< Previous | Contents | Next >

EFFECTIVE AND COST-EFFECTIVE MEASURES TO REDUCE ALCOHOL MISUSE IN SCOTLAND: A LITERATURE REVIEW

CHAPTER FOUR POLICY AND LEGISLATION

Introduction

4.1 Chapters 4-10 present the findings of the effectiveness and cost-effectiveness reviews by type of intervention, namely

• Policy and Legislation
• Enforcement
• Prevention
• Screening and Detection
• Brief Interventions
• Detoxification
• Relapse Prevention

This chapter concerns policy and legislative interventions that can impact on alcohol consumption and misuse. It covers:

• fiscal policy (alcohol taxes);
• drink driving controls;
• licensing provisions; and
• policy towards alcohol advertising.

These interventions have to be evaluated at a population level and the types of analysis employed include looking at trends over time, comparing outcomes of interest before and after the introduction of some change, and comparisons of different populations experiencing different policy or legislative frameworks. There are no cost-effectiveness studies in the literature covering this area, although some of the effectiveness evidence comes from other types of economic analysis. The studies reviewed are summarised briefly in table 4.2 at the end of the chapter.

Fiscal policy

Types of intervention

4.2 Governments can influence the price of alcohol through the level of general or specific taxation on alcohol products. Although governments can reduce disposable income through increases in income tax, this has never been considered as an intervention to reduce alcohol misuse. The remainder of this section concentrates on price effects.

Evidence of effectiveness

4.3 Statistical analysis of prices and alcohol consumption over time provides evidence of the effect of prices on the population level of alcohol consumption. Increases in price reduce alcohol consumption but estimates of the size of the effect vary considerably. Price effects are measured in terms of the price elasticity, which relates the change in consumption to the size of the price increase. Thus, a price elasticity of —1.0 implies that a 1% increase in price will produce a 1% reduction in consumption. If the absolute size of the price elasticity is less than 1 then any price increase will produce a less than proportional reduction in consumption. Table 4.1 summarises the available information on price elasticities. The upper part of the table gives the number of studies that have reported results in the ranges specified and the lower part of the table gives UK Treasury estimates used to forecast tax revenue.

4.4 These estimates cover different time periods and different countries. The latest data included are for 1992. Comparisons across the studies are not possible without access to the original data. However, there appears to be some consistency in the finding that elasticities are higher for wine and spirits than for beer (NIAAA 2000; Raistrick et al 1999). The impact of tax increases on targets for the National Alcohol strategy would depend, therefore, on which forms of alcohol are most associated with problem drinking.

Table 4.1 Summary of price elasticity data

Population groups

4.5 The evidence concerning the distribution of effects across the population is variable. One study reports that moderate drinkers are more price responsive than heavy drinkers. In another study, the largest effect was found in the 80%-90% band of the population when ordered by level of alcohol consumption (Raistrick et al 1999).

4.6 The evidence relating to under-age and youth drinking is unclear. Other interventions, such as drink drive penalties, may be more effective than price in determining the behaviour of college students (NIAAA 2000).

Quality and relevance of evidence

4.7 The methods and data vary between studies and the estimates are not particularly up to date. However, the direction of the price effect is clear and the main uncertainty relates to the size of the effect.

Drink drive legislation

Types of intervention

4.8 In the US, legislation has been enacted to introduce lower permitted blood alcohol levels for younger drivers and for persons with previous convictions for drink driving. There have also been reductions in permitted alcohol levels for all drivers. A number of laws have been passed to bring into force additional sanctions for drink driving, such as mandatory licence suspension and ignition interlock devices.

Evidence of effectiveness

4.9 The effectiveness of drink driving legislation depends upon enforcement and not just the enactment of the legislation. Evidence concerning the effectiveness of specific enforcement actions is considered in the next chapter. However, the levels of effectiveness indicated by studies attempting to evaluate the impact of the legislation will undoubtedly contain some element relating to enforcement.

4.10 The introduction of zero levels of permitted alcohol consumption for drivers under 21 has resulted in a 20% decline in crashes likely to involve alcohol; i.e. night-time single vehicle crashes (NIAAA 2000). Six studies have found a reduction in injuries or crashes following the legislation but half of these were not statistically significant (Zwerling and Jones 1999). One US State has introduced lower levels of permitted alcohol for drivers with a previous conviction and the effect of this was positive (NIAAA 2000).

4.11 Reducing blood alcohol levels from 0.10 percent to 0.08 percent for other drivers in the US has reduced alcohol related fatal crashes (NIAAA 2000). It should be noted that this has the effect of reducing permitted levels in the US to the current permitted level in the UK.

4.12 Legislation to allow automatic license suspension has reduced crashes and fatalities in some states (NIAAA 2000).

Population groups

4.13 The only specific groups considered are under-age drinkers and persons with previous convictions, covered above.

Intensity of intervention.

4.14 In one study of lower blood alcohol levels for younger drivers, there was some evidence of an increasing effect on night-time single vehicle fatalities as the permitted level of alcohol reduced:

• 7% reduction with .04%-.06%;
• 17% reduction with .02%; and
• 22% reduction with 0%.

