Table 6.4. Characteristics of the tobacco support program: flue-cured tobacco, 1975-2000 Year 197j I Y76 -... lY// -> I')/b 1 Y79 1 W) 1981 192 1983 1 w4 I')85 1986 1987 I 'MS IWY 1 Y90 1991 1992 1993 1994 1995 1996 1997 1998 1 YY9 2000 National marketing quota (million lbs.) National Real average average support support price price* (cents/lb.) (cents/lb.) No-net-cost assessment+ (cents/lb.) Producers Buyers 1,4Y I 1,26S 1,116 1,117 1,095 1,094 1,013 I ,o 13 y 10 804 775 7,Y 707 754 891 S/S 878 892 PI2 803 93s xi4 974 813 666 533 93.2 106.0 113.8 I21 .o 139.3 141.5 158.7 169.9 I hY.Y I W.C) 1hY.Y 143.8 143.5 144.2 14h.S I4X.S 152.8 133.0 157.7 138.3 139.7 160.1 163.1 162.8 163.2 164.0 173.2 186.3 lS7.8 185.6 178.1 171.7 174.6 176.1 170.6 lh3.5 137.Y$ 131.2 126.3 121.Y 118.4 113.8 112.3 111.2 109.1 106.8 104.8 102.0 101.0 99.9 98.0 95.6" 3.0 7.0 7.0 2.50 2.50 2.00 1.13 1.12 1 .oo 1 .oo I .oo 1.00 3.00 0.80 1 .oo 1 .oo 1 .oo 1 .oo 2.50 1.50 2.00 1.13 1.12 I .oo I .oo 1.00 3.00 5.00 I .80 1.80 1 .oo 1 .oo 1.00 2.50 *Real a\`erage support price is obtained by di\Gding the nominal support price by the national Consumer Price Index; the average of 1982-1984 is the benchmark. `No-net-cost assessment includes marketing budget deficit assessments from 1991 through 1998. IThe effective support price in 1985 Leas 165.0 cents/lb. by reduction of certain grades. `I'reliminarv estimate. Sources: Us. Department of Agriculture 1997b, IY99a,b. Table 6.5. Characteristics of the tobacco support program: burley tobacco, 1975-2000 1975 1976 977 978 979 1 980 981 982 1 983 983 985 986 987 1 988 989 Year 1990 1991 19Y2 1993 1994 1993 1996 1997 1998 1999 2000 National National Real average marketing average support quota support price* price+ (million lbs.) (cents/lb.) (cents/lb.) 670 96.1 178.6 635 109.3 192.1 636 117.3 193.6 614 124.7 191.3 614 133.3 183.6 614 145.9 177.1 660 163.6 180.0 680 175.1 181.5 647 175.1 175.8 582 175.1 168.5 524 148.8 138.3 493 14X.8 133.x 464 148.8 131.0 473 150.0 126.8 5x7 153.2 123.5 601 155.8 IIY.2 724 158.4 116.3 668 164.Y II 7.5 602 168.3 116.5 536 171.4 115.7 546 1/2..i 113.2 631 173.7 110.7 704 17h.0 109.7 635 177.8 1OY.l 4.51 178.9 7 07.3 247 180.5 105.2" No-net-cost assessment7 (cents/lb.) Producers Buyers 1.0 3.0 9.0 4.0 2.75 2.00 0.80 1.00 1.00 I .oo I .oo I .oo 4.50 1 .oo 1 .oo 1 .oo 3.00 3.00 3.00 1.25 2.00 0.80 1.00 1 .oo I .oo I .oo 3.50 3.60 I .oo 1 .oo 1 .oo 3.00 3.00 3.00 *The support price \vas reduced from 178.8 cents/lb. and the no-net-cost assessment was reduced from 30 cents/lb. bv Public La\v F-157, sec. 6 (IYH5). +Real average support price is obtained by dividing the nominal support price by the national Consumer Price Index; the a\.erage of IYP2-1984 is the benchmark. iNo-net-cost assessment includes nlarketin;: budget deficit assessments from 1991 through 1998. Vi-eliminary estimate. Sources: U.S. Department of Agriculture lYY7a; 1998a,b; 1999a. 2000. prices for domesticall\, grow\-ii tobacco are artificiallv high. Some estimates of the distortions resulting from the support program \1-cre pro\.ided by Sumner and Alston (1985) in their analysis of the economic conse- quences of renio\-ing the tobacco price support svstem. Their estimates 12.ere based on a detailed simultaneous equations model of the supplv and demand for tobacco and tobacco products (cigarettes) that allows for sub- stitution bet\\-een domestic and foreign tobacco in ciga- rette production. The authors estimated that domestic tobacco output K-ould rise by 5&100 percent or more if supplv restrictions \vere eliminated. This large in- crease in the quantitv of tobacco supplied should lead to sharp reductions iii- tobacco prices. As a result of the increase in output, tobacco prices I\Y~LI~CI fall b\. 3-30 percent, and the \-ariabilitv of tobacco prices ~vnulcl in- crease. Ho\\.el.er, o\.erall rel'enues from tobacco grow\-- ing ~vould rise bv 15-60 percent or more. Moreo\.er,this analysis predicted that the sharp drop in domestic tobacco prices that \\~oulcl follo\\~ the removal of supplv restrictions ~~0~11d lead domestic producers of cigaiettes and other tobacco products to use less foreign-gro\\,n tobacco. These estimates as- sumed the elimination of the program in 1983 and thus do not take into account the more recent changes in its operation. More recent estimates from Zhang and co- leagues (2000) suggest that the conclusions of Suninel and Alston (1985) still applv. For example, thev esti- mated that the price support program raised tobacco leaf prices by 36 cents a pound in 199-l. This price is about 21 percent above the estimated price in the ab- sence of the support program. The removal of the support program should also make domestic tobacco growers more competiti\.e in world markets. In the 198Os, U.S. tobacco prices ex- ceeded world market prices by JO-60 cents per pound (Warner 1988). Although part of the differential can be explained by the higher quality of U.S. tobacco, a significant factor is the U.S. tobacco support program. Sumner and Alston (1985) predicted that U.S. tobacco exports would have grown bv about 100 percent if the tobacco support program hai been eliminated in 1983. This change would have had an ad\rcrse impact on foreign tobacco growers, as producers of foreign ciga- rettes and other tobacco products increased their use of tobacco grown in the United States. Although the artificially high prices resulting from the support program tend to increase the income of small tobacco farmers, thev likely recei\re relativelv less benefit from the program than the tobacco quota owners. Because most small tobacco farmers rent some or all of their allotments from the quota o\vners at a significant cost (Watkins 1990), these farmers pav rents equix-alent to the excess value created by the support program. In the absence of the program, reduced in- come for these farmers lvould likely be offset by the resulting reduced rent they paid. Quota owners, on the other hand, have been estimated to lose about $800 million annually M'ere the support program eliminated (Sumner and Alston 1985). Despite the differing likely effects on quota own- ers and small tobacco growers, eliminating the tobacco support program \~ould probably not alter existing trends in the concentration of tobacco production into larger farms (Sumner and Alston 19%). Rucker and colleagues (1995) ha1.e estimated that eliminating the program's intercounty restrictions on the transfer of tobacco quotas I\-ould ha\ e little o\serall impact beyond redistributing lvealth from some tobacco growers and quota 011 ners to others. (Consequently, these research- ers suggest that the restrictions have remained in ef- fect not because the gains associated with them are large but because the political costs of removing them are.) Moreo\.er, remo\%~g supports would cause a mo\.ement a\vav from regions \vhere the costs of grow- in;; tobacco ari relativelv high torz-ard those where costs arc relati\.el>. Io~v. -The loss of income to quota o\\`ners \zo~ild lead to reductions in personal income of up to 2-3 percent for counties that are highly de- pendent on tobacco; larger losses would occur in the relati\.elv high-cost counties. However, total incomes would rise in areas that experienced a great expan- sion ~JI tobacco growing. In comparison, the effect of altering another government program would be con- siderable. Increases in cigarette excise taxes are also likely to bring significant losses to quota owners. Sumner and Wohlgenant (1985) estimated that dou- bling the federal cigarette excise tax in 1983 would louver quota oM.ners' lease income by an average of 13 percent, or about $44 million. As a result of the sharp drop in the price of to- bacco, cigarette prices could fall. Tobacco costs, how- ever, are a relatively small component of cigarette prices. Grise (1995) estimates that the 40- to 50-cent per pound drop in tobacco prices resulting from the elimination of the support program r~ould reduce ciga- rette prices by only l-2 percent. Zhang and colleagues (2000) estimaie an even smaller impact, concluding that cigarette prices are 0.52 percent higher than they would be in the absence of the support program. As noted by Sumner and Alston (19851, a reduction in cigarette prices would lead to a rise in U.S. cigarette exports. Moreo\.er, estimates of the price responsiveness of ciga- rette demand (described in "Effect of Price on Demand for Tobacco Products," later in this chapter) suggest that the reduction would lead to an increase of no more than 1 percent in cigarette smoking. At least part of the increase \voulcl come from increased smoking among young people. Opponents of the tobacco support program sug- gest that it can be removed w+th little impact on the farmers it is intended to benefit. For example, the less than 2-percent reductions in cigarette price that would result from eliminating the support program could be more than offset by an increased excise tax on ciga- rettes. A portion of the revenues generated from the tax hike could be used to help tobacco farmers diver- sify into other crops (through low-interest loans, grants, or other programs) or to purchase the farmer's tobacco base to retire it from tobacco growing (Northup 1993). Similarly, some of the funds could be used to develop nonfarm businesses, train farmers for other occupations, provide income support, and offer other economic support for local economies in transition (Womach 1994a). Critics also point out that the support program creates indirect political consequences: the depen- dence created by the support program results in a strong political constituency, composed of tobacco farmers and holders of tobacco allotments, that can impede legislation to reduce tobacco use (Taylor 1984; Warner 1988; Zhang and Husten 1998). In the absence of the support program, tobacco growing \vould likely become much more concentrated (Sumner and Alston 1985). Warner (1988) has observed that the reduction in numbers would lead to reduced political influence. Moreover, he describes the apparent