Paper for the conference “Alcohol: Evidence-based Impacts and Interventions”, sponsored by the Center for Alcohol Studies, of the Health System Research Institute and the Department of Mental Health, Ministry of Public Health, Thailand, held 13 -14 December, 2006 in Bangkok.
This paper is dedicated, with affection and respect, to the Canadian epidemiologist and public health expert, Dr Eric Single who has just retired. It will be evident from the bibliography that Eric has been involved in many of the key projects which form the intellectual foundations of this paper. Those projects would never have been completed without his inspiring, cheerful and energetic leadership. The sad fact is that drug abuse – alcohol, tobacco and the narcotics – is a pernicious part of the modern world. But it is the less so – and will be even less so in the future – that it might have been, is in part because of Eric’s efforts.
Depending on the cultural context and particular circumstances, the same drink of alcohol can generate a benign feeling of prosperity, or moroseness, or stupor. The health benefits may also be benign (or even beneficial), or the drink may result in harm – injury or death – in the short run from accident or in the long run from one of the diseases alcohol can precipitate. The consequences for others may also be benign or beneficial, or damaging or mortal from violence or collateral accident. Someone may get born as the result of intentional or unintentional impregnation. The loss of production due to poorer workplace productivity or non-attendance from drinking alcohol may cause financial loss to the drinker and possibly to others. Among the many sectors of the economy alcohol may, or may not, generate additional costs, especially in the criminal system, in the health system, and in the transport system. The national budget probably gains from the specific tax it levies on alcoholic beverages, but these levies may, or may not, cover its costs from the consumption of alcohol.
The impact of alcohol consumption is so multifarious that it is not easy to track all its social and economic impacts. This paper confines itself to the main ones the ‘social costs of alcohol misuse’ studies have paid attention to, although there is only a little on the burden of disease caused by alcohol. The framework used is that of the WHO report International Guidelines for Estimating the Costs of Substance Abuse (Single et al 2003), and related and subsequent studies (e.g. Collins & Lapsley 2006).
As the opening sentence reminds us, the social and economic impact of alcohol is affected by the cultural context, even though the physiological impact may not be. This means that each country – or culture within a country – has its own quantitative and qualitative particularities. While this paper, then, gives some general guidance, the challenge is for those working here in Thailand to adapt the generalities for their particular circumstances.
The main areas of impact are dealt with in alphabetical order, following a brief discussion on externalities.
The social cost of an activity is the total of all the costs associated with it, including both the costs borne by the economic agent involved and also all costs borne by society at large. It includes the costs of the resources (including labour and capital) embodied in the product. But it also includes and costs external to the firm’s private costs which are consequent on the production or consumption such as environmental degradation public costs (e.g. health care) and costs to others like a woman being beaten up by a drunk husband.
When the market economy is working effectively and there are no external social costs, the rational consumer will only purchase and consume the product if it is more valuable to her or him than the price paid for it. In this case society is better off since the consumer deems her or himself better off from the transaction, and society is no worse off since it has traded off the social costs of the consumption for the payment which embodies the ability to purchase other products with equivalent social costs.
There are all sorts of subtle assumptions embodied in this analysis: public health analysis often worries about the equity in the distribution of income – the power to purchase. But even if those assumptions hold (near enough), alcoholic beverages are n example where the purchase price is not generally equal social costs because the social cost of each unit of consumption varies greatly. Thus the social cost of the first drink in a session may reflect the private cost, but later drinks may generate additional social costs in poor health, material damage, public and private outlays which were ignored when the drinking decision was made, violence to others, and even death to the drinker or, in the case of motor vehicle accidents and the like, to innocent bystanders.
As a later section discusses, taxation and other public policy measures aim to reconcile in one way or another the inconsistency between the purchase price and social costs. Because of the variations from drinking situation to drinking situation, there is still likely to be a substantial difference for at least a number of significant ingestions even if the average is right.
A useful term is ‘misuse’ for occasions when the consumption generates outcomes which were unintended by the consumer and not taken into account in the purchase/drinking decision. It is sometimes preferred to ‘abuse’ in order to emphasise, that much alcohol consumption is useful, in the sense of its social impact is benign or even (mildly) beneficial.
