Addictive Behaviors 35 (2010) 10891093

Contents lists available at ScienceDirect

Addictive Behaviors

Revenue implications to the Vietnamese government of using taxes to curb cigarette smoking
Christopher M. Doran a,, Joshua M Byrnes a, Hideki Higashi b, Khoa Truong c
a b c

National Drug and Alcohol Research Centre, University of New South Wales, Australia School of Population Health, University of Queensland, Australia Department of Public Health Sciences, Clemson University, United States of America

a r t i c l e
Keywords: Tobacco Revenue Government Vietnam

i n f o

a b s t r a c t
This study explores the impact on government taxation revenue from increasing excise on cigarettes in Vietnam. A dynamic population model is used to estimate future patterns (both prevalence and consumption) of tobacco use in Vietnam, with and without changes to tobacco excise for the period 20062016. Three increases in the base case excise tax rate of 55% are modelled: 65%, 75% and 90%. Various price elasticities are used to examine variations in cigarette consumption while cross price elasticities are used to explore shifts from cigarette to other forms of tobacco. Revenue implications for the period 2006 2016 are reported as discounted net present values (NPV) in 2006 values. The model predicts that smoking rates in 2016, for both males and females, are marginally lower than base case estimates for all taxation excise options with higher price elasticities generating greater reductions in prevalence. In all cases, compared to base case estimates, the results indicate a fall in number of smokers, a reduction in amount of tobacco consumed and an increase in overall taxation revenue. The additional gain in government revenue, expressed in NPV terms, ranges from a low of VND 69,579 billion (or USD $4.35 billion) to a high of VND 108,492 billion (or USD $6.79 billion). Increases in tobacco excise provide an opportunity for the Vietnamese government to increase revenue at the same time as reducing tobacco consumption. Further research into the wider social and economic consequences of increasing tobacco excise in Vietnam is warranted. 2010 Elsevier Ltd. All rights reserved.

1. Introduction In Vietnam, almost 1 in 2 males (47%) and 1 in 50 females (2%) consume some form of tobacco (General Statistics Ofce, 2007a). Smoking prevalence is lower than a decade earlier when more than 50% of men and 3% of women smoked tobacco products, although most of this reduction appears to have taken place between 1993 and 1998 (General Statistics Ofce, 1994, 2000). Smoking prevalence appears to be evenly distributed between urban and rural areas although tobacco users in different areas appear to smoke different products. Waterpipe smoking is more prevalent in rural areas while cigarettes are more popular in urban areas (Ministry of Health, 2003). Among urban male smokers, 48.6% smoked cigarettes only and 3.8% smoked waterpipe tobacco only. This is in comparison to rural male smokers of whom 35.6% smoked cigarettes only and 16.0% smoked waterpipe tobacco only (Ministry of Health, 2003). The production of cigarettes has increased steadily since 2000 (Asian Development Bank, 2009). In the period 200006, total cigarette
Corresponding author. National Drug and Alcohol Research Centre, University of New South Wales, University of New South Wales, Sydney, NSW 2052, Australia. Tel.: +61 2 93850324; fax: +61 2 93850222. E-mail address: c.doran@unsw.edu.au (C.M. Doran). 0306-4603/$ see front matter 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.addbeh.2010.07.010

production increased by approximately 42% (Asian Development Bank, 2009). In 2007, production was estimated at 4320 million packs (Asian Development Bank, 2009). The introduction of tax stamps, and therefore more accurate reporting, is attributed to the majority of the 32% growth in cigarette output in 2000 (Kinh & Bales, 2003). The tobacco industry in Vietnam is largely under government control, with a few joint ventures with multinational companies. These joint ventures between the government and multinational companies are involved with all stages of production, from tobacco growing and processing (joint venture with British American Tobacco) to cigarette manufacturing (joint venture with Philip Morris) and accessory production (joint venture with New Toyo). The majority of cigarettes produced in Vietnam are made by the stateowned Vietnam Tobacco Corporation (Vinataba) and its subsidiaries, which currently own 11 of the country's 17 factories and produce more than 200 brands nationwide (Euromonitor, 2007). Vietnam has a tropical climate, which is very conducive to tobacco cultivation, though output is unstable (Kinh & Bales, 2003). Between 2000 and 2005 Vietnam produced between 23,000 and 33,000 tons of tobacco leaves annually (General Statistics Ofce, 2007b). Vietnam is both an exporter and an importer of tobacco leaves, and primarily an importer of various other inputs into the tobacco industry such as special papers and aromas added to tobacco to make cigarettes.

