Cover Image


View/Hide Left Panel

ures. We chose the year 2002 as our baseline year, which also had the advantage that there were no large changes in policies in recent years (since the large 1998–1999 price changes).

The basic SimSmoke model involves a population model, a smoking model, and policy modules. Following a discrete first-order Markov process, the entire population evolves through birth and death rates, and the number of smokers, never-smokers, and ex-smokers evolves through initiation, cessation, and relapse rates. Tobacco control policies change initiation and cessation rates through individual policy modules. Consequently, smoking rates over time depend on tobacco control policies and prior smoking patterns.

The version of the model used in this report is built on an Excel platform. This section presents a brief description of that model and a discussion of future policy scenarios. The data sources are summarized in Table J-1. A mathematical formulation and further description of the model can be found at pire_profile.pdf and referenced papers.

Population Model

SimSmoke is built first on a demographic model. The population, distinguished by age, starts in the year 2002. The population evolves over time with fertility (leading to births) and some portion of the population dying at each age. We do not consider immigration or changes in racial or ethnic composition for the purposes of this model.

Mathematically, the total population (Pop) is distinguished by time period (T) and age (A) (and is further distinguished in the model by gender and racial ethnic group). Mortality rates (MR) are distinguished by age and gender. The number of newborns depends on first-year death rates and fertility rates (Fert) of females by age, with equal birth rates for males and females. Births through the first year (age 0) for each gender are:

PopT,0 = 0.5*(1 − MR0)* ΣA (PopT,A,1 * FertA), where T = 1 … 20; A = 14 … 49

After the first year, the population evolves as:

PopT,A = PopT-1,A-1 * (1 − MortRateA)

Population data are obtained from the 2000 Census of Population, and projected forward to 2002. Fertility rates are from the U.S. Census Vital Rate Inputs Tables for the year 2002. Mortality rates are from the 2001 Multiple Cause-of-Death File compiled from death certificates, by the National Center for Health Statistics (NCHSU; The file includes information on all deaths in the United States in 2001.

Smoking Model

SimSmoke next divides the population in the base year into: (1) never-smokers, (2) smokers and (3) 16 categories of ex-smokers (n = 1 … 16+), corresponding to years since last smoking. After the base year, individuals are classified as never-smokers (designated by NS) from birth until they initiate smoking or die, according to:

NeversmokersT,A = NeversmokersT-1,A-1 * (1 − MortRateA,NS)*(1 − Initiation RateA)

Through age 24, the number of smokers (designated by S) is tracked as:

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement