Tax Avoidance and Evasion: Cigarette Purchases From Indian Reservations Among US Adult Smokers, 2010-2011

Abstract

Excise taxes are the primary public health strategy used to increase the price of cigarettes in the United States. Rather than quitting or reducing consumption of cigarettes, some price-sensitive smokers may avoid state and local excise taxes by purchasing cigarettes from Indian reservations. The objectives of this study were to (1) provide the most recent state-specific prevalence of purchases made on Indian reservations by non–American Indians/Alaska Natives (non-AI/ANs) and (2) assess the impact of these purchases on state tax revenues. We used data from a large national and state-representative survey, the 2010-2011 Tobacco Use Supplement to the Current Population Survey, which collects self-reported measures on cigarette use and purchases. Nationwide, 3.8% of non-AI/AN smokers reported purchasing cigarettes from Indian reservations. However, in Arizona, Nevada, New Mexico, New York, Oklahoma, and Washington State, about 15% to 30% of smokers reported making such purchases, resulting in annual tax revenue losses ranging from $3.5 million (Washington State) to $292 million (New York) during 2010-2011. Strategies to reduce the sale of non- or lower-taxed cigarettes to non-AI/ANs on Indian reservations have the potential to decrease smoking prevalence and recoup lost revenue from purchases made on reservations.

Increasing the price of tobacco products, including cigarettes, is the most effective public health intervention to reduce cigarette consumption, prevent smoking initiation, and promote smoking cessation.^1–7 Excise taxes are the primary public health strategy used to increase the unit price of cigarettes.^1–8 However, rather than quitting or reducing consumption of cigarettes, some price-sensitive smokers in the United States may avoid state and local excise taxes by purchasing their cigarettes from nearby Indian reservations. The federal government recognizes Native American tribes as sovereign nations; thus, tribal members are exempt from state excise tax on cigarettes sold on reservations. Although US Supreme Court rulings conclude that state taxes can be collected on cigarettes sold to nontribal members on reservations,^9–12 in practice, not all states are able to collect such taxes. Such avoidance and evasion may undermine the impact of cigarette unit price increases that are intended to reduce smoking and improve public health.

Most studies on tax avoidance and evasion focus on cross-border purchases or combined purchases from any lower- or nontaxed sources.^13–18 Data on cigarette purchases from Indian reservations are limited. One study reported that between 2003 and 2009, 19% of smokers in New York State always purchased cigarettes on Indian reservations.¹⁹ National estimates of reservation purchases were much lower. For example, based on data from the International Tobacco Control Policy Evaluation Project, the percentage of US respondents who made their most recent cigarette purchase from a reservation ranged from 2.6% in 2002 to 4.4% in 2011.²⁰ Using data from the 2009-2010 National Adult Tobacco Survey, 1 study reported that 7.4% of US smokers purchased cigarettes from Native American sources in the previous year.²¹

This study used data from the 2010-2011 Tobacco Use Supplement to the Current Population Survey (TUS-CPS),²² a national and state-representative survey that collects self-reported measures on cigarette use and purchases. Our objectives were to (1) evaluate the state-specific prevalence of Indian reservation purchases and (2) assess the impact of these purchases on state tax revenue. To our knowledge, this study provides the most recent state-specific estimates on cigarettes purchased by non-AI/ANs on Indian reservations and the associated tax revenue losses.

Methods

Data Source

A total of 229 456 civilian, noninstitutionalized adults aged ≥18 were interviewed in the 2010-2011 TUS-CPS in 3 waves: May 2010, August 2010, and January 2011. The survey included self-respondents and those who responded by proxy. We excluded proxy respondents because they were not asked if they had purchased cigarettes from Indian reservations. Thus, the analytic sample was restricted to current cigarette smokers (n = 27 611) among self-respondents who were eligible for all supplement questions. We defined current cigarette smokers as those who reported having smoked at least 100 cigarettes (lifetime) and who currently smoked “every day” or “some days.” We excluded respondents if they did not report whether their most recent cigarette purchase was made from an Indian reservation (n = 2511), their state of residence (n = 129), or the average daily number of cigarettes smoked (n = 511). The final analytic sample included 24 460 current cigarette smokers.

We performed all analyses in April 2014 using Stata version 13.²³ We weighted all estimates to be nationally and state representative using the appropriate TUS-CPS self-response adjustment weights. When combining data from the 3 waves of TUS-CPS, we divided the self-response adjustment weights by 3, as suggested by the survey documentation.²⁴

Measures

Individual annual cigarette consumption

For each respondent, we calculated monthly cigarette consumption by multiplying the number of days that the respondent smoked in the past 30 days by the number of cigarettes that the respondent smoked on smoking days. To derive annual cigarette consumption, we assumed that respondents’ cigarette consumption was constant throughout the months of a year. We calculated individual annual consumption of cigarette packs by assuming that each pack contained 20 cigarettes.

