THE DEPUTY SECRETARY OF THE TREASURY                                                          WASHINGTON                                                        May 12, 1998

The Honorable Orrin G. Hatch
Chairman
Committee on the Judiciary
United States Senate
Washington, D.C. 20510

Dear Mr. Chairman:

At my testimony of April 30, you asked for a more detailed presentation of the Treasury analysis of Senator McCain's bill (S. 1415) and other tobacco legislation. You specifically asked for information on how the following eleven issues are treated in our analysis: state and local excise taxes, smuggling, markups, youth lookback penalties, elasticity assumptions, non-price effects of the settlement, attorneys' fees, the effect of state settlements, liability exposure of the companies, export charges and the functional form of our demand specification.

Following this letter is an analysis that responds to your request. We have included a detailed technical description of our methodology, providing a step-by-step analysis of how we draw our conclusion that the McCain bill will raise prices to $3.19, in real terms, by the year 2003 and in so doing, addresses each of the specific issues you raised.

I hope that this letter makes clear the Administration's detailed analysis of the implications of S. 1415. I also hope that we can work together towards our jointly held goal of passing the tobacco legislation that will most effectively reduce youth use of tobacco in America.

Sincerely,

/s/

Lawrence H. Summers

Enclosure

cc: Ranking Member Patrick J. Leahy

Measuring the effect of comprehensive tobacco legislation on the price of cigarettes in the U.S. is an analytically difficult task that incorporates a large number of uncertainties. It is impossible to measure the effect of such legislation without making a number of assumptions and projections about an environment that is without historical parallel in the U.S. This document lays out in specific detail the steps that the Administration took in assessing the effect of S. 1415 on cigarette prices in the U.S. between 1999 and 2003.

This report is organized as follows. Section I describes the underlying demand analysis that translates price increases into quantity reductions. In Section II we discuss the Administration's baseline forecast of cigarette consumption and prices. This is important because our price estimate will depend in part on current law levels of cigarette consumption and prices. In Section III we discuss in some detail our assumptions about behavior under S. 1415, assumptions that influence how prices respond to comprehensive legislation.

I. The Demand for Cigarettes

S. 1415 legislates a fixed stream of payments from 1999 to 2003. In order to convert this fixed stream of payments to a stream of prices per pack, we must assess the relationship between prices and quantities of cigarettes consumed. Given this relationship, along with an assumption about the baseline path of prices and quantities, we can estimate the incremental price per pack (and the number of packs) that correspond to this stream of fixed payments.

The underpinnings of the Administration's analysis build on recent empirical work in the academic literature and use a semi-logarithmic demand function.¹ Parameters of this demand function are chosen to be consistent with an initial elasticity estimate, as well as with the forecast of federal cigarette excise tax revenue in the Administration's FY 1999 budget. Our functional form is well suited for analyzing large price changes because it is consistent with the intuition that more casual, less addicted and less enthusiastic smokers are the first to leave the market as the price of cigarettes rises. It specifies that a given increment to price (e.g. 10 cents) generates a constant percentage reduction in demand, which implies that as the price rises, there is a smaller and smaller reduction in the number of packs that are smoked.

There is a fairly wide range of estimates of the responsiveness of smokers to changes in the price of cigarettes. Recent reviews of the literature bracket the range of plausible price

_______________
¹See, for example, Harris, Jeffrey (1994), "A Working Model for Predicting the Consumption and Revenue Impacts of Large Increases in the U.S. Federal Cigarette Excise Tax," Working Paper No. 4803, National Bureau of Economic Research, Cambridge, MA; or Wasserman, Jeffrey et al. (1991), "The Effects of Excise Taxes and Regulations on Cigarette Smoking," Journal of Health Economics 10, pp. 43-64.
_______________

elasticities between -0.25 and -0.50.² That is, a 10 percent increase in the price of cigarettes is likely to reduce the quantity of cigarettes sold by somewhere between 2.5 percent and 5.0 percent. The elasticity that we use is the total elasticity. It reflects the combined effect of changes in consumption due to changes in the proportion of the population that smoke and changes in consumption due to changes in the number of cigarettes smoked by each smoker.

As mentioned above, our demand specification reflects the realistic assumption that the elasticity could change in the face of price increases. The analysis is calibrated so that the initial price elasticity of demand is -0.45 at the current level of cigarette prices. This estimate is consistent with results from high-quality academic studies and is consistent with elasticities used by some Wall Street analysts. A lower initial elasticity estimate would imply smaller settlement induced price changes.