For reference, the UK legal limit corresponds to .08%.

Quality and relevance of evidence

4.15 Wagenaar et al (1995) specifically reviewed the quality of the literature and found it to be limited by weak study designs and reports failing to include basic data required for meta-analysis. There was some evidence of publication bias in the US literature. Studies with comparison groups reported smaller effects than before and after designs. US studies using descriptive statistics only were likely to report larger effect sizes. The longer the follow-up, the smaller was the estimated intervention effect.

4.16 The US literature is particularly dominant in the evaluation of legislative changes as researchers are able to conduct robust inter-State comparisons of different legislation or legislation introduced at different times. However, the results from the US must be considered in the context of falling total consumption. Cross-national studies may be affected by cultural differences that make results difficult to interpret. Caution should be exercised in translating results from one country to another. Some States have introduced more than one legislative change at the same time, making the effects difficult to estimate separately.

Licensing

Types of intervention

4.17 Licensing controls can affect a range of issues, such as hours of operation for outlets selling alcohol, types of outlet permitted to sell alcohol, the density of outlets within an area and the age at which alcohol can be legally purchased or consumed. In some countries, such as the US, sales of alcohol may be controlled through public monopoly.

Evidence of effectiveness

Licensing hours

4.18 The UK evidence related to the effect of licensing hours on alcohol consumption and alcohol-related problems is unclear. Some studies from other countries have suggested that longer hours increase alcohol related problems (Raistrick et al 1999).

Outlet density

4.19 Higher outlet density may affect alcohol sales either through reducing the time and travel costs involved in purchasing alcohol or through competitive pressure leading to price reductions through promotional activities. There is some evidence that higher outlet density is associated with higher sales and increased fatal crashes (NIAAA 2000). Whilst this finding has been replicated in a number of countries, studies in the UK have produced mixed results (Raistrick et al 1999). Raistrick et al suggest that the evidence of an association between number of outlets and alcohol consumption needs to be interpreted with care. Effects appear to vary with type of alcohol product. Different types of outlet may have different influences particularly on alcohol related problems.

Outlet type

4.20 Evidence from Australia suggests that different licences are associated with different levels of alcohol related problems, with night-clubs and bars posing the highest risk (Raistrick et al 1999). In the US, however, bars had a lower risk for motor vehicle accidents than other licensed outlets. US evidence also shows a difference between outlet types in sales to under-age persons. Grocery stores were more prepared to sell to minors than other types of outlet. Factors associated with lower sales to minors included being part of a chain, membership of a trade association and having a high proportion of income from alcohol sales (Raistrick et al 1999).

Age limits

4.21 In the US, 20 of 29 studies on the effects of raising the drinking age to 21 showed significant decreases in traffic crashes and traffic fatalities (NIAAA 2000). These studies from the early 1980s showed a 10% - 15% drop in alcohol related traffic deaths among youth. There is some evidence of reduced alcohol consumption, with the proportion of high school seniors reporting drinking in the last month falling from 72% in 1980 to 51% in 1999. There is also a reported reduction in drinking among people aged 21-25 who grew up in States with a minimum legal drinking age of 21 (NIAAA 2000).

Quality and relevance of evidence

4.22 There are relatively few studies of licensing interventions and the results need to be treated with some caution. There are difficulties measuring the impact of age restrictions because of high levels of under-age drinking

Advertising controls

Types of intervention

4.23 Governments can potentially restrict the level of advertising and the content of advertising, either by legislative action or through voluntary agreements with the alcohol industry. There may also be controls on promotional activities.

Evidence of effectiveness

4.24 There is little evidence relating to the impact of specific advertising controls. In Canada, full or partial advertising bans in some provinces did not provide any evidence of an effect on total alcohol consumption. A comparison across 17 OECD countries did find an association between alcohol bans and reduced consumption (Raistrick et al 1999). However, statistical analysis of the relationship between expenditure on advertising and alcohol consumption has failed to find any strong association. Experimental studies have found either no effect on beliefs and behaviours or small or short-term effects (NIAAA 2000).

Population groups

4.25 Survey research on alcohol advertising and young people shows a small but significant association between exposure to and awareness of advertising and drinking beliefs and behaviours (NIAAA 2000). The effectiveness of media literacy training for children also provides at least indirect evidence of advertising effects on children (see Prevention).

Quality and relevance of evidence

4.26 The lack of evidence associating advertising and alcohol consumption may be the result of there being relatively little change in spending compared to the total level of spending. The effect of advertising bans in cross-country comparisons could be due to reverse causality; i.e. countries with low consumption being more likely to have alcohol advertising bans.

Further research

4.27 The evidence relating to the size of the price effect could be improved and updated but any further studies should also focus on the distribution of effects and the impact on particular problem behaviours. In the areas of drink driving and licensing, there is a lack of UK evidence and any new initiatives should be carefully evaluated. Longer-term follow up of the impact of legislative action is also required.

Table 4.2 Summary of Policy and Legislation Effectiveness Studies Reviewed

< Previous | Contents | Next >