1090

C.M. Doran et al. / Addictive Behaviors 35 (2010) 10891093

Between 2001 and 2004, Vietnam imported between 12,000 and 15,000 tons yearly (about 40 to 50% of local production) (General Statistics Ofce, 2007b). High-quality cigarette brands with lters tend to be produced using tobacco grown and processed abroad. Several types of cigarettes are produced with imported tobacco leaves processed domestically. The lower quality cigarettes rely primarily on domestic supply of tobacco leaf or illegal imports of tobacco from China (Kinh & Bales, 2003). The government's expansion policy for the tobacco-growing sector aims to stabilize annual production by 2010 at around 40,000 ha under cultivation and 80,000 tons of processed leaf, with a stated goal of domesticating tobacco supply by 2015 so that all cigarette production companies will use only domestically processed tobacco (Guindon, 2010). Tobacco cultivation employed about 72,000 full-time equivalent (FTE) workers in 2006, or about 0.2% of the workforce (0.4% of agricultural workers) (General Statistics Ofce, 2007b). Previous estimates by Kinh and Bales (2003) report the size of the workforce involved in cultivation and simple post-harvest processing to be as high 136,000 FTE workers. Tobacco farmers receive wide support from Vinataba in the form of seeds, capital and rural infrastructure. As of 2006, cigarette production employed around 18,000 workers or 0.05% of Vietnam's workforce (General Statistics Ofce, 2007b). Employment in cigarette manufacturing remains stable amidst increasing production suggesting gains in production efciency. In general, workers' incomes in cigarette factories are reported as high relative to the average worker income in the same localities (Kinh & Bales, 2003). Despite the presence of an ofcial licensing system, the number of people involved in cigarette distribution is largely unknown. The number of FTE employed in cigarette distribution has been estimated at approximately 23,000 (Kinh & Bales, 2003). In total, the tobacco industry accounts for approximately 0.3% of all jobs in Vietnam with the number employed in cultivation appearing to have been declining for some time. Tobacco control in Vietnam has over the past few decades undergone signicant transformations. The rst sign of commitment to tobacco control began in 1989 with the government of Vietnam introducing a law banning smoking in designated public places. The law was reinforced in 1991 by a government decree and in May 1991, the Ministry of Health establishes the Vietnam Committee on Smoking and Health (VINACOSH), a steering committee for Tobacco Control. In 2000, the government furthers its commitment to tobacco control by adopting a national tobacco control policy with the overall objective of reducing the male smoking rate from 50% to 20% and maintain the female smoking rate below 2%. Tobacco products in Vietnam are taxed under three separate taxation policies: a special consumption tax (tobacco excise), a valueadded tax and an enterprise tax. An excise tax rate of 65% is charged on the pre-tax factory price of cigarettes and cigars; VAT is charged at a rate of 10% of the pre-VAT retail price of cigarettes; and an enterprise tax rate of 28% is applied on the net prots of enterprises selling cigarettes. A plethora of research indicates that taxation is one of the most cost-effective strategies to reduce the burden of harm from tobacco use (Gallet & List, 2003; Jha et al., 2006). Higher taxes are relatively inexpensive to introduce and have the dual benet of addressing a serious public health problem at the same time as raising additional revenue. Higher taxes work by preventing the uptake of smoking by the young and creating an incentive for regular smokers to either quit or reduce smoking (Kinh et al (2006)). Governments achieve a net increase in revenue as the additional revenue collected from the higher tax paid by existing smokers is greater than foregone revenue from those smokers that quit or reduce their consumption. However, countries such as Vietnam, that have a vested interest in tobacco production, cultivation and use, may question the revenue implications of using tobacco taxation to impact on the industry. This study explores the impact on government revenue collected