inconsistency present when one arm of the federal government seem- ingly endorses tobacco production by continuing an economic support program even as another engages in numerous activities to reduce tobacco use (Warner 1988). Evolution of the U.S. Cigarette Industry Through much of the 19th century, most of the demand for tobacco products centered on smokeless tobacco and cigars (see Chapter 2). Cigarettes were relatively less popular, although the demand for them increased gradually during the middle of the century (USDHHS 1992). The watershed year for the cigarette, however, was 1881, when James Albert Bonsack an- nounced his de\,elopment of a machine that replaced hand-rolling as the primarir means of making ciga- rettes. The mechanization df production significantly reduced the costs of manufacturing cigarettes and, consequently, reduced cigarette prices. The steep de- clines in cigarette prices relati1.e to the prices of other tobacco products, due largely to Eonsack's cigarette machine, contributed significantly to the rapid rise in the popularity of cigarettes during the late 19th and early 20th centuries (Wagner 1971). James Buchanan Duke was the first cigarette pro- ducer to acquire rights to the new machines, which he- installed in 1884. Duke entered into long-term con- tracts with Bonsack to use the machines at a cost lower than Bonsack would make them available to other producers. Because of the resulting substantial cost advantage in production for his company, Duke successfully waged price wars with other producers while still earning relatively high profits. Over the next decade, the Duke family formed a holding company, which was composed of their firm and several corn-- petitors they had acquired. By 1889, as a result of its aggressive pricing and marketing strategies, the hold- ing company effectively monopolized U.S. cigarette markets (controlling more than 90 percent of the mar- ket), as well as portions of the markets for other to- bacco products. Eventually, in an attempt to avoid antitrust prosecution under the Sherman Act, the Dukes converted the holding company into The Ameri- can Tobacco Company. By 1901, The American Tobacco Company dominated all of the U.S. tobacco market< except cigars. The company was also a considerable presence in cigarette markets around the world. In response to allegations that The American To- bacco Company was abusing its market position, the U.S. Department of Justice charged the firm with vio- lating the Sherman Act. In 1911, the Supreme Court dissolved the company, thereby creating several new firms from the conglomerate, including a new Ameri- can Tobacco Company (which later became American Brands, Inc.), Liggett & Myers Tobacco Company, R.J. Reynolds Tobacco Company, and Lorillard Tobacco Company. The American Tobacco Company was also divested of its foreign holdings (Imperial Tobacco Ltd. and British-American Tobacco Company Ltd. [B.A.T. Company]). Imperial Tobacco Ltd. eventually rno; nopolized cigarette manufacturing in Great Britain, and B.A.T. Company concentrated on manufacturing in British colonies and elsewhere. Both companies ultimately resumed some operations in the United States (Johnson 1984). Although Imperial Tobacco Ltd. eventually dropped out of U.S. markets, B.A.T. Indus- tries PLC, the parent company of B.A.T. Company, owns Brown & Williamson Tobacco Corporation, a large U.S. cigarette manufacturer. R.J. Reynolds Tobacco Company (which had nc cigarette production after the breakup) soon devel- oped a new type of cigarette by using burley tobacco, which was quickly copied by the other producers. By the lY2Os, the cigarette producers were competing `tsgressi\.elv in promoting their main brand-for ex- ample, R.J. Revnolds Tobacco Companv's Camel, The .\merican Tobacco Compan>"s Lucky Strike, and Liggett Cy: Myers Tobacco Company's Chesterfield. In addition, firms on the competiti\,e tringe attempted to compete through price \vith their so-called IO-cent brands (Robert 1967). (For a more detailed discussion L,f the domestic operations of U.S. cigarette firms be- fore World War II, see the Surgeon General's report $iwX-i/7,y (712~1 H~wltl~ ir7 t/7(, A~7c~riin5 [USDHHS 19921). The U.S. Department of Justice e\,entually chal- lenged the four producers' coordinated \vholesale and retail pricing practices. In IY-ll, on the basis of con- juct starting as earl\ as 1933, these producers xvere I:harged \vith \,iolatin g the Sherman Act bv conspir- ing to restrain trade in an attempt to monopolize the Industry Their \vholesale tobacco-yurchasitig prac- tices Ivere deemed to be monopsonistic-that is, char- -1cteristic of a market situation tvhere one buver exerts .I disproportionate influence-and their retail pricing \\-a~ thought to reflect collusi\.e beha\,ior. In 1916, bas- ~ng its decision on the no\,el legal concept of "conscious parallelism," the Supreme Court upheld a jury deci- ,ion that found the firms guiltv. The uniformit\~ of ;>rices at both the \vholesale and ;he retail le\,el (a result that could occur in anv highly competiti\.e market), the near-SVnchronoLis increases in prices, and the rais- ing of 12-holesale prices \vheii labor costs irei-e falling .vere vieived bv the court as evidence of tacit collusion. As a result, the firms xvere fined up to $250,000 each, a relati\.elv minor penaltv compared \zith their profits. Jol;nson (lY83) and others have noted that the Court's decision \vas not supported by purely eco- nomic reasoning. There rvas little if any evidence that cigarette firms \vere jointly restricting output to raise cigarette prices and, consequently, profitability. Simi- larly, there \vas no evidence that the firms limited their \\rholesale purchases of tobacco to depress to- bacco prices and production costs and, consequently, to increase profits. The Court's decision had little impact on the sub- sequent structure of the U.S. cigarette industrv. The practical result has been that, from 1916 until today, the combined market shares of the six major firms (five after the merger of Bro\vn & Williamson and Ameri- can Brands, Inc.) has exceeded YY percent, although indi\,idual market shares ha\.e changed significantly (Table 6.6). More important in changing relati1.e market shares ivas the release of information during the 1950s and 1960s on the health consequences of cigarette smoking. In the lYX)s, Philip Morris Companies Inc., 1i.J. Re~~nolds Tobacco Company, and Lorillard Tobacco Companv aggressi\,elv marketed filtered cigarettes (Marlboro, Winston, a'nd Kent, respectii,ely), xvhich lvere percei\,ed as less dangerous than standard unfiltered cigarettes; The American Tobacco Companv and Liggett & Mi,ers Tobacco Company \vere not as rable 6.6. Domestic market shares of U.S. cigarette firms, selected years R.J. Philip Brown & American Year Reynolds Morris Williamson Brands Lorillard 1913 0.2 NA* NA 35.3 77 1 &k. 1925 31.6 0.5 NA 31.3 1.9 i 9-10 21.7 9.6 7.8 29.5 5.4 1955 23.8 8.5 10.5 32.9 6.1 1970 31.8 16.8 16.9 19.3 8.7 1975 32.5 23.8 17.0 14.2 7.9 IY80 32.8 30.8 13.7 10.7 9.8 1985 31.: 35.8 11.X 7.4 8.2 1991 27.8 43.4 11.1 7.0 7.3 1996 24.6 47.8 17.2 NA x.-l Liggett & Myers 34.1 26.6 20.6 15.6 6.5 4.4 2.2 5.0 3.4 1.9 Total 91.7 91.8 94.6 99.4 100.0 99.8 100.0 99.9 100.0 99.9 -X.4 = Not available. %xwces: Tennant 1950; O\wton 1981; Clarifeld lY83; Standard & I'oc)r's lYXc), 1993; Federal Trade .`Ornmission 1997. s~~cc~ssful in marketing their competing brands (Johnson 1981). Similarly, after the 1964 release of the U.S. Surgeon General's first report on the health conse- quences of cigarette smoking, and after the Federal Trade Commission's (FTC) publishing of tar and nico- tine content in the late 1 Y6Os, Philip Morris Companies Inc. and R.J. Reynolds Tobacco Company introduced and aggressively marketed lo\v-tar and lobv-nicotine cigarettes (again, products percei\,ed as healthier than existing cigarettes), whereas the other companies \vere less successful. As a result of the brand loyalty these two firms Lt'ere able to establish at this time, they came to dominate cigarette markets; in 1996, the two firms had a combined market share of 73.3 percent. Another notable change in the tobacco industry, beginning in the lY6Os, was the di\.ersification of the cigarette-manufacturing companies. Perhaps in part to offset the impact that the campaign to reduce tobacco use had on the industrv's profitability, the six major domestic cigarette producers acquired or merged !2-ith U.S. firms in a \,ariety of nontobacco markets, includ- ing food, alcoholic beverages, and transportation. Both U.S. and international cigarette producers significantly expanded their international acti\-ities. Di\.crsification was relativelv easv because of the high profitabilitv from _ - cigarettes and the lo~v long-term debt of these firms (Overton 1981). By 1972, no major domestic cigarette companv ivas completelv dependent on tobacco for its revenue (Johnson lY84). Durin, 0 the l%Os, diversifica- tion strategies and successes among the six firms var- ied markedly; some firms returned to a focus 011 cigarettes and other tobacco products, \\.hereas others di\.ersified further. B\, the late l%Os, a three-tiered classification of \Vorld cigarette producers, based on their international activities, had emerged: those in- volved in most global tobacco markets (Philip Morris Companies Inc., B.A.T. Industries PLC, 1i.J. Reynolds Tobacco Company, and Rothmans International Tobacco Ltd.); those \\ith some international, but not global, activities (including American Brands, Inc.); and smaller firms concentrating primaril!, on their domestic mar- kets (including Liggett & Myers Tobacco Company and Lorillard Tobacco Company) (L'SDHHS lYY2). Economic Implications of Concentrated Tobacco Production The concentration of production among relati\d) few firms in the cigarette industry has implications for cigarette pricin,, ~7 marketing, product de\.elopment, and other activities. Clearly, the cigarette industrl, is an oligopoly; no more than six firms ha1.e controlled vir- tually all cigarette