Scottish Government

8 COST TO PRODUCTIVE CAPACITY OF SCOTTISH ECONOMY

Summary

• This section estimates the cost of alcohol misuse to the productive capacity of the Scottish economy in 2007.
• The biggest impact was due to premature mortality, which significantly reduced the potential working years of those dying due to alcohol-related causes before the age of 65 years.
• Excluding the protective effects of alcohol consumption on mortality, the estimated cost of premature mortality to the Scottish economy was estimated to be £287.9 million - £348.0 million (depending on the source of the data on premature mortality), with a mid-point of £318.0 million.
• Including the beneficial effects of alcohol in preventing some premature deaths gives a cost of premature mortality to the Scottish economy of about £313.3 million.
• Presenteeism (range: £177.0 million - £193.1 million; mid-point: £185.1 million), absenteeism (range: £109.4 million - £273.5 million; mid-point: £191.5 million) and unemployment (range: £150.9 million - £191.5 million; mid-point: £171.2 million) due to alcohol misuse were each estimated as having mid-point cost impacts of similar magnitudes in 2007.
• The estimated cost to the productive capacity of the Scottish economy in 2007 due to the effects of alcohol misuse was £725.2 million - £1,006.1 million, with a mid-point of £865.7 million.

Introduction

8.1 Various studies have shown that many people who misuse alcohol are in employment ( e.g.HSE, 1996; Single et al., 2003).

8.2 The international guidelines for estimating the costs of substance misuse (Single et al., 2003) state that the impact on and costs to the productive capacity of the economy under consideration should be included in the calculations. This requires " robust estimates of premature mortality and morbidity that can be attributed to substance abuse".

8.3 Studies of the costs associated with alcohol misuse do, however, differ with regard to the elements that should be included in the analysis to capture this impact. The following three possibilities for reducing the productive capacity of an economy are identified in studies of alcohol misuse:

• Absenteeism ( i.e. taking paid time off work due to health-related and other problems, some of which may be directly or indirectly due to alcohol misuse);
• Unemployment ( i.e. not being in paid employment because of health-related or other problems directly or indirectly due to alcohol misuse);
• Premature mortality ( i.e. dying before the official age of retirement where alcohol is the underlying or a contributory cause of death).

In addition, presenteeism ( i.e. the reduced activity and productivity of those who misuse alcohol but who are at work) is included in some studies.

8.4 According to Single et al. (2003), prior studies have found that " the largest part of morbidity/lost and reduced productivity costs is not due to measurable lost days of work, but from impaired productivity while on the job". Nevertheless, due to methodological and measurement concerns, presenteeism is not considered in all of the recent studies of the costs of alcohol misuse undertaken in the UK⁵².

8.5 Presenteeism and absenteeism, which both reduce the productivity of the workforce, impact directly upon employers, whilst unemployment and premature mortality in the working population affects employers and individuals.

8.6 It can be argued ( e.g. Single et al. 2003) that if labour markets worked perfectly (from an economist's perspective), any reduction in a worker's productivity due to substance misuse would result in a reduction in the wages that the employer was willing to pay that person. Thus such costs would be private costs borne by the employee rather than by the employer. However, labour markets generally do not function perfectly and wages and wage structures are much more rigid.

8.7 This section explores the impacts of these four elements identified in paragraph 8.4 on Scotland's productive capacity. The economic costs are estimated using a human capital approach (Scottish Government, 2008a), which involves applying a cost, based on average wages, to lost productive time.

8.8 Tables F.1 - F.3 in Appendix F summarise the employment-related data for Scotland in 2007 that underpin many of the subsequent calculations. Table F.1 shows the employment status of those aged 16+, whilst Tables F.2a and F.2b in Appendix F summarise data on average earnings in Scotland in 2007. Table F.3 considers the numbers of workers who are employees and self-employed and who are full-time and part-time workers.

