Does the Presence of Very Young Children and/or Older Minor-aged Children in the Home Reduce Cigarette Smoking? Panel Data Evidence for the United States

Abstract

Within the context of the Period Fixed-Effects Model, this study of cigarette consumption in the US uses a state-level panel dataset to investigate a new hypothesis. This hypothesis argues that the presence in the home of minor-aged children, that is, children under 18 years of age, reduces the percentage of the population that smokes cigarettes. The eclectic model includes the levels of federal plus state cigarette excise taxation along with a number of other established explanatory variables. The empirical estimates in this preliminary study find support for the hypothesis proffered here, that is, it is found that the percentage of the population that smokes cigarettes is a decreasing function of the percentage of households with minor-aged children residing in the home. Moreover, aside from providing further insight into smoker behaviour, the results from this study also raise the question that since this phenomenon has been ignored in previous related studies involving such public policies as cigarette taxation and cigarette smoking bans that those studies may suffer from omitted variable bias.

JEL Classification: R22, I18, H71, H29

Keywords

Percentage of the Population that Smokes Cigarettes Presence of Minor-aged Children in the Home Cigarette Excise Taxation Per Capita Income Unemployment Education

1. INTRODUCTION

The impact of excise cigarette taxes and other factors on cigarette consumption has been studied extensively. The need to reduce cigarette consumption remains a pertinent public-policy issue (White, 2010), primarily because the adverse health effects resulting from cigarette smoking directly account for an estimated 443,000 deaths in the U.S. annually, a number that disregards the impacts of secondary smoke on morbidity (Centers for Disease Control and Prevention, 2009a; 2009b). Cigarette smoking is also linked to myriad other adverse health effects, including heart disease (Centers for Disease Control and Prevention, 2009a; 2009b). The usual finding is that higher cigarette taxation leads to a reduction in the number of packs of cigarettes consumed, although other factors significantly influencing cigarette consumption have been identified (Cebula, 2010; Chaloupka, 2004; Chaloupka and Warner, 2000; Connelly et al., 2009; DeCicca et al., 2002; Evans and Farrelly, 1998; Farrelly et al., 2004; Forster and Jones, 2001; Goel, 2008, 2009; Holt and Laury, 2002; Koch and Cebula, 1992; Lien and Evans, 2005; Meier and Licari, 1997; Showalter, 1998; Tan, 2012; Tauras, 2006). The present study seeks to add to this rich literature, after accounting for cigarette excise, cigarette taxation and various other standard variables argued to potentially affect the level of cigarette consumption, by testing two heretofore effectively unstudied hypotheses. The first hypothesis is that, the presence of very young children, that is, those under five years of age, in the household reduces the number of packs of cigarettes smoked. The second hypothesis is that the presence in the home of other minor-aged children, that is, those over five years of age but under 18 years of age, reduces the number of packs of cigarettes smoked.

This study adopts a state-level panel data series spanning the period 2002 through 2007; log–log estimation results from period fixed-effects estimates are provided to facilitate interpretation. An eclectic model inclusive of the two hypotheses given above is developed in Section 2 of this study. In Section 3, the empirical model and data are described. Section 4 of the study provides the empirical results, the robustness tests adopting alternative levels of educational attainment, a summary and conclusions.

2. AN ECLECTIC MODEL

Cigarette smoking is measured in this study as the per capita number of packs of cigarettes smoked annually in each state (CONSPC). As suggested above, a number of previous studies have explored the determinants of cigarette consumption (Adda & Cornaglia, 2006; Anderson & Mellor, 2008; Cebula, 2010; Chaloupka & Warner, 2000; Connelly et al., 2009; Cox, 2010; DeCicca et al., 2002; Evans & Farrelly, 1998; Farrelly et al., 2004; Forster & Jones, 2001; Goel, 2008, 2009; Holt & Laury, 2002; Koch & Cebula, 1992; Lien & Evans, 2005; Lovenheim, 2008; Meier & Licari, 1997; Showalter, 1998; Tan, 2012; Tauras, 2006). Relevant contents of a few of these studies are briefly considered here.

