GRUBER: EXPERT REPORT ON FORWARD-LOOKING REMEDIES TO COMBAT YOUTH SMOKING

March 21, 2005 11:48 pm by Gene Borio

EXCERPT

In this report, I have developed a remedy to reduce actions by tobacco manufacturers that attract youth to smoking. This outcome-based remedy provides a strong financial incentive to manufacturers to reduce youth use of their product, by charging each manufacturer an assessment at least as large as the money they would make from attracting a youth to their product. Therefore, this remedy can play a vital role in ensuring that the tobacco manufacturers take affirmative actions to reduce actions which attract youth to their products.

END EXCERPT

Text of PDF file of Gruber Expert Report follows:

**———————————————————

EXPERT REPORT ON FORWARD-LOOKING REMEDIES TO COMBAT YOUTH SMOKING

Jonathan Gruber

Professor of Economics, MIT

Research Associate, National Bureau of Economic Research

**———————————————————

1

I. QUALIFICATIONS

1. My name is Jonathan Gruber and I am a Professor of Economics at the Massachusetts Institute of Technology, where I have worked for twelve years. I am recognized as a leading expert in the fields of health economics and public economics. I am a member of the Institute of Medicine, was the recipient of the 1995 American Public Health Association Kenneth Arrow Award for the Outstanding Health Economics Paper of 1994, and was one of 15 scientists nationwide to receive the National Science Foundation Presidential Faculty Fellowship in 1995.

2. In 1997-1998, I was the Deputy Assistant Secretary for Economic Policy at the United States Treasury Department, where one of my primary responsibilities was work on tobacco issues. I played a lead role in drafting the Administration’s response to the June 1997 Proposed Resolution, and was a key member of the Administration team working with Senator John McCain to draft his comprehensive tobacco legislation. During that time I also worked on a wide variety of other issues, including global warming, children’s health insurance, personal bankruptcy, electricity deregulation, and the student loan program.

3. Since returning to MIT, I have been on the editorial board of the Journal of Health
Economics, the leading journal in the field of Health Economics, and the Journal of
Public Economics, the leading journal in the field of Public Economics. I also direct a research program on children’s economic issues at the National Bureau of Economic Research, a nonpartisan economic research organization.

4. My past research has focused on a variety of topics in these fields, including but not

**———————————————————

2

limited to: the impact of health insurance on the labor market; the impact of public insurance programs and tax credits for insurance on insurance markets, health care utilization, and health; the impact of the Unemployment Insurance program on the wellbeing of unemployed workers; the impact of the Disability Insurance program on the labor supply of older workers; the impact of the Social Security program, both in the U.S. and around the world, on retirement; and the economics of tobacco regulation. I have written over 90 scholarly articles, edited four volumes, and written a textbook.

II. OVERVIEW OF REPORT

5. The goal of this report is to discuss a remedy to reduce actions by tobacco manufacturers that attract youth to smoking. The design of this remedy begins with the recognition that tobacco companies can influence youth smoking through a wide variety of avenues, including but not limited to the list price of their products, discounts off that price, other promotions, advertising in print and other media, and the truthfulness of the health information provided about smoking. Since it would be difficult for the Court to monitor each and every individual avenue of influence comprehensively, it is therefore difficult to use an “input-based” strategy to reduce manufacturers’ efforts to attract youths to their product. A more effective route to prevent actions that attract youth to smoking is by using an “outcomes-based” incentive.

6. This report outlines the design of such an incentive. The starting point for this remedy is the targeted reduction in youth smoking agreed to by the major tobacco manufacturers (Philip Morris Incorporated, R.J. Reynolds Tobacco Company, Brown and Williamson Tobacco Corporation, Lorillard Tobacco Company, United States Tobacco Company) as part of the June 1997 Proposed Resolution (which was never implemented). I develop

**———————————————————

3

targeted reductions from 2003 youth smoking levels to achieve those same ultimate goals, albeit on a slower timetable. The incentive to meet the targets specified in this report is created by imposing a financial assessment on manufacturers if they miss these targets.1 The assessment is designed to reflect the future amount that a tobacco manufacturer would make by recruiting a youth to smoke their product. In this way, the monetary assessment removes any incentive for a tobacco manufacturer to attract youth to their products.

7. If youth smoking is reduced to the level proposed as part of their June 1997 Proposed Resolution, then tobacco manufacturers will not be subject to any monetary assessment.

If, however, youth smoking is not reduced to that level, then manufacturers will pay a monetary assessment in order to remove the amount they could have made from attracting youth.

III. TARGETS

8. The starting point for the design of this outcomes-based remedy is a set of targets for reductions in youth smoking that must be met by tobacco manufacturers. This section lays out the details of such targets.