from excise and value-added tax as a consequence of increasing the cigarette excise tax rate in Vietnam. 2. Method In order to estimate the revenue implications of higher tobacco taxes, the number of cigarettes sold without government intervention (i.e., no tax rise) is compared to the number of cigarettes sold with government intervention (i.e., tax rise). The excise tax is applied to the factory price of manufactured cigarettes, which was derived by dividing the value of output reported by the General Statistics Ofce (2007b) by the volume of cigarettes produced reported by the Asian Development Bank (2009). Value-added tax is calculated on the retail price of cigarettes. The average retail price of cigarettes was estimated by taking the weighted average of low grade cigarettes and high grade cigarette prices based on market share. Retail and factory prices are assumed to remain constant in real terms. Tax increases are only applied to the legal sales of cigarettes, approximately 91% of the cigarette market. Further, the proportion of illegal cigarette sales is assumed to remain constant and unresponsive to increases in the price of legal cigarettes. The year 2006 is used as our base year with future projections over the subsequent 10 years. A constant discount rate of 3% is applied to future taxation revenue to provide a present value estimate for comparison and evaluation. 2.1. Base case The base case or business as usual scenario estimates the agespecic smoking uptake and cessation rates by tting the parameters to the observed prevalence data between 1993 and 2006 from the Vietnam Living Standard Surveys (VLSS) (General Statistics Ofce, 1994, 2000, 2007a). Weighted least square method is used for estimation. The estimated age-specic uptake and cessation rates are then used to project the future prevalence of smoking in Vietnam without the proposed interventions. This projection utilises a dynamic population model thereby accounting for population growth. The future rates of smoking prevalence in Vietnam are then disaggregated by tobacco product, namely, cigarettes and waterpipe. The proportion of cigarette smokers for 1998 was taken from the VLSS. The future proportion of cigarette to waterpipe smokers was projected using International Monetary Fund (2009) forecast for Vietnam growth in GDP and the switching income elasticity from rustic tobacco to cigarettes measure from Laxminarayan and Deolalikar (2004). 2.2. Increase in tobacco excise tax The increase in excise tax is modelled as an increase in the tax rate applicable in 2006. The base tax rate is 55% for 2006. Three increases are modelled, increasing the 2006 excise rate to 65%, 75% and 90%. A single price elasticity estimate is used for the whole population. The price elasticity estimate derived by Kinh et al. (2006) for Vietnam is utilised to calculate the impact on participation (number of smokers) and quantity (smoked by those who still continue to smoke). Kinh et al. (2006) elasticity estimates for participation and quantity are 0.94 (dividing the estimated log coef cient of cigarette price by the smoking prevalence in 2002, i.e. 0.529/0.56) and 0.469 respectively. Kinh et al. (2006) calculate that using these estimates for participation and quantity results in a total price elasticity estimate 1.41. However, we calculated the price elasticity for participation by dividing the coefcient by the smoking prevalence observed in 1998 (i.e. 51%) to make it consistent with the year of data the coefcient was estimated from. Within the literature, total price elasticity estimates vary to some extent. As such, participation estimates of 0.446, 0.206 and quantity estimates of 0.192 and

C.M. Doran et al. / Addictive Behaviors 35 (2010) 10891093 Table 1 Price effect on participation and quantity. Price elasticity estimate 1 2 3 Participation 1.039 0.446 0.206 Quantity 0.469 0.192 0.088

