Productivity losses
4.1.1 Productivity loss due to disability
4.1.2 Lost lifetime earnings from premature mortality
A loss in productivity of an individual due to sight loss will only equate to a loss in productivity to the economy under fairly strict conditions. These are:
- the economy is at full employment so any reduction in hours worked due to sight loss, or any permanent reduction in labour force participation through early retirement or death, cannot be replaced by employing or increasing hours of other workers; and
- the income of an individual is proportional to the total value added to production.
The first condition will fluctuate over time as the economy moves into, and out of, full employment. A reduction in labour when labour is scarce will have a greater impact on productivity compared to an economy with an abundant labour supply. In this situation, a temporary or permanent reduction in working hours due to partial sight and blindness cannot be replaced by another worker.
Consequently, a loss in productivity due to VI and blindness is expected to represent a real cost to an economy operating at a low level of unemployment.
The second condition will occur if there is a perfect labour market such that the marginal benefit from an additional hour of work (the value added) is equal to the marginal cost (the wage). In reality, labour markets are imperfect for a number of reasons, for example asymmetric information in the market, and labour market restrictions imposed by government regulation and natural barriers. In addition, synergy created between labour, capital and land means a reduction in working hours may also impact the productivity of other factors of production. Consequently the value of productivity from labour will be larger than the wage provided to an individual so using lost income from partial sight and blindness as a proxy for lost productivity will tend to underestimate the true cost. It is likely that in the absence of sight loss, people with VI and blindness would participate in the labour force and obtain employment at the same rate and average weekly earnings as others. The implicit assumption is that the numbers of such people would not be of sufficient magnitude to substantially influence the overall clearing of labour markets, and average wages remain the same.
In this report, productivity losses are estimated using the lower than average employment rates for people out of work due to a seeing disability, and lost lifetime earnings due to premature death attributed to VI.
Productivity losses may also occur as a result of higher absenteeism, and lower productivity at work (‘presenteeism costs’). However, these components could not be estimated due to lack of available data for the ROI and similar countries. Thus, productivity losses presented in this report are conservative, and do not reflect the full magnitude of lost productivity from VI and blindness.
Deloitte Access Economics adopts a human capital approach to the estimation of productivity losses in developed countries such as the ROI, which is most consistent with the first condition above of an economy at, or close to, full employment. It is assumed the ROI operates at sufficiently low unemployment to incur a permanent loss in productivity. The unemployment rate is defined as the proportion of the labour force (people employed plus people unemployed but seeking work) without jobs. Unemployment averaged approximately 5.3% from the beginning of 2000 to the end of 2009 (CSO, 2010c). However, the current seasonally adjusted, standardised unemployment rate for the ROI is 13.5%. This may or may not be sufficiently low to incur a permanent productivity loss, and the productivity losses presented in this report should therefore be interpreted with caution.
Calculation of productivity losses is restricted to people aged between 15-64 years in the ROI.
4.1.1 Productivity loss due to disability
The productivity loss due to a disability is dependent on a number of factors:
- the likelihood of someone with that disability being employed;
- the likelihood of someone with that disability needing time off work due to the disability; and
- the value of lost production.
The ‘employment gap’, that is, the lower employment of people with seeing disabilities that can be attributed to their seeing disability, can be estimated directly from the ROI National Disability Survey (NDS) data. The NDS was a country-wide, cross-sectional survey conducted in the ROI in 2006 by the CSO with the main sample drawn from people who had indicated they had a disability in the 2006 Census (approximately 8% of the ROI population in 2006). From this population, 12,661 people were interviewed for the NDS. The results of the NDS sample were grossed-up by the CSO to the full population of 325,800 people who indicated they had a disability in 2006, using a weighting system.
The results of the NDS were used to estimate that 50,600 people in the ROI had a seeing disability in 2006. Disabilities were self-reported to the survey interviewer by participants, and the NDS included no formal testing of VA. This may lead to an overestimation or underestimation of VI when defined as a VA below 6/12.
Of these 50,600 people, it was estimated that 13,200 were adults living in private households. In total the results of the NDS estimated that 2,368 adults in private households were unable to work due to a seeing disability in 2006 (CSO, 2010d).
In the absence of age group specific data from the NDS, it was assumed that all adults reporting they are unable to work due to a seeing disability are primarily of current workforce age (18 to 64 years). Therefore, growth rates of the ROI population aged 18 to 64 years from 2006 were applied to this figure to estimate the numbers of adults unable to work due a seeing disability in 2010, 2015 and 2020 (CSO, 2008).