Goal: Match the Targets in the 1997 Proposed Resolution

9. In the June 1997 Proposed Resolution with the states, the tobacco manufacturers— including defendants in this case—agreed to a mechanism which would penalize them to the extent that youth smoking exceeded certain target levels. These levels were based on reductions from the base level of youth smoking over the 1986-1996 period, as measured

—-

1 Details of the calculations can be found in the Technical Appendix.

**———————————————————

4

by the Monitoring the Future (MTF) survey (or equivalent). From those base levels, tobacco manufacturers faced targeted youth reductions of 30% by year 5, 50% by year 7, and 60% by year 10. In each year that youth smoking exceeded the target level, there would be an assessment. For example, after 10 years, if youth smoking exceeded 40% of its 1986-1996 average level, there would be assessments levied on the tobacco manufacturers.

10. Since the tobacco manufacturers voluntarily agreed to these targets as part of the Proposed Resolution, I use them as a basis for designing a forward-looking remedy. In particular, I compute the further reduction—based on 2003 smoking levels—that would be necessary to meet the agreed upon targets found in the Proposed Resolution.

11. During the period from 1986 to 1996, the proportion of youth who on average smoked at least one cigarette per day, as measured by MTF, was 15.20%.2 In 2003, this figure was 10.69%.3 This means that by 2003 there had already been a reduction of approximately 30% relative to the 1986-1996 baseline. But it also means that additional reductions are still needed to meet the targets laid out in the Proposed Resolution. In particular, to reduce youth smoking to 60% below its 1986-1996 baseline, an additional reduction of 43% from 2003 levels is required. This forms the basis for the targets.

—-

2 Based on the requirements specified in the Proposed Resolution, the base percentage of youth who, on average,
smoked at least one cigarette per day is the average of the percentage for the following three age groups, weighted
by relative population size of such groups in 1995 as determined by the U.S. Census Bureau: (a) average percentage
of 12th graders (ages 16 and 17) smoking, on average, at least one cigarette per day from 1986 to 1996 MTF; (b)
average percentage of 10th graders (ages 14 and 15) smoking, on average, at least one cigarette per day from 1991 to
1996 MTF; and (c) average percentage of 8th graders (age 13) smoking, on average, at least one cigarette per day
from 1991 to 1996 MTF.

3 Based on the requirements specified in the Proposed Resolution, the 2003 percentage of youth who, on average,
smoked at least one cigarette per day is the average of such percentage for 12th graders, 10th graders, and 8th graders
from the 2003 MTF, weighted by relative population size of such groups in 1995 as determined by the U.S. Census
Bureau.

**———————————————————

5

Data

12. The data for measuring the performance of tobacco manufacturers relative to these targets, going forward, is the National Survey on Drug Use and Health (NSDUH). This survey was designed in the wake of the proposed McCain legislation to have the type of large sample sizes and brand-specific smoking information necessary to implement an outcomes-based remedy such as the one discussed here.

Manufacturer-Specific Targets

13. The proposed targets and incentives described here apply on a manufacturer-specific basis. That is, each manufacturer will be responsible for reducing the share of youth that are smoking their product by 43% of their 2003 baseline level, by 2013. In this way, each manufacturer has an incentive to undertake whatever actions it has at its disposal to reduce youth use of its product.

14. Manufacturer-specific targets and monetary assessments avoid a flaw inherent in industry-wide targets and monetary assessments, by giving each individual manufacturer an incentive to reduce the number of youths using their products. Industry-wide monetary assessments of the type in the Proposed Resolution do not provide as strong an incentive to reduce youth use of a manufacturer’s product. This is because the monetary amounts are apportioned across all manufacturers; hence, any one manufacturer would bear only a portion of the monetary amount for attracting a youth to smoke. This leads to a “free rider” problem, whereby manufacturers, particularly small manufacturers, can increase youth use of their product, knowing that they will bear only a share of the industry-wide assessment (potentially only a very small share) for missing the youth smoking targets. This share of the assessment may well be below the full benefits of

**———————————————————

6

having a consumer dedicated to your product for many years.

Details of Targets

15. The proposed pattern of targets is laid out in Table 1 of this report. Manufacturers would have no required reduction until 2007. Youth smoking would then be targeted to fall by 6% from 2003 levels, by the end of 2007. The targeted reductions from 2003 levels would then increase by 6% each year until reaching 42% by the end of 2013. The targeted reduction would remain at 42% below 2003 levels from 2014 onwards (a slightly smaller reduction amount than the target suggested in the original Proposed Resolution).

16. The next row of Table 1 shows how these reductions compare to the base in the 1997 Proposed Resolution. In 2003, actual youth smoking had fallen by 30% from the 1986- 1996 base. By the end of 2007, the targeted reduction rises to 34% of the 1986-1996 base, and by the end of 2013, the targeted reduction of 59% is essentially the same as the 60% reduction in the Proposed Resolution.