The next few sections summarise the most salient causes of a divergence between private and social costs.
Pernanem et al (2002) proposes four models (‘modes’ might be a better word) to account for alcohol associated crime.
1. The pharmacological model in which intoxication encourages the commission of crimes which otherwise would not have been committed.
2. The economic means model in which crimes are committed to fund alcohol consumption.
3. The systemic model involving the illegal economy, as in unlawful brewing or distilling or sale of liquor.
4. The substance-defined model, where actions are defined as being criminal by laws which regulate drug use, such as drunkenness in a public place, supplying underage or drunk people (if that be illegal) and drink-driving.
This crime causes social costs, both directly via damage to property and individuals, consequently with the costs to the public purse of policing, justice and corrections, and indirectly when it adds to concerns about community safety and security.
Estimating the attributable fractions (the proportions which alcohol causes) for criminal activity is much less advanced than for disease. Because law, practices of enforcement and cultural vary by country, fractions cannot be easily borrowed from other jurisdictions. (Indeed care is also required in generalising research findings on the relationship between crime and alcohol.)
It is for an epidemiologist to detail the disease consequences of alcohol misuse on the health of the consumer. For economic purposes the summary is that misuse of alcohol shortens the life and decreases the quality of life experienced by the living.
Clearly these are detriments and should be incorporated into any assessments of the economic consequences of the alcohol misuse. The direct losses of production are discussed below, but life itself also has a value which must be taken into consideration.
The valuation of life has proved contentious, There is widespread agreement that the appropriate measure is some notion of quality life years including those of the living, rather than the deaths by themselves. (Commonly the actual measure is QALYs, quality adjusted life years, although there are a some who use DALYs, disability adjusted life years.) Not to do so would lead to very distorted public decisions – such as terminating everyone’s life at their retirement, since their net cost to society – excluding the benefits of life itself – would be negative. .
The acknowledgement of the relevance of the quality of life can have a dramatic effect of the relative balance between alcohol and smoking. Deaths from smoking tend to occur towards the end of a normal life whereas many deaths from alcohol misuse occur to the young as when adolescents die from alcohol induced road accidents. Allowing for the many more years of life which a young death curtails raises the relative losses from alcohol misuse.
Putting a monetary value on life proves even more problematic. A common valuation is based on surveys which ask people how they value life – their ‘willingness to pay’ – in terms of material products they are willing to give up. The apparent difficulty with such valuations is that the aggregate annual cost of alcohol misuse seem to be a very substantial proportion of annual GDP. This arises because of the high value that people place on life, while that life itself is not measured in GDP: only material production. As a result it may be better to report the costs of alcohol misuse in two components: the tangible cost (the reduction in material product available) as a proportion of GDP, and the intangible cost (the reduction in the quality life) relative to the total value of life. Thus a study might usefully report that the cost of alcohol misuse is 1 percent of GDP and 2 percent of the total quality of life (through death and poorer quality living).
Another complication is the treatment of the beneficial effects from alcohol consumption, since there is epidemiological evidence that modest quantities may prolong life. At least one study concluded that these small beneficial gains across the population as a whole offset the substantial losses as a result of misuse. (Collins and Lapsley 2002) The technical resolution involves careful definition of the counterfactual – the alternative scenario against which costs are evaluated. More practically, if the focus is on alcohol misuse, rather than all alcohol consumption, then the beneficial effects of alcohol assumption may be ignored. In effect, the resulting estimate reflects the gains from preventing misuse, while the population still enjoys consuming moderately.
Harm to Others
As well as affecting the health of its drinkers, alcohol can affect the health of others. These include:
1. Children born as the result of an unplanned conception as a result of intoxication;
2. Those suffering Foetal Alcohol Syndrome and Foetal Alcohol Effects, said to be the major source of mental illness and behavioural disorders, and generating on average a large public cost during their life time; Harwood (2000)
3. Those close to the drinker who may experience pain and suffering, abuse, violence, injury and death;
4. The public at large who may experience pain and suffering, violence and death including from motor vehicle accidents.
There are rarely good estimates of these phenomena, and the incidence and costs vary from country to country. However, it may be that these total costs (especially including item 3 which is very poorly measured) are almost comparable with the harm alcohol misusers do to themselves.