As shown in Table 4.1, it is estimated that in 2010, 2,576 adults were unable to work due to their seeing disability. This increases to 2,890 people by 2020.
Table 4.1: Estimated adults in private households unable to work due to seeing disability
| 2010 | 2015 | 2020 | |
|---|---|---|---|
| Working age (18 to 64 years) population growth rate from 2006 | 8.8% | 16.2% | 22.1% |
| Estimated adults unable to work due to seeing disability (a) | 2,576 | 2,751 | 2,890 |
Source: Deloitte Access Economics calculations using CSO (2008; 2010d)(a) Estimated by applying population growth rate of the population aged 18 to 64 years from 2006 onwards to the number of adults unable to work due to a seeing disability in 2006 from the NDS (CSO, 2010d).
The estimates above for 2015 and 2020 are not implicated by the ROI State Pension age being increased to 66 years in 2014 (the current transition pension at 65 years is due to be abolished then), since the numbers are based on the NDS responses for all adults. A person of any age who wishes to work but cannot work due to a seeing disability (as assessed in the NDS) constitutes a productivity loss regardless of their age.
People who report being unable to work due to a seeing disability may not necessarily have been employed even in the absence of that disability. This response in the NDS is interpreted as the individual being unable to seek work. Therefore, the number of people unable to work due to a seeing disability is multiplied by the employment-to-population ratio (employment probability). For the ROI working age population, this ratio in quarter two of 2010 (the most recent quarter for which data were available) was 60.4% (CSO, 2010e). The employment-to-population ratio is assumed to remain constant to the year 2020.
Salaries are applied as a proxy for the value of lost production. Estimated average weekly earnings in the ROI in quarter two of 2010 (the most recent quarter for which data were available) were €690.48 (CSO, 2010f). This equates to average annual earnings of €35,904 in the year 2010.
In the absence of age-specific data for people unable to work due to disability, total productivity losses for the ROI due to VI were estimated as the product of people unable to work due to a seeing disability, the employment-to-population ratio, and average annual earnings for the entire ROI working-age population. This produced an estimated productivity loss of €55.8 million due to seeing disabilities in 2010. To estimate productivity losses in 2015 and 2020, constant wages were conservatively assumed and the employment-to-population ratio in 2010 was applied up to the year 2020.
The estimated productivity losses due to seeing disabilities in the ROI for years 2010, 2015 and 2020 are presented in Table 4.2.
Table 4.2: Estimated productivity losses due to seeing disabilities in the ROI
| 2010 | 2015 | 2020 | |
|---|---|---|---|
| Number of people unable to work due to a seeing disability (a) | 2,576 | 2,751 | 2,890 |
| Employment-to-population ratio | 60.4% | 60.4% | 60.4% |
| Average annual earnings (b) | €35,905 | €35,905 | €35,905 |
| Total estimated productivity losses due to seeing disabilities (€'000) | €55,872 | €59,664 | €62,682 |
Source: Deloitte Access Economics calculations using CSO data (2008; 2010d; 2010e; 2010f) (a) Calculated by applying estimated population growth from 2006 onwards (CSO, 2008) to estimated people unable to work due to seeing disability in 2006 (2,368 people). (b) For quarter two, 2010.
4.1.2 Lost lifetime earnings from premature mortality
VI and blindness are associated with a higher than average risk of mortality because they are correlated with a higher risk of falls, fractures, motor vehicle accidents and depression. However, only a proportion of the additional deaths in the vision impaired can be specifically attributed to low vision and blindness rather than associated co-morbidities. Premature death due to VI and blindness results in a future stream of productivity losses due to lost potential earnings.
The number of deaths due to low vision and blindness in the ROI were estimated using the following data:
- the number of people with mild VI, moderate VI and blindness in the ROI (Chapter 2);
- mortality rates in the general ROI population;
- the relative risk of death in people with VI or blindness; and
- the aetiological fraction – the proportion of additional deaths in people with VI specifically attributable to VI rather than co-morbidities.
As described below, the relative mortality risk was estimated using studies in people with VA less than 6/12. Therefore, in this study, deaths due to VI are estimated for all people with mild VI, moderate VI or blindness.
In estimating the increased risk of mortality with VI and blindness, it is important to control for age and gender (Anstey et al, 2001; Globe et al, 2005) as well as comorbidities. Klein et al (1995) reported that people with specific vision conditions had an increased mortality risk of 1.57 times for the presence of sight loss and of 1.28 times for any cataract. However, when accounting for the presence of cardiovascular disease none of the conditions causing sight loss showed a statistically significant odds ratio for decreased survival.