Presenteeism

8.9 A survey undertaken in 2004 by reed.co.uk suggested that workers turn up at work with a hangover on an average of two and a half days per year (InfoScotland.com website ⁵³). These workers thought that they were 27% less efficient on these days, leading the Info.Scotland.com website to suggest that the productivity lost to hangovers could therefore account for an additional 1.62 million lost days, at a cost of £154 million.

8.10 The Scottish Government's 2008 update of earlier work (Scottish Government, 2008a) suggests that if workers lose the equivalent of an average of 0.68 days annually ( i.e. 27% of 2.5 days, as stated in the previous paragraph), this equates to approximately 1.7 million days per year, at a cost of about £203 million (in 2006/07).

8.11 A survey carried out for Norwich Union Healthcare in December 2007 (Institute of Alcohol Studies, 2009 ⁵⁴) found that a third of employees admitted to having been to work with a hangover. Almost one-in-six (15%) reported having been drunk at work. One-in-ten employees reported hangovers at work once a month, whilst one-in-twenty did so once a week. Work problems resulting from hangovers or being drunk at work included difficulty concentrating, reduced productivity, tiredness, and mistakes. However, these data do not include an estimate of the lost productivity associated with hangovers and drunkenness at work.

8.12 Table F.3 in Appendix F shows that 2,551,000 persons were in employment in Scotland in 2007 (1,340,000 men and 1,211,000 women) ⁵⁵.

8.13 Assuming that each worker (irrespective of employment status) lost an average of 0.68 days per year (see paragraph 8.10) due to alcohol-related inefficiencies at work, a total of more than 1.7 million days of work (1,734,860) were lost during 2007.

8.14 The estimated median gross cost per day for employers is shown in Table F.2b. In this Table, the gross weekly earnings of full-time employees in April 2007 (see Table F.2a) have been uplifted by 4.6% to reflect the increase in weekly pay between April 2007 and April 2008, based on the findings of the 2008 Annual Survey of Hours and Earnings ⁵⁶. Uplifts of 10% and 20% ⁵⁷ have also been included to reflect the estimated additional costs, such as National Insurance contributions and pension contributions, incurred by employers on behalf of some employees. This gives values of £116.18 per day (with 10% uplift) and £126.73 (with 20% uplift).

8.15 If it is assumed that all workers (irrespective of whether they are employees of self-employed and whether they work full-time or part-time) 'lose' the equivalent of 0.68 work days per year due to alcohol misuse, presenteeism reduced output by about £202 million - £220 million during 2007/08, as shown in Table 8.1.

8.16 However, if it is assumed that the 620,000 part-time workers only 'lose' an average of 0.34 days per year due to alcohol misuse, this results in lost output due to presenteeism of about £177.0 million - £193.1 million during 2007/08, as also shown in Table 8.1.

Table 8.1: Costs of Alcohol-related Presenteeism to the Scottish Economy in 2007

8.17 It should also be noted that there may be some positive benefits associated with drinking with workplace colleagues and business associates, such as increased productivity due to positive 'networking' effects gained from such socialisation (Cabinet Office Strategy Unit, 2003).

8.18 The quantifiable cost to the Scottish economy in 2007 of alcohol-related presenteeism is estimated to be £177.0 million - £193.1 million, with a mid-point of £185.1 million.

Absenteeism

8.19 Several national and international estimates have been made of the extent of alcohol-related absenteeism and its costs to the economy. A UK study published in 1981 ⁵⁸ estimated that alcohol caused 3% - 5% of all absence from work, equating to about 8 million - 14 million working days in the UK each year.

8.20 However, alcohol consumption has increased considerably since the early 1980s, and a study undertaken in 2001 across the UK workforce calculated that over 176 million working days were lost due to sickness and absenteeism, with 6% - 15% of this aggregate ( i.e. about 11 million - 17 million days) attributed to alcohol-related sickness (Leontaridi, 2003, quoted in Scottish Government, 2008a).

