Unrecorded Alcohol and Alcohol-Related Harm in Rural Sabah ...

Alcohol and Alcoholism Advance Access published February 21, 2016 Alcohol and Alcoholism, 2016, 1–6 doi: 10.1093/alcalc/agw005 Article

Article

Unrecorded Alcohol and Alcohol-Related Harm in Rural Sabah, Malaysia: A Socio-economically Deprived Region with Expensive Beer and Cheap Local Spirits Wendy Diana Shoesmith1,*, Naing Oo Tha1, Khin Saw Naing1, Roslee Bin Haji Abbas2, and Ahmad Faris Abdullah1 1

Department of Community and Family Medicine, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia, and Department of Social Science, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia

2

*Corresponding author: Department of Community and Family Medicine, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia. Tel.: +60-162389571; E-mail: [email protected] Downloaded from by guest on February 28, 2016

Received 14 September 2015; Revised 19 January 2016; Accepted 21 January 2016

Abstract Aim: To investigate recorded and unrecorded alcohol and the relation to alcohol-related harm in a region with high taxation, economic deprivation and cultural use of alcohol. Methods: Two participants per household were systematically sampled from 12 different villages chosen using stratified random sampling in the North of Sabah, Malaysia. Participants were asked about each type and amount of drink consumed; price paid, whether tax was paid, number of days sick in the last year and whether they had experienced various health problems. A brief screen for mental disorders (PHQ) and an alcohol disorder screening test (AUDIT) were completed. Village heads were also interviewed about alcohol-related problems at village level. Results: 470 people were interviewed. The most commonly drunk beverages were beer and Montoku (a local distilled beverage), which had average prices of RM3.85 and RM0.48 per standard drink respectively. Montoku was more likely to be drunk by problem drinkers. Only 3.1% of alcohol drunk was believed by respondents to be taxed. Men with an AUDIT score of more than 15 were more likely to have had a sick day in the last year and have a female household member with symptoms of mental disorder on PHQ. Conclusions: Change in the taxation structure needs to be considered to reduce alcohol-related harm. Most alcohol consumed in rural Sabah is smuggled or informal. The low price of local spirits is likely to be contributing to alcohol-related harm. Differential effects on minority populations need to be considered when designing alcohol policy.

INTRODUCTION Alcohol is the third leading risk factor leading to death and disease in the world, contributing 4.6% of disability adjusted life years and leading to approximately 1 in 16 male deaths (Mathers et al., 2009). Malaysia is a country where alcohol problems do not affect the majority of the population, who are Muslim and therefore forbidden from drinking. There are three main ethnic groups in Peninsular Malaysia

and more than 80 ethnic groups in East Malaysia, on the island of Borneo. Although the majority Malay ethnic group (around 50% of the population in Malaysia) do not drink, many of the other ethnic groups do drink and there is concern that this may be causing considerable harm (Assunta and Valbuena, 2001; Lasimbang et al., 2015). The overall per capita alcohol consumption is estimated to be only 0.5 l (recorded) and 0.33 l unrecorded. Those who do drink are

© The Author 2016. Medical Council on Alcohol and Oxford University Press. All rights reserved

1

2

METHODS This study was part of a mixed methods study to assess attitudes towards drinking in Northern Sabah, Malaysia. Only the quantitative methods and results are described here. Systematic sampling was used, by selecting three villages in each area at random from a list of accessible villages in the area, approaching alternate households, and using a form of modified Kish sampling to select two household members. Face to face interviewing was used to ask participants about each beverage that they drunk. They were asked how much they consumed of each beverage in a typical drinking session, by pointing to the side of a glass. For drinks that were consumed communally, such as Tapai, they were asked how many jars/bottles they shared with how many people and the size of the jar/bottle. We also used the AUDIT scale (Babor et al., 2001) and the graduated-quantity frequency method (Department of Mental Health and Substance Dependence Noncommunicable Diseases and Mental Health Cluster, 2000), to determine whether people had an alcohol use disorder and if they were heavy episodic users. They were asked whether it was commercial or