Cigarette consumption attributable to purchases made on Indian reservations

We estimated the percentage of cigarette consumption from Indian reservations by state by dividing the cigarette consumption purchased from Indian reservations by the annual total cigarette consumption in that state. For each state, we estimated the annual total cigarette consumption by multiplying the total number of smokers in the state by the average individual annual cigarette consumption. We calculated the annual total cigarette consumption among smokers who purchased cigarettes from Indian reservations similarly among those who reported their race/ethnicity as non-American Indian/Alaska Native (non-AI/AN). We tabulated the state cigarette consumption attributable to Indian reservation purchases by quartiles.

Tax revenue loss attributable to Indian reservation purchases

Reservation purchases could be made in the state of residence or in other states. We estimated tax revenue loss attributable to non-AI/AN purchases from in-state reservations as the difference between the tax revenue that would have been collected if the full state cigarette excise tax had been applied to such sales and the portion of tax revenue collected on in-state reservation sales. To estimate the tax revenue collected on in-state reservations, we determined state tax on cigarettes sold on Indian reservations to non-AI/ANs from May 2010 to January 2011 (to be consistent with the time frame of the 2010-2011 TUS-CPS) from state tax statutes and by contacting state departments of revenue (online-only Table 1). We estimated revenue loss to out-of-state reservation purchases as the loss of full state tax. We did not calculate tax revenue losses in states with <10 in-state or out-of-state reservation purchases (n = 32) among non-AI/ANs or when the state cigarette excise tax was assumed to have been collected on such sales (n = 8); thus, we calculated revenue losses for 11 states.

Results

Among 24 460 US cigarette smokers during 2010-2011, approximately 3.8% (95% CI, 3.5%-4.0%) of non-AI/ANs reported purchasing cigarettes from an Indian reservation, ranging from 1.7% (95% CI, 1.1%-2.4%) in Texas to 30.0% (95% CI, 25.8%-34.3%) in Oklahoma, among the 20 states with a sufficient number of purchases made on reservations (Table). In contrast, 4.2% of total cigarette consumption came from Indian reservations. This finding is consistent with our estimate that smokers who purchased cigarettes on reservations tend to have higher consumption than smokers who did not.

Table.

Percentage of non-AI/AN smokers^a who purchased cigarettes from Indian reservations, nationally and by state, United States, 2010-2011^b

State	Non-AI/AN Smokers Who Purchased Cigarettes From Indian Reservations, % (95% CIc)
Alabama	1.8 (0.3 to 3.2)d
Alaska	0.7 (–0.1 to 1.4)d
Arizona	21.0 (15.8 to 26.2)
Arkansas	3.8 (1.7 to 5.9)
California	1.9 (1.0 to 2.7)
Colorado	0.5 (–0.1 to 1.1)d
Connecticut	0.6 (–0.1 to 1.3)d
Delaware	1.0 (–0.2 to 2.1)d
District of Columbia	0.3 (–0.3 to 0.9)d
Florida	3.4 (2.3 to 4.6)
Georgia	1.4 (0.2 to 2.5)d
Hawaii	0.8 (–0.2 to 1.8)d
Idaho	7.2 (4.1 to 0.3)
Illinois	0.1 (–0.1 to 0.2)d
Indiana	1.5 (0.4 to 2.5)d
Iowa	2.4 (1.1 to 3.6)
Kansas	5.1 (3.1 to 7.0)
Kentucky	0.4 (–0.2 to 0.9)d
Louisiana	0.4 (–0.3 to 1.1)d
Maine	0.5 (–0.1 to 1.1)d
Maryland	0.5 (–0.2 to 1.2)d
Massachusetts	1.6 (0.2 to 3.0)d
Michigan	0.7 (0.1 to 1.4)d
Minnesota	2.4 (1.2 to 3.6)
Mississippi	0
Missouri	0.3 (–0.1 to 0.6)d
Montana	3.8 (1.5 to 6.0)
Nebraska	4.9 (2.8 to 7.0)
Nevada	15.0 (11.6 to 18.4)
New Hampshire	1.1 (0.2 to 2.0)d
New Jersey	0.7 (–0.1 to 1.5)d
New Mexico	25.1 (18.3 to 31.9)
New York	16.4 (13.8 to 19.1)
North Carolina	0.9 (0.1 to 1.8)d
North Dakota	5.8 (3.3 to 8.4)
Ohio	0.5 (0.0 to 1.0)d
Oklahoma	30.0 (25.8 to 34.2)
Oregon	0.8 (–0.1 to 1.7)d
Pennsylvania	0.4 (–0.1 to 0.9)d
Rhode Island	2.0 (0.3 to 3.8)d
South Carolina	0.4 (–0.2 to 1.0)d
South Dakota	2.5 (1.2 to 3.7)
Tennessee	0.2 (–0.2 to 0.5)d
Texas	1.7 (1.1 to 2.4)
Utah	1.8 (–0.8 to 4.4)d
Vermont	5.0 (2.8 to 7.3)
Virginia	0.4 (–0.2 to 1.0)d
Washington	18.3 (14.3 to 22.3)
West Virginia	0.2 (–0.2 to 0.7)d
Wisconsin	6.8 (4.7 to 8.9)
Wyoming	4.9 (2.8 to 6.9)
United States	3.8 (3.5 to 4.0)

Abbreviation: AI/AN, American Indian/Alaska Native.

^aExcluded 156 smokers who were AI/AN and made their most recent pack of cigarettes purchases from an Indian reservation.

^bData source: Tobacco Use Supplement to the Current Population Survey, 2010-2011.²²

^cActual 95% CIs were reported, even if the lower value was zero, because of low statistical power due to the small sample size.

^dUnweighted sample size <10.

The highest percentages of annual cigarette consumption reported from purchases made on Indian reservations by non-AI/ANs were in Oklahoma (37.0%), New Mexico (29.2%), Arizona (23.4%), Washington State (22.5%), New York (21.5%), and Nevada (18.6%; Figure). The lowest percentages of annual cigarette consumption reported from purchases made on Indian reservations were in California (1.5%), Texas (1.9%), Minnesota (2.7%), South Dakota (2.8%), and Iowa (3.2%). State-specific annual excise tax revenue losses associated with cigarette purchases by non-AI/ANs made on Indian reservations ranged from $0.8 million in Wyoming to $292.3 million in New York (online-only Table 2).

Figure.

Percentage of annual cigarette consumption reported from purchases made on Indian reservations by non–American Indians/Alaska Natives in quartiles, by state of purchaser’s residence, United States, 2010-2011. Nationwide, 4.2% of annual total cigarette consumption was reported from purchases made on Indian reservations. Percentages in states where the unweighted sample size was <10 (n = 30 states) were suppressed. Data source: Tobacco Use Supplement to the Current Population Survey, 2010-2011.²²

Discussion

Using self-reported measures on cigarette purchases made on Indian reservations and cigarette consumption, we estimated that a small proportion of cigarette smokers reported purchasing cigarettes from Indian reservations nationwide during 2010-2011. However, in 6 states, about 15% to 30% of smokers reported purchasing cigarettes from Indian reservations, resulting in considerable excise tax revenue losses.

Among the states with the highest percentage of annual cigarette consumption reported from purchases made on Indian reservations (Arizona, Nevada, New Mexico, New York, Oklahoma, and Washington State), none collected a tax equal to or greater than the full state excise tax on Indian reservation sales to nontribal members during 2010-2011 (online-only Table 1). Of these states, and consistent with other studies,²⁵ New York lost the most revenue because of tax avoidance and evasion.

Although states such as New York and Oklahoma have made substantial changes to regulate the sale of cigarettes on Indian reservations,^26,27 other states have not. State governments might start to consider or continue to consider policies aimed at reducing the sale of lower- or nontaxed cigarettes on Indian reservations,²⁸ such as agreeing to a tribal-state compact that allows tribes to collect state excise taxes in return for remitting part of the revenues for tribal benefit or distributing a limited quota of free tribal stamps to sales made to tribal members. States may benefit from further guidance on identifying and developing effective strategies to reduce the sale of nontaxed cigarettes on Indian reservations, although the policy impact on state tax revenue may vary by state.

Limitations

This study had several limitations. First, respondents’ annual cigarette consumption was based on their consumption in the month prior to the interview, and cigarette consumption may vary by month. However, interviews were conducted in January, May, and August. Thus, the seasonal effect on estimates of state average annual consumption was likely to be small. Second, the estimated tax revenue losses attributable to purchases made on reservations must be interpreted with caution. These estimates relied on self-reported data on consumption rather than data on actual tax-paid sales. Research comparing data on tax-paid sales with self-reported data on consumption suggests that smokers may underreport their consumption.²⁵ Additionally, self-reported purchases may not include those made through intermediaries, such as bootleggers. The use of self-reported data may underestimate levels of tax avoidance and evasion and the associated tax revenue losses compared with the use of a pack collection approach.^13,18 Third, because of limited information on tribal tax rates and tribal consumption, we made several assumptions (online-only Table 1) to estimate state revenue losses to reservation purchases. Lastly, the findings of this study were based on the 2010-2011 TUS-CPS and the corresponding years of state tax policies on cigarette sale on Indian reservations. As such, results may not be generalized to other years.

Conclusion

The availability and purchase of lower-priced cigarettes undermine the impact that cigarette excise tax increases otherwise would have on reducing cigarette use and improving public health.²⁹ Strategies to reduce the sale of nontaxed cigarettes on Indian reservations to non-AI/ANs have the potential to reduce the prevalence of smoking and recoup lost revenues from purchases made on Indian reservations.