Principal investigator: Bruno Benevit
Original title: Cigarette taxes and smoking among sexual minority adults
Authors: Christopher S. Carpenter and Dario Sansone
Location of the Intervention: United States
Sample Size: 3,3 million adults
Sector: Public Sector Economics
Primary Variable of Interest: Cigarette consumption
Type of Intervention: Increase in imposts
Methodology: DID, DDD
Summary
Taxing cigarettes has historically been used as a tool to encourage a reduction in their consumption among the population. However, little attention has been paid to the impacts of this policy on the LGBTQ+ population. In this sense, this study conducted a quasi-experiment using data from 1996 to 2018 to evaluate the effect of applying taxes on cigarette consumption in the United States on same-sex couples. The results revealed that the impact of taxation on men in same-sex couples was considerably greater compared to that observed for men in heterosexual couples. No significant effects of replacing cigarettes with other risky behaviors were found, nor were any impacts identified on more recent generations.
- Policy Problem
Cigarette consumption is a major vector of chronic diseases for the general population, causing consequences for both consumers and the health systems of countries. Therefore, the application of taxes on cigarettes is commonly adopted as a way to discourage smoking among the population. In general, the economic literature corroborates the effectiveness of this public policy strategy (DECICCA; KENKEL; LOVENHEIM, 2022), although the magnitude of this effect varies among different studies (CARPENTER; SANSONE, 2021).
The analysis of social strata also becomes relevant, taking into account the specificities associated with each reality of different demographic subgroups in society. In addition to young people, a subgroup of particular interest in understanding smoking since most individuals begin smoking before the age of 18, researchers have also analyzed the impacts of this type of policy on the behavior of older adults, pregnant women, and different racial and ethnic minorities (CARPENTER; SANSONE, 2021). In this context, little attention has been given to the effect of increased cigarette taxation on the LGBTQ+ minority population.
Sexual orientation can influence the effectiveness of cigarette taxation due to several factors. Sexual minorities, such as gay men, lesbian women, and bisexual individuals, have significantly higher smoking rates, often linked to "minority stress" caused by stigmatization and discrimination (CARPENTER; SANSONE, 2021). This can lead these populations to use cigarettes as a coping mechanism, making taxation less effective in reducing consumption. Furthermore, people in these communities have less access to health insurance and smoking cessation services, hindering the adoption of health-related policies. Factors such as lower income among gay men or higher educational levels among lesbian women can also influence how these populations respond to tobacco tax policies, requiring greater attention from policymakers to this population.
- Policy Implementation Context
Taxation on cigarettes is applied with the goal of reducing tobacco consumption by increasing the cost for consumers. As the price of cigarettes rises, the idea is that individuals will reduce or quit smoking due to the greater financial impact. This type of policy is part of a set of tobacco control measures aimed at minimizing harm to public health. The effectiveness of this strategy depends on the price elasticity of demand, that is, how much consumption varies in response to price increases. However, its effectiveness can vary depending on socioeconomic and demographic factors, being directly influenced by dynamics related to the price elasticity of demand and income.
In the United States, cigarette taxes are levied at the federal, state, and municipal levels. Consistent with most previous literature, and given that publicly available Behavioral Risk Factor Surveillance System (BRFSS) data do not contain detailed geographic information below the state level throughout the sample period, this analysis focuses on state taxes on cigarette consumption. Furthermore, tobacco tax increases have been implemented in nearly every state: during the period considered in our main empirical analysis, only two states—Missouri and North Dakota—did not increase their tobacco tax. The other forty-eight states, plus the District of Columbia, have passed 160 cigarette tax changes since 1996. Most states have increased their taxes more than once.
According to data from the United States Centers for Disease Control and Prevention (CDC, 2018), gay men, lesbian women, and bisexual women have significantly higher smoking rates (20,3%) than heterosexual adults (13,7%). The difference in smoking rates among sexual minorities, at 6,6 percentage points (pp), exceeds that observed between men and women (3,6 pp), between young people aged 18 to 24 and adults over 65 (2,2 pp), between white and black adults (0,3 pp), between the regions of the United States with the highest (Midwest) and lowest (West) smoking rates (5,9 pp), and between single and married adults (2 pp).
- Evaluation Details
The database used in the study was the BRFSS, a nationally representative health survey conducted by the U.S. CDC. The BRFSS collected information through annual telephone interviews with over 400.000 non-institutionalized adults, aged 18 and older. The interviews addressed health risk behaviors, chronic conditions, and the use of preventive services. In addition, the BRFSS included demographic data such as age, ethnicity, education, and marital status. Starting in 2011, the BRFSS began including calls to cell phones, which increased the representativeness of the sample. Methodological changes in the calculation of sample weights were also implemented to ensure the accuracy of the results. The study used BRFSS data from 1996 to 2018 to assess the impact of taxation on cigarette consumption among same-sex and heterosexual couples.
The identification of same-sex couples in the BRFSS was carried out indirectly, based on information provided about the number of adult men and women in the household. When there were exactly two adult men and no women, the individuals were classified as belonging to a male same-sex couple. Similarly, female same-sex couples were identified when the household had exactly two women and no adult men. Heterosexual couples, on the other hand, were defined by the presence of one adult man and one adult woman in the household. This approach allowed for the identification of same-sex couples without requiring respondents to explicitly declare themselves as belonging to a sexual minority, which helped to avoid problems of selective disclosure.
