«Introduction The most common measure by which the public sector at local, state or national level has affected the economic availability of alcoholic ...»
Pricing of alcohol
The most common measure by which the public sector at local, state or national level has affected
the economic availability of alcoholic beverages is taxation in its different forms. These include,
among others, setting excise duties or value added taxes on alcoholic beverages. Historically, the
most common reason for taxing alcoholic beverages has been to acquire financial resources for the
public sector (Babor et al., 2010). However, the effects of price changes on alcohol consumption and related harm are the same regardless as to whether the changes in taxes – leading to price changes – are motivated by fiscal, social order or public health interests.
Besides taxing alcoholic beverages, there are other measures affecting the economic availability of alcohol such as minimum prices for alcoholic beverages or regulation of discount prices. Until recently, however, the primary research and policy attention has been directed towards tax levels, and the effects of tax and price changes are often not clearly distinguished.
The impact of changes in prices of alcoholic beverages on alcohol consumption and related harm has been more extensively studied than any other potential alcohol policy measure (Anderson & Baumberg, 2006). When other factors remain unchanged, an increase in alcohol prices generally leads to a decrease in alcohol consumption, and a decrease in alcohol prices usually leads to an increase in alcohol consumption (Anderson & Baumberg, 2006). As a rule, the effect of a change in prices of a certain magnitude has different effects on the consumption of different kinds of alcohol, for example, distilled spirits, wines and beer. Usually elasticity values also vary between countries, and their values may change within a country as time passes (see, for example, Bruun et al., 1975; Ornstein & Levy, 1983; Leung & Phelps, 1991; Edwards et al., 1994; Österberg, 1995; 2001; Leppänen, Sullström & Suoniemi, 2001). The addictive nature of alcohol implies that the short-term price elasticity of alcoholic beverages is smaller in absolute value than the long-term price elasticity (Anderson & Baumberg, 2006; Box 1).
Box 1. Price elasticity of alcoholic beverages In econometric studies, the term “price elasticity” is used to express the effect of price changes on alcohol consumption. Negative own price elasticity value means that changes in prices and consumption are in the opposite direction: a rise in price leads to a decrease in consumption and a reduction in price leads to an increase in consumption. The numerical value of price elasticity gives the strength of the effect of a price change on consumption.
Alcoholic beverages are said to be price-elastic if the elasticity has an absolute value greater than one, which means that the percentage change in the amount of alcohol consumed is greater than the percentage change inprice. If the price elasticity has a value of -1.5, it means that a 1% rise in alcohol price will reduce alcohol consumption by 1.5%.
If the elasticity has an absolute value smaller than one, alcoholic beverages are said to be price-inelastic. This means that the percentage change in the amount of alcohol consumed is smaller than the percentage change in price. If the price elasticity has a value of -0.5, it means that a 1% rise in alcohol price will decrease alcohol consumption by 0.5%.
Price inelasticity does not mean that consumption is not responsive to price changes. Rather, it means that the proportional change in consumption is smaller than the proportional change in price. Only if the price elasticity value is 0.0 will price changes have no effect on consumption.
A wide range of studies have shown that increasing the price of alcohol reduces both acute and chronic harm related to drinking among people of all ages. This kind of evidence indicates that heavy or problem drinkers are no exception to the basic rule that alcohol consumers respond to changes in alcohol prices (Babor et al., 2010).
Studies have found that increases in prices of alcoholic beverages disproportionately reduce alcohol consumption by young people, and also have a greater impact (in terms of alcohol intake) on more frequent and heavier drinkers than on less frequent and lighter drinkers (Anderson & Baumberg, 2006). Changes in alcohol prices have also been found to influence drinking to the point of intoxication.
Summary of recent evidence Since 2006, three meta-analyses of the effects of changes in alcohol prices and taxes on drinking have been published (Table 1). The latest one, published in 2009 by Wagenaar and colleagues, is based on 112 studies reported in English (Wagenaar, Salois & Komro, 2009). The meta-analysis by Fogarty (2006) is based on elasticity estimates from those studies dealt with in Alcohol Policy and the Public Good, published in 1994 (Edwards et al., 1994; Fogarty, 2006). The analysis by Gallet (2007) includes 132 studies of alcohol price, income or advertising elasticities from 24 countries.