II. Baseline Prices and Cigarette Consumption

In this section, we describe briefly our calculation of the quantity and price of cigarettes which would occur in the absence of S. 1415. This baseline forecast forms the jumping off point for our analysis of S. 1415. Consequently, some understanding of our baseline assumptions are crucial to an understanding of how the world will look should S. 1415 be enacted.

Our baseline forecast begins with a statistically estimated time series on the number of packs of cigarettes sold each year. We then calibrate our revenue estimating demand function (discussed above) to be consistent with this baseline quantity forecast. To do this calibration, however, we require an estimate of the annual price of cigarettes. Our working assumption is that, except for scheduled increases in the federal excise tax on cigarettes, the baseline price of cigarettes will be constant in real terms. This working assumption has implications for how we handle the major components of the price of a pack of cigarettes: federal taxes, state taxes and private cost/ profit margins. We discuss these issues in some detail below.

_______________
² See, for example, Chaloupka, Frank and Henry Wechsler (1997), "Price, Tobacco Control Policies and Smoking Among Young Adults," Journal of Health Economics 16, pp. 3 5 9-3 73; or Congressional Budget Office (1997), The Proposed Tobacco Settlement: Issues From a Federal Perspective, Congressional Budget Office, Washington, D.C.; or Gravelle, Jane (1997), "The Proposed Tobacco Settlement: Effects on Prices, Smoking Behavior, and Income Distribution," CRS Report for Congress No 97 -995E, November 5, Congressional Research Service, Library of Congress, Washington, D.C.; or U.S. Department of Health and Human Services (1994), Preventing Tobacco Use Among Young People: A Report of the Surgeon General, U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta, GA.
_______________

Federal Excise Taxes

The most straightforward component of the baseline price path is federal excise tax, which is scheduled to increase in the year 2000 (a 10 cent increase) and again in the year 2002 (an additional 5 cent increase). But these are nominal increases, which deflate over time. For this reason, the increase in 2002 is only 4 cents in real terms (row 7).

State Excise Taxes

In calculating our baseline revenue stream, we assume that state and local excise tax rates are constant in nominal terms (rows 8 and 18), and so fall over time in real terms. Our assumption is consistent with long-term historical trends; for example, state taxes have declined by about 11% in real terms since 1975. It also is consistent with the standard Treasury policy of refraining from guesses about the evolution of state policy when making baseline estimates. Nonetheless, we recognize that in recent years, states have increased their cigarette excise taxes at a rate slightly higher than inflation.

Our baseline assumption implies that, in the absence of a federally brokered tobacco settlement, states would not separately try to achieve the benefits of the settlement by increasing their excise tax rates, and engaging in additional litigation and further regulatory changes. If states proceeded in this way and these changes were reflected in our baseline, the settlement would simply substitute a federally induced price increase (compared to 1998 prices) for a state induced price increase, leaving the actual price observed in the market largely unaffected. Assuming that states would "reverse engineer" a settlement in the absence of Federal legislation would be contrary to historical evidence and states acting alone cannot put in place the coordinated regulatory and enforcement regimes that are contemplated by the federal legislation.

Manufacturer's Share and Retail & Wholesale Markup

The final component of the price of cigarettes is the private cost/rrofit margin. This margin includes the manufacturer's margin, as well as the retail and wholesale markup. Row 9 shows our assumption that the private margin increases slightly over time. Our assumption is consistent with a modest increase in the real manufacturer's margin, as is suggested by the historical record. For example, over the last decade the real manufactures' prices increased by roughly 2.2 percent. For two reasons, however, the rate of increase shown in row 9 is somewhat lower than this historical rate. One, it may be difficult for manufacturers to further increase their margins in the face of secularly declining cigarette sales. Two, row 9 shows the total private margin, and so includes retail and wholesale margins which seem less likely to grow in real terms.

Putting the Pieces Together: the Baseline Price Series

Putting together the manufacturer's price, markups and state and federal excise taxes results in our assumed baseline price of a pack of cigarettes, shown in row 3. In the absence of legislative changes, we expect the price of cigarettes to be $2.09 per pack in 2003, in constant 1998 dollars.