17. This display in Table 1 allows for a ready comparison of the timetable proposed in this report and the timetable included in the 1997 Proposed Resolution, which is shown in the final row of Table 1 (assuming that legislation enacting the Proposed Resolution had passed, which it did not). The timetable given in this report requires considerably slower reductions by the tobacco manufacturers. For example, under the Proposed Resolution, manufacturers would have been required to have reduced youth smoking by 60% by 2008, whereas under the schedule developed here the reduction in 2008 is only 38% off of the same base. Manufacturers have five extra years to reach roughly the same ultimate

**———————————————————

7

target.

18. These targets represent the percentage of youths who on average smoked at least one cigarette per day. In each year the NSDUH asks how many days a youth smoked in the past month, and the number of cigarettes smoked per day on those days.4 For each youth who on average smoked one or more cigarettes per day, the NSDUH survey can be used to identify their most frequently consumed brand of cigarette, and these brands can then be combined to assess the share of youth that smoke cigarettes from each manufacturer that is a defendant in this case.

19. To determine the target youth smoking rate for each manufacturer starting in 2007, the target reduction in that year is applied to the manufacturer-specific youth smoking rate calculated from the 2003 NSDUH. The actual youth smoking rate for each manufacturer from 2007 onward (used to see how well the manufacturer did compared with the targets) can also be calculated using the NSDUH from that particular year.

Age Range

20. The targeted reductions in youth smoking, and the assessments for missing those targets, will apply to youth ages 12-20 years old. This age range is chosen because, as the United States alleges in their RICO case, the defendants have fraudulently denied that their marketing affects smoking behavior, including initiation (which happens primarily in adolescence) and has fraudulently promised the public that they do not market to young people including those under 21, while continuing to market to those under 21 in order to sustain and increase the market for cigarettes. Therefore, the remedy described here

—-

4 The product of the information from these two questions, divided by the number of days in the month, is a measure
of whether, on average, the youth smoked one or more cigarettes per day.

**———————————————————

8

applies to those age 12-20, inclusive.

IV. INCENTIVES

21. If a manufacturer misses its target, that is, if the percentage of youth smoking their products declines by less than the targets presented in Table 1, then that manufacturer will pay an assessment. In this section, I discuss the structure of such assessments. The assessments are designed so that tobacco manufacturers no longer gain from attracting youth to their product. I then present an illustrative calculation of such assessments for future years.

Number of Misses

22. The computation of the assessment begins by computing the number of youth smokers by which each manufacturer misses its target. This is done by multiplying the percentage of youth by which the manufacturer misses their target, times the size of the youth population. The size of the youth population from age 12 to 20 can be measured using population estimates from the U.S. Census Bureau. For example, if 5% of youth smoked the manufacturer’s brands of cigarettes in 2003, then this remedy requires that this share fall to 4.7% by 2007. Suppose that, in fact, 5% of youth continue to smoke that manufacturer’s brands of cigarettes. Then the manufacturer will have missed their target by 0.3% of youth. If there are 40 million youth in the age range 12-20, then the manufacturer has missed its target by 120,000 youth smokers.

Assessments

23. Having computed the number of youth smokers by which a manufacturer has missed its target, the next step is to compute the assessment to be paid for each of these misses. The

**———————————————————

9

goal of this forward-looking remedy is to remove any incentive for the manufacturer to attract youth to their product. Thus, the assessment must be large enough that manufacturers have no incentive to promote youth to use their product. An assessment of $2,4005 per youth smoker (by which the target is missed) is sufficiently large to remove the profit motive for attracting that youth; this is because $2,400 is larger than the present discounted value of the amount made from cigarette sales over the lifetime of that smoker.

24. The computation underlying this conclusion proceeds in several steps, which are reviewed here. First, for any current youth smoker, compute the probability that the smoker will still be alive and smoking in each future year. The odds that a current youth smoker would quit in each future year can be calculated using the 2001-2003 National Health Interview Survey (NHIS), a nationally representative survey of the civilian noninstitutionalized population in the United States. The odds that a current youth smoker would still be alive in each future year can be calculated using mortality rates computed from the 2001 life tables provided by the Trustees of the Social Security Administration and relative risks of mortality for smokers computed from the Cancer Prevention Study (CPS) data.6 These various data sources can be used to compute, for an individual who on average smoked at least one cigarette per day as a youth, the odds that that individual will be alive, and if so will still be smoking, at every future age.