output in the United States for the past 80 vears (Table 6.6). Economic theory suggests that fir& in oligopolistic industries have substantiat market pobver in that their production decisions wilt have a significant impact on price. Moreover, these firms recognize their interdependence. That is, each firm recognizes that its pricing and marketing strate- gies have a significant impact on the sales and profit- ability of its competitors, as well as on its own sales and profitability. Consequently, each firm understands that its competitors are likely to respond to any changes in its own pricing, marketing, or other strategies. Economic theory provides several possibilities regarding the conduct and performance of firms in an oligopolistic industry. At one extreme, if entry is easy and if sunk (nonrecoverable) costs are low, firms in an oligopolistic industry will behave competitively. That is, firms will have little market power (their output decisions will have little impact on market prices), prices will reflect the costs of production, and firms will not earn excessive profits. At the other extreme, firms could behave collusively, jointly restricting out- put, raising prices well above costs, and earning very high profits. Most theoretical models of oligopolistic industries suggest behavior between the two extremes: prices and profitability will be above and output will be below Ivhat would result from highly competitive behavior, and output will be higher and prices and prof- itability will be lolver than their levels in a monopo- lized or highly collusive industry. Casual empiricism suggests that cigarette prices ha1.e historically been bye11 above costs, thereby allow- ing cigarette producers to achieve a rate of return ~-41 above that earned in most other industries. Even after the health consequences of cigarette smoking became apparent, the U.S. tobacco industry led all U.S. indus- tries in profitability (Miles 1982). Moreover, in the t\vc major antitrust cases brought against the cigarette in- dustry in the 20th century, firms were found guilty ir 1911 of monopolization and in 1946 of a conspiracy tc restrain trade (collusion). Most industry analysts sug- gest that the primary source of market power in the cigarette industry is the entry barriers resulting fron? marketing efforts, tvhich create significant brand loy- alties that are nearly impossible for a nets producer tc overcome. High Tobacco Concentration and the Impact of Prevention Policies The high concentration of the cigarette industrk and the apparent market polver this concentratior engenders have implications for the effects of changes in cigarette taxes and other prevention policies on the pricing, marketing, and other strategies of cigarette firms. For example, the historically high profitabilitv of existing cigarette producers pro\.ides them Lvith thk resources needed to s~~ccessf~~llv de\-elop and market new' products, as leas seen in ;he de\.clopment and introduction of filtered cigarettes in the lY5Os and lots- tar and lolv-nicotine cigarettes in the 1YhOs in response to the initial reports linking cigarette smoking to lung cancer. More recently, in response to the increased alrareness of the harmful effects of en\-ironmental to- bacco smoke (ETS) on nonsmokers and the ~~idesprcad restrictions on smoking that ha\-e been designed to protect nonsmokers, R.J. Re\,nolds Tobacco Cornpan\, introduced its Eclipse brand in se\ era1 test markets beginning in mid-lYY6, and Philip Morris Companies Inc. is currentI>, testing its Accord brand in the Lnitccl States and Japan. Both are ostensibl\, "smokt~lt~ss" cig,3- ret&s, primarilv heating rather than burning tobacco; consequently, both generate less secondhancl smoke than con\.entional cigarettes. Economic theor\, can predict sonic effect\ of in- creases in excise tax&on price, output, and profitabil- itv. At one extreme, tax increases in a perfectl\ competiti\.e market \vith constant costs of production should result in price increases of the same magnitude rzith no impact on long-run profitabilitv. Reductions in output \\.ould depend on the effect that price has on demand. At another extreme, standard models for a monopolized market suggest that producers; and consumers \ynuld share the burden of the tax increase but consumers \vould pav a greater share of the tax, because demand is less s&iti\-e than production to price. Output and profitabilit!, t~ould fall, Ivith smallet reductions in both-again because demand is less sen- sitive to price. Recent advances in the theoretical and empirical study both of oligopolistic behavior and of the supplv of addictii.e goods haire vielded se\,eral interesting predictions. Perhaps most interesting is the possibility that prices ITi increase by more than the amount of the tax increase i~hen excise taxes are raisecl. Several early studies of these relationships pro- duced generally inconsistent conclusions concerning hoM' much cigarette prices rzould increase after an in- crease in cigarette taxes (Barzel 1976; Johnson 1978; Sumner 1981; Sumner and Ward 1981; Bulo\~ and Pfleiderer 1983; Bishop and \Ibo 1985; Sulli\,an lY85; Sumner and Wohlgenant 1983; Ashenfelter and Sullivan 1987). One general Meakness of these stud- ies w'as their failure to account for the dynamic inter- action of firms in an oligopolistic industrv. Instead, the studies generally assumed that rules fo; the firms' behavior \vere established, and then, \vith obselved prices and taxes, the studies lvorked back\\-ard to determine the degree of competition within the indus- trv (Harris 1987). More recent studies have addressed these weak- nesses. Harris (1987) used the estimates obtained from se\,eral studies of cigarette demand and supply to e\raluate the impact of doubling the federal cigarette excise tax in 1983; moreover, Harris' framework al- lo\ved the change in the tax to affect the interaction of firms in the industrv. Using data on wholesale and retail cigarette prices as well as the costs of produc- tion, Harris concluded that the ii-cent increase in the tax 14 to a 17-cent increase in the retail price of ciga- rettes. He further argued that the price increase above the tax hike could not be accounted for by increases in production costs. Instead, this increase was attributed to the recognized interdependence of cigarette firms in an oligopolistic industry; that is, the firms recog- nize that their profitability \vould rise if all could suc- cessiull~ restrict output and raise prices. However, because foi-ma1 agreements on output and prices are illegal, the firms are alert to other bases on which they can coordinate their behavior. Harris suggested that such a base \\`as the announced increase in the federal ta\, scheduled for January 1, 1983, lvhich served as a ioordinnting mechanism for a joint oligopolistic price increase. As Bnrnett and colleagues (1995) note, Har- ris' analvsis fails to account for existing trends in ciga- rette prices. Barnett and colleagues argue that Harris attributed too much of the coordinated rise in price to the increase in the federal tax, because the upward trend in prices predates the consideration of the tax hike. The authors suggest that producers used the in- troduction of discount cigarettes in 1981 to coordinate the earlier price hikes for premium brands, because the lolver-priced "generic brands" would keep more price-sensitil-e smokers in the market. The spirit of this argument is the same as Harris', because both sug- gested that certain e\,ents served as focal points allow- ing firms to engage in more collusive behavior without appearing to establish a formal agreement. Keeler and colleagues (Sung et al. 1994; Barnett et al. 19%; Keeler et al. 1996) used national- and state- lc\.el data to estimate the effects of cigarette tax in- creases on price. Their empirical models have been used to examine the interaction of cigarette supply and demand in determining cigarette prices. By using alternative assumptions about firm behavior, these studies formally account for the oligopolistic aspects of the cigarette industry in their empirical models of cigarette supplv. At least some of these models also account for the+addictive nature of cigarette demand. In a study using data on all U.S. states from 1960 through 1990, Keeler and colleagues (1996) conclude c~\\~la~~i thti li\.pnthL5i\ a5 follo~\3: iigdrette firms Lvit1-t nl,lrht't ~cJ\\ t't- ma\. s;ft relati\ el\, low prices to "hook" i~~ti5uiilt~r5 on thtsir ddciiiti\-e product, thus raising the tature ciemdnc-l for their (-igdt-ettes; policies (including tCl\ incrc~,tssrsJ that reduce future smoking also reduce the tit-tns' prufitabilitv of maintaining lo\v prices. Nev- tWhele+s, the relativelv lo\v prices of these for~vard- looking firms (comp,~;ed \vith those of more myopic firms) I\-ill still c~ceecl the marginal and average costs of pr~dustion and distribution. A similar hypothesis lids been ~tscci to explain studies that found Ihat ciga- rette producers appear to ad\,ertise beyond the profit- tn~~simi~ing le~,el (Sholtxlter 1991). These firms rnal be engC~ging in excessive ad\.crtising (i.e., more than can be recouped through brand switching among cur- rent smokers) to attract new consumers and hoping to I,itcr bcnetit from a hi,gher demand for cigarettes as `i- result of these tie\vlv addicted consumers. The rapid incr&ses in cigarette prices since the e,irl\ I %X)5, I\.liich are only pnrtlv explained bv incrcdst~s in taxes and costs, thus reflect I-7rofit-maximizing beha\-ior b>f ;i highly concttitr~~tcd cigarette industr\~ that anticipate5 dtueas.ed future ~lemand as xiditiotiC~l c'iforts to rcduw tobaccn ust' are implemented (Kecket c't ,il. 1-W). ;Itl empiric-al dppliiation of this model tc tlic ti5liips bc`t\x.twi industrv concentratioti~ tri5tt-iiti~~n5 c>ii iigdrctte ad\~ertising, cigarette pricr~. ,Incl mdrket po\\.er. One such analysis supports the ion\ twticIndl \\,isdoni that advertising is an important ic)tiipt`titi\ e strategy in de\~elopin, u and m;lintainin$ br,incl io\~dIt~~ for firms in the