As a general background, it is first observed that Anderson and Mellor (2008) find that risk aversion, as measured by subject choices in the Holt and Laury (2002) lottery choice experiment, is negatively and significantly associated with cigarette smoking, heavy drinking, being overweight or obese and seat belt non-use, as well as with several summary measures of risky behaviour. The results in the Anderson and Mellor (2008) study suggest, among other things, that individuals in many, although far from all, cases tend to be risk-averse when it comes to cigarette smoking and their health.

Raising cigarette excise taxes is commonly regarded as one of the most effective prevention and control strategies for reducing and/or limiting cigarette consumption; indeed, it has been thusly regarded for many years (Drayton, 1972). In a study by Lien and Evans (2005), reduction in smoking by pregnant women and subsequent improvements in birth weight occur almost immediately after a large cigarette tax hike has been implemented. A study by Forster and Jones (2001) measures the cigarette tax elasticity of cigarette consumption. Elasticity estimates are all within the range of −0.4 to −0.7, indicating that, for example, an increase in the cigarette tax of, say, 10 per cent, would lead to a decrease in the volume of cigarettes consumed to between 4 per cent and 7 per cent. A more recent study by Cebula (2010) provides a log–log estimate focusing only on state-level cigarette excise cigarette taxes in which the state excise cigarette tax elasticity of cigarette smoking is found to be −0.2. Largely similar results to these studies are found in Meier and Licari (1997), Showalter (1998), Lovenheim (1998), Evans and Farrelly (1998), Forster and Jones (2001), Farrelly et al. (2004), Lien and Evans (2005) and Connelly et al. (2009). Accordingly, this study includes the total (federal plus state) cigarette excise taxes in each of the states (CIGTX) among the explanatory variables pertaining to cigarette smoking, with the demand for cigarettes expected to be a decreasing function of the level of cigarette excise taxes, ceteris paribus.¹

Furthermore, following the suggestion in Koch and Cebula (1992), as an additional related factor that can arguably be associated with higher cigarette taxation, this study also includes the percent of the population in each state that is in poverty (POV), with the argument being that, given the limited resources of the poor, higher cigarette prices resulting from higher cigarette excise taxes simply makes smoking a more unaffordable habit for at least some portion of those in poverty, ceteris paribus.

Koch and Cebula (1992), Goel (2008) and Cebula (2010) argue and find that the per capita annual cigarette consumption in a state is an increasing function of the unemployment rate in that state (UNEMP), with the reasoning being that smoking, perhaps like alcohol, is a coping technique for many who are unemployed. It is also argued and/or found by Koch and Cebula (1992), Forster and Jones (2001), Anderson and Mellor (2008), Cebula (2010) and others that the greater the average educational attainment level among the adult population (EDUC)² and the higher the per capita income level (PCI), the lower the per capita annual consumption rate of cigarettes.

Finally, there are the two new hypotheses being investigated by this study. As stated above, the first hypothesis is that the presence of infants in the home reduces the number of packs of cigarettes smoked. Alternatively expressed, the higher the percentage of households in a state where there is at least one infant present, that is, at least one child under one year of age (YNGKIDS%), the lower the per capita number of packs of cigarettes smoked annually in the state, ceteris paribus. The second hypothesis is that the higher the percentage of households in a state having a presence in the home of one or more minor-aged children, that is, children over one year of age but under 18 years of age (OTHERKIDS%), the lower the per capita annual number of packs of cigarettes smoked in the state, ceteris paribus.