home-made, where they brought it, how much it cost and whether they believed it was taxed. In Malaysia, the high tax rate, together with an excise sticker stuck over the lid of taxed beverages mean that most people know if their beverage has been taxed. Volume of alcohol (ml) was converted to grams by multiplying by 0.79 (Babor et al., 2001). The total grams of alcohol drunk per week by the whole sample was calculated for each beverage by adding together the grams of alcohol per week from that beverage for each participant. The total grams of taxed alcohol for the whole sample was calculated in the same way, only including participants who normally paid tax on that beverage. The percentage of alcohol that had taxes paid was calculated by dividing the taxed grams of alcohol drunk per week for the whole sample by the total grams of alcohol drunk per week for the whole sample. A standard drink was considered to contain 10 g of alcohol. Per capita consumption is normally expressed in litres and this was calculated by dividing grams of alcohol by the conversion factor. Alcohol drunk in a high risk way was calculated from the graduated quantity frequency data, by summating the volume of alcohol drunk as part of a session where more than five standard drinks were consumed. Alcohol-related harm was estimated by a number of methods, including days off sick in the last year, number of days in hospital (lifetime), number of medical problems which have a relationship with alcohol, PHQ scores (Kroenke et al., 2001) and by asking the village headman about alcohol-related social problems (e.g. fights, domestic violence). Potential harm related to alcohol was measured at an individual level using chi-squared tests to look at whether AUDIT scores above 8 and 15, high average volume drinking (>5200 ml per year), and heavy episodic drinking were related to measures of alcoholrelated harm (sick days in the last year, hospital admissions, any mental disorder on PHQ and any other health problem). Harm at household level was measured using chi-squared tests to assess if the male household member’s alcohol measures were related to the female member’s measures of alcohol-related harm. Since few women drank, this was not done the other way around. Respondents were also asked if any member of their household had alcohol-related problems. The harm at village level was determined by asking the village head about whether five different alcohol-related problems (fights, arguments, child abuse, accidents and police being called) had occurred in the last year. The relationship between the village mean AUDIT score, volume drunk, volume drunk in a high risk way and the presence of three or more alcohol-related problems in the village was tested using Mann–Whitney U test. Respondents were also given a scale looking at attitudes to alcohol and attitudes to taxation. A separate paper is written about attitudes to alcohol and only attitudes to taxation is reported here. The study received ethical approval from the ethical review board at University Malaysia Sabah. All participants gave written informed consent.

RESULTS There were 470 respondents and 40 non-respondents. Of the respondents, 112 (23.8%) were current drinkers, 55 had an AUDIT score of over 8 (indicating high risk drinking), 22 had an AUDIT score above 15 (indicating harmful or dependent drinking), 31 had drunk more than 5200 ml pure alcohol in the last year and 65 had an episode of heavy episodic drinking in the last year (6 or more standard drinks on a single occasion). Fifty per cent of respondents were women, but only 11 of the drinkers were women. The most common occupations were fishing and farming and 55% of people had a household income of

Downloaded from by guest on February 28, 2016

thought to drink at quite high levels, and per capita alcohol consumption among drinkers has been estimated to be over 30 l per year (World Health Organization, 2011). The places with the highest levels of drinking in Malaysia are Kuala Lumpur, Sabah and Sarawak (Institute of Public Health Malaysia, 2011). Sabah is one of the two states on the island of Borneo. It is different, both culturally and economically from West Malaysia. Malaysia is a mainly Muslim country, but Sabah has large Christian populations. It has highest poverty rate in Malaysia, with 10 times more people below the official poverty line than West Malaysia (The World Bank, 2010; Hatta and Ali, 2013). It has the largest rates of income inequality, particularly between urban and rural populations (Leete, 2007). In 2012, 46% of rural households had no water supply (Department of Statistics Malaysia, 2012). Northern Sabah is largely rural and is believed to be one of the poorest parts of Sabah (The Borneo Post, 2012). In Sabah a large proportion of the alcohol drunk is believed to come from unrecorded sources, such as home produced beverages. Tapai is a traditional beverage, normally made from rice, which is consumed during festivals. It is not distilled and so the alcohol content is believed to be around 12%. This is made in clay jars, called Tajau, and is nearly all home-made. It is normally drunk communally, straight from the Tajau, using a straw made of bamboo. Montoku is another local beverage, which is made by distilling Tapai, so the alcohol content is much higher (17–30%). Montoku is both home-made and available commercially. In villages, home-made Montoku is sold informally and is a source of income for some families. The alcohol content of the commercially available Montoku is normally 17–20%. The content of home-made Montoku depends on whether it is the first Montoku that comes from the distilling apparatus or the later distillate. While making Tapai is legal, distilling without a licence is not legal (The Commisioner of Law Revision Malaysia, 1976). Alcohol taxation is one of the recommended policies to reduce alcohol-related harm, but may not be the best policy where there is a large informal market (World Health Organisation, 2008). The aim of this study was to find out about drinking in the Northern part of Sabah, including how much alcohol drunk was informal, how much was taxed and to better understand people’s attitudes towards alcohol and taxation. Detailed analysis of the attitudes and knowledge about alcohol is in an accompanying paper. This paper will focus on taxation and alcohol-related harm.