cigarette indt.tstr\ ~N~u\~~ii ~IW?). Another analvsis, using nn empiricA niodt>l that ;Illo\\5 firms in an oligopolistic indurtrv tc h;l\ t` some cicyt-et> of market polver, conclucies ;lial dd\.cdi5iti~q raist5 market po\vet- and, consequetitl~- pr~~tit,lbilit\ in the cigarcttt> industrv (TremblCtv anti Tretnbl,~\~ iW3). A likelv eupl,wati& of this ef&ct i> that b\, tostcritig lov;llt\T to existing brdncls, cigdrettt ad\ vr;isiii:g raises LGrri&s to other brand5 that try tc tlntct- the market and shCue in the profits. Se\-et-al studies (Porter 19X6; Mitchell ant? ~~lulherin lc)Sti; Eckard Ic)c)l) ha\,e ~oncl~~deci that bnn- niti;; ii ~arrtte acl\~crtising from television and radii h mdclv the industr\- c`\ cn less conipetiti\x!, thercbv fur tlitlr raising pt-nfit~iL%lit\,. One such stud\, atfribufeti the incrtuses in cigitdte prices after thedadL.ertisitis bdn to the rcduiecl cotnpt~tition taultin~ front the L7d1- (Porter I%%). This cotiilu5ion ~vas supported, tc) wnic Discussion effecti\-ely in foreign markets (this t),pe of association is exempt from antitrust lal~ under the Webb-Pomertne .4ct). As Grise (1990) notes, trade in tobacco and tobacco products would be even higher if not for general trade policies and, in particular, widespread agricultural and industrial policies that protect domestic tobacco grow- ers and producers of tobacco products. Numerous countries have policies that support domestic tobacco growing; in the United States, examples are the tobacco support program and the short-lived mandatory mini- mum content of domestic tobacco in domestic ciga- rettes. Likewise, both tariff and nontariff barriers to trade in tobacco and tobacco products ha\re been erected around the \vorld. These barriers include quo- tas, restricted product lists, exchange controls, prior deposits, mixing regulations, licensing requirements, and limits on advertising and other promotional ac- tivities (Grise 1990). Moreover, in se\,eral countries (in- cluding Japan, South Korea, and Thailand), various aspects of the manufacture and distribution of cigarettes have long been controlled by go\rernment monopolies that have largely prevented the import of foreign ciga- rettes (GAO 1992). When tariff and nontariff barriers to trade are used to protect domestic tobacco and tobacco products, to- tal supply of these products is usuallv louver than it would be others-ise, lvhereas domestic supplv is higher. In the case of tobacco products, this arrang&ent has public health benefits resulting from the generally higher prices and reduced consumption of the pro- tected products. Domestic suppliers benefit b!r sup piying more at higher prices. Foreign suppliers, howei'er, are likely to lose in this arrangement, because their access to these markets is limited and costs of sup- plying the markets are higher. In addition, restrictions on advertising and promotion in gi\ren countries are likely to make it difficult for nelz firms to successfully enter newlv opened markets lvhere existing brands are firmly entrenched (Chaloupka and Corbett 1998). Past Tobacco-Related Trade Policy In general, tobacco products exported from the United States are specifically exempted from federal laws and regulations concerning the export of poten- tially harmful products, including the Federal Hazard- ous Substances Act (Public La\v 86-613), the Toxic Substances Control Act (Public Law 94-469), and the Controlled Substances Act (Public La\v 91-513) (GAO 1992). Similarly, although federal regulations (1) re- quire that all cigarette packaging and advertising in the United States contain health rtarning labels and (2) prohibit television and radio cigarette advertising, there are no federal regulations or laws concerning the packaging or advertising of domestically produced cigarettes that will be exported (GAO 1992). Various U.S. policies and programs have been used to help domestic tobacco growers and cigarette companies expand into foreign markets (Connolly and Chen 1993). These policies include the USDA's Food for Peace Program, which sent more than $1 billion in domestically produced tobacco to developing countries in the 1970s and early 198Os, and the 1984 Export Credit Guarantee Program, which exported domestically grown tobacco and helped U.S. cigarette producers enter Mideast markets (including AIgeria, Egypt, Iraq, and Turkey) (Taylor 1984). Perhaps the most impor- tant, however, is Section 301 of the Trade Act of 1974 (Public Law 93-618) and its subsequent amendments. Section 301 of the Trade Act of 1974 The Trade Act of 1974 was initiated by the Nixon administration when it sought permission to begin the Tokyo Round of GATT. GATT, an international trade agreement honored by nearly 120 countries, governs various aspects of international trade. (GATT is dis- cussed in greater detail in "Multinational Trade Agree- ments," later in this chapter.) The first of these agreements was reached among 23 nations shortly af- ter the conclusion of World War II. Since then, seven rounds have occurred, including the Uruguay Round, which concluded in April 1994 after more than seven years of negotiations. The Trade Act of 1974 included in its final legis- lation various measures with the stated purpose of- promoting free trade. One of these measures was Sec- tion 301, \vhich gave the President the authority to in- vestigate cases where trade and other practices of foreign countries were considered unjustifiable, unrea- sonable, or discriminatory in that they limited the abil- ity of U.S. firms to sell their goods and services in foreign markets. Section 301 expanded the authority given to the President by the Trade Expansion Act of 1962 (Public La\v X7-794). That earlier legislation allowed for inves- tigations of unjustifiable trade sanctions (those that di- rectly violated GATT). Consequently, the act applied only to goods covered by GATT (which at the time ex- cluded agricultural products, including tobacco). Set-m tion 301 expanded presidential authority to include trade in all C.S. goods and services and allowed the in- vestigation of practices that were unreasonable but did not necessarily violate GATT. If negotiations were not- successful in ieducing or eliminating the unjustifiable or unreasonable limits on trade, Section 301 authorized the President to impose retaliatory trade sanctions. Ini- tially, Section 301 recei\,ed little attention, although it rz-ould later become a rvidely used tool of U.S. trade policy (Nivola 1993). Section 301 of the Trade Act of lY73 \vas strength- ened by the Trade and Tariff Act of 1984 (Public Lab\ 98-573) and the Omnibus Trade and Competiti\.eness Act of 1988 (Public La\1 100418). NOM. known as "Su- per 301," the section required the U.S. Trade Repre- sentati\,e to annually identify countries and their practices that consistently limited market access to ti.S. firms. More important, if negotiations failed to elimi- nate the unfair trading practices of these countries, mandatory retaliatory measures \vere to be imposed unless the President deemed these measures harmful to U.S. economic interests. Four Section 301 cases in the late 1980s dealt \j.ith cigarettes: cases against Japan in 1985 and Taiil-an in 1986 were initiated by the U.S. Trade Reprtsentati\,t at the President's request, and cases against South Korea in 1988 and Thailand in 1989 ivere the result of the U.S. Cigarette Export Association's petitioning of the U.S. Trade Representati\,e. Threats of retaliatorv sanctions under Section 301 led to agreements tvith each country; as a result, U.S. cigarette firms \vere per- mitted access to those markets. The opening of the markets resulted in aggressive tobacco ad\,ertising by U.S. firms (Roemer 1993). Each of the four newly "opened" countries has la\vs, regulations, and ordi- nances concerning cigarette advertising and promo- tion. The go\,ernments of some of the countries ha\,e alleged that U.S. cigarette companies have violated restrictions on advertising and promotion. A brief review of the four Section 301 cases fol- lows; more details are contained in reports from the GAO (1990, 1992), and an empirical analysis of their impact on cigarette smoking is contained in Chaloupka and Laixuthai (1996). Jayall The tobacco industry in Japan is largely monopo- lized by the company Japan Tobacco Inc. In 1979, Ja- pan was the subject of two Section 301 cases, one involving cigars, which was prompted by the Cigar Association of America, and a second related to pipe tobacco, which was initiated at the request of the As- sociated Tobacco Manufacturers. The two cases were resolved in an agreement with Japan, which reduced market restrictions and lowered import duties (GAO 1990). Before 1986, the domestic cigarette monopoly Leas protected from foreign competition through tar- iffs of 28 percent on all imported cigarettes and through Japanese distribution practices, which discriminated against imported cigarettes. The threat of Section 301 sanctions led to an October 1986 agreement that elimi- nated Japanese cigarette tariffs and changed excise tax payment procedures and other distribution practices that adversely affected imports of U.S. cigarettes. Ex- isting Japanese policies related to cigarette advertis- ing and other promotional practices were not affected by the agreement. The agreement resulted in a significant expan- sion of U.S. cigarette firms in Japan. Japanese imports of U.S. cigarettes more than tripled in 1987 alone and Fontinued to rise in 1988 and 1989, by which time the market share of U.S. firms leas more than 15 percent (Grise 1990). This growth appeared to have slowed or stopped in the early 1990s. Total U.S. cigarette exports to Japan ranged from 54.0 billion to 57.7 billion annu- alla during 1991-1993. A dotvn\vard trend during the 1970s and 1980s in per capita cigarette consumption in Japan appears to have re\,ersed itself after theJapanese cigarette markets \verc opened to U.S. firms. Overall per capita consump- tion appears to have remained steady or increased slightly in recent years. However, among Japanese \lomen, smoking prevalence rose from 8.6 percent in 1986 (before the agreement) to 18.2 percent by 1991. The 1991 rates \vere even higher among young adult women (27 percent) (Connolly and Chen 1993). Part of this increase may be the result of adver- tising and promotional activities by U.S. cigarette firms in Japan. Between 1987 and 1990, total expenditures on cigarette advertising and promotion by U.S. ciga- rette companies in Japan nearly doubled. Most of these expenditures tvere on television advertising, which is allowed in Japan (but subject to some restrictions). Before the agreement, the domestic monopoly did not engage in extensive advertising. Afterward, it signifi- cantly expanded its advertising and promotional ef- forts. As a result, cigarette advertising moved from 40th to 2nd place in total television advertising in Ja- pan (Sesser 1993). Tnizvafr Virtually all aspects of the tobacco industry in Tai- wan are controlled by a state-run monopoly. In 1986, the U.S. Trade Representative threatened Taiwan with retaliatory trade sanctions over several governmental policies that limited the market access of U.S. cigarette companies. These policies included quotas and tariffs 011 im}wlkd 'is< IrcttLls, `1 twii 011 the rt?ail salt ot ini- pc~rted cigarettes, and a ban on print ad\-t\rtisin;: of imported cigarettes. An agreement \\`a~ reached in De- cember 1986 that reduced tariffs and eliminated other barriers, thcreb!. allort+ng C.S. cigarette companies greater access to the TaiIvanese cigarette market. The agreement also contained se\-era1 restrictions relating to cigarette packaging (rzhich \vas required to haIre a specified health \f,ai-ning label) as ~~ell as ad\,ertising and promotional acti\-ities (e.g., the distribution of free samples IVES limited and poiiit-of-t7urchase promo- tions \vere restricted to licensed establishments). The agreement greatlv increased C.S. cigarette companies' access to the Tai\\anese cigarette market. In 1987 alone, total L'S, cigarette shipments to Taitvan increased 3+fold, and the market share of U.S. ciga- rette companies rose from 2 to 17 percent (Grist 1 YYO); by lYY7, the mark& share of imported cigarettes had risen to 30 percent (Hsieh ancl \r'in lYY8). Moreover, TaiLvan's imports of relati\.elv higher-quality U.S. to- bacco rose, as the portion of U.S. tobacco in Tairvanese cigarettes increased from 35 to 55 percent to better com- pete \z,ith imported cigarettes (Grise 1YYOi. Ho\:,ever, per capita consumption of cigarettes, after increasing somet\-hat during the 19705 and earlv IYHOs, tell from 1957 through 1996, due to public and $1 ate antismok- ing policies (Hsieh and \I'in 1998). Smoking pre\-alence among Tailvancse \~omen significantI\, increased in the late lY8Os and has remained stables throughout the 1990s (Hsieh and l'in lc)Y8). Adwrtising and promotion of U.S. cigarettes af- ter the agreement are likelv to ha\-<, contributed to the large rise in the market share of U.S. cisarctte compa- nies in T, but fcil somel\.hat in the next three \`ears. Ne\-ertheless, total spending rose b)' -l3.8 per&t from 1 Y87 to IYYO (GAO 1 YY2). Gi\,eii preagreement restrictions on ad\.ertis- iii:< and promotion. almost all of these r\peiiditurc>s rzould have been for poj,it-of-I-7Lirchase and magazine ad\.ertising. Xd\.ertising by the Tailvanese cigarette moiiapolv. ho\j.e\.er. ~vds limited elen further after the ageement. Authorities in TaiI\,an ha\-e alleged that point-of- purchase promotional acti\.ities bv L'S. cigarette com- panies ha1.e violated the terms of the lYX6 agreement (GAO 1992). The agreement limits these activities to li- iensed \~.holesale, distribution, and retail establishments, Lvhich the Taiivan Tobacco & Wine Monopoly Bureau defines as those with a permit registering them as profit-seeking enterprises. Taiwanese authorities con- tend that U.S. cigarette firms have distorted this defi- nition to include unlicensed retailers selling cigarettes, resulting in Midespread advertising and unauthorized sales of U.S. cigarettes (GAO 1992). After 1987, the government of Taiwan enacted several strong tobacco control policies, largely in re- sponse to the liberalization of cigarette trade resulting from the Section 301 agreement (Hsieh and Yin 1998). Many of these policies were initially rejected by the U.S. Trade Representative as unfair or discriminatory toward the tobacco industry and in violation of the 1986 agreement. One contentious issue pertained to the health warning labels proposed for cigarette ad- iwtising and packaging. The Taiwanese government initially proposed a set of strong, rotating health warn- ing labels that would appear on the front of cigarette packaging and on all advertising. In response to the U.S. Trade Representative's opposition, the content of the label M'as changed to "excessive smoking is dan- gerous to health," and the label was placed on the side of packaging (Hsieh and Yin 1998). Eventually, in 1992, the labels were changed to include six rotating warn- ings communicating more specific information about Ihe hazards of smoking. The dispute over the Smoking-Hazards Pre\:en- tion Act, introduced in 1991 \z.ith the stated aim of pro- tecting the public health by pre\,enting and controlling damage from tobacco products, \vas even more COP tcntious (GAO lYY2). The aim of the act \yould be ac- complished by prohibiting smoking bv those under IS vears of age, banning \w~ding machine sales of to- bacco products, limiting the tar and nicotine content of all cigarettes, rquiring that the packaging of all to- bacco products include not only health warning labels- but also tar and nicotine content in Chinese, and ban- ning all tobacco aj\.ertising and certain other promo- tional activities. The act \vas immediately challenged- by the U.S. Trade Representative as a unilateral \,iola- tion of the 1986 agreement that allowed U.S. cigarette companies to ad\,ertise in Taiwan (GAO 1992). Sesser (lY93) reports that a confidential position paper drafted by the U.S. Trade Representative in January 1992 stated that the proposal was an attempt to protect the Tai- ilanese cigarette monopolv from foreign competition and that the various measures proposed would have little impact on smoking. In July 1993, the Clinton administration's US. Trade Representative, Michael Kantor, stated that his office would not challenge the act if it was enacted (Sesser 1993). Six years after its introduction, the Smoking-Hazards Preirention Act l\.as finall\- enacted j\,ith compromise clauses that permit cigar&e ad\,ertising in magazines (Hsieh and l'in 19YS). South Karen South Korea's Tobacco & Ginseng Corporation controls all aspects of that countr\,`s tobacco grn\\-ing and production, l\.hich hacl traditionall\, been prw tected bv high tariffs imposed on foreign cigarettes. In lY82,-South Korea enacted and aggressi\.elv en- forced legislation making it a criminal offense t(, sell, bu),, or possess foreign cigarettes (Edd!, and Walden lYY3). Beginning in 1987 , almost all cigarette ad\.er- tising and other promotional activities \j.ere banned b!. the Tobacco Monopol?. Act. After petitioning b!z the U.S. Cigarette Export Association in Januarl. 1988, the U.S. Trade Representati\.c in\.estigatcd these prac- tices. In response to the threat of retaliator\, sanctions on South Korean textile eNports to the United States, a Record of Understanding \~as signed hi. the t1x.o coun tries in May lY88. This agreement opened South Ko- rt'an cigarette markets to C.S. firms b\. eliminating the ban on the sale of foreign cigarettes, I-educing tlic tar- iff on imported cigarettes, alIn\\-in g tht> distribution of free samples, and allolviii s some print acl\-ertising of cigarettes and the sponsorship of sporting e\wits. The agreement also prohibited ad\-ertising that targeted \vomen and voung people (smoking is prohibited in South Korea-for persons under 30 wars of age). Fi- nall\J, all cigarette packagin g and niagazine ad\.ertis- ins i\.ere required to include a health \\.arning label. Although cigarette smoking had been increasing steaclilv in South Korea during the 19X&, the rate of oroiztl; in smoking more than tripled ~vhen cigarette h markets M'ere opened to foreign competition (Roemer 1993). Much of the increase appeared to have been the result of dramatic increases in smoking pre\.alence among voung people. From 1988 to 1989 alone, smok- ing pre\,alence among male teenagers rose frotll I8 to 30 percent, and smoking pre\.alence among female teen- agers increased from 2 to 9 percent (Sesser 1993). hluch of the increase in consumption \~as accounted for bv the increased use of imported cigarettes. Import share in the market rose from 0.06 percent before the agree- ment to nearlv 8.5 percent in 1993 and continued to in- crease steadilv (U.S. Department of Commerce, Tobacco Export Task Force Analvsis, unpublished data, No\.em- ber 13,1995). Part of th; increase mavbe attributable to an increase in adjrertising bv U.S. cigarette companies in South Korea after the liberalization of cigarette trade. In late 1988, South Korea passed the Tobacco Business Act (effective January 1, 19891, xl-hich limited ad\,er- tising and promotional efforts to p"int-of-yurchasc ad\-crtising, magazine adlrertising, and sponsorship of public e\.ents (GAO 1992). In 1991, the Korea Tobacco Association (comprising the U.S. Cigarette Export As- sociation firms and the Korean tobacco monopoly) out- lined a self-regulating voluntary marketing agreement to comply \vith the Record of Understanding and the Tobacco Business Act. Ne\,erthelcss, the South Korean government in- dicates that some promotional activities of U.S. ciga- rette companies \,iolate the spirit of the Tobacco Business Act. These allegations concern distribution of free cigarettes, advertising placement for televised c\.t`nts sponsored bv U.S. tobacco firms, the distribu- tion of nontobacco. "gifts" bearing company trade- marks, r7nd the targeting of youth. Although no formal actions related to these I,iolations were initiated, the Koreans did begin renegotiating the