JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

The Market Adjustment to Taxation: A Case Study

Among Alternative Forms of Tobacco
Christopher K. Coombs1 and Timothy W. Vines2

ABSTRACT

In this paper we illustrate some basic economic principles against the

backdrop of the U.S. tobacco market over the last decade, when a great deal of
tax policy was being implemented at both state and federal levels. Specifically,
we illustrate and estimate the secondary (unintended) effects from tax
increases on factory-made cigarettes. The readjustment by policy-makers, in
order to realize their objective, is also apparent throughout this period and
leads once again to behavioral responses by producers and consumers. Finally,
we estimate the cross-price elasticity between the demand for “roll your-own”
tobacco and tax increases on factory-made cigarettes over the period.

Introduction
Economics is often defined as the study of how society makes decisions about the allocation of scarce
resources among unlimited possible uses and desires. Government is often seen as an agent for improving
upon the outcomes achieved from the aggregated decisions of private economic units. One approach is for
the government to pass laws that are designed to provide incentives to change economic unit behavior. A
frequent difficulty noted by economists and other social scientists is that when the government is able to
change behavior, often at least part of the behavior change is unanticipated or unexpected by the regulators.

In general, the unexpected behavioral change can be termed an unintended consequence. It may be
rational for some private economic units to participate in activities that have been termed “loophole
mining” in order to find ways to circumvent the primary incentive provided by government action. In
addition, the unintended, secondary behavior changes might completely cancel the primary incentive and
thus worsen the outcome as compared to before the regulatory change. In this case, government can re-
regulate in order to try to mitigate the secondary behavior changes and preserve the primary changes.

This process in which the passage of a new law results in innovative, unintended new behaviors by the
regulated economic units and then re-regulation has been termed the regulatory dialectic. 3 Kane (1983,
1984) uses this principal to explain changes in the structure of the delivery of financial services as well as
competition among and reorganization of the regulatory agencies themselves. 4 Similar stories can be told
about many categories of regulation. Changes in the behavior of economic units can often be traced to
efforts to avoid or lessen regulatory burdens, especially taxation. For example, Cebula and Coombs (2011)
find that an increase in the top marginal tax rates on income in the U.S. leads to increases in income tax
evasion. Another example of this process can be found in the way that tobacco usage is regulated and taxed
in the United States. For example, Cebula (2010) provides a contemporary analysis on the impact of
cigarette taxation on consumption.

1
Department of Economics and Finance, Louisiana State University – Shreveport, [email protected]

2
Department of Economics and Finance, Louisiana State University – Shreveport, [email protected]

3
See, e.g., E. J. Kane (1983, 1984).
4
ibid.

27
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

The purpose of this paper is to show advanced undergraduate and beginning graduate students in
microeconomics, public choice, and public finance some basic economic principles at work through a
series of recent tax changes and the markets’ response against the backdrop of the U.S. retail tobacco
market. Also, students in health care economics will find that this analysis underscores the potential issues
(micro and macro) with tobacco consumption, the new health care law, and the moral hazard problem
associated with insurance. A historical narrative is provided, along with straight-forward econometrical
modeling and testing that demonstrates the reactions of the buyers and sellers of tobacco. The reactions
include a significant substitution effect, changes in the marketing of tobacco products, and efforts by
various governmental jurisdictions to re-regulate. A simple procedure for the estimation of a national cross-
price elasticity is also provided.

Tobacco Taxation
During the American Civil War, the U.S. government began to use both income and excise taxes as
temporary means of financing the conflict. Although most of the taxes instituted at that time have been
removed, the income tax and the two most important excise taxes, those on alcohol and tobacco persist to
this day.5 Tobacco is commonly taxed at the state or local level as well. Proponents of the excise taxes on
tobacco often state that there is a dual purpose to this regulation. The first purpose is to raise revenues,
often for a specific program. A second purpose is to enhance public health by discouraging the use of
tobacco. This dual purpose, in and of itself, suggests a problem with unintended consequences. To the
extent that the revenue is dedicated to a specific program, usage of tobacco creates a socially desirable
outcome that funds the government activity. To the extent that the tax acts as a deterrent to tobacco usage,
funding for the desirable activity is diminished, creating the need for the government to either reduce the
level of services or seek alternative financing.

In 2009, Congress passed the Children’s Health Insurance Program Reauthorization Act (CHIPRA).
The provision of insurance to the covered population is financed by an increase in the tobacco excise tax.
Thus a government activity designed to expand access to health insurance and thus improve public health is
largely funded by the revenue from an activity that is harmful to the public health. However, the taxes on
different forms of tobacco were not raised uniformly, resulting in behavioral changes by tobacco sellers and
users that were likely largely unanticipated. A recent study by Gootnick (2012) and the U.S. Government
Accountability Office suggests that many tobacco users recently switched to the less highly taxed forms of
tobacco. Sales of pipe tobacco and large cigars have risen immensely. The less favorably (relatively higher)
taxed “roll your own” (RYO) and small cigars have experienced decreases in sales.

Consumers and sellers, in an effort to avoid the tax, changed their behavior. Remote transactions,
through mail order and the internet, grew at the expense of the more easily traceable and taxable face-to-
face transactions. Signed in 2010, the Prevent All Cigarette Trafficking Act (PACT) was designed to offset
the change in behavior. PACT imposes numerous new reporting requirements on sellers and shippers of
tobacco products. Although this eases the process of tax enforcement, CHIPRA and PACT may result in
other previously noted behavior changes. Evans and Farrelly (1998) and Farrelly, Nimsch, Hyland, and
Cummings (2004) indicate that higher taxes lead to increased sales of tobacco products that have higher
concentrations of nicotine and tar. Adda and Cornaglia (2006) find that smokers may consume more
nicotine and tar per cigarette in the wake of increased taxation. Finally, Laugesen, Epton, Frampton,
Glover, and Lea (2009) found that RYO smoking is associated with increased smoke exposure per
cigarette, and even with filters is apparently no less and possibly more dangerous than factory-made (FM)
cigarette smoking. Overall, however, the extent to which these behavior changes offset the decrease in
tobacco usage is undetermined.

Attempts to re-regulate the unintended consequences are not always successful. The Tobacco Tax
Parity Act was introduced in Congress in 2010, but failed to achieve passage. The intent of the law was to
equalize the tax treatment of RYO and pipe tobacco. A similar attempt in 2011, concerning smokeless
tobacco also failed. The present study provides observations of the regulatory dialectic that has occurred in

5
See Ratner, Soltow, and Sylla (1979).

28
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

the U.S. with respect to excise taxes on various tobacco forms over the 1995 – 2011 time period. Within
these observations lies the fundamental component of the regulatory dialectic, consumers’ (and sellers’)
behavioral response to the price (tax) increase. Although the sample period is considered relatively small,
the adjustments made by all levels of government, as well as consumers and producers of tobacco have
been very dynamic. Next, we provide a simple illustrative and empirical analysis of the key economic
variables over the sample period. We will also explain the behavioral responses and regulatory dialectic as
it occurs over the sample period. The last section provides future policy implications, some potentially
important uses of the estimates, and concludes the analysis.

Statistical Evidence
The Federation of Tax Administrators (FTA) and The Bureau of Alcohol, Tobacco and Firearms (ATF)
have been releasing data on many forms of taxes on tobacco and tobacco manufacturing for the last decade
or so.6 We start by providing an illustration on federal and state excise taxes on FM cigarettes over time. 7
Data from the ATF allows for Figure 1 which shows that from 1995 to 2001, taxes on FM cigarettes in the
United States as a whole were relatively stable; the aggregate excise tax on FM cigarettes increased
annually by 4.6 percent, on average over this period.8 During the second half of the sample period,
however, we observe a remarkably different scenario. Clearly, the slope of the trend line is steeper over the
2002 – 2012 period. In fact, the average annual growth rate on the aggregate excise tax on FM cigarettes is
almost double at 8.5 percent. Using data from the FTA, we are able to observe information on changes in
states’ excise taxes on FM cigarettes. Table 1 provides this information. 9 The main finding from Table 1 is
that in contrast to the previous two years, starting in 2002 there was an increase in the number of states that
implemented changes in the tax rates on FM cigarettes. 10

