andrew friedson and daniel rees teach economics at the University of Colorado, Denver.
Published August 25, 2020
State excise taxes on cigarettes have been trending upward to levels that would have been unthinkable when the Marlboro Man roamed the range and more doctors smoked Camels than any other brand. In 1980, the average per-pack tax was $0.21; by 2017, it had reached $2.61. Even after adjusting for inflation, this represents a five-fold increase, and the evidence is pretty conclusive that it contributed to one of the most important public health achievements of the past century: the steady decline in U.S. adult smoking rates, which continues to this day.
Teenagers seem to be especially sensitive to cigarette taxes, presumably because they lack the ready cash for impulse purchases and may not yet have settled into hard-core nicotine addiction. But it is not clear whether — and for how long — this sensitivity lasts. In 2019, Illinois increased its per-pack tax by $1.00. Will Illinois teenagers who were initially discouraged from smoking because of this increase eventually take up the habit? Ten years from now, when most of them have full-time jobs and have grown accustomed to paying $8 for an IPA at the corner bar, will they still smoke less than they would have, had Illinois not increased its tax?
This is, of course, a crucial question: If cigarette tax increases merely delay the onset of smoking, their public health benefits have been grossly overestimated.
What We Did
We measured the effect of cigarette taxes experienced by teenagers on their smoking behavior later in life. Our data came from the University of Michigan’s widely used Panel Study of Income Dynamics (PSID), which has been following a large group of Americans for half a century.
In 1986, the PSID added a battery of smoking-related questions. Respondents were asked, among other questions, how many cigarettes they smoked in a usual day leading up to being interviewed and at what age they began smoking. Combining their answers with data on state cigarette taxes from 1970 through 2005, we estimated the effects of taxes at ages 12-19 on smoking participation and intensity up through 2017, when the PSID respondents were 30 through 66 years of age.
Adjusting for the influences of age, gender, race, ethnicity and educational attainment, all of which could potentially be confounding factors, we leveraged within-state changes in real (i.e., adjusted for inflation) cigarette taxes.
Think of comparing the smoking of two PSID respondents as adults, both of whom lived in the same state as teenagers but were exposed to different cigarette taxes because they were born, say, a decade apart. During the period 1970-2007, there were a total of 482 changes in state cigarette taxes stemming from legislation. Most of these changes were modest, and might well have been lost in the background noise. But not all of them: there were 82 changes in the per-pack cigarette tax between $0.50 and $1.00 and 20 changes between $1.00 and $1.74.
Pricing teenagers out of the market makes them less likely to bear the financial and health burdens of smoking later on.
We found strong evidence that taxes faced by teenagers did, in fact, make a difference later in life. In general, taxes experienced at ages 14-16 had the biggest impact. For instance, we found that a $1.00 increase in the cigarette tax at age 14 led to an 8.6 percentage point reduction in the probability of smoking participation in 2017, while a $1.00 increase at age 16 led to a 5.3 percentage point reduction in the probability of participation.
We also examined the effect of cigarette taxes experienced by teenagers on smoking intensity. Again, taxes experienced early in life — at ages 14-16 — appear to have had the biggest impact. For instance, a $1.00 increase in the cigarette tax at age 14 led to a 7.8 percentage point reduction in the probability of smoking one pack per week in 2017, while a $1.00 increase at age 16 led to an 8.0 percentage point reduction in the probability of smoking two packs per week in 2017.
Finally, we examined the effect of cigarette taxes experienced by teenagers on smoking among mothers. Smoking during pregnancy in the United States has been slowly trending down for the past several decades. Nevertheless, many women still smoke for some portion of their pregnancy, risking premature delivery, low birth weight and congenital defects.
The PSID Childbirth and Adoption History File (CAHF) contains complete childbirth and adoption histories for PSID respondents surveyed between 1985 and 2017. We identified the first birth of every female head of household in the survey file, along with her state of residence when she gave birth to her first child, and the year in which the birth occurred. Then, using the answers to the smoking questions asked in 1986 and 1999-2017, we constructed two outcomes: an indicator for whether the mother smoked during the year in which her first child was born and an indicator for whether she smoked during the year before that first birth. (This circuitous approach was necessary because the PSID does not provide information on the month in which respondents began smoking or the month in which they quit, and thus we could not specifically measure smoking during pregnancy.)
We found that cigarette taxes experienced at ages 13-14 were quite effective at reducing the incidence of smoking among mothers later on. For instance, a $1.00 increase in the cigarette tax at age 13 led to a 0.057 reduction in the probability of smoking the year in which the birth occurred, and a 0.078 reduction in the probability of smoking the year before.
Having and Eating the Proverbial Cake
Economists have long been of two minds about cigarette taxes as a deterrent to smoking. On the one hand, it’s pretty clear that the taxes help spare a lot of people a lot of heartache. On the other hand, the taxes are regressive: most people who continue to smoke nonetheless have modest incomes, and a serious tobacco habit hits them hard in the wallet.
Our results about the delayed impact of cigarette taxes, then, can be thought of as good news on two fronts. First, cigarette taxes experienced by teenagers really do shift the lifetime trajectory of smoking behavior. This shift does not just occur on the “will I or won’t I” dimension of smoking; it is reflected in how much and for how long people smoke.
But there’s a bonus here, too, for anyone worried that the collateral damage from an otherwise worthwhile public policy is disproportionately borne by lower-income households. Cigarette taxes may not be so regressive after all when the distributional consequences are measured over lifetimes. After all, pricing teenagers out of the market makes them less likely to bear the financial and health burdens of smoking later on.