In subsequent sections of this study, we discuss in some detail the tax and enforcement histories of Michigan, New Jersey and California. We have focused on these states for several reasons: Each is populous; each borders other taxing jurisdictions that have imposed significantly different levels of cigarette taxation in the past; and each falls in a separate part of the country — East, West and Central.

Historical and anecdotal evidence indicates that each state has also experienced both casual and commercial smuggling. This evidence is consistent with the findings of our model.

We estimate that from 1990 through 2006, tax-induced casual and commercial smuggling imports in Michigan provided 6.0 percent and 11.6 percent respectively of the state's total cigarette consumption. Our casual smuggling estimate is thus similar to that of DeCicca and his colleagues in their review of 2003 survey data for Michigan (their study did not measure commercial smuggling activities). We also estimate that a small percentage of Michigan's total cigarette market involved smuggling of Michigan cigarettes into Canada (a figure not included in the percentages above), meaning that on balance, casual and commercial smuggling into the state represented 16.0 percent of Michigan's estimated total cigarette consumption from 1990 through 2006 — a little less than one in six cigarettes.

Our smuggling estimates for Michigan in 2006 alone are much higher, a natural result of the state's 50-cent per-pack tax hike in 2002 and its 75-cent per-pack tax hike in 2004. We estimate that tax-induced casual and commercial smuggling in Michigan were 13.6 percent and 22.5 percent respectively of the state's total cigarette consumption in 2006. Although we again calculate that a small amount of Michigan's potential cigarette consumption was diverted to Canada through smuggling, we estimate that the total number of cigarettes smuggled into Michigan amounted to 34.5 percent of the state's total cigarette consumption in 2006. This represents more than one in three cigarettes consumed in the state, or an amount equal to more than 52 percent of the state's legal cigarette sales.

Our calculations suggest that New Jersey also experienced high smuggling rates. We estimate that commercial smuggling imports represented 13.8 percent of New Jersey's total cigarette consumption from 1990 through 2006. During the first half of this period, however, New Jersey's tax rates were relatively low compared to surrounding states. Hence, it is not surprising that our estimates for the period also show that casual smuggling on balance moved a small number of cigarettes — an amount equal to about 0.6 percent of the state's total consumption — out of New Jersey (probably in part to high-tax New York). As discussed below (see "Case Study: New Jersey"), variations in New Jersey's cigarette tax and regulatory policies have meant that the state has served as both an illegal exporter and importer of cigarettes to surrounding states, making it an active commercial hub for cigarette smuggling.

In recent years, New Jersey has raised its cigarette excise tax to the country's highest, and our model no longer shows a net export of casually smuggled cigarettes. Our estimates indicate that casual and commercial smuggling imports in 2006 represented 9.3 percent and 30.1 percent of New Jersey's total cigarette consumption.[*] The total number of cigarettes smuggled into the state in 2006 appears to be about 40 percent — two in five of the cigarettes consumed in the state, or an amount equal to about 68 percent of the state's legal cigarette sales.

In California, we estimate that casual and commercial smuggling imports from 1990 through 2006 amounted to 5.9 percent and 7.4 percent respectively of the state's total cigarette consumption, but these figures do not include cigarette smuggling from Mexico. Our model estimates that illegal imports from Mexico represented an additional 10.1 percent of cigarette consumption in California during this period, meaning that smuggled imports of cigarettes represented 24.5 percent of the state's total cigarette consumption — about one cigarette in four.

Our estimates for California's 2006 smuggling rates are much higher. We find that the imports from casual and commercial smuggling in 2006 represented 10.7 percent and 8.0 percent respectively of California's total cigarette consumption, with smuggling from Mexico representing an additional 15.5 percent. On the whole, then, we estimate that smuggled imports of cigarettes accounted for 36.6 percent of California's total cigarette consumption in 2006 — more than one in three cigarettes, and an amount equal to about 58 percent of California's legal cigarette sales.

Graphic 8: Estimated Tax-Induced Smuggling in Michigan, New Jersey and California as a Percentage of Total Cigarette Consumption (Legal and Illegal)

Graphic 8: Estimated Tax-Induced Smuggling in Michigan, New Jersey and California as a Percentage of Total Cigarette Consumption (Legal and Illegal) - click to enlarge

Notes: Estimates computed based on the regression results presented in columns 3 and 4 of Table 3 of Appendix A. The sum of commercial, casual and Canada/Mexico smuggling does not equal the total presented in the final column due to the non-linear nature of the model.

As the discussion above suggests, we estimate that Michigan, New Jersey and California have some of the highest levels of smuggling — particularly commercial and international smuggling — in the country. The second section of the study recounts the three states' experiences with tax hikes and smuggling. The stories are memorable and sometimes almost comical,  but tax-induced smuggling has had serious impacts on the stated policy goals of improved health and reliable tax revenue.


[*] The New Jersey rate in our study is significantly different from the estimates made by Lovenheim and DeCicca et al. Several factors contribute to the differences. First, each of the authors uses different techniques and datasets for measuring smuggling in the state. For instance, Lovenheim's dataset involves Metropolitan Statistical Areas, whereas our data is statewide. DeCicca et al. use survey data. Moreover, each author uses different periods of study, another factor that would affect the output, since state tax rates fluctuate relative to one another. Perhaps most importantly, Lovenheim and DeCicca are reporting a percentage of consumers who smuggle or report cross-border shopping activities. This study's output measures a percentage of cigarettes smuggled.