Table 1.
Changes in State Excise Taxes on FM Cigarettes
Year # state changes Mean St. Dev. Min Max
2000 3 $0.53 $0.48 $0.04 $1.00
2001 3 $0.16 $0.11 $0.04 $0.26
2002 21 $0.43 $0.20 $0.07 $0.75
2003 18 $0.37 $0.18 $0.09 $0.70
2004 9 $0.38 $0.32 -$0.10 $0.80
2005 11 $0.56 $0.30 $0.10 $1.00
2006 5 $0.29 $0.30 $0.05 $0.82
2007 10 $0.56 $0.33 $0.20 $1.00

6
See, e.g., http://www.taxadmin.org/fta/link/default.php; http://www.ttb.gov/tobacco/tobacco-stats.shtml.
7
Each state can impose its own excise tax on FM cigarettes. In addition, the federal government has an excise tax on FM
cigarette manufacturing (or, consumption if we assume an equilibrium condition). To generate each year’s aggregate excise
tax on a pack of FM cigarettes we add to the federal excise tax each state’s excise tax on a pack of FM cigarettes in that year.
This “total” excise tax is what we will use for FMcigtx. It is not the actual level that we are interested in but the change in
taxes, or, marginal changes, that allow us to analyze the changes in behavior by the market participants.

 X  n 
1
8
The average annual growth rate is found using:  l
  1  100 where xf and xl are the first and last
 X f  
period’s values, respectively, and n is the total number of periods. In this case, x is the aggregate excise tax on a pack of FM
cigarettes
9
Unfortunately, data on tax changes in a state are only available from 2000.
10
Although the lack of observations before 2000 implies that we do not know what, if anything was occurring within each
state, with respect to excise taxes on FM cigarettes, the federal excise tax on FM cigarettes was $0.24/pack from 1993 to
1999. In 2000, the tax increased to $0.34/pack until 2002, when it increased to $0.39/pack. In 2009, the federal excise tax on
FM cigarettes jumped to $1.01/pack.

29
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

2008 5 $0.89 $0.40 $0.20 $1.25

2009 14 $0.59 $0.38 $0.10 $1.50
2010 6 $0.88 $0.43 $0.40 $1.60
2011 4 $0.22 $0.23 -$0.10 $0.40

To formally test whether the average annual growth rate is different between the two subsample periods
is equivalent to testing for different slopes. To do this, consider equation 1.

(1) FMcigtax = f(year, yr_dummy, year*yr_dummy)

Equation 1 expresses the U.S. aggregate tax on FM cigarettes (FMcigtax) as a simple function of time
(year), allowing for different intercepts for the two subsample periods by including a dummy variable
(yr_dummy) that equals 1 for years after 2001 and 0 otherwise, and also allowing for different slopes for the
two subsample periods by including an interaction term (year*yr_dummy) which is created simply by
multiplying the continuous variable year with the dummy variable yr_dummy.11

Estimating equation 1, using ordinary least squares to provide the “best-fitted” line through the actual
data, results in

(2) FMcigtax = -1942.12 + 0.982year –7641.5yr_dummy + 3.82year*yr_dummy

and the adjusted R2 = 0.99 and F-statistic = 910.58.12 All parameter estimates in equation 2 are statistically
significant and the F-statistic (p-value ≤ 0.0001) indicates that all variables belong in the model. Statistical
evidence therefore exists that the marginal effect (slope) is different between the two subsample periods.
For example, the estimated average annual increase from 1995 to 2001 is 0.982 which implies that each
additional year, from 1995 to 2001, the aggregate measure rose $0.98 on average, everything else constant.
The parameter estimate on the interaction term is 3.82, which implies that each additional year after 2001
(yr_dummy = 1), the aggregate measure increased by $4.80 (0.98 + 3.82) on average, everything else the

A simple transformation of the data allows us to present the previously found direct relationship
between time and the U.S. aggregate tax on FM cigarettes in percentage terms. By taking the natural log
ofthe dependent variable the coefficients on the explanatory variables will explain changes in the dependent
variable in percentage terms and therefore our results will be measured as growth rates. Equation 3,
therefore, includes the estimates from our log-linear specification and is

