Drink Driving

Drink Driving

How Effective?

The counter-measures against drink driving consist of a range of legislative measures and their associated enforcement and penalties, education, and media campaigns.

Many of the evaluations described below have been carried out in other countries, e.g. the U.S.A and Australia. Given the cultural and contextual differences between the UK and other countries, and lack of comparability of drink driving between countries, it is unclear how these findings would translate to the UK.

Drink drive legislation

Successful and effective drink drive legislation relies on publicity (people’s awareness of the law and consequences of not complying with it) and visible, rapid enforcement (to act as a deterrent). The drink drive limit in the UK was introduced based on the studies discussed previously that describe the relationship between alcohol consumption and risk of collision and injury (see Research Findings section).

A review of evidence conducted by NICE (Killoran et al, 2010) examined the effectiveness of potential measures, including laws to limit blood alcohol concentration (BAC) levels, on drink driving and its associated injuries. The review included international studies that had examined the effect of lowering the drink drive limit for drivers, e.g. from 100mg/100ml to 80mg/100ml, or from 80mg/100ml to 50mg/100ml.

The review concluded that:

  • Overall, there is sufficiently strong evidence to indicate that lowering the legal BAC limit for drivers does help reduce road traffic injuries and deaths in certain contexts.

(Killoran et al, 2010)

The NICE review highlights a particular high quality study by Albalate (2006), which examined the effect of reducing the drink drive limit using data from 15 European countries (by comparing countries who had reduced their limit with a group of countries who had not):

  • The study analysed the total fatality rates for the period 1991-2003 across 15 European countries.

  • Reducing the BAC limit from 80mg/100ml to 50mg/100ml decreased alcohol-related driving death rates by 11.5 per cent in young people aged 18-25, and by 5.7 per cent in men of all ages (the effect was not statistically significant for the whole population).

  • There was a time lag before the benefits of the reduction in limit were seen. The effects were evident after 2 years and increased over time, with the greatest impact between 3 and 7 years.

(Albalate, 2006)

There has been much debate regarding whether the UK should introduce a lower drink drive limit. The European Commission adopted a Recommendation in January 2001 that Member States should set prescribed limits at or below 50mg/100ml. Attempts have been made to estimate the number of deaths and injuries that would be avoided in the UK if the drink drive limit were lowered from 80mg/100ml to 50mg/100ml.

  • Assuming reducing the limit would produce the same relative effect on accidents and casualties as observed in other European countries, between 77-168 deaths and 3611-15832 injuries are estimated to be avoidable (if the limit were lowered to 50mg/100ml).

(Killoran et al, 2010)

  • Using UK data from Maycock (1997) and Reported Road Casualties Great Britain, and making certain assumptions about how drivers would behave if the limit were lowered (e.g. that the behaviour of those driving well over the existing 80mg per 100ml alcohol limit is unlikely to be affected by lowering the limit), it has been estimated that:

    • Approximately 43 deaths and 280 serious injuries would be saved by lowering the limit (based on 2008 drink drive road casualty data and estimated relationships between accident risk and driver’s BAC).

    • Approximately 65 deaths and 230 casualties would be saved by lowering the limit (based on 2003 drink drive road casualty data and estimated relationships between accident risk and driver’s BAC).

(Allsop, 2005; Allsop, 2010, cited in North Report, 2010)

The range of the estimates is large, and the estimates presented above are based on different modelling procedures. The Killoran et al (2010) estimates are based on an extrapolation of the effect of lowering the BAC limit in other countries, assuming a shift in drinking behaviour across the whole driving population (not just drivers above the limit), and does not relate specifically to drink drive deaths and injuries (as defined as those deaths and injuries where an involved driver/rider was over the drink drive limit, i.e. the estimates include deaths and injuries where the involved driver/rider is under the drink drive limit).

Drink-Drive Rehabilitation Scheme

Since 1 January 2000, the courts have been able to offer drivers who have been disqualified for a period of at least 12 months for a relevant drink drive offence, a referral to an approved Drink Drive Rehabilitation (DDR) course. The courses were developed as an educational intervention aimed at reducing the likelihood of further drink drive re-offending.