III. Translating Base Payments into Per-Pack Equivalents: Potential Effects of S. 1415.

Our goal here is to estimate the effect of S. 1415 on the price of a pack of cigarettes. This exercise is made more difficult by the fact that the legislation does not directly specify the per-pack payment it expects of cigarette manufactures. Instead, it specifies a stream of total annual payments that the industry must make. To translate this into the effect per pack of cigarettes requires a number of assumptions about how the market participants will react to this legislation. We have been and will continue to be mindful of the many uncertainties about how an increase of this kind will ultimately translate into retail prices. Because our primary goal in this endeavor is to advance public health through the reduction of teen smoking, we have been conservative in many of our calculations in order not to risk falling short of our public health goals. That is, we have deliberately attempted to avoid over-stating the price increase estimated for S. 1415 in order to avoid over-stating that bill's health benefits.

Payments by Cigarette Industry

The first thing we must do is to determine what share of the total payments called for from all tobacco manufacturers will accrue to cigarette manufacturers. It is only these payments which are likely to form the basis of an increase in the price of cigarettes.

The annual payments required under S. 1415 are shown in row 2. Not all of these payments, however, are likely to be the liability of cigarette manufacturers, as some will accrue to other tobacco manufacturers. Since cigarettes account for roughly 92 percent of dollar sales of tobacco products, our analysis has assumed that 92 percent of the payments made by the industry would fall on cigarettes; that is, price increases on cigarettes must account for 92 percent of the total settlement costs. A lower percentage borne by cigarettes would lead to smaller expected per-pack price increases, while a higher percentage would lead to a larger per-pack price increase.³

Incidence Issues

There is no a-priori reason that the cigarette manufacturers would necessarily pass exactly

______________
³ In its current version, S. 1415 would place more than 92 percent of the payment burden on the cigarette industry. The administration has made technical comments to the S. 1415 that would expand the scope of the proposed legislation to cover other tobacco products, including smokeless, and thus lead to price increases on tobacco products that are more proportional to sales. The Committee's report language supports such changes, at least in part, as it says:

Spit tobacco poses a substantial health risk to youth. It is the Committee's intent that the price of spit tobacco rise commensurately with cigarettes to effectively deter youth consumption. The Committee will work on a further amendment to repair this serious problem with the reported bill.

_______________

100 percent of the settlement cost through into the price of cigarettes. It is possible that part of the payment might be passed backwards onto the owners of cigarette companies, or onto their workers or suppliers. In this case, the price might rise by less than implied had 100 percent of the payment been passed through. Alternatively, the cigarette manufacturing industry might be able to use the payment as a focal point around which to coordinate a profit increasing price rise in excess of that implied by a 100 percent pass through of the settlement payment.

These issues, which relate to the incidence of the payments, are difficult to resolve definitively. They would complicate the analysis even if S. 1415 raised revenue through a relatively simple excise tax. The actual structure of S. 1415, in which the payments are fixed to the industry, but many vary across firms because allocated on the basis of market share, adds additional complications.

We follow many other analysts, however, in assuming the tobacco manufacturers will pass through into price 100 percent of the cost of the settlement. Indeed, the McCain legislation explicitly requires manufacturers to pass on the settlement payments to prices.

Effect on Retail and Wholesaler Markups.

Our incidence assumption is that consumers bear the burden of the payments called for by S. 1415. This means that, in our analysis, retailers and wholesalers do not increase their margins in response to S. 1415. The way that wholesalers and retailers set their markup is as an absolute dollar amount above their costs of business. The competitive nature of the wholesale and retail sectors make it impossible for these groups to increase this dollar amount when the prices of cigarettes rise because of legislative action. In these competitive sectors, any retailer or wholesaler that tried to increase their margins would soon be undercut by a neighboring store or another wholesaler.