Alcohol and Alcoholism

3

15.19 30.10 14.54 13.22 4.34 0.00 3.56 4.16 1.0% 0.0% 0.0% 0.0% 9.1% 0.0% 20.3% 40.3% 24% 2% 18% 17% 32% 6% 0% 0% 6.28 3.93

Beer (5%) Strong beer/stout (9%) Toddy/bahar (8%) Tapai (12%) Montoku (17%) Montoku (30%) Wine (13%) Spirits (40%)

83 4 17 32 49 1 5 5

4 0 0 3 21 0 3 3

2 1 17 31 21 1 0 0

3.85 4.07 0.98 1.75 0.48

3.95 7.39 14.79 7.58 9.01 85.32 0.57 1.06

Per cent of alcohol drunk that is believed to be taxed Contribution of total amount of alcohol drunk by whole sample from each beverage Number of people Number of people Number of people that Mean price per Mean number of that have drunk in that believed they reported drinking standard drink standard drinks per the last year paid tax home-made alcohol (RM) week per drinker of each beverage

There were 118 people (25.1%) who had a sick day in the last year and 123 respondents (26%) that had previously suffered from any of the alcohol-related health problems listed, with the most common being hypertension. Twelve people had PHQ-9 scores of five or more, indicating depression and seven people had a PHQ-7 score of five or more, indicating anxiety. No respondents were higher than the cut off for somatisation disorder or panic disorder. Number of sick days per year was related to having an AUDIT score above 8, AUDIT score above 15, high average volume drinking, being male and being older. Since alcohol use disorder and gender were also related and could be confounding, the analysis was done separately for men and women. Men with an AUDIT score above 15 were more likely to have

Table 1. Calculation of standard drinks per week, amount of alcohol taxed and spending on alcohol

Alcohol-related harm


below RM 600 (approx 180USD) per month. The annual per capita alcohol consumption for this sample is calculated to be 1.94 l, with a per capita consumption of 8.15 l per year when only drinkers are included. Twenty-seven per cent reported that other members of the household were drinkers and 1.5% reported that another member of the household had a ‘drinking problem’. The majority of alcohol drunk was bought commercially. Only 3.1% of alcohol consumed was believed by the respondents to be taxed. The mean price per standard drink was RM0.6 (0.18 USD). The price of each drink and the percentage believed to be taxed is shown in Table 1. Beer was the most common beverage consumed, and nearly all of it was bought commercially, but believed to be untaxed by the respondents. The mean price for a standard drink of beer was RM3.7 if untaxed, and RM5.6 if the respondent believed it was taxed. Montoku was the drink that most people believed that they paid tax on and was also the cheapest drink. There was no difference between the prices people said they paid for Montoku that was taxed compared to untaxed Montoku. There were significant differences in the main beverage drunk, depending on risk category. High average volume drinkers (>5200 g alcohol per year) (OR = 4.4, 95% CI (1.4–13.2), P = 0.06), heavy episodic drinkers (more than 5 drinks per occasion) (OR = 2.9, 95% CI (1.0–8.0), P = 0.04), very heavy episodic drinkers (drank more than 20 drinks on a single occasion) (OR = 9.9, 95% CI (1.9–52.2), P = 0.02) and AUDIT score more than 15 (OR = 3.7, 95% CI (1.1–12.1), P = 0.02) were more likely to drink Montoku than low volume drinkers. Thirty-two per cent of people who mainly drank Montoku and 40% of people who mainly drank Toddy (a local drink made from fermenting coconut water, approximately 8% alcohol) had AUDIT score more than 15, compared to 10.3% of beer drinkers. Beer (5%) was the main drink of 52% of low risk drinkers, 51% of hazardous drinkers and 27% of people with an AUDIT score of more than 15. Montoku was the main drink of 21% of low risk drinkers, 15% of hazardous drinkers and 41% of people with an AUDIT score of more than 15. Attitudes to taxation are shown in Table 2. The four attitudes to taxation items all correlated together, even when they were split according to religion, and a single dimension accounted for 77% of the variance on PCA so they were summed to form a ‘Positive attitudes to taxation’ scale. Women (F(1,460) = 15.43, P < 0.001), Muslims (F(1,460) = 50.71, P < 0.001), non-drinkers (F(1,460) = 87.25, P < 0.001) and non-smokers (F(1,458) = 17.19, P < 0.001) were all more likely to be positive about taxation. Being positive about taxation was negatively correlated with total AUDIT score (R = −0.42, P < 0.001), grams of pure alcohol drunk per year (R = −0.30, P < 0.001), amount of alcohol drunk in a high risk way (R = −0.31, P = 0.001) and age (R=−0.26, P < 0.001).