Record of Under- standing i\,ith the United States in 1995. In August lYY'i, the UniteJ States government agreed to modify the market access agreement \\rith the Koreans to al- ln\v them greater flexibilitv to impose nondiscrimina- tnr!, health-based measures that restrict the use of tobacco products, including limitations on tobacco product ad~.ertising. Perhaps the most publicized and contentious Section 301 dispute Ilas initiated by the U.S. Trade Representati\-e in response to petitioning by the U.S. Cigarette Export Association in April 1989 over Thailand's \,irtual ban on the import of cigarettes and complete ban on cigarette ad\,ertising and other pro- motional activities in that country. The complaint cited \-arious restrictions on the importation and sale of ciga- rettes and referred to discriminatory duties and taxes on cigarette imports (GAO 1992). `All aspects of the domestic tobacco markets in Thailand are controlled by a go\.ernment-run monopoly, which stopped its 0w.n cigarette advertising and promotion in April 1988. Ho\j.ever, foreign companies continued their activities, \vhich prompted a total government ban on cigarette ad\,ertising in Thailand in February 1989. The formal investigation began in May. After no agreement could be reached, the U.S. Trade Representative consented to submit the complaint to the GATT dispute resolu- tion process. The panel created by GATT investigated the U.S. complaint that the import barriers and advertising restrictions lvere a \riolation of the international agreement's principles. In October 1990, the GATT Council sustained the panel's recommendations and ruled that the ban on imports \\.as a \.iolation of the GATT treaty. However, the council upheld the high Thai cigarette excise taxes (applied to both domestic and foreign cigarettes) and the right of the government to restrict the overall supply of cigarettes. Regarding the Thai advertising ban, the council noted that GATT allows member nations to use lrarious policies to pro- tect public health if the policies are applied to both domestic and foreign products. A cigarette advertis- ing ban that made it difficult for new foreign firms to compete with existing domestic firms was ruled justi- fiable under the treaty, because alloying advertising could stimulate the demand for cigarettes, particularly among youth (Contracting Parties to the General Agreement on Tariffs and Trade 1991; Roemer 1993). This decision teas based on Article XX of GATT, which states that: Subject to the requirement that such measures are not applied in a manner which p1-ould constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions pre- vail, or a disguised restriction on international trade, nothing in this Agreement shall be construed to prevent the adoption or enforcement by any contracting parties of measures necessary to protect human health [or] necessary to secure compliance with 1aM.s or regulations which are not inconsistent with the provisions of this Agreement. The GATT ruling led to an agreement in No\,en- ber 1990 betlveen the Cnited States and Thailand that allowed the importation of U.S. cigarettes into Thai- land. Imported cigarettes were then subject to the same laM-s and regulations as those marketed by the Thai Tobacco Monopoly (GAO 1992). Thus, U.S. cigarettes would be taxed the same and subjected to the same supply restrictions, and the adlpertising and promo- tion of these cigarettes (including the use of cigarette company logos, trademarks, and other symbols on nontobacco products) \vould be prohibited. The Thai government, however, has indicated that U.S. cigarette companies have tried to circumvent the ban on pro- motional activities bv tactics such as sponsoring sport- ing events and placing cigarette logos or symbols in televised programming. No formal complaints ha\,e been filed. After its SLICC~SS in upholding the ban on adver- tising and promotion, the Thai government in 1992 enacted two la\vs restricting smoking: the Non Srnok- ers Health Protection Act and the Tobacco Products Control Act. The first act restricted smoking in desig- nated public places. The second \vas a comprehen- sive act that required that all tobacco products disclose their ingredients, allowed the Ministry of Public Health to determine all aspects of labeling, including health warnings, and banned the following: smoking by those under 18 years of age (imposing fines on viola- tors); vending machine sales; distributing free samples, exchanges, and gifts of cigarettes; tobacco advertising (including, under the Thai definition of advertising, the use of cigarette logos and other symbols on nontobacco products) except in international maga- zines and live telecasts originating outside Thailand; advertising products with the same name as tobacco products; producing, importing, advertising, and sell- ing products imitating tobacco products; and selling cigarettes not complying with th? labeling provisions (Roemer 1993). The cigarette trade agreement that opened the- Thai cigarette market to US. firms has led to a rise in imports from less than 1 percent of the market before- the agreement to about 4 percent in 1993. Because of- current trends, this change is likely to increase sub- stantially in the future (e.g., U.S. cigarette exports to Thailand rose by more than 56 percent from 1992 to 1993). Part of the increase may be the result of in- creased smoking prevalence among women and young people in Thailand (USDA 1994a). Multinational Trade Agreements The North American Free Trade Agreement In 1993, the United States approved the North American Free Trade Agreement (NAFTA), a compre- hensive agreement that eliminated most of the barri- ers to trade between the United States, Canada, and Mexico; implementation began January 1, 1994. This agreement further reduced already low trade barrier: betrveen the United States and Canada resulting fron an earlier free trade agreement. More important, the nect agreement substantially reduced existing trade barriers between the United States and Mexico b) eliminating all nontariff barriers to trade and by phas- ing out most tariffs. Mexican tariffs on U.S. tobaccc and tobacco products were initially set at 50 percent the 1998 rate was 25 percent. Supporters of the agree ment argued that it would lower prices, lead to a ne increase in jobs (particularly in export industries), ant spur economic growth in all three countries. Oppo nents countered that U.S. firms would have an incen tive to shift production to Mexico to reduce labor ant other operating costs, thereby leading to a net reduc - tion in C.S. employment. Before the agreement, some trends in tobacco production in the United States, Canada, and Mexico \vere similar. Total tobacco production and acreage de\.oted to tobacco grooving in 1990 lrere rvell belo\\ their 1981 levels in all three countries, but doIvnlvard trends in the United States had reversed b\r 1987. Sim- larl"; in recent years, tobacco production in Mexico has been expanding (USDA 1997cl). During the 1980s and earlv 199Os, cigarette consumption fell sharplv in both the United States and Canada but rose in Mixico. At least part of the increase in the Mexican demand fnt cigarettes resulted from increases in income, \vhich contributed to a shift to the consumption of higher- quality cigarettes among Mexican smokers (USDA 1992). Since 1991, ho\Vel.er, cijiarettc imports into Mexico have fallen as consumer purcliasins po\\.t't declined; no imports \vere expected in 19% (USDA 1997d). Trade in tobacco among the three countries \vas relati\,elv limited before the agreement. Mesican e\- ports of iobacco to the United States I\-ere about 3 per- cent of total esports, or less than 2 percent of total U.S. tobacco imports. Similarly, less than 1 percent of L.S. tobacco imports came from Canada, and about 7 per- cent of U.S. tobacco exports event to Canada. Finally, almost no tobacco \vas exported from the Cnited States to Mexico (USDA 1992). Trade in tobacco products (mainly cigarettes) \j.as even more limited before the agreement. In 1990, just over 0.1 percent of total U.S. cigarette exports \Vetit to blexico, and only 0.07 percent \2-ent to Canada. Simi- larly, there \vas no trade in cigarettes betlveen Canada and Mexico. The only exception \vas for exports of cigarettes from Canada to the United States, ~~hich \vere almost 64 percent of total Canadian cigarette ex- ports and almost 20 percent of total Canadian produc- tion (USDA 1992). However, as is discussed later in this chapter (see "International Tobacco Taxes"), most of these cigarettes M'ere reintroduced into a Canadian black market to evade the significantly higher Cana- dian cigarette taxes (Sweanor and Martial 1994). Because of the earlier free trade agreement be- tween the United States and Canada, NAFTA does not appear to have had a significant impact on trade in tobacco and tobacco products between the tw.0 coun- tries. If anything, the reduction in Canadian cigarette taxes in 1994 has led to a substantial reduction in Ca- nadian cigarette exports to the United States, as the smaller differential in cigarette prices reduced the in- centive to export cigarettes to the United States for bootlegging back into Canada. The agreement's elimination of Mexican import li- censes on tobacco and cigarettes, and gradual reduction in Mexican tariffs on tobacco and tobacco products, holyever, \vere expected to increase Mexican imports of both flue-cured and burley tobacco as well as ciga- rettes from the United States (USDA 1992). The elimi- nation of U.S. tariffs on Mexican tobacco and the impro\.ed quality of this tobacco wrere also expected to result in increased Mexican tobacco exports to the United States. I'ri\,atization of the unmanufactured tobacco industry in Mexico, however, has changed the nature of the industry and has led to an improvement in the quality of Mexican leaf tobacco (USDA 1997d). The slog\- elimination of tariffs and the improved qual- ity of domestically grobvn tobacco, coupled with the decline in the \-alue of the peso, appear