The rationale behind these two hypotheses is effectively identical. To begin with, if there are one or more very young children in the home and/or there are one or more other minor children in the home, smoking in the household exposes the very young and/or other children to a health hazard in the form of second-hand smoke from cigarette consumption. Clearly, this rationale implies that responsible parents have a clear reason to alter their cigarette consumption. In addition, to the extent that the very young children and/or other minor children in the home are witness to cigarette smoking, it serves as an example that cigarette smoking is acceptable if not condoned and thereby raises the likelihood that the offspring in question may themselves become future cigarette smokers. This constitutes another rationale for responsible parents to alter their cigarette consumption. Furthermore, to the extent that there is parental cigarette smoking, whether inside or outside of the home, the direct health risk from that behaviour to the parents, that is, one or both parents’ contracting lung cancer and/or some other serious illness as well as possibly succumbing to premature death, puts the minor offspring living in the home in the tenuous and risky position of having to cope both personally and financially with the adverse health consequences of that cigarette smoking on the parental units. Thus, the presence of infants and other minor children in the home creates disincentives to smoke cigarettes in the home on the one hand and to smoke cigarettes in general. Thus, the greater the percentage of families having young children and/or other minor children living in the home, the greater is the incentive to not smoke cigarettes. Clearly, the latter incentive can, at least for more responsible parents, create a compelling reason either to smoke less or to quit smoking altogether.

Based upon the above-cited studies as well as the two hypotheses proffered above, it follows that:

\begin{matrix} C O N S P C = f (C I G T X, P O V, U N E M P, E D U C, P C I, \\ Y N G K I S %, O T H E R K I D S %) \end{matrix}

(1)

where it is expected that:

f_{C I G T X} \leq 0, f_{P O V} \leq 0, f_{U N E M P} > 0, f_{E D U C} \leq 0, f_{P C I} \leq 0, f_{Y N G K I D S %} \leq 0, f_{O T H E R K I D S %} \leq 0

(2)

3. THE EMPIRICAL MODEL AND DATA

Based on the studies cited and discussed in Section 2 above and the two hypotheses proffered at the end of Section 2, this study adopts an eclectic model of cigarette consumption in which the determination of the annual per capita cigarette consumption in year t in state j is modelled as each of the following three similar models:

\begin{array}{l} \begin{matrix} \log C O N S P C_{t j} = a_{0} + a_{1} \log C O N S P C_{t ​ - ​ 1 j} + a_{2} \log C I G T X_{t j} + a_{3} \log P O V_{t j} \end{matrix} \\ \begin{matrix} + a_{4} \log U N E M P_{t j} + a_{5} \log H I G H S C H L_{t j} + a_{6} \log P C I_{t j} \\ + a_{7} \log Y N G K I D S S %_{t j} + a_{8} \log O T H E R K I D S %_{t j} + ε_{t} \end{matrix} \end{array}

(3)

\begin{array}{l} \begin{matrix} \log C O N S P C_{t j} = a_{0} + a_{1} \log C O N S P C_{t ​ - ​ 1 j} + a_{2} \log C I G T X_{t j} + a_{3} \log P O V_{t j} \end{matrix} \\ \begin{matrix} + a_{4} \log U N E M P_{t j} + a_{5} \log A S S O C I A T E S_{t j} + a_{6} \log P C I_{t j} \\ + a_{7} \log Y N G K I D S S %_{t j} + a_{8} \log O T H E R K I D S %_{t j} + ε_{t} \end{matrix} \end{array}

(4)

\begin{array}{l} \begin{matrix} \log C O N S P C_{t j} = a_{0} + a_{1} \log C O N S P C_{t ​ - ​ 1 j} + a_{2} \log C I G T X_{t j} + a_{3} \log P O V_{t j} \end{matrix} \\ \begin{matrix} + a_{4} \log U N E M P_{t j} + a_{5} \log B A C H E L O R S_{t j} + a_{6} \log P C I_{t j} \\ + a_{7} \log Y N G K I D S S %_{t j} + a_{8} \log O T H E R K I D S %_{t j} + ε_{t} \end{matrix} \end{array}