25. Second, for each of those future years that the youth will be alive and smoking, it is necessary to compute the intensity with which they will smoke. This can also be done

—-

5 The $2,400 is in 2007 dollars. The actual assessment amount in a particular year will be indexed from the 2007
amount.

6 The mortality rates and relative mortality rates were provided by Dr. Timothy Wyant of Decipher, Inc.

**———————————————————

10

using the 2001-2003 NHIS data, computing the intensity of smoking at each age, conditional on having smoked as a youth.7 Multiplying this by the probability of smoking, gives a measure of the expected total cigarettes per year that youth smokers will be smoking in each future year.

26. The third step is to compute the amount that manufacturers make on each cigarette smoked in each future year.

27. Finally, multiply the amount made per cigarette by the cigarettes smoked in each future year to obtain the amount made in each future year. These amounts are then discounted back to today’s dollars.

Avoiding Double-Counting

28. While the calculation described above is conceptually straightforward, it has the limitation that, if the assessment were paid on each youth smoker by which the targets were missed in each year, youth smokers who smoke in more than one year would be double-counted. For example, consider someone who smokes at age 12 in 2008 and again at age 13 in 2009. The method above could be used to compute the present discounted value of the amount manufacturers would make from that person, given that they smoke as a youth. But if the assessment were set equal to that amount in each year, the manufacturer would end up paying twice the present discounted value of future amounts – once when the youth is 12, and again when the youth is 13.

29. Therefore, in computing the assessment in years starting from the second year in which

—-

7 The 2001-2003 NHIS surveys only contain smoking intensity information for individuals age 18 and above.
Smoking intensity at age 18 is extrapolated to produce smoking intensity at younger ages based on the smoking
trend among youths in the 1992 Youth Risk Behavior Supplement.

**———————————————————

11

assessments are levied, the assessment amount is only for new misses from the target: the net change in the stock of youth smokers, adjusting for the natural attrition as youths move from age 20 to age 21. That is, if 2007 is the first year in which assessments are levied, then from 2008 onwards, the assessments are only on the net increase in the number of misses over the previous year, accounting for natural attrition.8 The assessment applies only to this net increase in the number of misses, in order to avoid double-counting.

30. The following example illustrates how this adjustment works. Suppose that in 2008 a manufacturer missed their target by 1 million youth smokers. Suppose that in 2009 they also missed by 1 million youth smokers. Finally, suppose that 20% of that manufacturer’s youth smokers in 2008 were 20 year olds. This implies that the 1 million by which the target is missed in 2009 includes 800,000 of those by which the target was previously missed, plus 200,000 new misses (to replace the 200,000 who became 21 years old). So the assessment would only be applied on the 200,000 new misses in the year 2009.

Results on Assessments

31. Table 2 presents an illustrative calculation of the types of assessments that would be imposed under this structure. For this illustrative calculation I assume that the youth population is constant at 40 million (row 1), and that the actual youth smoking rate stays at 12.5% (row 2). The target rate is shown in row 3, and the number of misses is shown in row 4. This is computed by the difference between the actual smokers and the target

—-

8 Natural attrition is the percentage of age 20 youth smokers in the total youth smoker population in the previous
year.

**———————————————————

12

rate, multiplied by the size of the youth population. Row 5 shows the new misses, after accounting for double-counting. Row 6 shows the assessments per miss, and row 7 shows the total payments.9

32. Under this structure, if there were no reduction in youth smoking, tobacco manufacturers would have payments that rise to a peak of $1,584 million in 2013, and then fall to $1,008 million thereafter. The decline is due to the double-counting adjustment, which removes payments for previous youth smokers on whose behalf assessments were already paid.

IV. CONCLUSIONS

33. In this report, I have developed a remedy to reduce actions by tobacco manufacturers that attract youth to smoking. This outcome-based remedy provides a strong financial incentive to manufacturers to reduce youth use of their product, by charging each manufacturer an assessment at least as large as the money they would make from attracting a youth to their product. Therefore, this remedy can play a vital role in ensuring that the tobacco manufacturers take affirmative actions to reduce actions which attract youth to their products.

—-

9 All of the monetary amounts in this table are given in 2007 dollars.

One Response to “GRUBER: EXPERT REPORT ON FORWARD-LOOKING REMEDIES TO COMBAT YOUTH SMOKING”

  1. tobacco observer Says:

    “Therefore, this remedy can play a vital role in ensuring that the tobacco manufacturers take affirmative actions to reduce actions which attract youth to their products.”
    ****

    Since marketing cigarettes to youth (especially “youth” over 18 who are legally entitled to smoke) isn’t actually a violation of RICO, or any other law, this remedy is impermissible and irrelevant.

    Nice try, though.

Leave a Reply

The primary purpose of this site is to provide information in a timely manner. Postings should be informative. The usual rules apply: No libel, no profanity, no personal abuse, keep it on topic, and short.

If you are scheduled as a court witness, CHECK with your lawyer before posting anything here!