Motor Vehicle Accidents
As well as injury and death from motor vehicle accidents caused by alcohol consumption there is damage to the cars and property, the costs of law enforcement and prevention (which may include road design), and the costs of administrating the insurance system.
Motor vehicle accidents are particularly onerous, but there are other alcohol induced accidents. Industrial accidents caused by alcohol are usually covered in the next section.
Drinking alcohol may result in production losses from the
1. Reduced workforce size as a consequence of death or premature retirement;
2. Absenteeism from sickness or injury;
3. Reduced on-the-job productivity from sickness or injury.
Alcohol misuse may also cause loss of production in the unpaid household sector and in other non-market activities. Although the informal economy is not included in the standard measures of market output (such as GDP of GNP) alcohol induced losses are significant to those who are affected by them.
Most of the above events described above, to some extent impact on public spending (or, for those countries which depend more upon private provision, on the private insurance system). Another impact may be a reduction of public revenue (taxation). Such impacts are classic externalities, because they are unlikely to be taken into account by the drinker.
Among the most significant sectors affected by additional spending are costs to the health system (at every level from prevention to advanced tertiary specialities, but also nursing care later in life). social security or other forms of public income maintenance, traffic management, and also policing, justice and corrections.
Are Drinkers Rational?
The standard theory of economic value requires that people value what they purchase at least at the price they pay and usually more. It would be irrational to exchange money for a product if the money was more valuable to the purchaser. Underpinning this is the assumption that purchasers are rational. But are alcohol drinkers so rational, especially when they are drunk or addicted? How are we to value the consumption if they are not?
Economics tries not to be judgmental on such matters but in recent years behavioural economics has moved away from the assumption of stark economic rationality and paid attention to time inconsistency in decisions. (Frederick, Loewenstein & O’Donoghue 2002) An example is when a man goes into a bar to have a couple drinks, he drinks more than he planned , and the following day he regrets the change of mind.
When the purchaser values the product at at least the purchase price, it is usual to offset this value against the costs of supply (which are often the purchase price if taxes and subsidies do not cause distortions). It is perfectly true that the production of alcohol (or any other product) uses capital, labour and land. Forgoing that consumption would release those resources for other economic purposes. whose use would (almost) be valued the same as the forgone consumption. Thus social cost estimates do not include the cost of the resources used in production and paid for by the purchaser/consumer since they are offset by the consumption.
Unplanned consumption which is subsequently regretted – where there is time inconsistent consumption – is a waste and should not be valued at the full purchase price. So some proportion of their outlays should be deducted when the value of consumption is being calculated. Given the degree of misuse of alcohol, this deduction can add considerably to quantum of net social costs.
Calculating Social Costs
As the paper has already indicated, economists not only identify the various economic aspects, but quantify them and value the quantities to get a sum total of the net social costs of alcohol misuse. It is a complicated and sometimes subtle exercise, involving such technical issues such as counterfactual hypotheses, opportunity costs, the valuation of life, discounting and the problem of rationality, only some of which are touched on here. That is why it has been necessary to set out a framework of International Guidelines for Estimating the Costs of Substance Abuse. (Single et al 2003; Collins & Lapsley 2006).
But should we? In a paper to the Kettle Bruun Society, Eric Single and I argued that while there are indeed many difficulties regarding the ‘bottom-line’ estimates of total economic costs attributable to alcohol use, economic cost studies are nonetheless worthwhile research exercises that should be undertaken. But the alternatives – anecdotes and heuristic guesstimates – are unattractive, while carrying out the systematic estimate exposes data gaps (such as a lack of knowledge of the impact of alcohol misuse on others’ quality of life) and changes our understanding of the relative importance of the various impacts. (Single & Easton 2001).
Additionally, their magnitude can be compared with the tax revenue to indicate whether the levy covers those costs, although I shall argue that the levy should be higher than that.
The International Guidelines are based on rigorous economic theory. Aside from giving the estimates an integrity, they are also compatible with cost benefit and cost effectiveness evaluations. This means that the same conceptual framework and data can be used when a particular policy option or treatment is being evaluated.