(3) lnFMcigtax= -97.46 + 0.05year – 80.29yr_dummy+ 0.04year*yr_dummy

and the adjusted R2 = 0.99 and F-statistic = 747.80 (p-value ≤ 0.0001).13 We can now present the marginal
effects as rates. For example, the estimated average annual increase from 1995 to 2001 is 0.05 which
implies that each additional year, from 1995 to 2001, the aggregate measure increased by 5 percent on
average, everything else constant. The parameter estimate on the interaction term is 0.04, which implies
that each additional year after 2001, the aggregate measure increased by 9 percent (0.05 + 0.04) on average,
everything else the same. Because the coefficient on the interaction term is statistically significant, we can
reject the hypothesis that the slopes are the same between the two subsample periods. These quantitative
results are similar to what was presented earlier in terms of our simple “average annual growth”
calculations.

11
Allowing for the potential for different intercepts between the two subsample periods is not germane to this analysis. The
interpretation would be similar to the expected differences in the aggregate tax at year = zero. The coefficient on this
variable, however, can be important in its use for other calculations.
12
The robust standard errors (absolute values) are 148.97, 0.075, 595.88, and 0.297 for the coefficients on the intercept, year,
yr_dummy, and year*yr_dummy, respectively. All parameter estimates are statistically significant (p-value ≤ 0.0001).
13
The robust standard errors (absolute values) are 8.86, 0.004, 17.79, and 0.009 for the coefficients on the intercept, year,
yr_dummy, and year*yr_dummy, respectively. All parameter estimates are statistically significant (p-value ≤ 0.0001).

30
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

The goal of this analysis is to provide chronologically the behavior (or changes in behavior) of those
participants (consumers/producers) in U.S. tobacco markets due to third-party intervention (government).
The change in behavior by market participants is in response to government intervention. The regulatory
dialectic continues in the form of government’s reaction to the unintended effects from changes in behavior
by the market participants when the ultimate goal is to internalize the negative externality associated with
the consumption of tobacco products and/or generate the optimal amount of tax revenue. First, we have
shown that many states have monotonically increased the excise taxes on FM cigarettes, especially after
2001. To the extent that substitutes exist and assuming purposeful behavior, some market participants
(consumers) will be expected to respond to the price increase (increase in taxes) on FM cigarettes by
changing their preferences toward relatively lower-priced substitutes. A popular substitute has turned out to
be RYO cigarettes. RYO “manufacturing” is relatively novel and perhaps this is why the ATF has
manufacturing data on RYO tobacco only as far back as 2000. However, RYO manufacturing has grown in
market share within the tobacco industry and now has the attention of big cigarette corporations. 14 Briefly
described, the market for RYO cigarettes consists of tobacco outlets, or “filling stations,” that include a
heavy duty all-in-one cigarette machine. Each machine, costing $40,000, is able to produce 200 cigarettes
in 10 minutes. Consumers rent the machines at a cost of about $10. Including tobacco, papers or tubes, a
carton of RYO cigarettes costs each consumer generally about half of what it costs for a carton of FM
cigarettes.

Figure 2 provides another illustration of a time trend. In this case, we plot the manufacturing of RYO
tobacco and aggregate excise taxes on FM cigarettes over time. To the extent that the production of RYO
tobacco proxies for the demand for RYO tobacco, and changes in excise taxes on FM cigarettes leads to
price changes in FM cigarettes, we can see that there is a substitute effect illustrated in the figure.
Specifically, over the 2001-2008 period, when aggregate excise taxes on FM cigarettes collectively were
increasing 10 percent, on average, the production of RYO tobacco was increasing. The substitute effect
implies, for example, that when the price of a good increases, consumers will substitute away from the
relatively higher-priced good to another relatively cheaper-priced good, one that consumers regard as a
substitute for FM cigarettes. In this case, RYO tobacco has served as a substitute for FM cigarettes.