Although an indirect technique, the identification of same-sex couples by the BRFSS was validated by comparisons with the optional modules on sexual orientation and gender identity (SOGI), introduced in 2014 in some states. These modules confirmed that men and women in same-sex households were more likely to identify as non-heterosexual. On average, 11% of women and 28% of men in same-sex households identified as belonging to a sexual minority, compared to only 1% in heterosexual households. Furthermore, men in same-sex households reported a higher frequency of HIV testing and condom use for disease prevention, corroborating the reliability of BRFSS data in research on health behaviors and sexual orientation.
- Method
The method used to estimate the impact of increased cigarette taxes on tobacco consumption was based on a technique called difference-in-differences (DID). This approach compares the variation in smoking behavior between two groups: one affected by the tax increase and the other not. The central idea is to verify how cigarette consumption changed in states after the tax was applied compared to states that did not experience this change, estimating differences between individuals and states before and after the increase. The constructed model controlled for fixed factors for each state and period, such as demographic characteristics of individuals (age, race, ethnicity, and education) and variations over time in states' population and employment rates. Additionally, the impacts were also verified for different marital structures and time horizons (1993–2018, 1996–2010, and 2011–2018), testing the hypothesis that younger generations of smokers are less sensitive to price increases.
Furthermore, the study also presents a triple-difference model (DDD). Similar to the DID method, the DDD method considers the relationship of differences between distinct groups of observations within the sample, adding a new group to identify causal effects through differences after the intervention. Therefore, the DDD model incorporated the difference between individuals in same-sex couples compared to different-sex couples. This approach allows us to identify whether cigarette taxes had a greater differential effect on individuals in same-sex families.
Finally, to verify the hypothesis that the greater risk-taking behavior among homosexual individuals stems from historical discrimination suffered by this population, the study also analyzed the presence of policy effects on a possible substitution mechanism between smoking and other risky behaviors. Thus, the effect of increased cigarette taxes on alcohol consumption, body weight, physical exercise, and HIV or AIDS testing was estimated.
- Main results
The results indicated that a one-dollar increase in cigarette taxes reduced the likelihood of daily smoking among individuals in same-sex households. For women, this reduction was 0,6 percentage points, but the effect was not statistically significant across all models. For men, the impact was more pronounced, with a 1,8 percentage point reduction in the likelihood of being a daily smoker, remaining statistically significant. Furthermore, the tax increase also decreased the likelihood of current smoking, especially among men. These results demonstrated that cigarette taxes were effective in reducing consumption, particularly among men in same-sex couples.
Regarding the effects considering different time horizons, the robustness tests performed confirmed that the impact of cigarette taxes was consistent for men. For men, the effects were strongest in the period from 1996 to 2010, indicating that the impact of taxes may have diminished in more recent years. When the analysis was restricted to individuals who were never married or who lived with unmarried partners, an even more pronounced effect was observed in men.
The results indicated that increasing cigarette taxes reduced daily smoking among both women and men in heterosexual households. The reduction was 0,6 percentage points for women and 0,4 percentage points for men, replicating previous findings on the impact of taxes on cigarette consumption in the general population. However, for men in same-sex households, the impact was more pronounced. The one-dollar tax increase was 0,9 percentage points more effective in reducing smoking among these men, suggesting that, without the tax increase, the difference in smoking rates between heterosexual and same-sex men would have been greater.
Regarding the effects on other risky behaviors, the increases in cigarette taxes did not generate significant impacts on alcohol consumption, exercise practices, body weight, or HIV testing among men in same-sex households. This indicated that the reduction in smoking was not replaced by other harmful behaviors. However, the increase in cigarette taxes significantly improved the self-reported health of these men. The likelihood of reporting excellent or very good health increased, while the chance of reporting fair or poor health decreased, suggesting an overall improvement in well-being.
- Lessons in Public Policy
In this article, the authors investigated how increased taxes on cigarettes influenced cigarette consumption among different groups of individuals, especially in relation to men and women in same-sex households. The results showed that the tax increase led to a significant reduction in daily cigarette consumption among men in same-sex households, with a more pronounced effect compared to men in different-sex households. For women in same-sex households, the reduction was less significant.
These findings indicate that increased cigarette taxes had a stronger impact on consumption among men in same-sex households, suggesting that taxation policies can be adjusted to better meet the specific needs of different groups. For policymakers, these results highlight the importance of considering variations in consumption behavior among different subgroups when designing tobacco control strategies. Implementing policies that take these differences into account could potentially increase the effectiveness of tobacco consumption reduction measures.
References
CARPENTER, CS; SANSONE, D. Cigarette taxes and smoking among sexual minority adults. Journal of Health Economics, 2021.
CDC, 2018, 2017. Tobacco Product Use Among Adults: United States. Morb. Mortal. Wkly. Rep. 67, 1-8
DECICCA, P.; KENKEL, D.; LOVENHEIM, MF The Economics of Tobacco Regulation: A Comprehensive Review. Journal of Economic Literature, v. 60, no. 3, p. 883–970, 2022.