This view is shared widely. For example, the FTC, in their analysis of the original Attorney's General settlement, assume in their "baseline", or central, calculation that there would not be any mark-up at the wholesale or retail level of the payments made by manufacturers.⁴ This view also is supported by virtually all of the relevant empirical evidence. A large number of economic studies have examined the impact of cigarette tax increases on the retail price of

_______________
⁴ As the FTC states: "The gains or losses from a pass-through different from 100 percent will accrue entirely to the manufacturing sector, so long as the wholesale and retail distribution of cigarettes is competitive, with distributors obtaining no more than a competitive rate of return for providing their services" (p. 26). Federal Trade Commission, Competition and the Financial Impact of the Proposed Tobacco Industry Settlement. Washington, D.C.: Federal Trade Commission, September 1997.
_______________

cigarettes, and they have uniformly concluded that there is no appreciable markup.⁵

State Response to the Settlement

It is possible that states may respond to the settlement by raising their excise taxes in an attempt to partially recoup lost excise tax revenue. Such a response is consistent with a casual reading of the historical record. However, the proposed legislation contemplates a large "pass back" of settlement receipts to the states that will substantially exceed the lost excise tax revenues. Thus, it is quite possible that states would lower state and local excises in response to this inflow. In our analysis we have assumed that states neither raise nor lower their nominal tax rates in response to the settlement.

Youth Lookback Penalty Payments

A central issue in examining price effects is whether teenagers will reduce their cigarette consumption sufficiently so that manufacturers can avoid the lookback penalties. It is clear that the combination of anticipated price increases and the tighter restrictions on youth access and marketing will lead to dramatic reductions in youth smoking. However, the many uncertainties involved in making these predictions underline the importance of incorporating in any legislation concrete targets for reducing youth smoking with significant penalties for missing the reduction targets.

The S. 1415 targets call for cutting youth smoking by 15 percent after 3 years, 30 percent after 5 years, 50 percent after 7 years, and 60 percent after 10 years. The S 1415 lookback penalties are levied on the basis of the percentage shortfall (non-attainment percentage) from the targeted reduction. The penalties are $80 million for each non-attainment percentage point of 5 or less; an additional $160 million per percentage point greater than 5 but less than 10; and an additional $240 million per percentage point beyond 10. The lookback penalty is capped at $3.5 billion per year.

In assessing the potential costs of this lookback penalty for the industry over the first five years of the settlement, it is necessary to estimate both the price and non-price impacts of the settlement on youth smoking:

_______________
⁵ For example, Barnett, P. et al. (1995), "Oligopoly Structure and the Incidence of Cigarette Excise Taxes," Journal of Public Economics, 57, p. 457-470; Merriman, D. (1994), "Do Cigarette Tax Rates Maximize Revenue?," Economic InQuiry, 32, 419-428; Sumner, D., "A Measurement of Monopoly Behavior: An Application to the Cigarette Industry," Journal of Political Economy, 89, 1010-1019. A number of states do have laws which try to limit predatory pricing by regulating that retailers cannot sell at below cost. But this is irrelevant for the question of whether additional federal payments will be marked up at the retail level. So long as there is some markup, there will be no predatory pricing finding - so that these laws provide no reason to hold markups constant in percentage terms.

______________

Price Effects: Treasury has developed an estimate of youth smoking reduction which draws on a sizeable literature on the effect of prices on youth smoking participation (whether youths smoke at all). We use a semi-log specification for the impact of prices on youth smoking rates, and so is similar in form to our estimates of the effect of prices on the number of packs sold. We use this youth analysis, however, to estimate the youth smoking participation rate (rather than total cigarettes consumed), since that is the figure relevant for the lookback penalty. Consequently, we calibrate our youth smoking analysis to an elasticity of -0.7, which differs from the elasticity in the revenue estimates (as it should since the analyses capture very different effects), but is well within the range of estimates suggested by the relevant economic literature. For example, the CBO recently summarized this literature by stating that most of the evidence points to participation elasticities ranging from -0.50 to -0.75.⁶

_______________
⁶ See: Congressional Budget Office (1997), The Proposed Tobacco Settlement: Issues From a Federal Perspective, Congressional Budget Office, Washington, D.C. See also: Chaloupka, F., and M. Grossman, "Price, Tobacco Control Policies, and Youth Smoking," NBER Working Paper #5740, 1996; Eugene Lewit, Douglas Coate, and Michael Grossman, "The Effects of Government Regulation on Teen Smoking," Journal of Law and Economics, Vol. 24, December 1981, pp. 545-573; Eugene Lewit and Douglas Coate, "The Potential for Using Excise Taxes to Reduce Smoking," Journal of Health Economics, Vol. 1, 1982, pp. 12 1 -145; Jeffrey Wasserman, et al., "The Effects of Excise Taxes and Regulations on Cigarette Smoking," Journal of Health Economics, vol. 10, 199 1, pp. 43 -64.