An improved level of statistical control was achieved in the Melbourne Visual Impairment Project (MVIP) where partial sight and blindness was found to be significantly associated with an increased mortality risk of 2.34 times (McCarty et al, 2001). The result accounted for the confounding presence of age and age-related co-morbidities, such as basic cardiac risk factors. Similarly, Wang et al (2001) report an increased mortality risk of 1.7 times with the presence of any sight loss. Their analysis accounted for co-morbidities such as a history of cancer, stroke, gout and diabetes, some of which result from basic cardiovascular risk factors such as hyperlipidemia and hypertension.
Following previous international Access Economics studies on the burden of disease of VI and blindness, and in the absence of Ireland-specific studies on the relative mortality risk with VI and blindness, the relative risk of death was derived from the MVIP and Australian mortality data (Access Economics, 2010a; 2008b; 2009; 2006; 2004). Using MVIP data, the odds ratio of mortality for the vision impaired (VA < 6/12) compared with the general population is estimated to be 2.34 (95% confidence interval: 1.03-5.32) for people aged 40 years and older, based on approximately five years’ follow-up for urban participants (McCarty et al, 2001) and standardising for age, male sex, smoking duration, duration of high blood pressure and arthritis.
The odds ratio of 2.34 was used to proxy the relative risk of death in the population with VI and blindness aged 40 years and older. The age/gender specific numbers of deaths in the VI and blind population in each year, were estimated as the product of the odds ratio, the most recent (2009) age and gender specific mortality rates for the general ROI population (CSO, 2010g), and age/gender specific population estimates for 2010, 2015 and 2020 (CSO, 2008).
As detailed above, incremental deaths in the vision impaired are not all due to VI and blindness but also due to co-morbidities. An aetiological fraction of 1.38% has previously been estimated using Australian Bureau of Statistics (ABS) mortality data for the years 2003 to 2008 (ABS, 2010; Access Economics, 2010a). These data show a ratio of 72.4:1 between people for whom ‘diseases of the eye and adnexa’ was one of multiple causes of death (3,040 deaths) and the underlying cause (42 deaths). Factors that combine with VI and blindness to cause death include conditions such as osteoporosis, events such as falls or motor vehicle accidents, and risk factors such as poor light or roads. Thus, 1.38% (1 ÷ 72.4) of deaths due to eye disease are specifically caused by the eye disease.
Deaths attributable to VI and blindness were estimated by applying the aetiological fraction of 1.38% to calculated total deaths in the VI and blind population aged 40 years and over.
Estimates of the numbers of deaths attributable to VI and blindness in the ROI for 2010, 2015 and 2020 are presented in Table 4.3. It is estimated that there were 193 deaths from VI and blindness in 2010, which is projected to rise to 319 deaths by 2020. The majority of deaths from VI and blindness are estimated to occur in people aged over 64 years.
Table 4.3: Estimated deaths attributable to VI and blindness in the ROI
| Age-gender group | 2010 | 2015 | 2020 |
|---|---|---|---|
| Males | |||
| 40-44 | 0.16 | 0.18 | 0.21 |
| 45-54 | 0.66 | 0.73 | 0.83 |
| 55-64 | 1.93 | 2.14 | 2.39 |
| 65-74 | 6.24 | 7.73 | 9.31 |
| 75-84 | 15.22 | 18.40 | 22.64 |
| 85-94 | 29.51 | 40.80 | 57.18 |
| 95+ | 20.93 | 33.50 | 50.10 |
| Total male deaths | 74.64 | 103.29 | 142.46 |
| Females | |||
| 40-44 | 0.14 | 0.16 | 0.18 |
| 45-54 | 0.69 | 0.76 | 0.84 |
| 55-64 | 1.60 | 1.77 | 1.99 |
| 65-74 | 4.06 | 4.95 | 5.94 |
| 75-84 | 14.22 | 15.55 | 17.96 |
| 85-94 | 51.30 | 61.27 | 72.41 |
| Total female deaths | 46.04 | 61.65 | 76.94 |
| Total deaths | 192.71 | 249.59 | 318.92 |
Source: Deloitte Access Economics calculations using ABS (2010), CSO (2008; 2009; 2010f), Department of Finance (2010) and McCarty et al (2001).
