Article Text

Tobacco and poverty: evidence from Vietnam
1. Debra Efroymson1,
2. Hoang Anh Pham2,
3. Lori Jones3,
4. Sian FitzGerald3,
5. Le Thi Thu2,
6. Le Thi Thu Hien2
2. 2HealthBridge Foundation of Canada in Vietnam, Hanoi, Vietnam
1. Correspondence to Debra Efroymson, Health Bridge Foundation of Canada, c/o House #49, Road #4/A, Dhanmondi Dhaka 1209 Bangladesh; debra{at}healthbridge.ca

## Abstract

This review examined existing evidence to investigate the link between tobacco and poverty in Vietnam, to assess the impact of tobacco control policies on employment related to tobacco consumption and to identify information gaps that require further research for the purposes of advocating stronger tobacco control policies. A Medline, PubMed and Google Scholar search identified studies addressing the tobacco and poverty association in Vietnam using extensive criteria. In all, 22 articles related either to tobacco and health or economics, or to the potential impact of tobacco control policies, were identified from titles, abstracts or the full text. 28 additional publications were identified by other means. PHA, LTT and LTTH reviewed the publications and prepared the initial literature review. There is extensive evidence that tobacco use contributes to poverty and inequality in Vietnam and that tobacco control policies would not have a negative impact on overall employment. Tobacco use wastes household and national financial resources and widens social inequality. The implementation and enforcement of a range of tobacco control measures could prove beneficial not only to improve public health but also to alleviate poverty.

• Tobacco use
• poverty
• Vietnam
• economics
• public policy
• taxation and price

## Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

## Introduction

International studies have demonstrated that tobacco use increases poverty both directly and indirectly. In Albania smokers spend 2 months' wages on cigarettes annually1; Bangladeshi smokers spend on average more than twice as much on cigarettes as on clothing, housing, health and education combined. If such expenditures were instead spent on food, 10.5 million malnourished Bangladeshi children would be sufficiently fed.2 The poorest Filipino households spend 2.5% of their income on tobacco, more than on clothing, education or healthcare.3 At least 15% of poor Indonesian households' disposable income goes to tobacco; in Burma (Myanmar), the rate is 5%. The annual Cambodian tobacco expenditure would cover the US$69 million national deficit.4 Poor adults and street children in Mumbai spend less on food than on tobacco.5 Smoking also contributes to disability and death, increased medical expenditures and reduced productivity, which in turn can increase poverty. In 1990, tobacco caused an estimated 3.7% of total deaths in developing countries. The Disability Adjusted Life Years (DALYs) lost on account of tobacco were 16.7 million, which is about 1.4% of total DALYs.6 In China, the economic costs of smoking in 2000 amounted to US$5 billion. The direct cost of smoking accounted for 3.1% of China's health expenditure in 2000.7 In Bangladesh, the total annual direct and indirect cost of tobacco-related illnesses was estimated at US$1.1 billion in 2004.8 Research on tobacco in Vietnam began in the mid-1980s but has evolved slowly given its low priority and a lack of resources. With a general poverty rate of 16% in 20069 and smoking rates among the highest in the world, greater awareness of what Vietnam-specific evidence exists about the tobacco-poverty link is crucial to inform tobacco control policy advocacy. A comprehensive tobacco control law has been under development since 2008. While the draft law is expected to be passed in early 2012, many loopholes remain. Even at 65% of manufacturing price, the tobacco excise tax together with the VAT and a raw material import tax represent only 41% of the retail price, compared to 66% recommended by the World Bank and the World Health Organization.10 Common misconceptions remain among policy-makers about tobacco economics—particularly about the tobacco industry's role in the local economy—and continue to hinder the adoption of a strong tobacco control law. ## Objectives This literature review aimed to: 1. Gather Vietnam-specific evidence on the relation between tobacco and poverty and the potential impact of tobacco control measures on employment and poverty. 2. Identify the current research gaps related to tobacco and poverty in Vietnam. ## Methods We conducted a no time-limit search of the Medline, PubMed and Google Scholar databases using the following criteria to search all fields: tobacco AND Vietnam AND (economics OR tax OR burden OR health cost OR control policies OR poverty). We screened all retrieved articles to ensure that they addressed either (i) the relation between tobacco consumption, expenditure or production AND health or healthcare costs, family or national economics, basic or necessary household expenditures, environmental degradation, standards of living, or standards of employment OR (ii) the potential impact of tobacco control policies on tobacco consumption, expenditure, or production and related employment, economics or health. Where retrieved articles were published in both local and international journals, only articles from the international journals were kept. Likewise, where an unpublished and published article existed for the same study, we kept only the published paper. We reviewed the references of all collected articles to identify further articles. Twenty-two articles were included in the review after screening.9 11–27 i The online search was supplemented with a manual library search and direct contact with known researchers and advocates. This provided three legislative and government publications related to tobacco control, four government of Vietnam national survey publications,28–30 two Vietnamese monographs (Monograph of research on tobacco control, period 1999–2005 and Study reports on tobacco smoking prevalence and related disease in Vietnam) which yielded five31 32 and four articles, respectively, meeting our search criteria), one project report and 11 articles, books or online publications that did not specifically address Vietnam but that provided background and contextual information about tobacco and poverty more broadly.1–8 10 33 34 PHA, LTT and LTTH read the 50 publications and categorised them thematically as follows: (1) smoking prevalence and health burden6 11–15 20; (2) economic burden of tobacco use and production2–5 7 8 16–19 21 22 31 32; (3) environmental impact of tobacco production23; (4) impact of tax and tobacco control policies on tobacco consumption and its consequences24–27; (5) general publications1 9 10 33 34; and Vietnamese policy and statistical documents.28–30 DE, LJ and SF edited and finalised the review drafted by PHA, LTT and LTTH (Pham HA et al Tobacco and poverty: evidence from Vietnam literature review, unpublished project report). PHA and LJ abstracted information from the 50 publications, which appears in appendix see online. ## Results ### Tobacco consumption Tracking smoking prevalence in Vietnam has become more regular in recent years through a number of different national surveys. The National Health Survey demonstrates that socioeconomic characteristics appear to be a major predictor of tobacco use.31 Likewise, the Vietnam Living Standard Surveys show that the lowest income groups have the highest smoking rates and the lowest quit rates.31 32 A small scale rural study using the WHO STEPwise surveillance approach found that smoking prevalence and risk of becoming a regular smoker were higher among low-income groups than higher-income groups (68.6% vs 41.8%; OR=1.37; 95% CI 1.06 to 1.78).11 People with lower levels of education are more likely to smoke than are people with higher levels of education (58.2% vs 48.1%). Daily smokers in the highest-income group had a greater chance of quitting than did those with low incomes (RR=2.8, 95% CI 1.72 to 4.56). ### The health burden of tobacco use A verbal autopsy approach demonstrated that cardiovascular diseases were the leading cause of death in 1999, at 20.6%.12 WHO estimated that, in 2002, 66% of deaths in Vietnam were caused by chronic diseases.33 The proportion of hospital admissions attributable to chronic diseases increased from 39% in 1986 to 68% in 2003.28 29 In 2002, the International Agency for Research on Cancer found that lung cancer was the most prevalent cancer in Vietnamese males.34 The estimated national economic loss caused by chronic diseases in 2005 was approximately US$20 million, not including losses for families and individuals.13

In 2006, the tobacco control policy simulation model SimSmoke was used to predict deaths attributable to smoking in Vietnam; a male smoking prevalence of 55% and a RR of total mortality of 1.35 were used, based on similar Chinese studies. The model predicted that about 35 000 deaths would occur in 2004 and about 37 500 in 2008.14

In 2008, a cross-sectional survey among a representative sample of 2500 rural adults found that tobacco use, alcohol drinking, age, low education levels and economic status were associated with higher probability of having at least one chronic disease.15