Total spent on beverage per week per drinker of each beverage


4


Table 2. Attitudes to taxation

Alcohol should be taxed at a high rate by the government The government taxes alcohol to make money The government taxes alcohol for religious reasons The government taxes alcohol to improve health

Factor loading

Non-drinkers

Drinkers

Total

0.832 0.891 0.908 0.881

83% 73% 80% 85%

39% 46% 54% 68%

73% (341) 67% (313) 74% (347) 81% (380)

DISCUSSION This study showed that alcohol was causing some harm in the communities studied, in that men with an AUDIT score of more than 15 were about three times more likely to have had a sick day in the last year and around 17 times more likely to have a female member of the household with symptoms of mental disorder on PHQ. There was also an effect of the volume of alcohol drunk in a high risk way on problems at a community level. It appears that most alcohol drunk in this region is not taxed. This is unsurprising for home-made beverages, such as Tapai, but it is more surprising for commercial beverages, such as beer. The normal price of a beer with taxes paid at that time was around RM8, so it is highly likely that the respondents were correct in saying that their beer was untaxed. The most likely explanation for this is that a large proportion of the beer that was sold was smuggled. The Northern part of Sabah is close to the Philippines and it is easy to find beer that has been smuggled from there, or the nearest duty free island, Labuan. Although more people drank beer, most alcohol was drunk as Montoku, because the volume of alcohol drunk was much higher. Low risk drinkers and hazardous drinkers were most likely to be drinking beer, whereas people with an alcohol use disorder were more likely to be drinking Montoku. There is a consistent relationship between affordability of alcohol and alcohol consumption (Wagenaar et al, 2009) as well as alcoholrelated problems in most societies in the world (Jernigan, 2002). Taxation policy is an important way of curbing alcohol-related problems