(5)

where:

log CONSPC_tj = the natural log of cigarette consumption in year t in state j, measured as the number of packs of cigarettes purchased annually per capita;

log CONSPC_t–1j = the one-year lag of the dependent variable, to serve as a de facto trend variable for the study period;

a₀ = the constant term;

log CIGTX_tj = the natural log of the average total federal plus state cigarette excise tax per pack in year t in state j, adjusted for inflation over the study period;

log POV_tj = the natural log of the percentage of the population in year t in state j that was at or below the federally defined poverty level;

log UNEMPLOY_tj = natural log of the average percentage unemployment rate of the civilian labour force in year t in state j;

log HIGHSCHL_tj = the natural log of the percentage of the population in year t in state j over age 25 that has completed at least a high school diploma;

log ASSOCIATES_tj = the natural log of the percentage of the population in year t in state j over age 25 that has completed at least an Associates (two-year) college degree;

log BACHELORS_tj = the natural log of the percentage of the population in year t in state j over age 25 that has completed at least a bachelors (four-year) college degree;

log PCI_tj = the natural log of per capita income in year t in state j, adjusted for inflation over the study period;

log YNGKIDS%_tj = the natural log of the percentage of the households in year t in state j that had at least one very young, that is, child under the age of five years, present in the home;

log OTHERKIDS%_tj = the natural log of the percentage of the households in year t in state j that had at least one other minor child, that is, child over the age of five years but under the age of 18 years, present in the home;

and

ε_t = stochastic error term.

For the interested reader, descriptive statistics for the variables in the model are provided in Table 1, whereas the correlation matrix among the explanatory variables in Equation (3) is provided in Table 2. As shown in Table 2, overall there is no evidence of serious multi-collinearity. In the eclectic model above, it is expected that:

a_{2} \leq 0, a_{3} \leq 0, a_{4} > 0, a_{5} \leq 0, a_{6} \leq 0, a_{7} \leq 0, a_{8} \leq 0

(6)

The study uses state-level panel data for all 50 states over the period 2003 through 2007, with Washington, DC excluded because it is not a state. Equation (3) was estimated by panel least squares (PLS), first using the random-effects model and then using the fixed-effects model. In this specification, a Hausman specification test (Hausman, 1978) was performed, and it generated a t-statistic with a p = 0.0391, so that the study adopted the fixed-effects model, namely the period fixed-effects model. The model is expressed in log-log form in order to simplify interpretation.

Table 1

Descriptive Statistics

Variable	Mean	Standard Deviation
CONS	75.99	29.56
POV	11.78	2.85
PCI	23,769	2,652
CIGTX	108.1	54.3
UNEMP	5.1	1.1
ASSOCIATES	22.16	3.70
INFANT%	7.28	1.05
KIDS%	31.49	2.51

Table 2

Correlation Matrix among Explanatory Variables in Equation (3)

	CIGTX	POV	UNEMP	ASSOCIATES	PCI	INFANT%	KIDS%
CIGTX	1.00
POV	−0.371	1.00
UNEMP	−0.012	0.361	1.00
ASSOCIATES	0.295	−0.510	−0.329	1.00
PCI	0.256	−0.511	−0.253	0.461	1.00
INFANT%	0.012	0.098	−0.113	−0.073	0.081	1.00
KIDS%	0.079	−0.028	0.289	−0.089	0.006	0.486	1.00