to have lim- ited the impact of NAFTAon trade between the United States and Mexico in tobacco and tobacco products. This ma\ change, ho\\-e\-er, as tariffs are further re- duced and, el,entually, eliminated and if the peso con- tinues its recent strengthening against the dollar. This latest GATT agreement, which concluded in April 1991, in\rol\,ed 117 countries, and many other nonmembers have agreed to abide by its provisions. Formal appro\.al of the agreement by the U.S. Con- gress came at the end of 1994. Se\,eral basic principles are outlined in GATT: a commitment to achieving free trade by limiting and eventually eliminating tariff and nontariff barriers to trade, the notidiscriniinator); application of any restric- tions on trade to all member countries, the compensa- tion of trading partners for any damages resulting from changes in trade barriers, and the negotiated settle- ment of any trade disputes through an orderly pro- cess rather than through retaliation. However, GATT has had no enforcement power. Since the conclusion of its first round in 1947, GATT has led to sharp reductions in tariffs and other impediments to trade in manufactured goods. Before the most recent round, GATT had not been applied to trade in agricultural commodities or services. The 1994 Uruguay Round, however, significantly expanded GATT's coverage to include trade in agricultural prod- ucts, ser\,ices, and more. Moreover, the new agree- ment created the World Trade Organization, a permanent forum for GATT members to address trade- related issues among member countries. This forum strengthened GATT's ability to resolve trade disputes. Supporters of the GATT treaty have argued that it Lvill lead to a substantial increase in world trade to the economic benefit of all countries inv-olved. For example, President Bill Clinton stated in the introduction to the L~LI~LI~!. Round .~gret~niciits .,\ct that the treat), ~\.htw fullv implemented, ~~.ou]d add 5100-X1(1 billion to the L.Sl econoni\. annuall\, and \j,ould create hundreds of thousands of ne\v jobs. He \\-ent 011 to note tll`lt be- cause the United States is the \rorld's largest trading nation, it \VOLII~ be the biggest bencficiari~ of the treat\ CC .S. Congress 1 YYl). The Urugua!, Round of GATT \j'as expected to benefit the U.S. tobacco industry by reducing the his- torically high tariffs on tobacco and tobacco products imposed in numerous countries and bv reducing other lvidelv used nontariff barriers to trad-e. For example, the European Communitv \\uLII~ reduce tariffs on ci- gars by 50 percent, tariffs & cigarettes and other manu- factured tobacco products bv 36 percent, and tariffs on unmanufactured tobacco bv 20 percent, and the Philippines \\.auld reduce tariffs on leaf tobacco, ci- gars, and cigarettes bv 10 percent (USDA 1YWb). Sim- larlv, foreign access to U.S. markets \\.ould rise, as U.S. tariifs on cigar lvrappers lvould be eliminated. At the same time, U.S. tariffs on cigar filler and binder to- bacco, cigars, and most cl+ `crirettes 1~0uld be reduced by JS percent; tobacco stems and refuse LX, 20 percent; and other unm~~nuiacture~l toL?acco and smoking to- bacco bv 15 percent (USDA 39Y4b). More important, Section 472 of the Lrugua!, Routid Agreements Act allon.ed the President of the United States to \\.ai\.e Section I lOri(a) c)f the Omnibus Budget Reconciliation Act of lYY3 if he determined that this action \vas necessarv or appropriate to conipl\~ \t.ith international trade aqeements that include thi> United States. As noted pre\.iousl>., the IYY3 lcgisla- tion recluiring that cigarettes manufactured in the United States include a minimum of Z percent dc)- niesticall\, gro\\`n tobacco or face penalties \\.a.5 \\.ai,.ed b>, President Clinton's tariff rate-quota proclamation in September 199-I. The redactions in t~~bacco-rcl`ite~l trade barriers achie\.ed in the Uruguay Round appear to ha\.e had a dramatic impact on global trade in tobacco and tobacco products (Chaloupka and Corbctt IYY8). From IYW to 1997, for example, there \~as a 12.5percent increase in unmanufacturecl tobacco exports globall\~, follows ing a decade of almost no grol\.th; similarli; global cigarette exports rose bv 12 percent from 199.3 to I YYh, l\Iiile glo- bal cigarette c&sumption rose by, 5 percent (Chaloupka and Corbett 1998). As discussed pre\~iousl~~, ho\Ve\,er, the GATT Council's resolution of the tobacco-related dispute bet\\-een Thailand and the Cnited States clearl\* indicates that the adoption and implementation of strong tobacco control policies aimed at improving public health is consistent L\.ith the liberalization of trade. Discussion and Recent Developments The threat of retaliatory trade sanctions undel Section 301 of the Trade Act of 1974 has successfu]]\- opened some foreign markets to U.C cigarette manL]- facturers, thereby significantly expanding trade in to. bacco products between the United States and these countries. Chaloupka and Laixuthai (1996), in their empirical examination of these agreements, concluded that the market share of L.S. cigarette companies in the affected countries 1~2s 600 percent higher, on a\.- erage, in 1991 than it should have been in the absence of these agreements. More important, they concluded that o\.erall cigarette smoking rose as a result of the Section 301 agreements. Chaloupka and Laixuthai (I 9%) estimated that per capita cigarette consumption in 19Yl \vas 10 percent higher, on al'erage, in the four countries than it should ha\,e been had the markets remained closed to U.S. cigarettes. They attributed the increase in smoking to gwater competition in the cig- rette markets, resulting in 1oIrer cigarette prices and increased cigarette ad\,ertising. In addition, they pre- dieted that similar actions in other historically clowd countries ~~-0uld lead to similar increases in cigarettt smoking:. SimilarI>., the imylemcnt~itioli of multinationa algrwnients liberalizing trade, including trade in tr> bacco and tobacco products, is likely to further increase L`.S. clports of tobacco and tobacco products to COLIU tries around the \\wrld. A probable consequctic~~ 0 this incrt>ase is that the prices of cigarettes and othe tc>bacco products \\.ill fall as trade barriers are reduce-c or eliminated and competition is enhanced. As is di\ cusst~l in dt,tail later in this chapter (see "Effect of Prim4 on Demand for Tobacco Products"), reductions in prier \\.ill stimulate the uw of cigarettes, particulnrl\r ~llll~i~l~ `lclolt+xcYlts `17ld \ oun g adult5 Bcc~lLlse ot the sLlL~~t~ discour- dge the esecuti\.c branch from assisting U.S. tobacco conipan\i efforts to open forei, `~711 tobacco mdrket5 t Knenier~ lYY3J. Later that \sear. as ;I result of the U.S. Trade Reprewnt,lti\ c'\ in\ cstigdtion ot Thdil~iiid'~ trdde practices, `3 public hedrin, l' 011 the i`l5C' L\ `15 IlCld Sumerous congressmen, public health officials, and others (including former L;.S. Surgeon General C. E\,erett Koop) testified against tobacco-related U.S. trade policies (Eddy and Walden 1993). Although nei- ther effort Leas successful (the bill did not pass, and the hearing produced ii0 change in trade policy), both linhcd the issue vf the health consequences of tobacco LIW to U.S. trade polis\,. The 1990 GAO report, for ex- Ll~npl~, \\.as the direct ;csult of the failed 1989 bill. More rcientlv, interagencv discussions betlveen the cjtfic`e of the U.S. Trade Ripresentative and offi- ci,jls from the USDHHS ha\,e pursued the harmoniza- tion of trade and health policv Lvhile representatives tram the LSDHHS ha\ e participated in recent nego- tiations \\.ith T;lii\-an, South Korea, and others concern- in;: cigart>ttc trade issues (Holzman 1997). Moreover, the U.S. Trade Reprtsentati\~c has shown greater sen- siti\ it\, to public health concerns and has not opposed nr,nililcrimin~t~~r~, tobacco control legislation in other c~~untrics (Bloom .I YYX; National Cancer Policy Board IYYS). This position has been formalized as pait of the Doggt'tt Amendment to the Department of Commerce and Relattd Aseiicies Appropriations Act, 1998, that allo\\,s for the uw of Section 301 ill \`ery limited cir- cumst,inces. Specifically, the Doggett Amendment, spconsored b\, Lloyd Doggett (D-TX), states that: None of the funds pro\icled bv this Act shall be a\.ailable to promote the sale or export of tobacco or tobacco products, or to seek the reduction or remo\-al by anv foreign country of restrictions on the marketing of tobacco or tobacco products, ex- ccpt for restrictions \1.hich are not applied equally to all tobacco or tobacco products of the same type (Public Lar\~ 105119, Section 61X). Similar glidelines rz.ere Distributed by the Clinton administration to all diplomatic posts in February 1998. These guidelines state that: In light of the serious health consequences of to- bacco use, the U.S. Government \vill not promote the sale or export of tobacco or tobacco products or seek the reduction or removal by any foreign country of iiondisci~iiiiiiiator~ restrictions on the marketing of tobacco or tobacco products. At the same time, the U.S. Government will continue to seek elimination of discriminators trade prac- tices and lvill stri1.e to ensure that U.S. firms are accorded the same trcatnient in fowign countries ~5 that countrv's o\vn firms and firms from other countries (The National Economic Council and The N~ltion,ll Styurit\. Council of the White House, Final Guiclelincs on IHealth, Trade, and Commer- cial Issues, facsimile transmission to all diplomatic and consular posts, February 16, 1998). Moreo\,er, as part of the guidelines, U.S. diplomatic "posts are encouraged to assist and promote tobacco- control efforts in host countries." Several important issues remain unresolved. Perhaps most important is the opening of Chinese ciga- rette markets to U.S. and other multinational tobacco companies as part of China's World Trade Organiza- tion accession. With more than 300 million cigarette smokers (67 percent of men but only 7 percent of lvomen), China is a particularly attractive market for international cigarette