(Casswell and Thamarangsi, 2009). There are different forms of taxation policy, including taxing as a percentage of total cost and differential taxing depending on whether alcohol is sold in places such as bars and clubs or for consumption at home (World Health Organization, 2010). Taxation policy can have differential impact on different groups and some policies have a greater effect on alcohol-related harm than others, depending on the local environment (Meier et al., 2010). Excise duty on alcohol can be charged in two main ways: (a) at a specific rate per litre of alcohol or beverage or (b) ad valorem where alcohol is taxed according to value. The specific rate per litre of alcohol method is better at reducing alcohol consumption, particularly among people from lower socioeconomic groups (Mäkelä et al., 2015; Vandenberg and Sharma, 2015). Ad valorem taxation is better for revenue collection, but not as effective at reducing consumption and can lead to downgrading of quality (Sornpaisarn et al., 2015). In Malaysia a combination of taxation methods is used. Most beverages are taxed per litre of beverage, with some spirits charged per litre of alcohol, depending on alcohol content and type of spirit (Attorney General’s Chambers of Malaysia, 2012). Ad valorem tax is 15% and there are also import duty and goods and service taxes (5%), which add significantly to the cost of high value items, such as wine and imported spirits. This taxation structure means that beer and high value imported drinks are relatively expensive, but locally produced low value spirits are relatively cheap. It costs around RM8 for a 330 ml can of beer (5%) and RM200 for a 1 l bottle of mid-range imported spirits, but a 750 ml bottle of Montoku (20%) costs less than RM5 (with excise sticker at the supermarket). It may be that a ‘one size fits all’ taxation policy it not appropriate for a country as diverse as Malaysia. The rural drinking culture of Sabah could not be more different to the Kuala Lumpur night scene, where wealthy middle class drinkers predominate. Although the population we surveyed were generally supportive of the taxation of alcohol, the price of taxed beer (approximately RM8 per beer) is a lot for people earning RM200–600 per month. It is unsurprising that smuggled beer and locally produced cheap spirits predominate. The amount of excise paid on a beer in Malaysia is 11 times higher than the excise paid on a beer in the Philippines. This is likely to be adding to the cross border smuggling of alcohol (Preece, 2012). The solutions could include better controls to reduce the smuggling of beer, for example by increasing policing resources in Sabah, upgrading fiscal marking systems and increasing penalties for smuggling. Although this may reduce the consumption of beer, it may not significantly reduce alcohol-related harm, since people that drank mainly beer were less likely to be drinking in a harmful way. There is also evidence that when the price of a beverage increases, consumption of that beverage reduces, but consumption of other beverages will increase (Jiang and Livingston, 2015). Increasing controls on beer smuggling is likely to increase the consumption of Montoku, which can be bought legally for less than RM1 per standard drink. Changing the taxation policy to increase the price of locally produced spirits (such as increasing specific rate taxation or including a minimum price per standard drink) is likely to reduce the consumption of commercial Montoku. This may


had a sick day in the last year (OR = 3.24 (95% CI 1.35–8.02), P = 0.006). The relationships with AUDIT score above 8, high average volume drinking, or heavy episodic drinking were not significant. There was no significant relationship for women. Chi-squared tests showed that there was no significant relationship between AUDIT score above 8 or 15, high average volume drinking and number of days admission to hospital, total number of health problems, or the presence of mental disorders on PHQ. Those respondents that reported that there was another member of the household with ‘alcohol problems’ were more likely to have any mental disorder on PHQ (OR = 13.82 (95% CI: 2.50–77.93), P = 0.018 (Fisher’s exact test)). When male household members’ drinking was compared to that of female members’ measures of alcohol-related harm (151 male-female pairs), there was a significant effect of AUDIT score above 15 (OR = 17.02 (95% CI: 3.00–99.45), P = 0.005 (Fisher’s exact test)) and heavy episodic drinking in the last year (OR = 7.42 (95% CI: 1.30–42.4), P = 0.025 (Fisher’s exact test)) on any mental disorder on PHQ. At village level, 8 out of 12 village heads mentioned one or more alcohol-related problems in the village in the last year. Five village heads said that there had been three or more. There was a significant relationship between volume of alcohol drunk in a high risk way and having three or more alcohol-related problems in the village (Mann– Whitney U test exact significance of 0.042). There was no relationship with the other measures of drinking pattern at village level.

5


LIMITATIONS This study was based on self-report and it is likely that there was under-reporting of actual drinking and alcohol-related harm. The levels of mental health morbidity in this study were very low, which meant that the study was underpowered to measure some of the differences in mental health morbidity.

CONCLUSIONS Very little alcohol that is drunk in Northern Sabah is taxed, so taxation is unlikely to be significantly affecting drinking practice. The cheapest standard drink in Northern Sabah is Montoku, which contains a relatively high proportion of alcohol and is more likely to be drunk by people with alcohol use disorders. This can be legally bought for around RM0.5 per standard drink. Harmful/dependent drinking patterns are likely to be causing harm at an individual level, whereas both harmful/dependent drinking patterns and heavy episodic drinking are causing harm at a family level. The current system where alcohol in the form of beer is taxed at a higher rate than alcohol in stronger drinks needs to be reconsidered. Consideration needs to be given to

increasing the price of Montoku and other locally produced beverages, perhaps through a minimum price per standard drink or by increasing specific taxation rate per litre of alcohol. Consideration also needs to be given to increasing enforcement and encouraging communities to produce locally enforced by-laws to reduce alcohol-related harm. Alcohol policy also needs to consider differential effects on minority populations and it may be more appropriate that some control mechanisms are local rather than central.