Another eventual use is if we have comparable data across countries and through time we can trace better the economic and social impact of alcohol misuse and hence evaluate the gains from measures to reduce it. Would it not be a major step forward to know which countries were the more successful at limiting the damage from alcohol misuse?
Here are two reliable estimates of tangible costs as a proportion of GDP calculated using the current International Guidelines. They do not cover the intangible costs, arising from death injury and poor quality life which alcohol misuse generates (avoiding, among other things, the problem of valuing of life). Nor do they include the impact on the informal non-market economy, because there is no measure of the aggregate value of its output (the equivalent of GDP).
In each case the costs are estimated to be near 1 percent of GDP, the implication being that without alcohol misuse effective GDP – that available for useful consumption and productive investment would be 1 percent greater. However the increase in GDP per capita would be lower, depending on the mortality impact of the misuse on the population.
(TANGIBLE) SOCIAL COSTS OF ALCOHOL AS A PROPORTION OF GDP
Note however, that despite using the same international guidelines, there are considerable differences between the subtotals, only some of which may be explained by classification differences or by the different cultures and institutional arrangements of Australia and Canada..
Even so, while other countries may have different social costs of alcohol misuse – the waste from drinkers not taking into consideration the social consequences of their decisions– the magnitude per unit drink is likely to be large, reenforcing the common sense conclusion of a need for public policies to reduce alcohol misuse .
Whether Thailand is ready for a complete estimate of the social costs of alcohol misuse depends upon the quality and extent of its data. At the very least, a preliminary survey would be useful, indicating the gaps in knowledge and some approximate orders of magnitude.
Economic Contributions to Regulating Alcohol Misuse
Given that alcohol misuse generates public damage it is a reasonable goal of public policy to minimise or eliminate that damage by reducing the misuse. The use of economic instruments is but one of a a number of means of doing so. The interventions can cover such things as taxing alcohol and prohibiting or restricting its consumption in some, or all, circumstances, and prohibitions or restrictions on advertising and promotion. Of these, taxation is the most obvious for an economist to contemplate.
Historically, specific taxes on alcohol was imposed because it could be readily collected at brewery, distillery or port. (There are greater difficulty collecting excise from small vineyards.) Later in some jurisdictions the justification for special duties shifted from convenience to a notion of the propriety of taxing ‘sinful’ activities.
More recently in some jurisdictions, alcohol taxes have been justified by the arguments that raising the cost of alcohol better reconciles the private cost and social cost of alcohol consumption, internalising externalities.
The specific alcohol taxes deal with the externalities in two ways. First, the higher price signals to the imbiber the higher costs of consumption to the economy than the cost of production. That may discourage him from additional consumption which adds to the costs. Second, to the extent it does not, the additional public revenue may be used to offset the costs.
What is an appropriate taxation regime? The design problem is that with which this paper opened. In different circumstances the same drink can generate different social costs. In principle then, the specific tax on alcohol should vary according to those circumstances: in practice that is totally impractical. The tax on a drink will have to be the same whether it is a benign or harmful. That means that some drink will have too much tax levied on them, while others may not have enough.
Thus it is not obvious that the total revenue from alcohol specific taxation should exactly cover all the fiscal costs (or the total public costs). The economic intuition is that the revenue should be higher, in order to send a stronger signal to the marginal drink that the social costs of their consumption is very high.
This conclusion has been reinforced recently by analysis of optimal taxation when there is time inconsistent behaviour. Even if there are no externalities, there may be welfare gains from taxes, insofar as the following morning one may be pleased that the higher price of alcohol from the duty discouraged one from drinking too much the previous night. (O’Donoghue & Rabin 2006).
Just how far we this approach should be applied is unclear since it could apply to all consumption goods and services – since we are all inclined to indulge in instant gratification on almost everything (except saving for the future). Perhaps there are only a limited number of products for which the consequences of time inconsistency are so great that it is appropriate to apply the principle. Almost certainly one of the selected few would be alcohol.
Moreover the consumption of teenagers and binge drinkers (but not alcoholics) is more sensitive to the price of alcohol, and more likely to cut back if faced higher prices. Both of these groups are more likely to generate higher social costs than the average drinker.
In any case, what we may conclude is that it seems likely that an optimal taxation regime on alcohol is likely to generate more revenue than the immediate costs to the exchequer – or even to the public at large.