8.21 The Scottish Government (Scottish Government, 2008a) used CBI/ AXA data on absenteeism for 2006 to calculate that alcohol-related absence cost the Scottish economy about £200m during that year ⁵⁹.

8.22 According to the CBI/ AXA Absence Survey ( CBI, 2008), the average employee took about 6.7 days of sick leave during 2007 ⁶⁰.

8.23 Based on an employed workforce of 2,551,000 people (see Table F.3) and each employee (irrespective of employment status) having an average of 6.7 days of sick leave, there were a total of about 17 million (17,091,700) days of sick leave in Scotland in 2007.

8.24 However, if it is assumed that the 620,000 part-time workers have an average of 3.35 days of sick leave each year, there would have been slightly over 15 million (15,014,700) days of absence in Scotland in 2007. Table 8.2 summarises the associated costs under different assumptions about the proportion of sick days due to alcohol ( i.e. 6% and 15%), based on an average cost per employee of £116.18 per day and of £126.73 per day (see Table F.2b).

Table 8.2: Costs of Alcohol-related Absenteeism in Scotland, 2007

8.25 Table 8.2 suggests an annual cost to the Scottish economy in 2007 of £109.4 million - £273.5 million due to absences from work resulting from alcohol misuse, with a mid-point value of £191.5 million.

8.26 A recent study in Australia has used data collected as part of the 2001 National Drug Strategy Household Survey to consider the extent and cost of alcohol-related absenteeism in the Australian workforce (Pidd et al., 2006; Roche et al., 2008). Respondents suggested that 3.5% of people had taken absence from work for one day or more in the previous three months as a consequence of their alcohol consumption, compared with 39.7% because of illness/injury not due to alcohol (Roche et al., 2008) ⁶¹. The analysis also found that young employees and males were more likely to report alcohol-related absenteeism than older workers and females.

8.27 However, although the emerging picture of drinking patterns in Roche et al. (2008) is interesting, it is a previous paper by the same team of researchers (Pidd et al. 2006) that is potentially of greater relevance to this current analysis. Using the same data set ( i.e. from the 2001 Australian National Drug Strategy Household Survey), the researchers estimated that the mean number of days of alcohol-related absenteeism during the previous 12 months per employee had been 0.350 (95% CI: 0.254-446).

8.28 If every person in employment in Scotland in 2007 had had 0.35 days of alcohol-related absenteeism, this would have resulted in 892,850 such days at a total cost of about £104 million to £113 million ⁶². However, these values assume that drinking and absenteeism patterns in Scotland in 2007 are the same as those in Australia in 2001, which seems unlikely.

8.29 A recently-published Norwegian study (Norström and Moan, 2009) addressed the relationship between per capita alcohol consumption and sickness absence in Norway. It was based on a study (Johansson et al., 2009) using Swedish time-series data that had shown that a one litre increase in consumption was associated with a 13% increase in sickness absence among men. Although the Norwegian study strengthened the conclusion from the Swedish study, it used data for manual employees for 1957-2001, which limits its relevance to this study.

8.30 However, the Norwegian study does cite two previous Norwegian studies (Grimsmo and Rossow, 1997) and (Hammer, 1999) that had estimated that between 14% and 50% (depending on assumptions) of total short-term sick leave ( i.e. of 1-3 days) could be attributable to alcohol. Such data further reinforce the limited knowledge of the extent of the relationship between alcohol and sickness absence during the 1990s, which does not seem to have improved significantly during the 2000s.

8.31 Therefore the most reliable estimate of the annual cost to the Scottish economy in 2007 of absences from work resulting from alcohol misuse is £109.4 million - £273.5 million, with a mid-point value of £191.5 million.

Unemployment

8.32 The relationship between alcohol misuse and unemployment is not clear cut, although there is some evidence (Cabinet Office Strategy Unit, 2003) that excess drinking is negatively associated with employment.