### Tobacco wastes scarce financial resources

The estimated total number of cigarettes consumed in Vietnam in 1998—2.34 billion packs—was calculated by totalling the number of locally produced cigarettes with the estimated number of smuggled cigarettes. At an average retail price of US$0.16 per pack, the estimated total expenditure was US$435.6 million; these funds could have instead purchased 1.6 million tonnes of rice, enough to feed 10.6 million people for one year.31

Poverty levels are underestimated when spending on tobacco is included in measures of wealth. After separating tobacco spending from total household expenditures, 1.5% of the handouts whose living standards used to be above the food poverty lineii fell into the category of food-poor. If the amount spent on tobacco was instead used to purchase food, 11.2% of current food-poor households could emerge from poverty.31

The poverty-related effect of tobacco consumption was measured using the Gini coefficient, which is a measure of inequality in society, with 0 representing perfect equality and 1 total inequality. As the poor spend proportionately more on tobacco, tobacco use contributes to inequality; thus, after separating spending on tobacco from total household expenditure the Gini coefficient increases from 0.34 to 0.43 in urban areas and 0.27 to 0.32 in rural areas.31 iii

Several studies evaluated the economic impact of tobacco expenditure on poor households using data from the Vietnam Living Standard Survey (VLSS). In one, households were classified into five groups based on per capita household expenditure, with quintile 1 having the lowest expenditure. Households falling into quintiles 1 and 2 were defined as poor, with those in quintile 1 representing the food-poor population. Tobacco expenditure was calculated by type of tobacco (cigarettes and waterpipe/chewing tobacco) and compared with total and ‘essential’ household expenditures (food, education, health and rent). Households using pipe and chewing tobacco spent about 1.2% of total household expenditure on tobacco, whereas ‘cigarette user’ households spent as much as 5.3% of their total expenditure on tobacco.32 While richer households spent more money on tobacco than poorer households, the proportion of income spent on tobacco was highest among the poorest households. Further calculations demonstrated that an average poor cigarette smoker ‘burnt away’ an amount equal to between 19.1% (quintile 2) and 24.8% (quintile 1) of total per capita average expenditure, which could have purchased about 850 calories of rice or similar food daily. This was sufficient to bring one (in quintile 1) or two (in quintile 2) members of poor smoking households into food sufficiency.

Among cigarette-smoking households in quintiles 1 to 4, higher per capita tobacco expenditure correlated with lower other expenditure when compared with non-smoking households in the same quintile. The poorest households spent 2.2 times more on cigarettes than on education and 1.6 times more than on healthcare. These ratios are presented in table 1.

Table 1

Ratio of cigarette expenditure versus expenditure on health and education by income groups32

Another study examined the opportunity cost of smoking for low-income families. Only households that spent money on education were included. As households without tobacco expenditures were also included, the results represented an average across all households, not only those containing smokers.iv The poorest group spent 2.6 times more on tobacco than on education for small children, versus 1.2 times for the average household in the highest quintile. The ratio of tobacco to health expenditures showed little variation between quintiles, with the poorest families spending 1.5 times the amount on tobacco versus health and the wealthiest spending over twice on tobacco what they spent on health.16 More details about these ratios are provided in table 2.

Table 2

Ratio of tobacco expenditure versus education and health expenditures16

A further study compared tobacco spending with expenditures on basic needs such as education, healthcare, shelter and food between rural and urban households.17 Rural households had a higher ratio of tobacco spending to all other types of spending than did urban households. Table 3 presents the results of the survey.