If an offender opts to take up the referral opportunity and satisfactorily completes a course, their period of disqualification can be reduced. In the case of a 12-month period of disqualification, the reduction will be 3 months. For longer periods of disqualification, the period of reduction will be up to one quarter, as determined by the court. Responsibility for managing the DDR scheme lies with the Driving Standards Agency (DSA). The DSA set out the syllabus for DDR courses and approve courses offered by course providers. The DDR course is built on a behaviour change model and has specified learning outcomes and assessment criteria (see DSA (2011) for the syllabus).

  • In 2008, approximately 60-65 per cent of drink drive offenders were referred to DDR courses in Great Britain.

(North Report, 2010)

An evaluation of the DDR courses in Great Britain has been conducted, based on offenders who were convicted of a drink drive offence and referred to a DDR course provider between 2000 and 2002. By comparing those offenders who attended the course with those who were referred but did not participate in the course, analysis of DVLA offending data indicated that:

  • Up to 2 years after the initial drink drive conviction, offenders who did not attend a DDR course were 2.6 times more likely to be convicted for a subsequent drink drive offence compared with offenders who had attended a course.

  • Overall, the study found that attending a DDR course reduced the likelihood of re-offending for all offenders, regardless of social status, age or gender.

  • An extended period of evaluation (over 5 years) indicated that, in the longer term, those who do not attend the course are about 1.75 times more likely than attendees to be re-convicted for a drink-drive offence.

(Smith et al, 2004: Inwood et al, 2007)


It is difficult to isolate the effect of campaigns from the numerous other factors, e.g. enforcement, penalties, that can affect drink drive behaviour. Campaign evaluations use a variety of outcome measures, and these can range from subjective measures of campaign awareness and attitudes towards drink driving, to objective measures of behaviour.

In the UK, national drink drive campaigns are conducted through the Department for Transport’s THINK! campaign. The most recent evaluation of the THINK! Drink Drive campaign was conducted in 2009, and examined campaign awareness and attitudes towards drink driving.

  • Eighty-one per cent of respondents recalled seeing or hearing something in any of the campaign sources for the Christmas Drink Drive campaign.

  • The ‘Moment of Doubt’ drink drive TV ad was recognised by 78 per cent of respondents. Thirty-six per cent of respondents agreed the ‘Moment of Doubt’ TV ad ‘sticks in my mind’, and 28 per cent agreed ‘It made me think about the dangers of driving even after a small amount of alcohol’.

  • Young male drivers aged 17-29 were more likely to agree that the ad ‘had made me drive more carefully’ (10 per cent versus 5 per cent of all drivers).

  • The acceptability of driving after 2 pints changed little between all campaign stages among all adults (monitored since July 2007 following the initial burst of the campaign).

  • The proportion who thought it was very likely that they would get a criminal record if they were caught drink driving increased pre to post campaign (from 55 per cent in July 2007, to 62 in January 2009).

(THINK! Road Safety Campaign Evaluation, 2009)

Systematic reviews and meta-analyses of published evaluation studies (that have attempted to control for publication bias) looking at the effect of campaigns on behavioural outcomes, such as drink drive collisions, suggest that drink drive campaigns do reduce collision and injury numbers when combined with enforcement.

  • In a systematic review of 8 studies (all non-UK, published between 1975 and 1998), and using crashes and measured BAC levels as outcome measures, a median decrease in all crashes of 13 per cent was evident following the implementation of mass media campaigns. The median decrease in injury crashes was 10 per cent.

  • However, there was concern that not all relevant factors that could affect the outcome measures were controlled for. Most of the campaigns took place in areas with relatively high levels of enforcement and other activities to prevent alcohol-impaired driving.

  • None of the studies provided unequivocal evidence for the effectiveness of mass media campaigns.

(Elder et al, 2004)

  • A meta-analysis examined the effect of drink driving campaigns on behaviour (all non-UK studies). Studies included utilised different types of media, e.g. internet, TV, radio, cinema, and some campaigns were combined with enforcement measures.

  • The results of the meta-analysis show that significant reductions in the number of injury accidents were found for drink driving campaigns (a decrease of 14 per cent). The effect of the campaigns was only evident when combined with enforcement.