FUNDING FRGS grant from the Malaysian Ministry of Higher Education.

CONFLICT OF INTEREST STATEMENT Dr Wendy Shoesmith works with a local NGO (Mercy Malaysia) to reduce alcohol-related harm in the community.

REFERENCES Assunta M, Valbuena JP. (2001) The alcohol problem in Malaysia. The Globe [online] (Special issue 4) p. 4–8. http://www.ias.org.uk/What-we-do/ Publication-archive/The-Globe/Issue-4-2001-amp-3-2001/The-alcoholproblem-in-Malaysia.aspx. Attorney General’s Chambers of Malaysia. (2012) Federal Government Perintah Duti Kastam 2012. Excise Duties Order 2012. Federal Government Gazette [online] (October). http://www.customs.gov.my/ms/pg/pg_co1/ Pengedaran Perintah Bil 19, 20 dan 21(2012)/pua_20121030_Perintah Duti Eksais 2012.pdf. Babor TF, Higgins-Biddle JC, Saunders JB, et al. (2001) Appendix C. Translation and Adaptation to Specific Languages, Cultures and Standards. In: The Alcohol Use Disorders Identification Test: Guidelines for Use in Primary Care, Second Edition [online]. World Health Organisation. pp. 32–3. http://apps.who.int/iris/bitstream/10665/67205/1/WHO_MSD_ MSB_01.6a.pdf. Casswell S, Thamarangsi T. (2009) Reducing harm from alcohol: call to action [online]. Lancet 373:2247–57. (25 November 2010, date last accessed). Department of Mental Health and Substance Dependence Noncommunicable Diseases and Mental Health Cluster. (2000) Estimating levels and patterns of alcohol consumption from national surveys. In: International Guide for Monitoring Alcohol Consumption [online]. World Health Organization. pp. 58–9. http://apps.who.int/iris/bitstream/10665/66529/1/WHO_MSD_ MSB_00.4.pdf. Department of statistics Malaysia. (2012) Household Income and Basic Amenities Survey Report 2012: Tables of basic amenities [online]. http://catalog. ihsn.org/index.php/catalog/5431. Esser MB, Jernigan DH. (2014) Assessing restrictiveness of national alcohol marketing policies [online]. Alcohol Alcohol 49:557–62. Hatta ZA, Ali I. (2013) Poverty reduction policies in Malaysia: trends, strategies and challenges [online]. Asian Cult Hist 5:48–56 (21 February 2014, date last accessed). Institute of Public Health Malaysia. (2011) National Health and Morbidity Survey. Jernigan D. (2002) Alcohol in developing societies: a public health approach. Summary [online]. World Health Organization. Jiang H, Livingston M. (2015) The dynamic effects of changes in prices and affordability on alcohol consumption: an impulse response analysis [online]. Alcohol Alcohol 50:631–8. Kroenke K, Spitzer RL, Williams JB. (2001) The PHQ-9: validity of a brief depression severity measure [online]. J Gen Intern Med 16:606–13. Lasimbang H, Shoesmith W, Singh J, et al. (2015) Private troubles to public issue: empowering communities to reduce alcohol-related harm in Sabah, Malaysia. Health Promot Int. Advance Access. Epub Sept 2015:1–8. Leete R. (2007) Sabah’s Human Development Progress & Challenges [online]. http://dspace.africaportal.org/jspui/bitstream/123456789/20417/1/Sabahs Human Development Progress and Challenges.pdf.