1091

0.088 are used to supplement the ndings. Table 1 presents the participation and quantity estimates employed in this analysis. Laxminarayan and Deolalikar (2004) provide evidence that an increase in the real price of cigarettes not only has an initial impact on the decision to smoke or not but a continuing effect on individuals' initiation into smoking. Laxminarayan and Deolalikar (2004) nd a 1.175 own price elasticity on the initiation into smoking cigarettes. A sensitivity analysis is conducted to assess the impact on our nal estimates of taxation revenue from incorporating the continuing impact of higher cigarette prices to dissuade initiation into smoking. As was mentioned earlier, waterpipe or rustic tobacco in Vietnam is not taxed. Furthermore, illegally smuggled cigarettes are estimated at up to 9% of the total cigarette market in Asia (Framework Convention Alliance, 2008). As such, an increase in the price of legal cigarettes comparative to other tobacco products provides incentives for smokers to change their tobacco preferences. This has the potential to reduce the taxation revenue collected from tobacco in Vietnam. Laxminarayan and Deolalikar (2004) provide an estimate of the impact of cigarette price increases on smokers switching from cigarettes to rustic tobacco. This estimate suggests that a 10% increase in the nal price of cigarettes results in a 13.95% switch from cigarettes to rustic tobacco. A sensitivity analysis is performed to measure the impact on taxation revenue from incorporating the potential effect of cigarette smokers moving from cigarettes to waterpipe tobacco. 3. Results Results of the impact upon the prevalence of smoking and the impact upon taxation revenue (excise and VAT) are presented for the base case and the tobacco excise interventions. 3.1. Base case The base case or business as usual scenario predicts the prevalence of smoking, the number of manufactured cigarettes legally sold and the resulting taxation revenue that would occur in the absence of government intervention. Table 2 presents a summary of the results for the base case. It can be seen from this
Table 2 Base case results: 20062016. 2006 Prevalence Male Female Cigarette smokers ('000s) Male Female Total Cigarette quantity (million packs) Male Female Total Sales (VND billion) Taxation (VND billion) Excise VAT 49.5% 1.61% 2007 48.6% 1.51% 2008 47.8% 1.43% 2009 47.1% 1.35%

table that despite falling smoking prevalence rates, the number of cigarette smokers is predicted to increase over time. For example, male smoking prevalence falls from 49.5% in 2006 to 45.4% in 2016 whilst the number of male cigarette smokers increases from 11,862,000 males to 13,547,000 males. This result reects the population growth and the effect of rising incomes on smokers' preference for cigarettes over rustic tobacco. The total taxation revenue derived from tobacco excise tax and value-added tax for the period 2006 to 2016 is VND 113,979 billion. Applying a 3% discount rate to future revenue collected, the net present value (NPV) in 2006 of the tax revenue from 2006 to 2016 is VND 95,518 billion, equivalent to USD $5.97 billion. 3.2. Increase in tobacco excise tax results Table 3 presents a summary of the 2016 smoking prevalence rates, number of smokers, quantity of cigarettes consumed, sales, taxation receipts and net present value of these taxation receipts for the period 20062016. The table summarizes these results for each of the three taxation policy options using the varying price elasticity estimates. Smoking prevalence rates in 2016, for both male and females, are marginally lower after increasing the excise tax rate to 90% compared with other taxation excise options and the base case estimates. It is important to note the impact on prevalence of using a higher price elasticity estimate (i.e., price elasticity 1) compared with a lower price elasticity estimate (i.e., price elasticity 3). For example, increasing the excise rate to 75% using price elasticity 1 results in an estimated 2016 male smoking prevalence of 41.8% compared with a male smoking prevalence rate of 44.7% using price elasticity estimate 3. In all cases of higher taxation, the results indicate a decrease in number of total smokers and amount of tobacco consumed, compared with the base case. These decreases are achieved in conjunction with higher overall tobacco taxation receipts. Although the amount of revenue generated by VAT decreases, as a consequence of a decrease in packets sold, the increased revenue obtained from excise more than offsets this. The NPV of total taxation receipts under all options is substantially higher than base case estimates. The additional gain in government revenue, expressed in NPV terms, ranges from VND 69,579 billion (or USD $4.35 billion) to VND 108,492 billion (or USD $6.79 billion) for a 90% excise rate price elasticity 1 and 3, respectively. 3.3. Sensitivity analysis The results after the inclusion of a continuing impact upon initiation (Scenario 2) are compared with the results produced from accounting for a switch from cigarettes to rustic tobacco (Scenario 3);

2010 46.6% 1.28%

2011 46.3% 1.21%

2012 46.1% 1.14%

2013 46.0% 1.09%

2014 45.8% 1.03%

2015 45.6% 0.98%

2016 45.4% 0.93%

11,862 413 12,275

12,053 404 12,457

12,242 394 12,636

12,392 383 12,775

12,537 371 12,908

12,710 359 13,069

12,905 348 13,253

13,096 337 13,432

13,273 326 13,600

13,423 316 13,739

13,547 305 13,852

2503 61 2564 21,120

2544 60 2603 21,441

2583 58 2642 21,757

2615 57 2672 22,004

2646 55 2701 22,242

2682 53 2735 22,527

2723 52 2775 22,854

2764 50 2814 23,171

2801 48 2849 23,467

2833 47 2880 23,715

2859 45 2904 23,916

5078 4620

5155 4690

5231 4759

5291 4813

5348 4865

5417 4928

5495 4999

5571 5069

5643 5133

5702 5188

5751 5232

1092 Table 3 Excise rate results for 2016. Base case Prevalence Male Female Cigarette smokers ('000s) Male Female Total Cigarette smokers ('000s) Male Female Total Sales (VND billion) Taxation (VND billion) Excise VAT Net present value (VND billion) Net present value (USD billion)*