producers. In recent years, U.S. and other multinational tobacco companies have en- tered the Chinese tobacco markets through joint ven- tures with the Chinese government's tobacco monopoly, the China National Tobacco Corporation (Holzman 1997). Economic Impact of the U.S. Tobacco Industry Tobacco grooving played a key role in the devel- opment and groll-th of the U.S. economv. Throughout much of the 20th century, holve\,er, the-importance of tobacco to the overall economy has diminished sig- nificantly, although its regional and local importance in some areas remains high. Several recent studies provide more detailed c\.idence concerning the eco- nomic importance of tobacco to the U.S. economv. A recent study by American Economics Cr&p, Inc. ([AEG] 19%), lrhich i1.a~ funded bv the tobacco industry, provides some information concerning the impact of tobacco on the U.S. economy in 1994. The report updates similar pre\,ious reports by other firms, including that by Price Waterhouse (lYY2). AEG di- vided the macroeconomic effects of tobacco into those affecting the core sector, \\,hich includes tobacco pro- duction and distribution, and those affecting the sup- plier sector, ivhich consists of industries producing and distributing intermediate goods for the core sector (in- cluding the goods and services used in cigarette prc>- duction). The analysis also separately considered expenditure-induced impacts, Lvhich depend on the multiplier effects associated with spending by those in the core and supplier sectors, and tobacco-related tax revenues, including those raised by tobacco taxes, general sales taxes on tobacco products, and income and other taxes on tobacco industrv emplovees and firms. The studv estimated that in 1494, mori than I .8 million persons were employed, earning $54.3 billio in wages and benefits, as a result of the tobacco busj ness in the United States. Total estimated tax revenue from tobacco were almost $36 billion in 1994. The rt port concluded that tobacco made a significant contr bution in every state and the District of Columbia. Several recent studies, however, have indicate, that these estimates significantly overstated the ccc nomic impact of tobacco on the U.S. economy. At th request of the Coalition on Smoking OR Health (CSH Arthur Andersen Economic Consulting (1993) rc viewed the Price Waterhouse estimates for 1990. The concluded that, as a result of several methodologic: flaws, the Price Waterhouse "employment and job lo: figures are grossly inflated" (p. 1). For example, of th 681,351 jobs Price Waterhouse attributed to tobacco i its core and supplier sectors, only 259,616 were direct1 related to tobacco growing, manufacturing, warehou: ing, and wholesaling. Of the difference, 166,791 wer retail jobs and 254,944 were supplier jobs, most of whit Lvere not devoted full-time to tobacco. Thus, statin that these jobs depended on tobacco was inaccurate. Other studies questioned the Price Waterhous assumption that every one job that is dependent o tobacco creates, through the multiplier effect, an add tiona12.35 jobs throughout the economy. This assume effect would result because those who purchase tobacc products would generate income for those who produc and those who distribute tobacco, who in turn woul spend this income on other goods and services-thereb generating income for others, as this effect spread eve further. Warner (1994) and Arthur Andersen Econom Consulting (1993) noted that this multiplier effect likely to significantly overstate the impact of tobaccf because it implicitly makes the incorrect assumptio that money spent on tobacco would not be spent else Lvhere in the absence of tobacco. Instead, those func not spent on tobacco would be spent on other gooc and services, creating jobs and generating income th; ~rould also be spent. Warner and Fulton (1994) addressed these issue by using a macroeconomic model to consider the nl impact of tobacco on the economy of one state, Mich gan. The Price Waterhouse study had estimated thi direct tobacco-related employment in Michigan w: 7,721 in 1YYO and that all tobacco-related employmel in Michigan totaled 69,575. Warner and Fulton (198 estimated that in 1992 in Michigan, 7,843 jobs direct1 depended on tobacco but that only an additional 11,28 jobs lvere either indirectly related to tobacco or induce by spending from those whose jobs were dependeI on tobacco. (This estimate for indirect tobacco-relate jobs did not consider [as the Price Waterhouse estimai did] the impact of income deri\,ed from tobacco pro- duction and distribution in the rest of the nation and spent on products produced in Michigan.) These re- searchers further estimated that, in the absence of to- bacco, total employment in Michigan should ha\.e risen b!, about 5,600 because of a redistribution of spending aLcay from tobacco products to other goods and ser- I-ices, including those more integral to the Michigan economv. As a result of the changes in employment, total incbmes in Michigan ~\fould have been $226 mil- lion higher in 1992 in the absence of tobacco. This amount resulted not only from incomes associated xvith nelv jobs but also from higher incomes for those \lith existing jobs (in part because of a change in job mix from lolver-income to higher-income jobs in the absence of tobacco). Warner and colleagues (1996) extended this analysis to examine the impact of tobacco on the re- gional economies of the United States. The research- ers examined the effects of reducing or eliminating domestic expenditures on tobacco on nine regional economies (the eight regions defined bv the L'.S. Dc- partment of Commerce, Bureau of Economic Anal\,- sis, subdi\.iding the Southeast into tit-o parts based on tobacco grooving and producing). The\, estimated that the elimination of spending on tobacco products in 1993 IVould ha\,e led to 303,000 fexver jobs in the South- east tobacco region, 12-bile increasing jobs in all othci regions bv about the same amount. B\, the \`ear 2000, they estimated that, under this scenario, the ioss in jobs in the tobacco region ~~ould fall to about 222,000 as the regional economy adjusts, \Vhile the net impact nationally \vould be an increase in jobs of 133,000. A more realistic scenario-one that doubles the recent rate of decline in tobacco use-is estimated to ha1.e smaller effects on employment. Warner and colleagues (1996) estimated a loss of 36,600 jobs in the tobacccl region by the year 2000, an amount equal to 0.2 per- cent of total regional emplovment. They concluded that the industry's claims concerning job losses resulting from reduced tobacco use are significantly overstated and that the impact of tobacco on employment should not be a primary concern, given the magnitude of the toll it takes on health. The AEG and Price Waterhouse reports \vere lim- ited also because thev presented static estimates of the economic impact of tobacco (Arthur Andersen Economic Consulting 1993). That is, the reports ignored underly- ing trends in the domestic demand for cigarettes, trends in the import and export of tobacco and tobacco prod- ucts, and changes in agricultural and manufacturing technologies that themselves are reducing employment in tobacco grooving and manufacturing. Warner and Fulton (199-I) considered these factors by predicting the net impact that eliminating tobacco-related revenues t~,ould have on the Michigan economy if existing down- \\,ard trends in tobacco sales continued: by 2005, the loss of re\`enue from tobacco in Michigan would yield a net gain of 1,500 jobs in the state. A similar issue was considered in two recent re- ports of the USDA (1993, 1997~). The reports noted that the large declines in tobacco production through- out the 1980s had a relatively minor impact on the macroeconomics of major tobacco-growing regions. Indeed, total personal income, adjusted for inflation, gre\l- bv 13-57 percent from 1979 through 1989 in the nine major regions analyzed; the average growth in all U.S. tobacco-grooving counties was 28 percent (USDA 1993). This phenomenon M'as attributed to the relatil-elv small share of tobacco in these diverse re- gional economies (on average, less than 1 percent of total income ivas accounted for by tobacco in tobacco- cTro\\.ing counties). E\,en though acreage devoted to b tobacco grooving has declined over time, rising prices ha\,e helped to keep gross income from tobacco grow:- in;: relati\.clv stable, lvhile clearlv reducing the share of tobacco iti local economies (USDA 1997~). Critics of higher cigarette excise taxes and other policies to reduce tobacco USE have argued that the ni~lcroeconomic consequences of these policies would be significant, particularly for some state and local economies. For example, economist Dwight R. Lee predicted that the 75-cent increase in the federal ciga- rette excise tax included in the proposed 1993 Health Securitv Act should lead to a loss of about 82,000 jobs and 51 .G billion in incomes in the tobacco sector, which \j-ould cause an additional loss of 192,000 jobs and an attendant loss of income throughout the economy (U.S. House of Representatives 1994). He further noted that southern states lvould be particularly hard hit by this tax increase. Similar arguments, based on the AEG and Price Waterhouse analvses, were made in the recent debate over proposed national tobacco legislation. For rea- sons noted pre\Tiously, predictions based on these es- timates are almost certain to substantially overstate the effects of higher tobacco taxes and stronger preven- tion policies on the U.S. macroeconomy. As discussed previously, Warner and colleagues' (1996) regional analysis of the economic role of tobacco concluded that tobacco has a negative net economic impact in all but the most tobacco-dependent region. Thus, it ap- pears inappropriate to raise concerns about adverse economic impact in opposing policy measures that ~~~ould cliscourage tobacco use. Moreo\w, nidn\. supporters ot legislation calling for increases in the ciiarettc excise ta\ ha\ e urged that measures be included to mitigate the p