8.33 The University of Sheffield's systematic review of the effects of alcohol pricing and promotion (Booth et al., 2008) concluded (Evidence statement 22) that:

" no recent systematic reviews or meta-analyses were identified that examined the effects of alcohol … on employment-related outcomes such as unemployment or absenteeism".

8.34 The modelling work undertaken as part of the Sheffield study for the Department of Health (Brennan et al., 2008) refers to a study by MacDonald and Shields (2004), which showed that "problem drinking" (measured by a combination of psychological and physical symptoms or in terms of quantity and frequency of alcohol consumption) was negatively associated with the probability of being in work. The study, which analysed data from the Health Survey for England (1997-1998) and focused on males aged 22-64 years old, showed that being a "problem drinker" led to a reduction in the probability of working of between 7% and 31%. Figures were not reported for females.

8.35 This current study uses two approaches to estimate the costs to the Scottish economy due to alcohol-related unemployment, but recognises the considerable limitations associated with each approach.

Approach 1: Based on Study by Catalyst/Scottish Executive 2001

8.36 The Scottish study by Catalyst (Scottish Executive, 2001) stated that data on the number of people unemployed in Scotland due to alcohol misuse are not available. Catalyst used an approach (after Meltzer, 1995) to calculate the unemployment rate among those with an "alcohol dependency" (which was not specifically defined in the report) using the prevalence rate for alcohol dependency stratified by employment status. These calculations suggested that 10% of males and 3% of females who were alcohol-dependant were unemployed in Scotland. Subtracting the general unemployment rates at that time of 7% for men and 2% for women indicated an "excess" unemployment rate for alcohol dependents of 3% for men and 1% for women. The Catalyst study calculated that that there were 3,536 unemployed individuals (3,393 males and 138 females) in Scotland in 2001 due to alcohol dependency, at a cost to the Scottish economy of £84 million ⁶³.

8.37 Due to the absence of more recent research data on the relationship between alcohol and unemployment, an attempt was made to replicate the Catalyst approach in this study (whilst recognising that the data need to be treated with caution). Data shown in Table 2.1, derived from SHeS 2008 (Scottish Government, 2009a), estimate that there were a total of about 230,000 (141,000 males and 88,000 females) adults drinking at harmful levels (defined as more than 50 units per week for males and more than 35 units per week for females) in Scotland in 2007. These data have been used to as a proxy for the numbers of "alcohol-dependent" males and females in Scotland for this element of the analysis ⁶⁴.

8.38 In terms of working age adults ( i.e. those aged 16-64), there were a total of 212,350 (131,301 males and 81,049 females) such persons.

8.39 Table F.1 in Appendix F shows that 84.0% of males and 77.5% of females of working age were economically active in Scotland in 2007. These percentages therefore equate to 110,293 economically-active, alcohol-dependent males and 62,813 economically-active, alcohol-dependent females in Scotland in 2007. Table F.1 also shows that 4.5% of all economically active people aged 16 and above were unemployed (4.5% for males and 4.6% for females) in 2007.

8.40 If 10% of alcohol-dependent males in Scotland were unemployed in 2007, these data suggest that the 'excess' unemployment amongst men due to alcohol dependency in 2007 was 5.5% ( i.e. 10%-4.5%), which equates to 6,066 unemployed males in 2007 due to alcohol dependency. However, these does not appear to be any 'excess' unemployment amongst women in 2007 due to alcohol dependency (as 4.6% exceeds 3%).

8.41 Based on average annual gross earnings for males in Scotland of £26,300 in 2007, the resultant annual cost to the Scottish economy of male unemployment due to alcohol dependency would be about £160 million ⁶⁵.