The productivity loss, or future stream of lost income, is assumed to be incurred until the assumed retirement age of 65 years. Transition and Contributory State Pensions are available for people aged 65 years, and 66 years and older, respectively. Thus, productivity losses were not estimated for people aged 65 years and over. In the ROI, average life expectancy exceeds 64 years and was therefore not incorporated into these calculations (CSO, 2009).
The annual productivity loss due to death was valued using 2007 average annual earnings data by workforce age group (CSO, 2010h) adjusted to 2010 using the change in average weekly earnings between 2007 and 2010 (CSO, 2010i).
Age/gender specific wage data were applied since, in contrast to the productivity loss calculations in Section 4.1.1, age/gender specific data were available for deaths.
It is important to discount future streams of costs and benefits to a present value when expressing results for a specific year such as 2010. For cost effectiveness studies, there have been numerous international debates regarding both the discount rate and whether the same rate should be applied to both costs and health benefits (Drummond et al, 2005). This study does not synthesise economic costs and DALYs, and the issue of differential discount rates for costs and benefits is not relevant. A 4% discount rate was applied to future productivity losses in this study; this rate is commonly used in cost benefit and cost effectiveness analyses of public sector projects in the ROI (Department of Finance, 2010).
Similar to productivity losses from disability, productivity losses from premature mortality also need to be adjusted by the likelihood that someone with VI or blindness is employed. This is estimated using gender-specific employment-to-population ratios for quarter two of 2010 in the ROI (CSO, 2010e), being 64.5% for males and for 56.4% for females.
Productivity losses from premature mortality were calculated as the product of deaths attributable to VI or blindness in each age group (Table 4.3), the gender-specific employment-to-population ratio (CSO, 2010e) and estimated (discounted) remaining lifetime earnings for that age group.
Table 4.4 shows that five people aged between 40 and 64 years are estimated to have died due to VI and blindness in 2010. Of these, three would have been employed. Total lost lifetime earnings from the premature mortality of these people were estimated to total over €847,230 in 2010.
Table 4.4: Estimated cost of premature mortality from VI and blindness in the ROI, 2010
| Age group | No. of people with VI and blindness | No. of people who die due to VI and blindness | No. who would have been employed (a) | Assumed years to retirement (b) | Lost lifetime earnings per person € © | Total cost € |
|---|---|---|---|---|---|---|
| Male | ||||||
| 40-44 | 2,731 | 0.16 | 0.10 | 23.0 | 676,041 | 67,827 |
| 45-54 | 6,175 | 0.66 | 0.42 | 15.5 | 526,738 | 223,079 |
| 55-64 | 7,158 | 1.93 | 1.25 | 5.5 | 218,220 | 272,178 |
| Female | ||||||
| 40-44 | 4,785 | 0.14 | 0.08 | 23.0 | 445,051 | 35,425 |
| 45-54 | 9,744 | 0.69 | 0.39 | 15.5 | 336,176 | 130,797 |
| 55-65 | 9,504 | 1.60 | 0.90 | 5.5 | 130,572 | 117,923 |
| Total | 40,097 | 5.18 | 3.14 | n/a | n/a | €847,230 |
Source: Deloitte Access Economics calculations using ABS (2010), CSO (2008), CSO (2010e; 2010g; 2010h; 2010i), Department of Finance (2010) and McCarty et al (2001).
Abbreviations: n/a = not applicable. (a) Estimated by applying the gender-specific employment-to-population ratio in the ROI in 2010 (CSO, 2010e) to the number of people who die due to mild/moderate VI and blindness. (b) Estimated as the difference between the midpoint of the age group and 65 years. © Discounted to present day at a 4% discount rate (Department of Finance, 2010).
Estimated productivity losses account for the foregone future streams of income due to death are substantially lower than productivity losses due to disability, primarily because of low attributable mortality risk. Although the risk of death in the vision impaired is 2.34 times that of the general population, only 1.38% of the additional deaths is attributed to VI in the model. Furthermore, general mortality is low in the ROI, and the greatest additional risk of death due to VI and blindness is therefore experienced by those people aged over 65 years to whom productivity losses are not applied.
Productivity losses due to mortality were also estimated for the years 2015 and 2020, accounting for population growth (CSO, 2008) and increased subsequent deaths in the VI population (Table 4.3), and multiplying this by current annual earnings and gender-specific employment-to-population ratios (CSO, 2010e; 2010h). Estimates for 2015 and 2020 assumed constant wages and the same employment-to-population ratio as 2010.
The summary of productivity losses for each year is presented in Table 4.5. Productivity losses from premature mortality from VI and blindness of those aged between 40 and 64 years are estimated to rise to over €1.0 million by 2020.