Interestingly, however, the substitute effect does not occur after 2008, even when aggregate excise taxes
on FM cigarettes are still increasing at a 10 percent annual rate, on average. In fact, the relationship appears
exactly opposite; as taxes on FM cigarettes are increasing, the quantity demanded of RYO tobacco is
decreasing. Because FM cigarettes and RYO are not complementary goods, something else must explain
this anomaly.

To determine the factors that explain why we don’t observe a substitution effect throughout the sample
period we first consider Figure 3. The figure illustrates the time trend of federal tax rates ($) on RYO and
pipe tobacco. Although both are relatively stable to 2008, federal tax rates on RYO and pipe tobacco
increased by 2159 and 159 percent, respectively, in 2009. That is, the tax on RYO tobacco jumped from
$1.0969/lb. to $24.78/lb. and the tax on pipe tobacco increased from $1.0969/lb. to $2.83/lb. In addition to
this being government’s attempt to counter the change in consumer behavior due to federal and state
increases on taxes on FM cigarettes, Congress also levied the tax increases on RYO and pipe tobacco in
anticipation of price-sensitive teens switching from cigarettes to what had been the lower-taxed tobacco
alternatives. Interestingly, at this time when the federal government increased the excise tax on pipe
tobacco by 157 percent, the federal excise tax on FM cigarettes increased by almost the same rate (or, $0.39
to $1.01). The revenue-generating side of government was making adjustments to make sure that the
expansion of SCHIPS (reauthorization of CHIPRA) would be covered and the health and safety side of
government was in part considering the unintended effects of legislation.

However, although the federal government increased tax rates on pipe tobacco, as shown in figure 3, the
rate increase comparatively speaking was not as pronounced as the rate increase on RYO tobacco. Because
cigarettes and pipe tobacco are not considered perfect (or close) substitutes, what happened next was
clearly unanticipated. Figure 4 illustrates the time trend of the production of RYO and pipe tobacco.

14
Large tobacco corporations and convenience store chains have actively lobbied state and federal legislatures to more
strictly regulate RYO shops. This is a form of rent-seeking; another topic of inefficiency.

31
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

Assuming once again that production is a proxy for demand we see that after 2008 the quantity demand
decreased for RYO tobacco and increased for pipe tobacco. In fact, since the tax increases took effect on
April 1, 2009, according to the U.S. Government Accountability Office (GAO), RYO tobacco sales have
decreased by 74 percent while pipe tobacco sales have increased from 3.2 to 30.5 million pounds a year. Is
it that clear that consumers are switching to pipe smoking? Not according to the GAO study, that concluded
the shift is mostly due to consumers switching from using RYO tobacco to pipe tobacco in the RYO
machines. Moreover, according to industry experts, government officials and nongovernmental
organizations, to help facilitate the switch, manufacturers of RYO tobacco have simply re-branded their
products as pipe tobacco making minimal changes other than packaging and labeling. 15 Therefore, the
result is that consumers are switching from the relatively higher priced RYO tobacco to the relatively less
expensive pipe tobacco in the RYO machines when there is arguably no actual difference in the tobacco.

In this case, we would expect to see a substitution effect where the price of FM cigarettes is increasing
(due to federal and state tax increases) over the entire sample period and the response by consumers (with
the help of manufacturers) is an increase in the demand for RYO and “pipe” tobacco. First, however, the
demand for pipe tobacco must be disaggregated into actual pipe tobacco and the portion that is actually
RYO, referred to here as “pipe” tobacco. To do this, we assume that the demand for actual pipe tobacco
follows the same linear time trend in 2009-2011 based on years prior to 2009. Observing pipe tobacco
demand again in figure 4 illustrates the linear trend prior to 2009. We are simply assuming that holding
everything else constant, the demand for pipe tobacco would have continued in this linear path if tax rates
on RYO tobacco did not increase. Violating the ceteris paribus rule, however, is the previously mentioned
fact that tax rates on actual pipe tobacco increased as well in 2009 (from $1.10/lb to $2.83/lb). To the
extent that smokers of actual pipe tobacco altered their behavior by demanding less pipe tobacco after the
tax increase, our estimates of “pipe” tobacco demand (what will be added to RYO tobacco for 2009-2011)
should be considered conservative. 16

Consider the results of a linear regression of pipe tobacco demand expressed as a function of time for
the years 1999 – 2008:

(4) Pipe tobacco = 567,000,000 – 280775.8year

The results in equation 4 are used to forecast quantity demand for the years 2009 – 2011 and are shown in
figure 5. For each year after 2009, the vertical distance between reported pipe tobacco and the forecasted
pipe tobacco is considered “pipe” tobacco (which is actually RYO tobacco) and is added to RYO tobacco
to make up real or all RYO tobacco. That is, for each year, we take the difference between the reported pipe
tobacco demand and the linear-forecasted pipe tobacco demand. The difference for each year is then added
to RYO tobacco demand to give us the quantity demanded of real (all) RYO tobacco.

The responsiveness of smokers to substitute RYO tobacco for FM cigarettes over the sample period can
be thought of as a national cross-price elasticity of demand where the tax increase on FM cigarettes leads to
an increase in the demand for RYO tobacco. Illustratively, In Figure 6 this amounts to simultaneously
observing a movement along the FM tax rate curve (change in the independent variable) and a movement
along the Real RYO tobacco curve (change in the dependent variable) for a particular year (or, period). If
we are interested in estimating the national cross-price elasticity during and after 2009, when federal taxes
on FM cigarettes increased 159 percent, first consider equation 5

(5) RYO tobacco = f(FMcigtax)

15
Some manufacturers label the “pipe” tobacco in easy to familiarize (and substitute) terms. That is, if a sample of “pipe”
tobacco is labeled “MRD,” consumers are informed that this tobacco tastes like the Altria Group’s Marlboro Red.
16
We estimated a simple linear regression of pipe tobacco demand as a function of the pipe tobacco tax rate from 1999 to
2008. With the results, which were statistically significant, we rearranged the equation to solve for the pipe tax rate that
would result in no quantity demand for pipe tobacco. The estimated “choke” price is $1.97. This rate is lower than the actual
increase ($2.83) implying once again that our estimates of “pipe” tobacco demand for 2009-2011 should be considered
conservative. The results of the auxiliary regressions are available upon request.

32
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

where the coefficient on FMcigtax is expected to be positive. 17 To estimate a cross-price elasticity at a

particular aggregate excise tax on FM cigarettes, we specify equation 5 in linear functional form.

Estimating equation 5, using ordinary least squares to provide the “best-fitted” line through the actual
data, results in

(6) RYO tobacco = -935089.4 + 372200.8FMcigtax

and the adjusted R2 = 0.80 and F-statistic = 21.13 (p-value ≤ 0.0001). The parameter estimate on
FMcigtax in equation 6 is statistically significant at less than one percent significance level (p-value ≤
0.0001) and interpretation of the coefficient is as a marginal effect. That is, a one-dollar increase in FM
cigarette taxes (from any and all states as well as at the federal level) increases the demand of RYO tobacco
by 372,200 pounds, everything else constant. 18 Once again, we aggregate each state’s excise tax on FM
cigarettes and add this to the federal excise tax in FM cigarettes to get the total (or, national) excise tax on a
pack of FM cigarettes. We do this so that we can assume that although the excise tax on FM cigarettes
changes, the total change can be distributed among many states and at the federal level. This is what
ultimately (in theory) affects the national (aggregate) demand for RYO tobacco. For example, in 2009
when the federal excise tax on FM cigarettes increased to $1.01 per pack (from $0.34 per pack in 2008), 15
states also increased their own excise tax on FM cigarettes. In 2009, the total excise tax was $69.1956. This
is an increase of $8.4516 from 2008 ($60.744) and the change implies the increase in the federal excise tax
as well as the total from the 15 states that increased their own excise tax on FM cigarettes.