Some have cited the recent study from Cornell University as refuting the previous literature (Philip DeCicca, Donald Kenkel, and Alan Mathios, "Putting Out the Fires: Will Higher Taxes Reduce Youth Smoking?" Cornell University, December 1977). In fact, this study finds higher estimates than the remainder of the literature when standard estimation techniques are used on their full sample of 8th - I 2th graders. The study does find smaller effects when they choose the particular sample of 12th graders who weren't smoking in 8th grade, and try to model whether they start smoking. But there is no obvious explanation for this anomalous result; after all, removing from their model a population that is more addicted to cigarettes - individuals smoking from 8th to 12th grade - should raise, not lower, the elasticity estimate. It is troubling that dropping only 5% of their sample - the 5% of teens that are most addicted - reduces their estimate so dramatically. The likely explanation for this anomalous finding is problems with their empirical methodology. A recent re-analysis of their data by Professors Thomas Dee of Georgia Tech and William Evans of Maryland has found these results to be very sensitive to the particular sample restrictions imposed by the Cornell authors. As these experts note, "The results appear to be purely an artifact of the way that the authors constructed the analysis sample." When a broader sample of observations is used, there is a very significant effect of taxes on youth smoking in their onset model - indeed, the results are quite comparable to the previous literature. This partly explains why the results of the Cornell study are so statistically imprecise. For example, in this particular model, they estimate that a $ 1. 10 price rise would reduce smoking onset by only 7%. However, given the level of statistical imprecision in their model, their findings would be equally consistent with a reduction in youth onset of 50% or more from this $ 1. 10 price increase - a range which encompasses the Treasury estimate, as well as their own estimates using a more straightforward methodology.

_______________

As discussed below, we estimate that without a youth smoking penalty, S. 1415 would increase the real price of cigarettes by $ 1.10. Our youth smoking analysis implies that a $ 1.10 increase in the real price of cigarettes would reduce youth smoking by 32%. (Our analysis also predicts an additional 4% reduction in youth smoking from the scheduled 15 cents increase in the Federal excise tax on cigarettes, a reduction which would occur without S. 1415, but which nonetheless counts for achieving S. 1415's youth smoking target.)

Nonprice Effects: We also estimate a 15 percent reduction in the remaining base of smokers from the non-price elements of the legislation, particularly the youth access and advertising restrictions. This reflects a rather conservative reading of a number of case studies where well-enforced access restrictions lowered youth smoking by 44-69 percent, as well as a cross-state comparison of existing access restrictions (which are much lighter than those contemplated by S. 1415) which found that comprehensive access restrictions would lower youth smoking by 18 percent.⁷

Since youth smoking is already reduced by 36 percent from price increases in our analysis, this non-price effect adds another 10 percentage points (15 percent of 64 percent), for a total reduction in youth smoking of 46 percent in 2003. In 2001, when the first youth lookback target is presented, we estimate a reduction in youth smoking of 39%. Thus, for the first five years of the legislation, we estimate youth smoking reductions far in excess of the required targets. As a result, we do not add any youth lookback assessment to the price effects of S. 1415 (row 10).

Smuggling

It is not possible to reach definitive conclusions about the risks of smuggling given the wide range of changes contemplated by comprehensive tobacco legislation. Incentives to smuggle may well be sensitive to details of tobacco legislation, including price changes, the way in which assessments are levied and other specifics. Nonetheless, the Treasury Department believes that the creation of a sound regulatory system - one that will close the distribution chain for tobacco products - will ensure that the diversion and smuggling of tobacco can be effectively controlled

_______________
⁷ See: Jason, L. A., P. Ji, M. Anes, S. Birkhead, "Active Enforcement of Cigarette Control Laws in the Prevention of Cigarette Sales to Minors," Journal of the American Medical Association, Vol. 266, no. 22, December 11, 1991, pp. 3159-3161; DiFranza, J.R., R.R. Carlson, R.E. Caisse, "Reducing Youth Access to Tobacco," Tobacco Control, 1992; Chaloupka, F., and R.L. Pacula, "Limiting Youth Access to Tobacco: The Early Impact of the Sinar Amendment on Youth Smoking," Working Paper, University of Illinois-Chicago, January, 1998.

_______________

and will not defeat the purposes of comprehensive tobacco legislation.