C.M. Doran et al. / Addictive Behaviors 35 (2010) 10891093

Excise rate = 65% PE = 1 43.6% 0.9% PE = 2 44.6% 0.9% PE = 3 45.0% 0.9%

Excise rate = 75% PE = 1 41.8% 0.8% PE = 2 43.8% 0.9% PE = 3 44.7% 0.9%

Excise rate = 90% PE = 1 39.2% 0.8% PE = 2 42.7% 0.9% PE = 3 44.1% 0.9%

45.4% 0.9%

13,547 305 13,852

12,801 285 13,086

13,154 295 13,449

13,297 299 13,595

12,180 267 12,447

12,888 287 13,176

13,175 295 13,470

11,238 241 11,479

12,489 276 12,765

12,991 290 13,282

2859 45 2904 23,916

2637 41 2678 22,056

2749 43 2792 22,994

2793 44 2838 23,368

2448 38 2485 20,468

2667 42 2709 22,306

2755 43 2799 23,048

2173 33 2206 18,169

2545 40 2585 21,289

2698 42 2741 22,572

5751 5232 95,518 5.97

6268 4825 165,521 10.35

6534 5030 172,343 10.78

6641 5112 175,069 10.95

6711 4477 166,545 10.42

7314 4879 181,063 11.32

7557 5042 186,937 11.69

7149 3974 165,097 10.33

8376 4657 192,648 12.05

8881 4938 204,010 12.76

*VND: USD exchange rate = 15,988:1 (XE currency converter); PE refers to price elasticity and values correspond to those reported in Table 1.

the results derived from including both effects (Scenario 4) and not including the effects (Scenario 1). All scenarios account for an increase in the excise rate to 75% using price elasticity 2. A comparison of the 2016 male and female smoking prevalence rates, the 2016 number of cigarette smokers and the NPV of taxation revenue over the period 2006 to 2016 using a 3% discount rate is presented in Table 4. It is clear from Table 4 that the inclusion of a continuing impact upon initiation rates from an increase in price (Scenario 2) has only a marginal effect upon both prevalence and taxation revenue estimates when compared with Scenario 1. However, accounting for the potential for continuing cigarette smokers to switch their tobacco preferences has a signicant impact upon the taxation revenue estimate with no effect on the overall smoking prevalence for males or females. 4. Discussion The purpose of this study was to explore the impact upon government taxation revenue collected from excise and valueadded tax from increasing the excise tax rate applicable to cigarettes in Vietnam. The results of this analysis suggest that taxation increases are an effective policy option that can be used by the Vietnam government to simultaneously curb tobacco use and raise revenue. Although this is not surprising given the international evidence around the cost-effectiveness of tobacco taxation, these results are useful for governments of developing countries contemplating the implementation of more effective tobacco control policies.

From the results presented in this study it is evident that assumptions regarding the level of smokers' responsiveness to price inuence the net impact on prevalence rates and subsequent revenue from increases in tobacco taxes. Although a range of price elasticities were applied in this study, each rate was applied for the whole population. International evidence suggests that the responsiveness to price is likely to differ by age, income and geographic location. Another potential limitation of our analysis is the lack of attention to non-taxed tobacco products such as waterpipe tobacco. The option of substituting cigarettes for water pipe tobacco or other types of black market tobacco may have an impact on government revenue from increased cigarette excise rates and undermine the effectiveness of taxation strategies on reducing smoking prevalence rates. This study also did not consider any potential increase in smuggling of illegal cigarettes into Vietnam as a result of increases in the legal cigarette price. Although it may be argued that price is only one of many factors that contribute to smuggling activity, increases in the retail price of legal cigarettes may be expected to have some effect. Further, our analysis did not take into account other tobacco control policies and factors that can affect the predicted smoking prevalence. In spite of the potential caveats, the results of this study demonstrate that increases in tobacco excise provide an opportunity for the Vietnamese government to increase revenue at the same time as reducing tobacco consumption. However, as noted above, taxation implications are but one consequence of tobacco policy. Further research into the wider social and economic consequences of increasing tobacco excise in Vietnam is warranted.
Role of funding sources The researchers would like to acknowledge the University of Queensland and Atlantic Philanthropies for nancial support. The funders had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