Using the results from equation 6 and the 2009 total excise tax of $69.1956 and the formula for point
elasticities, we estimate a national cross-price elasticity for RYO tobacco equal to

(7) EXR = 372200.8 * 69.1956/24819568.28 = 1.04.19

The positive sign implies that FM cigarettes and RYO are considered substitutes. Moreover, similar to
butter and margarine, they are considered rather close substitutes. For example, a 10 percent increase in
excise tax rates on FM cigarettes elicits a 10.4 percent increase in the quantity demand for RYO tobacco,
everything else constant. Consider another example and suppose an increase of five dollars (across all
states and at the federal level) in the aggregate excise tax on FM cigarettes. The 6.6 percent increase
($75.3306 to $80.3306) in excise taxes will lead to a 6.89 percent increase in the demand for RYO tobacco
(including RYO tobacco labeled as pipe tobacco), everything else constant. With our national cross-price
elasticity estimate (1.04), the percentage increase from our example (6.89%), and the 2011 demand for
RYO tobacco (36.5 million pounds), this implies an increase in 2,514,850 pounds of RYO tobacco.
Because it takes 0.5 pound of RYO tobacco to roll the equivalent of 10 packs of FM cigarettes, this
compares to over 50 million packs of RYO cigarettes.

Conclusion

The results presented herein clearly indicate a significant pattern of regulatory change leading to
secondary and important unanticipated behavioral changes by market participants. In turn, the changes in
behavior initiated additional efforts by lawmakers to re-regulate. The increase in taxes on FM cigarettes at
any level of government has prompted consumer/producer substitution of tobacco products with relatively
lower taxes. The constant back and forth on these and other related important issues continues. For
example, headlines in 2012 included that voters in California recently narrowly defeated Proposition 29, a
new excise tax on cigarettes to fund cancer research; Illinois is using a new tobacco tax to offset Medicaid

17
Additionally, compared to the previous year’s increase of 5, in 2009 there were 15 states that increased the excise tax on
FM cigarettes. The average increase among the 15 states was $0.59 and at least one state increased the tax by $1.50. Once
again, we aggregate each state’s excise tax on FM cigarettes and add this to the federal excise tax in FM cigarettes to get the
total (or, national) excise tax on a pack of FM cigarettes assuming every state sells FM cigarettes.
18
The robust standard error for the parameter estimate is 80977.6.
19
See Wooldridge (1999).

33
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

cuts; and finally, several jurisdictions are struggling with the regulation of the RYO shops, including the
Federal government. The transportation bill signed into law in July 2012 included a little known
amendment that classifies RYO retailers as cigarette manufacturers, apparently offsetting much of the tax
advantage to pipe tobacco. Several questions, however, remain. Will this change lower government
revenues? Will consumers/producers find another unanticipated loophole to avoid the new regulation?

The results in this analysis can be discussed and used by educators and students in microeconomics,
health care economics, and public finance in a variety of contexts. For example, the paper illustrates how to
use readily available data to demonstrate a significant change in the average and growth rate of time series
data. A simple forecasting technique is also demonstrated and using the parameter estimates from equation
6, a national cross-price elasticity is estimated. Assuming the consumer (retailer bears none) bears 100
percent of the burden of all tax increases on FM cigarettes, the results show that FM cigarettes and RYO
tobacco in the U.S. are indeed close substitutes. Policy-makers can use the estimate to forecast changes in
tax revenues on RYO tobacco given changes in taxes on FM cigarettes. Moreover, those in public finance
(e.g., state operating budgets) should find this information useful when forecasting revenues. For example,
in Louisiana the governor is proposing the elimination of its state income tax that will be offset in part by
an increase in taxes on FM cigarettes. In order for this proposal to be revenue neutral, the governor and his
advisors must understand the substitution that will occur if the taxes on FM cigarettes increase while
holding the taxes of other forms of tobacco constant.

To the extent that the goal is to reduce tobacco consumption in general, policy-makers should consider
the potential shock to the nation’s overall health stock when this analysis has shown that substitution has
mitigated at least some of the intended effect. Finally, although RYO cigarettes have increased in
popularity, their comparative potential toxicity is uncertain. To the extent that RYO tobacco is relatively
more dangerous to one’s health, this paper has presented a troubling outcome for health care officials to
contemplate. This should be considered even more important when bearing in mind the new health care law
in the U.S. which essentially requires everyone (with some exceptions) to have health insurance by 2014.
Because of the moral hazard problem that can exist within health care from those insured, these results
suggest that individuals/consumers who will have health insurance might have less of an incentive to quit
or less of a disincentive to start.

