that were audited both times (N = 29) was insufficient to assess price change over time longitudinally. Second, we had only two points of data collection (2015 and 2019/20), which may not fully reflect the effect of implementation of tobacco policies on price changes. After the second wave of data collection as of June 2021 (Holmes et al., 2022), six additional cities in Alameda County enacted comprehensive local flavor ordinances and two cities (Berkeley and Oakland), eliminated exceptions in their policies, and terminated sales of all flavored tobacco products, including menthol (Holmes et al., 2022). Third, differences in the brand of cheapest cigarettes and variations in the products reported as the “cheapest vape pens” may have affected the reported prices of these products. However, to avoid overfitting of regression models, we did not control for the numerous product brands. In addition, our measures of e-cigarette pricing do not reflect the major changes in the e-cigarette market during the study (King et al., 2018) with JUUL taking over 75% of the market in 2017–2018 (Huang et al., 2019; Nielsen Total US xAOC/Convenience Database & Wells Fargo Securities, LLC, 2018). The price specifically of the JUUL brand was not measured in 2015 retail audits; hence, the study could not compare JUUL pricing in 2019. In 2019, “vape pen” devices had been largely replaced by JUUL and other “pod vapes” decreasing the relevance of this measure (Ali et al., 2020). Fourth, our sample was limited to SF and Alameda Counties and may not generalize to other areas or other studies in California (Boettiger & White, 2020; Gunadi et al., 2021; Henriksen et al., 2019). Fifth, accounting for users’ preferences or sales data of purchased products was beyond the scope of our study, and thus, we could not parse out whether changes in prices were attributed solely to tax over-shifting or were related to changes in demand. Finally, data on wholesale prices were not available and we could not assess whether differential tax shifting varied by retailers in the sample.

Despite these limitations, this is the first study to examine changes in tobacco prices in SF Bay Area Cities before and after state and local policy implementation, to compare cigarette price increases in two counties by different flavor policies, and to include non-cigarette tobacco products such as e-cigarettes. Although the current study has rather descriptive pre- to post-policy analysis design, our data suggest that the implementation of flavor policies could have indirectly impacted tobacco prices in the two counties in addition to the $2 California excise tax increase, and that increases in Cigarette Litter Abatement fee and implementation of comprehensive flavor policies in SF may have had synergistic, positive spillover effects (Venkataramani et al., 2017), and contributed to increased prices far exceeding that expected from the statewide tax.

Conclusion

The present study of a descriptive pre- to post-policy analysis design provides evidence of significant increases in cigarette prices in SF Bay Area counties after the implementation of California’s Proposition 56 and local price- and non-price-related policies such as restrictions on sales of flavored tobacco. We found a higher increase in cigarette prices in SF versus Alameda County, and complete elimination of menthol (e)cigarettes among SF retailers. Average prices for the cheapest cigarette pack increased significantly more for cities with comprehensive and partial restrictions on sales of flavored tobacco compared to cities without such restrictions. Local tobacco control policies may augment the impact of state taxes on prices.