(Elvik, 2009)

  • A more recent meta-analysis estimated that road safety campaigns (defined as using organised communications involving specific media channels within a given time period) coincide with a 10 per cent reduction in accidents (or a 9 per cent reduction when controlling for publication bias and the variation in study outcomes between studies). Most of the campaigns included in the analysis were accompanied by enforcement measures (all were non-UK studies).

  • Meta-regression of these evaluation studies showed that campaigns may be more effective in the short term if the message is delivered with personal communication in a way that is proximal in space and time to the behaviour targeted by the campaign.

(Phillips et al, 2011)

Designated driver programs

There has been limited evaluation of designated driver programs. The review described below considered studies from the U.S.A and Australia.

  • A campaign to promote the concept and use of designated drivers reported a 13 per cent increase in survey respondents ‘always’ selecting a designated driver, but no significant change in self-reported alcohol-impaired driving.

  • Incentive programs based in drinking establishments to encourage people to act as designated drivers showed an increase of 0.9 in the number of patrons who identified themselves as designated drivers after the program was implemented.Ø All outcome measures had limited value in assessing the potential injury prevention benefits of the programmes.

  • There is insufficient evidence to determine the effectiveness of either campaign or incentive designated driver programmes for reducing alcohol-impaired driving and crashes.

(Ditter et al, 2005)

Alcohol ignition interlocks

An alcohol ignition interlock (alcolock) requires a driver to perform a breath test in order to start the vehicle. If the device detects alcohol in excess of the threshold value (can be set at required value), the vehicle will not start.

There is provision within the Road Safety Act 2006 to offer drink drive offenders the opportunity to participate in an alcolock programme. This would be done at the offender’s expense, and a reduction in the period of disqualification would be offered in return. This provision has yet to be brought in to force however. The North report (2010) notes that interlocks are not part of the ‘sentencing toolkit’ of courts in Great Britain, but that some parts of the passenger transport industry use them.

Studies have demonstrated the potential effectiveness of alcolocks in preventing drink driving whilst they are fitted to the vehicle, but there do not appear to be any long term effects on re-offending once the device has been removed (Clayton and Beirness (2008), Willis et al (2009)). Where programmes are voluntary, there have been issues regarding low participation and compliance.

  • The installation of ignition interlocks (in non-UK programs) was associated with large reductions in re-arrest rates for alcohol-impaired driving. Following removal of the interlocks, re-arrest rates reverted to levels similar to those for comparison groups. Limited evidence from studies that used crash rates as an outcome measures suggests that alcohol-related crashes decrease while interlocks are installed in vehicles.

(Elder et al, 2011)

  • Interlock programmes have been shown to be effective in reducing drink-driving recidivism for both first-time and repeat offenders while the device is installed. However, there is little, if any, residual effect in preventing impaired driving after the device is removed.

(Clayton and Beirness, 2008)

  • In a sample of largely first time drink drive offenders who had been subsequently re-licenced following disqualification, a 12-month interlock programme conducted in Great Britain reported 328 recorded BACs over 80mg/100ml corresponding to 172 potential trips.

  • The drop out rate for the interlock programme was high, with 43 per cent of participants in the interlock group failing to complete the 12 month programme.

(Beirness et al, 2008)

  • A review published in 2004 concluded that:

    • More studies of good quality are needed to confirm the effectiveness of alcolocks in reducing recidivism.

    • The participation rates for interlock programmes were too low for devices to have had much impact on the drink driving population as a whole.

(Willis et al, 2004)

Low BAC laws for young drivers

Evaluations of the effectiveness of low drink drive limits for young drivers have been carried out in the U.S.A and Australia:

  • A review of 6 studies concluded that there was sufficient evidence that lower BAC laws were effective in reducing crashes among young or inexperienced drivers. The studies reported reductions in crashes of between 4 and 24 per cent, depending on the study outcome employed (e.g. fatal crashes, non-fatal injury crashes).

(Shults et al, 2001)


  • Date Added: 03 Apr 2012, 08:10 AM
  • Last Update: 26 Jan 2017, 03:43 PM