References

Adda, Jerome, and Francesca Cornaglia. 2006. Taxes, Cigarette Consumption, and Smoking Intensity. The
American Economic Review (96) 4:1013-28.

Cebula, Richard J. 2010. A Preliminary and Contemporary Panel Data Analysis of the Consumption and
Impact of Cigarette Taxation. Journal of Public Finance and Public Choice 1:39-59.

Cebula, Richard J., and Christopher K. Coombs. 2011. What Happens When You Increase Taxes on the
Rich? Tax Notes (131) 3:299-308.

Evans, William, and Matthew Farrelly. 1998. The Compensating Behavior of Smokers: Taxes, Tar, and
Nicotine. Rand Journal of Economics (29) 3:578-95.

Farrelly, Matthew, Nimsch, C.T., Hyland, A. and M. Cummings. 2004. The Effects of Higher Cigarette
Prices on Tar and Nicotine Consumption in a Cohort of Adult Smokers. Health Economics (13) 1:49-58.

Gootnick, 2012, “Large Disparities in Rates for Smoking Products Trigger Significant Market Shifts to
Avoid Higher Taxes,” GAO 12-475, Apr 18, 2012.

34
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

E. J. Kane. 1983. Metamorphosis in Financial-Services Delivery and Production. Strategic Planning of

Economic and Technological Change in the Federal Savings and Loan, San Francisco: Federal Home Loan
Bank Board, 1983:49-64.

E. J. Kane. 1984. Technological and Regulatory Forces in the Developing Fusion of Financial Services
Competition. Journal of Finance (39) 3:759-772.

Laugeson, Murray, Epton, Michael, Frampton, Chris MA, Glover, Marewa, and Rod A. Lea. 2009. Hand-
Rolled Smoking Patterns Compared with Factory-Made Cigarette Smoking in New Zealand Men. BMC
Public Health 9:194.

Ratner, S., Soltow, James, and Richard Sylla. 1979. The Evolution of the American Economy. New York:
Basic Books.

Jeffrey M. Wooldridge. 1999. Introductory Econometrics: A Modern Approach. South-Western College

Publishing, U.S.

http://www.taxadmin.org/fta/link/default.php

http://www.ttb.gov/tobacco/tobacco-stats.shtml

35
 JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

Figure 1

Aggregate FM Cigarette Taxes over Time

80
60
FM cigarette excise tax

(in U.S. dollars)

40
20

2000 2005 2010

Year

Figure 2

RYO Tobacco Demand and FM Cigarette Excise Taxes over Time

80
00
00
00
34
or received from Puerto Rico

FM cigarette excise tax

Pounds produced

60
(in U.S. dollars)
00
00
00
24

40
00
00
00
14

20
0
00
00

2000 2005 2010

40

Year
RYO tobacco FM cigarette excise tax

36
 JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

Figure 3

RYO and Pipe Tobacco Excise Tax Rates over Time

25
20
Federal tax rates
(in U.S. dollars)

15
10
5
0

2000 2005 2010

Year
Pipe tobacco tax RYO tobacco tax

Figure 4

RYO and Pipe Tobacco Demand over Time

00
00
00
34
or received from Puerto Rico

Pounds produced

00
00
00
24

00
00
00
14

0
00
00

2000 2005 2010

40

Year
Pipe tobacco RYO tobacco

37
JOURNAL OF ECONOMICS AND FINANCE EDUCATION ∙ Volume 12 ∙ Number 2 ∙ Winter 2013

Figure 5

Reported Pipe Tobacco Demand and Estimated Pipe Tobacco Demand

00
00
00
38

00
Pipe tobacco

00
in pounds

00
28

00
00
00
18

0
00
00
80

2000 2005 2010

Year
pipe tobacco pipe_lb_hat

Figure 6

Aggregate FM Cigarette Excise Tax Rates and Real RYO Tobacco Demand over Time
80
07
e+
80
3.

FM cigarette excise tax

Real RYO tobacco

60
in U.S. dollars
07
in pounds

e+
80
2.

40
07
e+
80
1.

20
0
00
00

2000 2005 2010

80

Year
Real RYO tobacco FM cigarette excise tax

38