Table 3

Spending comparisons, by quintile (%)17

A cross-sectional survey in five provinces investigated the opportunity costs of tobacco use by examining household expenditure patterns of a representative sample of 478 smoking and 680 non-smoking households.18 Opportunity costs were defined as expenditures on tobacco that could have been used to improve the family's quality of life (better nutrition, shelter, healthcare, education and access to basic household amenities). Consistent with the other studies, households without smokers spent more on education per student than households with smokers. Smoking households classified as very poor spent 2.3 times more on tobacco than on education per pupil. Reallocation of tobacco expenditures to food expenditures could potentially raise 11.3% of all smoking households above the food poverty line.18

### Tobacco use increases healthcare expenditure

The costs of hospitalisation for three smoking-related diseases (lung cancer, chronic obstructive pulmonary diseases (COPD) and ischaemic heart disease) were measured by surveying Hanoi's major hospitals.19 The study did not examine the opportunity cost of premature deaths, outpatient expenditures or other smoking-related diseases, probably resulting in an under-estimation of the actual costs. On average, the cost for one inpatient episode was US$2093, US$824 and US$250 for ischaemic disease, lung cancer and COPD, respectively. A further macro-level analysis revealed that about 72.5% of the social costs related to the treatment of these three diseases could be attributed to smoking, which equalled US$77.5million, or 4.3% of total healthcare expenditures and 0.22% of 2005 GDP. These costs fall most heavily on the government, which bears 51% of smoking-related costs. Families and the insurance sector bear the remaining cost. The data indicated that Vietnam might be in the early stages of a tobacco epidemic, meaning that these costs will rise rapidly with economic growth.19

### Impact of tobacco growing on farmers

One argument used frequently in the media against tobacco control is that tobacco production generates jobs and that tobacco is a ‘poverty eradicating’ crop. To evaluate the validity of this argument, some studies investigated the impact of tobacco growing on Vietnamese farmers. None of the studies found that tobacco farming was beneficial.

Vietnam ranks among the top 20 tobacco-producing countries, at 4–4.5 billion packs annually.20 Between 2000 and 2005, Vietnam produced 23 000–33 000 tonnes of tobacco leaf annually.30 Tobacco cultivation employed about 72 000 fulltime equivalent workers in 2006, or 0.2% of the workforce, while cigarette production employed a further 17 9000 workers.21 Employment in tobacco trading was not included in these figures, as the number of people involved is unknown.

Tobacco cultivation in one northern and one southern commune brought little benefit to farmers compared to control communes in the same areas.21 To minimise labour costs, family members undertook most of the work. When the labour opportunity cost was taken into consideration, meaning the amount of income not earned by family members and using US$2 per day as an average manual labour rate of pay, the ‘advertised’ net annual income of US$275 was significantly reduced to US\$32 in southern communes, while in northern communes a net loss was realised. In half of the participating areas, 17–30% of households reported that tobacco cultivation resulted in debt.

Tobacco farmers also had more illnesses than farmers who grew other crops; they showed significantly higher rates of nine of 16 investigated symptoms, including fatigue/weakness, nausea, increased perspiration/sweating, chill, poor appetite, itchiness and rashes.21

The involvement of children in tobacco production was common and more intensive in tobacco farming than for other crops. Most children started to work at the age of 10; others began as early as 6 years of age. The majority of children in tobacco farming families were not paid for their work; they worked before or after school, at weekends and during school breaks. Children were engaged mostly in sticking tobacco leaves onto bamboo sticks for sun drying. Women performed 60–70% of the total work involved in tobacco production. Tobacco farming burdens were exacerbated where husbands worked outside the home and village.21

A case study in Phong Lai, a commune with a long history of tobacco growing that has recently switched to other crops,22 suggested that tobacco was labour-intensive and unlikely to be profitable. The commune replaced tobacco cultivation with higher value products such as peanuts, chilli peppers and fish. Young women formerly employed in tobacco found that they gained time to start lucrative small businesses, while others were able to seek better livelihoods elsewhere. This case study suggests that even in communities where tobacco is traditionally grown, more remunerative possibilities may exist.