Table 4 Sensitivity analysis results: 2016. Scenario 1 Scenario 2 Scenario 3 Scenario 4 Prevalence Male Female Cigarette smokers ('000s) Male Female Total 43.8% 0.9% 43.0% 0.9% 43.8% 0.9% 43.0% 0.9%

12,888 287 13,176

12,652 286 12,937 102,980 6.44

11,800 263 12,063 93,344 5.84

11,583 262 11,845 92,544 5.79

Contributors All authors participated in the conception, analysis and writing of the manuscript. In addition Doran was responsible for submitting the manuscript to Addictive Behaviors. All authors have reviewed the nal manuscript submission and have approved the submission of this manuscript to Addictive Behaviors. Conict of interest No authors have any actual or potential conict of interest.

Net present value (VND billion) 103,869 Net present value (USD billion) 6.50

C.M. Doran et al. / Addictive Behaviors 35 (2010) 10891093 Acknowledgements The authors would like to thank Professor Jan Barendregt (University of Queensland) for assistance in forecasting smoking prevalence and the cost-effectiveness team at Health Strategy and Policy Institute.

1093

References
Asian Development Bank (April 2009). Key indicators for Asia and the Pacic 2008. Accessed online. www.adb.org/Statistics Euromonitor (2007). Baisha group. Local company prole series. London: Euromonitor International. Framework Convention Alliance (2008). How big was the global illicit tobacco trade problem in 2006? Geneva: Framework Convention Alliance. Gallet, C. A., & List, J. A. (2003). Cigarette demand: A meta-analysis of elasticities. Health Economics, 12, 10. General Statistics Ofce (1994). Vietnam living standard survey (VLSS), 19921993. Hanoi: Statistical Publishing House. General Statistics Ofce (2000). Vietnam living standard survey (VLSS), 19971998. Hanoi: Statistical Publishing House. General Statistics Ofce (2007a). Result of the Vietnam household living standards survey 2006. Hanoi: Statistical Publishing House.

General Statistics Ofce (2007b). Statistical handbook of Vietnam, 2006. Hanoi: Statistical Publishing House. Guindon, G.E., Nguyen TT Hien, Hoang V Kinh, McGirr E, Dang V Trung, Nguyen T Lam (2010). Tobacco Taxation in Vietnam. Paris: International Union Against Tuberculosis and Lung Disease. International Monetary Fund and International Development Association. (2009). Joint IMF/World Bank debt sustainability analysis 2008. Jha, P., Chaloupka, F. J., Moore, J., Gajalakshmi, V., Gupta, P. C., Peck, R., et al. (2006). Tobacco addiction. In D. Jamison, J. Breman, A. Measham, G. Alleyne, D. Evans, P. Jha, A. Mills, & P. Musgrove (Eds.), Disease control priorities in developing countries, Second Edition. . New York: Oxford University Press Chapter 46. Kinh, H. V., & Bales, S. (2003). Tobacco in Vietnam: The industry, demand, control policies and employment. Economic, social and health issues in tobacco control. Report of a WHO international meeting. Kobe: Centre for Health Development, World Health Organisation. Kinh, H. V., Ross, H., Levy, D., Minh, N. T., & Ngoc, V. T. (2006). The effect of imposing a higher, uniform tobacco tax in Vietnam. Health Research Policy and Systems, 4, 6. Laxminarayan, R., & Deolalikar, A. (2004). Tobacco initiation, cessation and change: Evidence from Vietnam. Health Economics, 13, 11911201. Ministry of Health (2003). Vietnam national health survey (VNHS), 200102. Hanoi: Ministry of Health.