8.42 However, using data in Table F.3, adjustments can be made to reflect that 89.2% of employed males work full-time, whilst 10.8% work part-time. Assuming that these proportions would apply to the estimated 6,006 males who are unemployed due to their alcohol dependency, and that part-time workers have half the earnings of full-time workers, the resultant annual cost to the Scottish economy due to unemployment in 2007 due to alcohol dependence would have been about £150.9 million ⁶⁶. This total disregards any other impacts of lower levels (below dependence/harmful) of alcohol misuse on male and/or female employment and unemployment. It should also be noted that the Scottish macroeconomic context in 2007 was very different from that of 1995, which reinforces the need to treat these figures with caution. However, the lack of relevant recent research in this area precludes more accurate estimates.

Approach 2: Based on Study by Cabinet Office Strategy Unit 2003

8.43 The Cabinet Office Strategy Unit (2003) used a different methodology from Catalyst, which was based on the findings of MacDonald and Shields (2004) ⁶⁷. When this approach was applied to Scotland by the Scottish Government (Scottish Government, 2008a), it suggested that the value of alcohol-related unemployment to the Scottish economy in 2001 was more likely to have been about £146 million (rather than the £84 million for 2001 derived by Catalyst, based on Approach 1).

8.44 When calculating the effect of alcohol misuse on unemployment, the Cabinet Office (Cabinet Office Strategy Unit, 2003) used data showing that male heavy drinkers ( i.e. males drinking 50 plus units per week) spent an average of 11.4 days per annum out of employment. This estimate accounted for full-time and part-time male employment rates in England in the early 2000s.

8.45 It also used an estimated average of 8.1 days per year out of employment for female heavy drinkers (who were not included in the study by MacDonald and Shields, 2004). This estimate accounted for lower female participation rates and higher rates of part-time employment.

8.46 The Cabinet Office study also linked these estimated days out of employment for heavy drinkers to a reduction in the probability of working for "problem drinkers" of 6.9%. This was the most conservative estimate in the work by MacDonald and Shields (2004), and was based on a definition of problem drinking as "drinks every day" when asked about the frequency of drinking in the Health Survey of England.

8.47 Applying these estimates of 11.4 days per annum out of employment to the assumed 110,293 economically-active, alcohol-dependent males and 8.1 days per annum out of employment to the assumed 62,813 economically-active, alcohol-dependent females in Scotland gives an estimated cost to the Scottish economy of unemployment due to alcohol dependence in 2007 of £191.5 million ⁶⁸.

8.48 However, both of these estimates should be treated with considerable caution, as they are derived from data from England from the late 1990s and early 2000s (when the macroeconomic context was different) and rely on a variety of assumptions about alcohol dependency, its effects on rates of employment and unemployment, and the proportions of full-time and part-time employees.

8.49 Nevertheless, although both of these approaches have significant limitations, there is a lack of relevant recent published research on the effects of alcohol misuse on employment. Based on the above calculations, the estimate of the annual cost to the Scottish economy in 2007 of unemployment resulting from alcohol misuse is £150.9 million - £191.5 million, with a mid-point value of £171.2 million.

Premature Mortality

8.50 A different approach is required to calculate the effect of premature mortality due to alcohol misuse on the Scottish economy. The number of potential years of working life lost directly or indirectly due to alcohol misuse needs to be calculated. Then the earnings associated with these years (discounted to reflect their present value ⁶⁹) have to be calculated to reflect the cost of this lost productivity to the Scottish economy due to premature alcohol-related deaths in 2007.

8.51 There are two sources of data for the number of alcohol-related premature deaths in 2007:

• Alcohol Statistics Scotland 2009 ( NHS National Services Scotland, 2009a);
• Alcohol attributable mortality and morbidity ( NHS National Services Scotland, 2009b).

Approach 1: Using Data from Alcohol Statistics Scotland

8.52 The figures for alcohol-related deaths in 2007 in Alcohol Statistics Scotland 2009 are based on data from the General Register Office for Scotland ( GROS). These deaths are recorded using the World Health Organisation's International Classification of Diseases 10th Revision ( ICD10) ⁷⁰.