Data for the variable CONSPC were obtained from the Centers for Disease Control (2008, Table 8), using fourth-quarter data for each year in the study period. The data on CIGTX were obtained from Orzechowski and Walker (2008, Table 11). It should be emphasised that the variable CIGTX is the sum of the federal cigarette excise tax and the state excise tax. Adding these two excise taxes reflects the finding by Showalter (1998, p. 1118) that ‘Federal excise taxes per se do not appear to be more effective than state excise taxes in terms of reducing cigarette smoking.’ Hence, we do not differentiate between federal excise taxation of cigarettes and state excise taxation of cigarettes. Data for nominal per capita income were obtained from the US Census Bureau (2004, Table 653; 2006, Table 662; 2009, Table 659). The nominal per capita income, as well as the cigarette tax were adjusted to reflect inflation using the GDP deflator from the US Department of Commerce: Bureau of Economic Analysis (2008, Table 1.1.9). Data for the variable UNEMP were obtained from the US Census Bureau (2004, Table 592; 2006, Table 615; 2009, Table 609). The POV data were obtained from the US Census Bureau (2004, Table 688; 2006, Table 691; 2007, Table 687; 2009, Table 687; 2010, Table 693). Data for computing the educational attainment variables HIGHSCHL, ASSOCIATES and BACHELORS were computed from data from the yearly American Community Survey (2008, Table C15002) and the US Census Bureau (2004, Table 216; 2006, Table 218; 2007; Table 218; 2009, Table 225; 2010, Table 225). Data for the variables YNGKIDS% and OTHERKIDS% were obtained from the US Census Bureau (2003, Table 20; 2004, Table 20; 2006, Table 21; 2007, Table 21; 2008, Table 16; 2009, Table 16).

4. FINDINGS

The period fixed-effects estimations of Equations (3), (4) and (5), after adopting the White period heteroskedasticity correction, are provided in columns (a), (b) and (c) of Table 3, where terms in parentheses are t-values.

In columns (a), (b) and (c) collectively of Table 3, all of estimated coefficients on the non-trend explanatory variables exhibit the expected signs. Of the explanatory variables, 19 are statistically significant the 1 per cent level, three are statistically significant at the 5 per cent level and two statistically significant at the 10 per cent level. All three of the F-statistics are statistically significant at far beyond the 1 per cent level, attesting to the robustness of the model. The coefficient of determination (R²) and adjusted R² values all imply that the model explains more than nine-tenths of the variation in the dependent variable.

Table 3

Period Fixed-effects Estimation Results

Dependent Variable: log CONSPC_tj
Explanatory Variables	Estimate (a)	Estimate (b)	Estimate (c)
Constant	4.364	3.207	3.361
log CONSPC_t–1j	0.847***	0.819***	0.832***
	(30.57)	(38.28)	(44.42)
log CIGTX_tj	−0.198***	−0.197***	−0.194***
	(−11.15)	(−10.21)	(−9.05)
log POV_tj	−0.261***	−0.311***	−0.281***
	(−12.07)	(−19.30)	(27.32)
log UNEMP_tj	0.099***	0.085**	0.091**
	(2.90)	(2.22)	(2.23)
log ASSOCIATES_tj	–	−0.17**	–
		(−1.99)
log BACHELORS_tj	–	–	−0.099*
			(–1.94)
log HIGHSCHL_tj	−0.05*	–	–
	(−1.71)
log PCI_tj	−0.41***	−0.332***	−0.334***
	(−22.74)	(−6.58)	(−3.92)
log YNGKIDSS%_tj	−0.12***	−0.17***	−0.15***
	(−3,81)	(−10.31)	(−12.62)
log OTHERKIDS%_tj	−0.217***	−0.232***	−0.213***
	(−3.80)	(−3.98)	(−4.05)
R ²	0.941	0.943	0.942
Adj. R²	0.938	0.939	0.938
F	273.83***	280.99***	276.88***

Notes: ***statistically significant at the 1% level; **statistically significant at the 5% level; *statistically significant at the 10% level.

The elasticity values on the federal plus state cigarette excise tax variable (CIGTAX), which from −0.194 to −0.198, are negative and statistically significant at the 1 per cent level. This indicates that the higher the cigarette excise tax, the lower the aggregate consumption of cigarettes. In particular, a 10 per cent across-the-board (all states) increase in the cigarette excise tax reduces aggregate cigarette consumption by 2 per cent.³ As for other results, cigarette smoking is found to be a decreasing function of poverty, per capita income and higher educational attainment while being an increasing function of the unemployment rate. Finally, all six of the coefficients on the variables YNGKIDS% and OTHERKIDS% exhibit negative elasticities and are statistically significant at the 1 per cent level in all cases. Thus, these findings are strongly supportive of the two basic hypotheses being investigated in this study and thereby provide further insight in the behaviour of cigarette smokers.