The regulatory regime we envision has three essential elements. First, all entities in the distribution chain for tobacco products - manufacturers, wholesalers, exporters, importers, distributors and retailers - should be required to hold a license or a permit. Licensing of retailers could be done at the state level. Licenses would be issued based on certain clearly specified criteria and could be revoked or suspended for certain specified violations. Those conducting business without a license would be subject to penalties. Licensed entities should only be authorized to sell tobacco products to other licensed entities. Second, legislation should require the marking, branding and identification of packages of tobacco products intended for domestic distribution and for export so that they may not be diverted or smuggled in circumvention of the legitimate channels of distribution. Third, any regulatory proposal should include penalty and administrative provisions that will allow for effective, efficient and uniform enforcement of controls over distribution.

With the necessary regulatory provisions in place to deal with potential smuggling, we assume there will not be an increase in smuggling for several reasons. First, the "closed" distribution scheme would limit drastically smugglers' ability to enter products into a legitimate distribution channel. Potential black marketeers will not be able to move products through legitimate wholesalers or distributors. Nor will they be able to sell products to retail consumers at the local convenience stores or other licensed retail outlets. Instead, without a way to place contraband products in the market legally, smugglers would have to sell cigarettes outside channels of legitimate distribution. This would be a risky proposition and one we do not believe will represent a significant problem. Second, U.S. cigarette manufacturers would have great incentives not to become complicit in any smuggling operation, as they would encounter enormous legal risks (such as the possibility of losing their license or, as the McCain bill provides, losing their cap on liability risk) and public opprobrium. Indeed, it is hard to imagine that large scale smuggling could occur without the manufacturers' knowledge. Third, the U.S. Customs Service has the expertise and the experience to deal with imported contraband products and has already made a substantial investment in the currently planned introduction of non-intrusive inspection systems and other equipment needed to detect smuggling of contraband. The organic nature of tobacco and the distinctive shape of cigarettes makes them readily detectable by equipment that Customs currently has in place.

Liability Exposure and Attorneys' Fees

Liability costs should only be reflected in prices to the extent that they are industry-wide, and apportioned by current market share. Firm-specific costs are unlikely to be passed on to prices, since there is a high likelihood of brand substitution if there are large price differentials among premium brands. The legislation does not provide that any liability costs outside of the base payment should be shared on an industry-wide basis. We therefore assume that there is no incremental impact on the price of cigarettes from the liability cap.

It is difficult to determine the magnitude, or even the direction, of the effect of the legislation on attorneys' fees. S. 1415 establishes a process to determine the level of such fees, which would appear to be paid by the states, not the firms. As a result of this process, and as a result of their likely small magnitude relative to base revenues, we have assumed that attorneys' fees will have no effect on price.

Our assumptions on liability costs and attorneys' fees are reflected in row 11 of the Table.

Export Charges

S. 1415 includes a licensing fee of $1 per thousand cigarettes. This fee, however, would be creditable against federal tobacco excise taxes, and so effectively would apply only to exported cigarettes. We also understand that this provision is likely to be dropped from the final version of the bill. We therefore assume that this provision has no effect on the price of domestic cigarettes.

Calculating the Price-Per-Pack Equivalent

Using the demand specification and other salient assumptions described above, we calculate the price increase that leads to the required settlement payments, accounting for any reduction in demand that occurs as the result of the price increase. For example, 92 percent of the 2003 settlement amount of $23.6 billion is $21.7 billion. The baseline price - that is, the price that would occur in the absence of tobacco legislation - for our analysis is $2.09 in 2003, as shown in row 3 of the Table. A price increase of $ 1.10 per pack (row 4) in 2003 will reduce cigarette demand by 22.5 percent (row 1), and raise the designated settlement amount of $23.6 billion (row 2), assuming that other tobacco products bear 8 percent of the settlement. Price per in constant 1998 dollars would be $3.19 in 2003 under S. 1415. In nominal terms, as shown in the bottom panel of the Table (row 22), S. 1415 raises the price per pack to $3.57 in 2003.

Notes:
1/   Change taken relative to Administration Baseline Quantities.
2/   Statutory payments required to hit Administration proposal. Incorporates effects of loss Federal Excise tax receipts and the normal indirect income tax offset.
3/   Deflated relative to Administration's CPI-U forecast.
4/   October 1,1998 effective dates. Assumes Cigarettes bear 92% of settlement costs.
5/   Totals may not add due to rounding.