8.53 As shown in Table F.4a in Appendix F, there were 1,093 alcohol-related deaths (754 male and 339 female) of people aged 15-64 years with alcohol as the underlying cause in Scotland in 2007. There were also an additional 560 alcohol-related deaths (429 male and 131 female) amongst the same age group where alcohol was identified as a contributory cause of death. There were therefore a total of 1,653 alcohol-related deaths (1,183 male and 470 female) amongst people of working age in Scotland in 2007 according to this data source.

8.54 These data are presented by 5-year age bands in Table F.4a. Assuming all of deaths occur at the mid-point within each age band and that both men and women retire at 65 years of age, Table F.4b shows that total of 22,279 years of potential working life (15,764 male and 6,515 female) were lost due to the alcohol-related deaths in 2007. Table F.4c shows that 15,234 years of potential working life were lost due to deaths where alcohol was given as the underlying cause of death, whilst Table F.4d shows that 7,045 years of potential working life were lost by those for whom alcohol was a contributory cause of their death.

8.55 However, not all of these people who died prematurely due to alcohol misuse would have been in full-time employment, and adjustments need to be made to the numbers to reflect an assumed future employment rate of 80.2% for males and 73.8% for females (based on the 2007 rates shown in Table F.1). The average full-time annual earnings of £26,300 for males and £20,835 for females in 2007 also need to be adjusted to £24,880 for males and £16,751 for females to reflect that only 89.2% of the male workforce and 60.8% of the female workforce were in full-time employment. It is assumed that those who were employed part-time earned half the amount of full-time workers.

8.56 The resulting numbers were discounted at 3.5% (the rate currently recommended by HM Treasury) and are summarised in Table 8.3. The estimated cost to the Scottish economy in 2007 associated with the total loss of future earnings due to premature alcohol-related deaths in the working population in 2007 was £287.9 million. This total comprises £194.0 million due to those with alcohol as the underlying cause of their death and £93.9 million for those with alcohol as a contributory cause of their death.

Table 8.3: Estimated Costs to Scottish Economy in 2007 due to Alcohol-related Premature Mortality in 2007 (Approach 1)

8.57 This amount is slightly lower than the £328 million calculated for 2006 (Scottish Government, 2008a), which was based on 2,372 alcohol-related deaths in Scotland in 2004.

Approach 2: Using Data from Alcohol Attributable Mortality and Morbidity

8.58 The data reflecting premature mortality presented in the Alcohol Attributable Mortality and Morbidity report have been calculated by ISD Scotland by applying their age- and sex-specific alcohol population attributable fractions ( PAFs) to the number of deaths for each cause by age and by gender to estimate mortality that was wholly or partly attributable to alcohol consumption in Scotland in 2003 (as recorded by GROS). Only those causes of deaths which have an alcohol attributable condition recorded as the underlying cause were included in the analysis. Mortality data were extracted for 2003 because that year corresponded to the latest prevalence estimates of population alcohol consumption in Scotland available at the time from the Scottish Health Survey ( SHeS) 2003 (Scottish Executive, 2005) ⁷¹.

8.59 Table F.5a in Appendix F shows that, according to this data source, there were a total of 1,865 alcohol-attributable deaths (1,276 males and 589 females) amongst those aged 16-64 years in Scotland in 2003.

8.60 Given that there were a total of 11,390 deaths (7,091 males and 4,299 females) amongst this age group in 2003, these data suggest that about one-in-six of these deaths (16.4%) were attributable to alcohol (18.0% for males and 13.7% for females).

8.61 This PAF-based total of 1,865 alcohol-attributable deaths is higher than the total of 1,653 deaths (based on the GROS data for 2007 in Alcohol Statistics Scotland 2009) where alcohol was mentioned as either the underlying cause or a contributory factor (see Table F.4a).