### Negative impact of tobacco on the environment

Little information is available on the environmental costs of tobacco in Vietnam. A 2008 study found that tobacco farmers often used highly toxic pesticides, which can lead to acute poisoning; 1,3-D is known to cause cancer. The fertiliser commonly used by tobacco farmers, maleic hydrazide, can cause skin and eye irritation. Most tobacco farmers used coal or wood to cure tobacco leaves; 75% of the farmers reported that they took wood for curing tobacco from the forest.15 In 1999, Geist estimated that in Vietnam, 1.4% of forest area had been destroyed because of tobacco cultivation; he noted that Vietnam ranked among countries with ‘heavy to medium level of deforestation due to tobacco’.23

### Estimated impact of tobacco control policies on tobacco-related employment

Higher tobacco taxes would not necessarily lead to employment losses in Vietnam. The reasons for this include: (i) shifts in spending away from tobacco products would generate new employment in other sectors, probably with a net positive impact; (ii) employment in tobacco cultivation and manufacturing accounts for a very small share of total employment; and (iii) population growth and rising incomes will probably offset any negative impact that higher taxes might have on overall tobacco employment. Furthermore, industry restructuring—which has resulted in the closure of seven factories— might have already had a stronger impact on tobacco employment than tax changes would in the near future.24

The likely employment consequences of implementing comprehensive tobacco control policies are not significant.25 Employment in four sectors was studied: tobacco cultivation, processing, manufacturing and distribution. Input-output analysis was used to analyse the impact of tobacco control on employment. In 2000, the tobacco industry provided about 122 470 jobs, accounting for 0.32% of total employment.v Of these, 97 600 were in tobacco cultivation, 12 400 in tobacco manufacture and 12 470 in distribution. The effects of adapting a comprehensive set of tobacco control policies (a 50% increase in excise tax, a strong clean air policy, a moderate-sized media campaign, an advertising ban, introduction of strong health warnings and reduction of youth access to tobacco) were estimated using the VNSimsmoke computer model. The implementation of tobacco control policies was deemed likely to lead to two effects: (i) a decline in tobacco consumption, production and jobs, assuming no change in technology, and (ii) money formerly spent on tobacco reallocated to other goods and services, leading to an increased total output and employment.25 Tobacco control policies might thus result in a reduction in tobacco employment but a larger increase in overall employment; concerns about potential negative impacts of tobacco control on employment and output in the economy are therefore unfounded.

### Impact of tobacco control policies on tobacco consumption

The SimSmoke model was also used to examine the potential effect of tobacco control policies on future smoking rates in Vietnam. If the government implemented a combination of policies, representing a 100% tax increase, comprehensive and enforced smoke-free policy, a high-intensity media campaign, strong enforcement of the total advertising ban, strong health warnings and strict youth access controls, smoking prevalence would drop by an estimated 29.6% for males and 22.4% for females. By 2033, smoking prevalence would drop by 38.5% for males and 31.8% for females relative to 2002. If the number of lives saved each year were added for all years, more than 231 500 (221 000 male and 10 500 female) lives would be saved by the year 2033, using a RR of 1.35.14

A cross-sectional household survey data was used with expenditure information to derive the spatial price elasticity of cigarette demand in Vietnam. The estimated direct price elasticity of cigarette demand was −0.53. This suggests that tobacco taxation is likely to have a significant impact on cigarette consumption. Since the estimated price elasticity is less than 1, the introduction of cigarette taxation would also generate additional government revenue.26

A sample of smokers and non-smokers was used to examine the association between decisions to initiate or quit smoking and tobacco prices.27 Cigarette price changes were significantly and negatively associated with decisions to initiate cigarette smoking (elasticity: −1.175), while waterpipe tobacco price changes would not significantly affect decisions to initiate this form of smoking. Cigarette price changes were significantly and positively associated with decisions to switch from cigarette to waterpipe smoking. Income changes were found to be significantly and negatively associated with decisions to initiate waterpipe smoking and to switch from cigarette to waterpipe smoking.27

## Areas for future research

More research needs to be conducted on tobacco use prevalence, particularly on socioeconomic variables, in a consistent way over time to allow for comparisons.