Of course, a reasonable question to raise is whether these findings have actual or at least potential implications in terms of public policy implications. Clearly, influencing, or attempting to do so, the number of minor-aged children living at home is not a plausible policy. However, there is an indirect and potentially important policy implication of these findings that does warrant acknowledgement, at least for the US. Namely, the results from this study raise the question that since the phenomenon investigated in this study has been effectively ignored in previous related studies involving public policies such as cigarette excise taxation and cigarette smoking bans, that those studies may suffer from omitted variable bias. To the extent that this bias is relevant, the dependability and accuracy of these earlier policy studies might be questioned. Indeed, the results from these earlier studies arguably must be regarded with a degree of caution.

Footnotes

Notes

References

Adda

, & Cornaglia

(2006). Cigarette taxes, cigarette consumption, and smoking intensity. The American Economic Review, 96(4), 1013–28.

Anderson

L.R.

, & Mellor

J.M.

(2008). Predicting health behaviors with an experimental measure of risk preference. Journal of Health Economics, 27(4), 1260–74.

Cebula

R.J.

(2010). A preliminary panel data analysis of the consumption impact of cigarette taxation. Journal of Public Finance and Public Choice, 28(3), 39–59.

Centers for Disease Control. (2008). Cigarette consumption, state monitoring system. Retrieved 18 January 2012, from http://apps.nccd.cdc.gov/StateSystem/stateSystem.aspx?selectedTopic=602&selectedMeasure=10005&dir=leg_report&ucName=uclegsmkfreesummaryexcisetax&year=2006_4&excel=htmlTable&submitBk=yhttp://www.cdc.gov/mmwr/preview/mmwrhtml/mm5745a3.htm

Centers for Disease Control and Prevention. (2009a). Health, United States, 2009. Retrieved from http://www.cdc.gov/nchs/data/hus/hus09.pdf

Centers for Disease Control and Prevention. (2009b). Morbidity and Mortality Weekly Report. Retrieved from http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5745a3.htm

Chaloupka

F.J.

(2004). The impact of nicotine replacement therapies on cigarette demand, Journal of Economics and Finance, 28(3), 345–369.

Chaloupka

F.J.

, & Warner

K.E.

(2000). The economics of smoking. In Culver

Anthony J.

& Newhouse

Joseph P.

(Eds), Handbook of health economics (pp. 1539–627). Amsterdam: Elsevier Science, North-Holland.

Connelly

, Goel

, & Ram

(2009). Demand for cigarettes in the U.S.: Effects of prices in bordering states and contiguity with Mexico and Canada. Applied Economics, 41(18), 2255–60.

10.

Cox

R.A.

(2010). Will high cigarette taxes make smokers quit? State Tax Notes, 56(13), 1005–1006.

11.

DeCicca

, Kenkel

, & Mathios

(2002). Putting out the fires: Will higher cigarette taxes reduce the onset of youth smoking? Journal of Political Economy, 110(1), 144–69.

12.

Drayton

(1972). The tar and nicotine cigarette tax: Pursuing public health through cigarette tax incentives. The Yale Law Review, 81(8), 1487–516.

13.

Evans

W.N.

, & Farrelly

M.C.

(1998). The compensating behavior of smokers: Cigarette taxes, tar, and nicotine. RAND Journal of Economics, 29(3), 578–95.

14.

Farrelly

M.C.

, Nimsch

C.T.

, Hyland

, & Cummings

K.M.

(2004). The effects of higher cigarette prices on tar and nicotine consumption in a cohort of adult smokers. Health Economics, 13(1), 49–58.

15.

Forster

, & Jones

A.M.