Because of the variable incidence of the social costs of imbibing, there are proposals to vary the taxation according to circumstances. The most common is the claim that different forms of alcohol – wine, beer, spirits – induce different levels of social costs (or just health effects), and the more benign should be taxed at a lower rate. Evidence for these differences are largely anecdotal, and – unsurprisingly – tend to be most promoted by those whose have a self interest in the lower tax on the drinks they produce or imbibe. But whatever the opinions, there is no evidence that the magnitudes of the effects are at all great.
On the basis of the available evidence the logic seems to be that the level of alcohol specific taxation should be related to the quantity of absolute alcohol rather than the value or type of drink. That would mean a levy that was related to the amount of absolute alcohol in the drink not its price (as in a sale tax).
This would impose more heavily on low value drinks which tend to be the most harmful, since it is cheaper to get drunk. The logic here is that it is cheap drinking where the heaviest social costs occur – by teenagers and binge and heavy drinkers. We would be targeting them, rather than all drinking. That is probably what the tax policy instrument does best towards reducing alcohol harm.
A further step would be to target the minimum cost of alcohol. This might mean that spirits, which are cheaper to supply, are levied more heavily than wine or beer. (Easton 2002)
However no country has an economically rational alcohol taxation regime, and tax revenue and administration considerations remain important in its design as do cultural particularities. Another key factor may be political. Even ignoring the liquor lobby moderate drinkers make up a formidable group resistant to higher taxation on their pleasures, even if it means lower taxation elsewhere. (One of the uses of social cost estimates is to convey to them that they are paying substantially for the external costs generated by others.)
More fundamentally, while taxes are a useful means of reducing alcohol misuse, they target clumsily, and so cannot be relied upon by themselves. There are a range of other measures including those that are targeted at specific social cost generating activities such as drink driving. There is a tendency to promote policies for which there is only tenuous evidence of their effectiveness. The attractiveness of the tax instrument is that if it fails to reduce alcohol misuse, it finances the consequences of the misuse.
Conclusion: Cultural Change Strategies
However, it may be that we focus too much on such economic strategies to deal with alcohol abuse, and dont think enough about those matters which economists take as given. In particular there is enough cross-cultural evidence to indicate that drinking practices are heavily influenced by cultural factors. Economists tend to take to take culture as given, but other disciplines may envisage strategies for cultural change.
As an economist I have to be cautious about such things. But I can report that Australasia, led by New Zealand, have embarked on programs of changing the drinking culture. The New Zealand theme is ‘It’s not what you have been drinking, it’s how you have been drinking’.
While there have been some immediate gains – getting really drunk is becoming less socially acceptable – the strategy is a generational one.
John F. Kennedy, walking through the gardens of the White House, observed some gardeners had dug the hole for a tree, but were finishing for the day before the tree was planted. Remonstrating, he was told that it would be planted the following day. It did not matter because the tree would take twenty years to grow. To which Kennedy replied ‘then it should have been planted yesterday.’
A cultural change strategy is so long term it is easy to leave it until tomorrow and the day after. But it should have been started yesterday. However, until it becomes fully effective, we will have economic and social costs from alcohol misuse, and it will be necessary to use taxes and other economic instruments to limit these costs.
(This includes some references not directly referred to in the body of the paper, but which are a part of its background.)
Andlin-Sobocki P. and Rehm J. (2005), “Cost of addiction in Europe”, European Journal of Neurology, 12(s1): p.28-33.
Babor, T. Caetano, R. Casswell, S. Edwards, G. Giesbrecht, N. Graham, K. Grube, J. Gruenewald, P. Hill, L. Holder, H. Homel, R. Österberg, E. Rehm, J. Room, R. and Rossow, I. (2003), Alcohol: no ordinary commodity. Research and public policy. Oxford and London; Oxford University Press.