To better document the benefit of tax increases, a reliable and consistent mechanism to monitor the price, market share and affordability of tobacco products must be established. More research on tax equity is needed, including on such issues as the impact of tax increases on household tobacco expenditures, keeping in mind that the lowest-income groups are likely to benefit the most from a tax because they are more likely to quit as a result (and thus both save money and improve health). For the minority who are unable to quit and whose tobacco expenditures increase (who do not compensate for higher prices by reducing consumption or switching to lower-priced products), programmes to decrease costs in other areas could lessen any negative effect of a higher tax.vi

More and larger studies on the impact of tobacco growing on farming households are needed. Future studies should investigate the question of ‘net benefit’ of tobacco growing given its high labour costs and negative health impacts.

Additional work is required on the cost-effectiveness of tobacco control measures. This type of research is especially important for government decision-making.

## Conclusion

Findings from studies conducted in Vietnam confirm many of the findings from other countries about the link between tobacco and poverty:

• Smoking prevalence and the risk of becoming regular smokers are higher among the poor and those with lower education; these groups are also less likely to quit.

• Tobacco use widens social inequality.

• At the household level, tobacco use increases poverty by diverting scarce money from basic needs such as food, education and healthcare.

• Tobacco growing often fails to bring anticipated benefits to growers owing to the high labour demand. Meanwhile the labour burden falls heavily on women and children, who are exploited as unpaid workers.

• Taxation and the enforcement of other strong tobacco control policies would encourage the reduction of tobacco consumption, thereby reducing the tobacco-related health and economic burdens on society and households and expanding alternative and possibly more remunerative employment.

• Several international studies have demonstrated that tobacco use increases poverty by redirecting money required for basic needs to tobacco and by contributing to disability and death, which in turn increases medical expenditures and reduces productivity. A comprehensive body of research highlighting the relation between tobacco and poverty in Vietnam has yet to be undertaken, given the low priority given to tobacco-related issues, a lack of resources and the persistence of common misconceptions among Vietnamese policy-makers about the tobacco industry's role in the local economy.

• This literature review suggests that extensive evidence does exist that tobacco use contributes to poverty and inequality in Vietnam and that tobacco control policies would not have a negative impact on overall employment. Tobacco use wastes household and national financial resources and widens social inequality. The implementation and enforcement of a range of tobacco control measures could therefore prove beneficial not only to improve public health but to alleviate poverty.

## Acknowledgments

The authors wish to acknowledge the technical insight provided by a number of people, most notably Nguyen Tuan Lam and Sarah Bales, and the administrative support provided by Pham Thu Ha. We also wish to thank all those people whose work, both published and unpublished, provided the foundation upon which this review was undertaken.

• ## Web Only Data

Files in this Data Supplement:

## Footnotes

• Funding This literature review was funded by the Bloomberg Initiative to Reduce Tobacco Use (BI) through a grant from The International Union Against Tuberculosis and Lung Disease (The Union). Neither the BI nor The Union played any role in the design of this review, in the collection, analysis and interpretation of the data, in the writing of the report or in the decision to submit this paper for publication.

• Competing interests None.

• Ethics approval This study did not involve human subjects and as such ethics approval was not required.

• Provenance and peer review Not commissioned; externally peer reviewed.

• i All articles and publications not cited in this paper are included in appendix see online.

• ii Food poverty level is the level of household expenditure required to ensure that the household can buy a ‘basket’ of food to provide 2100 kcal per person per day.

• iii Poverty in Vietnam is measured by total household spending, in which tobacco is included. However, since tobacco is not welfare-enhancing, it should not be included; when it is taken out the number of poor increase.

• iv 65% of households in the survey have at least one smoker. The inclusion of households without smokers is one limitation of the study.

• v These numbers are slightly different from those used above, which were full-time equivalent jobs rather than jobs per se.

• vi Simply put, is it better to subsidise health and other social programmes or tobacco consumption?