Table 1. Own-price elasticities for alcoholic beverages in three recent meta-analyses
The explanation for the low absolute value of own-price elasticity for beer could be that beer is a common beverage consumed during everyday leisure activities or with meals in countries where the studies in these meta-analyses originate. In these countries beer may be viewed as a necessary commodity among beer drinkers. In some other countries, beer is more of a luxury item with higher price elasticity (Edwards et al., 1994). This point is highlighted in the case of wine, which seems overall to have a higher absolute value for own-price elasticity than beer. In countries where it is an ordinary beverage with meals, wine can, however, be quite priceinelastic, with a low absolute value (Edwards et al., 1994). On the other hand, not so many decades ago, wine was quite a rare luxury item in the Nordic countries, where it was used mostly on festive occasions. It is, therefore, no surprise that at that time wine had an own-price elasticity of -1.6 in Sweden, and in later years the price elasticity of wine has been -1.5 in Norway and -1.3 in Finland (Sundström & Ekström, 1962; Horverak, 1979; Salo, 1990).
As Wagenaar and colleagues emphasize, price elasticities are not inherent properties of alcoholic beverages (Wagenaar, Salois & Komro, 2009). In the same manner as different uses of alcoholic beverages are reflected in price elasticity values, so the substitution between different alcoholic beverages as well as other commodities mainly depends on the uses of alcoholic beverages (Bruun et al., 1975). For instance, in a country where wine is used as a beverage with meals, a substantial increase in wine prices could increase the consumption of bottled water but it would hardly increase the use of home-distilled spirits or illicit drugs. In countries where alcoholic beverages are mainly used as intoxicants, increases in alcohol prices are more likely to lead to an increase in the consumption of home-distilled or -brewed beverages or even illicit drugs than of bottled water or milk.
Besides the price elasticity for all alcoholic beverages of -0.50 in the short term, Gallet also found a long-term elasticity of -0.82 (Gallet, 2007). Furthermore, he examined the importance of income elasticities for the demand of alcoholic beverages. In his meta-analysis income elasticity for all alcoholic beverages is 0.50, meaning that a 1% increase in consumers’ incomes leads to a 0.5% increase in alcohol consumption (Gallet, 2007).
One of the results in the meta-analysis by Wagenaar and colleagues is that price changes affect all types of beverage and all kinds of drinker, from light to heavy drinkers. According to their analyses, price and tax changes affect heavy drinking significantly, but the magnitude of the effect on heavy drinkers was less than on overall drinking (Wagenaar, Salois & Komro, 2009).
In Switzerland, over four waves of panel data, Gmel and colleagues found that heavier drinkers increased their consumption more sharply in the short term but declined to the level before the tax change in the long term (Gmel et al., 2008). Furthermore, recent studies of the effects of tax changes on problem indicators provide strong evidence that changes in alcohol taxes do influence the rates of problem drinking (Babor et al., 2010). For instance, the reduction in alcohol taxation in Finland in 2004 had substantial effects on alcohol-related sudden deaths, overall alcohol-related mortality, and criminality and hospitalizations (Koski et al., 2007;
Herttua, 2010; Mäkelä & Österberg, 2009). In Alaska, United States, excise duty increases in 1983 and 2002 were associated with substantial reductions in alcohol-related disease mortality (Wagenaar, Maldonado-Molina & Wagenaar, 2009).
Affordability The term “alcohol affordability” is nowadays also frequently used. This refers to people’s ability to buy and consume alcohol, and it is a function of alcohol price and consumers’ income (Rabinovich et al., 2009). According to Rabinovich and colleagues, affordability of alcohol increased between 1996 and 2004 in almost all EU member states. Their analysis also indicated that across the EU, 84% of the increase in alcohol affordability in the period 1966–2004 was driven by increases in income, and only 16% was driven by changes in alcohol prices (Rabinovich et al., 2009). This is because while incomes have increased considerably across the EU countries, the relative prices of alcoholic beverages have remained relatively stable or fallen (Rabinovich et al., 2009).