then have the knock-on effect of increasing home production. Distilling Montoku at home is illegal and potentially dangerous, but is quite common. It is normally sold locally, among friends and neighbours, so it may be difficult to control through central mechanisms such as fiscal marks. Community based participatory action would be one way of reducing this, through education and locally enforced village by-laws. A community empowerment approach has been effective in Sabah and some villages have produced local alcohol related by-laws as a result of this (Lasimbang et al., 2015). These by-laws have included rules around underage drinking, and the way alcohol is served at functions, but have not yet included rules around home-distilling or selling untaxed beverages. The attitude towards taxation of alcohol was generally positive. Despite drinking alcohol that was mainly untaxed, most drinkers appeared to have a positive attitude towards taxing alcohol. This may be because hardly anyone was actually paying the tax and so it didn’t affect them. It could be because people in this region are generally positive about the government and criticizing the government (including taxation) is not the cultural norm. Surprisingly, the item ‘The government taxes alcohol for religious reasons’ correlated with the other items reflecting a positive attitude to taxation, even among nonMuslims. This perhaps reflects the generally positive attitude to religion, showing that taxing alcohol for religious reasons is not seen as unreasonable in Malaysia and protecting religion is seen as being as much part of government function as protecting health. Since heavy episodic drinking appears to be the main cause of alcohol-related harm, health campaigns should centre on the effects of this, rather than the effects of dependent drinking. Drink driving is not currently a problem in these villages, since car ownership is rare but because heavy episodic drinking is the main pattern it may become a problem in the future, as people become more prosperous and car ownership increases. Drink driving campaigns should start now, before drink driving becomes the norm. If alcohol becomes more affordable as society becomes more prosperous then it is likely that alcohol-related problems will increase, since participants reported that it was mainly the price that was limiting their consumption. Malaysia is relatively permissive towards alcohol advertising (Esser and Jernigan, 2014) and increasing advertising restrictions may also be effective.

6 Mäkelä P, Herttua K, Martikainen P. (2015) The socioeconomic differences in alcohol-related harm and the effects of alcohol prices on them: a summary of evidence from Finland [online]. Alcohol Alcohol 50: 661–9. Mathers C, Stevens G, Mascarenhas M. (2009) Global health risks: mortality and burden of disease attributable to selected major risks [online]. Available from: http://www.who.int/healthinfo/global_burden_disease/ GlobalHealthRisks_report_full.pdf. Meier PS, Purshouse R, Brennan A, et al. (2010) Policy options for alcohol price regulation: the importance of modelling population heterogeneity [online]. Addiction 105:383–93. (19 June 2012, date last accessed). Preece R. (2012) Excise taxation of key commodities across South East Asia : a comparative analysis ahead of the ASEAN Economic Community in 2015. World Cust J 6:3–16. Sornpaisarn B, Kaewmungkun C, Rehm J. (2015) Assessing patterns of alcohol taxes produced by various types of excise tax methods—a simulation study [online]. Alcohol Alcohol 50:639–46. The Borneo Post. (2012) Poverty Level in Sabah Drops [online]. The Borneo Post. 11 July. http://www.theborneopost.com/2012/07/11/poverty-levelin-sabah-drops/.

Alcohol and Alcoholism The Commisioner of Law Revision Malaysia. (1976) Laws of Malaysia, Act 176, Excise Act 1976. Percetakan Nasional Malaysia BHD. The World Bank. (2010) Malaysia Economic Monitor: Inclusive Growth [online]. http://documents.worldbank.org/curated/en/2010/11/14893214/ malaysia-economic-monitor-inclusive-growth. Vandenberg B, Sharma A. (2015) Are alcohol taxation and pricing policies regressive ? Product-level effects of a specific tax and a minimum unit price for alcohol. Alcohol Alcohol. Advance Access. epub Nov 2015:1–10. Wagenaar AC, Salois MJ, Komro KA. (2009) Effects of beverage alcohol price and tax levels on drinking: a meta-analysis of 1003 estimates from 112 studies [online]. Addiction 104:179–90 (3 January 2015, date last accessed). World Health Organisation. (2008) Strategies to reduce the harmful use of alcohol. In: Sixty-First World Health Assembly [online]. pp. 1–6. http:// www.who.int/mediacentre/events/2008/wha61/issues_paper3/en/. World Health Organization. (2010) The WHO global strategy to reduce the harmful use of alcohol [online]. http://www.who.int/substance_abuse/ alcstratenglishfinal.pdf?ua=1. World Health Organization. (2011) Global status report on alcohol and health. Country Profiles: Malaysia [online]. http://www.who.int/substance_abuse/ publications/global_alcohol_report/profiles/mys.pdf.