(2001). The role of tobacco taxes in starting and quitting smoking: Duration and analysis of British data. Journal of Royal Statistical Society, 164(3), 517–47.

16.

Goel

R.K.

(2008). Unemployment, insurance, and smoking. Applied Economics, 40(20), 2593–99.

17.

Goel

R.K.

(2009). Smoking restrictions and spillovers across jurisdictions. Applied Economics Letters, 6(16), 1587–92.

18.

Hausman

J.A.

(1978). Specification tests in econometrics. Econometrica, 46(6), 1273–91.

19.

Holt

C.A.

, & Laury

S.K.

(2002). Risk aversion and incentive effects. The American Economic Review, 92(5), 1644–55.

20.

Koch

J.V.

, & Cebula

R.J.

(1992). A note on smoking health costs and public policy. American Journal of Economics and Sociology, 41(2), 159–60.

21.

Lien

D.S.

, & Evans

W.N.

(2005). Estimating the impact of large cigarette tax hikes: The case of maternal smoking and infant birth weight. The Journal of Human Resources, 40(2), 373–92.

22.

Lovenheim

M.F.

(2008). How far to the border? The extent and impact of cross-border casual cigarette smuggling. National Tax Journal, 61(1), 7–32.

23.

Meier

K.J.

, & Licari

M.J.

(1997). The effect of cigarette taxes on cigarette consumption. American Journal of Public Health, 87(7), 1126–30.

24.

Orzechowski

, & Walker

(2008). Tax burden on tobacco. Retrieved 15 December 2011, from http://www.tobaccoissues.com/getobject.aspx?folder=root/tobaccoissues/excisetaxes&name=Taxburden_2008_vol_43.pdf

25.

Showalter

M.H.

(1998). The effect of cigarette taxes on cigarette consumption. American Journal of Public Health, 88(7), 1118–19.

26.

Tan

A.K.G.

(2012). Distinguishing between non-smokers, casual smokers, and compulsive smokers. Atlantic Economic Journal, 40(2), 173–84.

27.

Tauras

J.A.

(2006). Smoke-free air laws, cigarette prices, and adult cigarette demand. Economic Inquiry, 44(2), 333–42.

28.

Thun

M.J.

, Lally

C.A.

, Flannery

J.T.

, Calle

E.E.

, Flanders

W.D.

, & Heath

C.W.

(1997). Cigarette smoking and changes in the hystopathology of lung cancer. Journal of the National Cancer Institute, 89(21), 1580–86.

29.

US Census Bureau. (2003). Statistical abstract of the United States, 2003. Washington, DC: US Government Printing Office.

30.

US Census Bureau. (2004). Statistical abstract of the United States, 2004. Washington, DC: US Government Printing Office.

31.

US Census Bureau. (2006). Statistical abstract of the United States, 2006. Washington, DC: US Government Printing Office.

32.

US Census Bureau. (2007). Statistical abstract of the United States, 2007. Washington, DC: US Government Printing Office.

33.

US Census Bureau. (2008). Statistical abstract of the United States, 2008. Washington, DC: US Government Printing Office.

34.

US Census Bureau. (2008). American Fact Finder, the Census Bureau’s Data Website. Retrieved from http://factfinder.census.gov/servlet/DatasetMainPageServlet?_program=ACS&_submenuId=&_lang=en&_ts=

35.

US Census Bureau. (2009). Statistical abstract of the United States, 2009. Washington, DC: US Government Printing Office.

36.

US Census Bureau. (2010). Statistical abstract of the United States, 2010. Washington, DC: US Government Printing Office.

37.

US Department of Commerce: Bureau of Economic Analysis. (2008). ‘Implicit price deflators for gross domestic product’. Retrieved 17 January 2012, from http://www.bea.gov/national/nipaweb/SelectTable.asp?Selected=N

38.

White

N.M.

(2010). New York governor approves highest cigarette tax in nation. State Tax Notes, 56(13), 1007.