In the period 1995–2010, developments in excise duty rates were not at all uniform. In some countries (mainly the Nordics), alcohol excise duty rates were lower in nominal terms in 2010 than in 1995. In some countries, Germany being the most important example, alcohol excise duty rates were held constant in nominal terms in the 1955–2010. In most of the countries belonging to the EU before May 2004, the nominal values of alcohol excise duty rates were increased but by less than the rate of inflation, meaning than even in these countries the real values of excise duties fell. Only in a few countries, such as Greece and Italy, the nominal values of excise duty rates were increased so much that the excise duty rates also increased in real terms (Österberg, 2011).
Countries that joined the EU in 2004, and later countries such as Bulgaria and Romania, had to increase their alcohol excise duty rates considerably before or when they joined the EU. In almost all new EU member states since 2003, the nominal and real values of alcohol excise duty rates increased between 2004 and 2010. The exceptions are Cyprus, with a constant nominal rate, and Malta, with a constant excise duty rate for beer and a 50% decrease for distilled spirits.
Despite increases in alcohol excise duty rates in the new EU member states, the lowest excise duty rates were still found among them in 2011, Bulgaria and Romania being the clearest examples. Low excise duty rates for beer can also be found among the older member states (Germany, Luxembourg and Spain) (Österberg, 2011).
By 2011, no EU member states had moved from a zero excise duty rate for wine to a positive excise duty rate. In fact, during the creation of the single market in 1993 or in the process of joining it later, four countries (Bulgaria, Hungary, Luxembourg and Romania) abandoned their former positive duty rate for wine.
Minimum prices A complementary measure to tax increases, and one which manages any lack of pass-through of tax to price, is to set a minimum price per gram of alcohol, a policy option with an impact on heavy consumers far in excess of that on light consumers. The impact has been tested in Canada (Stockwell et al., 2012), and modelled in the United Kingdom in England (Purshouse et al.,
2009) and Scotland (ScHARR, 2010), where there is currently a law before the Scottish Parliament to introduce a minimum price for alcohol of €0.07/g alcohol (Scottish Government, 2012).
Minimum alcohol prices in British Columbia, Canada, were adjusted intermittently over the years 1989–2010. Time-series and longitudinal models of aggregate alcohol consumption with price and other economic data as independent variables found that a 10% increase in the minimum price of an alcoholic beverage reduced its consumption relative to other beverages by 16.1% (Stockwell et al., 2012). Time-series estimates indicated that a 10% increase in minimum prices reduced the consumption of spirits and liqueurs by 6.8%, wine by 8.9%, alcoholic sodas and ciders by 13.9%, beer by 1.5% and all alcoholic drinks by 3.4%.
In England, 59% of the alcohol sold for consumption elsewhere (“off trade”) and 14% of the alcohol sold for consumption on the premises (“on trade”) is sold for less than 5 pence (£0.05/€0.06) per gram of alcohol (Purshouse et al., 2009). Modelling estimated that setting a minimum price of 5 p/g (€0.06/g) would reduce overall consumption by 2.6% (a 3.4 g reduction per week), affecting heavy drinkers far more (25 g/week) than moderate drinkers (0.01 g/week).
It was estimated that annual deaths would decline by 157 in the first year and by 1381 after 10 years. Annual hospital admissions would fall by an estimated 6300 in the first year and by 40 800 after 10 years. The intervention would also lead to an estimated decline of 16 000 criminal offences during the 10 years modelled. During the same period, the study predicted that there would be 12 400 fewer unemployed people and 100 000 fewer sick days. The study estimated the value of these reductions in harm to society as £5.4 billion (€6.2 billion) over 10 years. The estimated value of this minimum price policy for the first year included National Health Service savings (£25 million/€29 million), the value of quality-adjusted life-years (QALYs) gained through better health (£63 million/€72 million), savings related to the costs of crime (£17 million/€19 million), the value of crime QALYs gained (£21 million/€24 million) and employment-related benefits (£312 million/€356 million). Again, the cost impact of this policy on consumers varied substantially among different groups of drinkers. It would cost drinkers an estimated £22 (€25) per year, ranging from £106 (€121) for heavy drinkers down to £6 (€7) for moderate drinkers. If no changes were made to consumption, it would cost heavy drinkers an estimated additional £138 (€157) per annum and moderate drinkers an estimated additional £6 (€7).