Collins, D.J. and Lapsley, H.M. (2002), Counting the cost: estimates of the social costs of drug abuse in Australia in 1998-9, Canberra, National Drug Strategy Monograph Series No. 49,
Collins, D. Lapsley, H. Brochu, S. Easton, B. Perez-Gomez. A. Rehm, J. Single, E. (2006) International Guidelines for the Estimation of the Avoidable Costs of Substance Abuse, Ottawa, Canadian Centre on Substance Abuse. http://hc-sc.gc.ca/ahc-asc/activit/strateg/drugs-drogues/nat-res-rech/cost-cout-abus/index_e.html
Easton, B. (1997), “The Social Costs of Tobacco Use and Alcohol Misuse”, Public Health Monograph No. 2, Department of Public Health, Wellington School of Medicine.
Easton, B. (2002) Taxing Harm: Modernising Alcohol Excise Duties, Wellington, Alcohol Advisory Council of New Zealand, http://www.eastonbh.ac.nz/?p=641 or http://www.alac.org.nz/InpowerFiles/Publications/CategorisedDocument.Document1.1081.751cbc77-3b52-4fb0-88ae-84b3cf562076.pdf.
Frederick, S., G. Loewenstein & T. O’Donoghue (2002) ‘Time Discounting and Time Preference: A Critical Review’ Journal of Economic Literature Volume: 40: 2 June 2002, p.351-401.
Gmel, G, Rehm, J. Room, R. and Greenfield, T. (2000) “Dimensions of alcohol-related social and health consequences in survey research”. Journal of Substance Abuse 12:113-138.
Harwood, H. (2000). Updating estimates of the economic costs of alcohol abuse in the UnitedStates: Estimates, update methods, and data, Report prepared by the Lewin Group for the National Institutes on Alcohol Abuse and Alcoholism, Washington.
Haworth, A. Simpson, R. (2004) Moonshine Markets, New York and Hove: Brunner Routledge.
Heather, N. Peters, T. and Stockwell, T. (eds.) (2001) International Handbook of Alcohol Dependence and Problems. Chichester: John Wiley & Sons.
Huakau, J. Asiasiga, L. Ford, M. Pledger, M., Casswell, S. Suaalii-Sauni, T. and Lima, I. (2005) “New Zealand Pacific peoples’ drinking style: too much or nothing at all?” New Zealand Medical Journal 118:1491-.
Klingemann, H. and Gmel, G. (eds.) (2001) Mapping the Social Consequences of Alcohol Consumption. Dordrecht: Kluwer Academic Publishers.
Mathers, C. Bernard, C. Iburg, K. Inoue, M. Fat, D. Shibuya, K. Stein, C.; Tomijima, N. and Xu, H. (2003), Global Burden of Disease in 2002: data sources, methods and results, Global Programme on Evidence for Health Policy Discussion Paper No.54, Geneva, World Health Organization.
Pernanen, K. Cousineau, M-M. Brochu, S. and Fu Sun (2002), Proportions of Crimes Associated with Alcohol and Other Drugs in Canada, Ottawa, Canadian Centre on Substance Abuse, www.ccsa.ca/docs/crime2002.pdf.
Österberg, E. and Karlsson, T. (eds) (2002) Alcohol Policies in EU Member States and Norway: A Collection of Country Reports, Helsinki, STAKES.
Rice, D. Kelman, S. Miller, L. Dunmeyer, S. (1990) The Economic Cost of Alcohol and Drug Abuse and Mental Illness, 1985 (DHHS Publication No. ADM90-1694). San Francisco: Institute for Health and Ageing, University of California.
Room, R. Babor, T. and Rehm, J. (2005), “Alcohol and public health”, Lancet, 365(9458): 519-30.
Single, E. and Easton, B. (2001), Estimating the economic costs of alcohol misuse: why we should do it even though we shouldn’t pay too much attention to the bottom line results, paper presented at the annual meeting of the Kettil Bruun society for Social and Epidemiological research on Alcohol, Toronto, May.
Single, E. Collins, D. Easton, B. Harwood, H. Lapsley, H. Kopp, P. and Wilson, E. (2003), International Guidelines for Estimating the Costs of Substance Abuse-Second Edition, Geneva, World Health Organisation.
Single, E. Robson, L. Xie, X. Rehm, J. (1998) “The economic costs of alcohol, tobacco and illicit drugs in Canada, 1992”, Addiction 93:7: 983-998.
Skog, O.J. (1985), “The collectivity of drinking cultures: a theory of the distribution of alcohol consumption”, British Journal of Addiction, 80(1): 83-99.