8.62 Undertaking similar adjustments ( e.g. to reflect employment rates in 2007) and equivalent calculations with these data gives an estimated total loss of earnings due to premature alcohol-related deaths in the working population of £348.0 million (£274.1 million for males and £73.9 million for females), as shown in Table 8.4.

8.63 This value of £348.0 million is considerably higher than the loss calculated using the other data source for all alcohol-related deaths, but only slightly higher than the loss of £328 million calculated by Scottish Government for 2006 (Scottish Government, 2008a).

8.64 The above calculations take no account of the protective ( i.e. beneficial) effects of alcohol consumption ( e.g. by reducing cardiovascular disease). Alcohol attributable mortality and morbidity ( NHS National Services Scotland, 2009b) includes estimates of deaths prevented as a result of alcohol consumption by age and gender. These estimates for the working age population are presented in Table F.6a in Appendix F, which shows that an estimated 254 deaths were prevented in 2003 in those aged 16-64 (216 for males and 38 for females).

8.65 Repeating the above calculations for Approach 2 using the adjusted numbers of deaths (see Table F.6b) gives an alternative estimated total loss of earnings due to premature alcohol-related deaths in the working population (adjusted to include the protective effects of alcohol consumption) of £313.3 million (£242.5 million for males and £70.8 million for females), as shown in Table 8.4.

Table 8.4: Estimated Costs to Scottish Economy in 2007 due to Alcohol-related Premature Mortality in 2007 (Approach 2)

8.66 Both data sources for premature mortality have limitations. For example, the data based on attributable fractions relates to deaths in 2003, and determining the appropriate fractions for attributing is not an exact science. However, the GROS data are based on a more restrictive set of ICD10 codes. Both approaches are also dependent on the quality of the decisions identifying the recorded causes of death. Excluding any protection effects of alcohol, the costs to the Scottish economy in 2007 due to the premature mortality of those of working age were £287.9 million - £348.0 million, with a mid-point of £318.0 million. This mid-point value is also similar to the value using attributable fractions after adjusting for the protective effects of alcohol consumption.

Limitations and Summary of Costs

8.67 The above calculations have a number of limitations, due to the shortcomings of some of the data and a number of the underlying assumptions. For example, there is a lack of recent published studies relating to several of the aspects, which in turn has restricted the formulation of appropriate assumptions.

8.68 For example, the data on presenteeism and absenteeism are largely based on estimates that may be inappropriate for the working environment in Scotland in 2007. Calculations of lost earnings are based on average values, which would be over-estimates if workers with alcohol-related problems tend to be concentrated in relatively low-paid jobs.

8.69 A lack of information about the effects of alcohol misuse on unemployment means that the calculations relating to the impact of unemployment are also based on estimates that may be inappropriate for the Scottish context in 2007.

8.70 The calculations about premature mortality are more robust (assuming that people who consume alcohol have similar employment prospects as the total working population), but are dependent on the accuracy of the estimated numbers of premature deaths attributable to alcohol amongst those of working age. Such numbers are highly dependent on the accuracy of the recording of the cause(s) of death.

8.71 However, using the alcohol-attributable mortality derived from PAFs (as presented in NHS National Services Scotland, 2009b) requires using PAFs that have been applied to 2003 mortality figures (which are now out-of-date). PAFs could also change over time due to changes in population prevalence, age and sex and of drinking at particular levels ( NHS National Services Scotland, 2009b).

8.72 Nevertheless, the estimates presented in this section give an indication of some of the potential costs of alcohol misuse to the productive capacity of the Scottish economy in 2007, and of their comparative magnitudes. These are summarised in Table 8.5, which shows that premature death generates the largest economic impact. The estimated cost to the productive capacity of the Scottish economy in 2007 due to alcohol misuse is £725.2 million - £1,006.1 million. The "Best Estimate", which is based on sum of the mid points of the calculated ranges, is £865.7 million.

Table 8.5: Estimated Costs (£ million) to Productive Capacity of Scottish Economy Due to Alcohol Misuse in 2007