All data used in this study were collected from a work environmental survey in the main plant of the Swedish company Sandvik Materials Technology (SMT) during May 2015. The collected dataset included data about work capacity, work performance, and work quality in addition to data on lost working hours due to sickness absenteeism and presenteeism. The employer's efforts to improve the work environment are described in Table 1 in terms of psychosocial, ergonomic, and physical factors.

A total of 49 workers participated anonymously in the study. Most of them were involved with manufacturing and were sufficient qualified for and experienced in their tasks. Some of the collected data have been used previously in other studies [1, 2, 17]. The questionnaire, which was distributed to the workers by the SMT safety unit during April 2015, was based on the Harvard questionnaire (http://www.hcp.med.harvard.edu/hpq/ftpdir/absenteeism%20presenteeism%20scoring%20050107.pdf), (http://www.hcp.med.harvard.edu/hpq/ftpdir/survey_clinical_7day.pdf), knowledge of Swedish workplaces and the Swedish labour market, and findings from the literature. As the employer's multifactorial efforts to improve the work environment (EE) was difficult to assess directly, workers' satisfaction with the employer's efforts was used as an indicator of the level of these efforts in SMT. In the survey, the workers were asked to evaluate their physical, ergonomic, and psychological work environment. The workers were also asked to answer how they perceived their employer's multifactorial efforts to improve the work environment. The answers would be in percent. To avoid missing and confusing answers, the participants answered the questions while the supervisor was in attendance. Then, the employer's efforts were identical to the workers' satisfaction with these efforts. For instance, if the workers were 70% satisfied with their employer's efforts to improve the ergonomic work conditions, the efforts were assessed to be 70% of their potential level.

The other basic concepts were:

To assess the workforce availability, the workers were asked about their actual working hours compared to the scheduled working hours; to assess the work performance (i.e. the actual output divided to the standard output), the workers were asked about the achieved quantitative requirements of work in percent; and to assess the work quality (i.e. the accepted output divide to the total output), the workers were asked about the achieved qualitative requirements of work in percent. Both the company's standard and the workers' own judgements were considered in the assessments of the performance and the quality of work.

Five regression equations were used to analyse the socio-economic consequences of labour ineffectiveness and the employer's efforts to improve the work environment in SMT from May 2014 to April 2015. The socio-economic consequences were assessed in terms of important econometric issues. First, two regression models were used to assess the economic cost of labour ineffectiveness, which is measured as 1 – OLE, and the economic benefit of the employer's efforts to improve the work environment. Then, three regression models were employed to assess the marginal effects of the employer's efforts on labour effectiveness, informal impairments, and illness-related lost working hours.

The economic cost of labour ineffectiveness (LIE) in terms of labour productivity loss (LPL) was assessed by:

where LPL is expressed in terms of money, LIE is measured in percent, the intercept α₁ is interpreted as the level of LPL independent of LIE, and β₁ is the marginal cost of LIE (i.e. the cost of an additional percentage unit of LIE).

The economic benefit of the employer's efforts to improve the work environment (EE) in terms of saving/decreasing labour productivity loss (LPL) was assessed by:

where the intercept α₂ is interpreted as the level of LPL independent of EE, and β₂ is the marginal benefits of EE in terms of reducing LPL (i.e. the economic benefit of increasing EE by one precentage unit).

The impact of the employer's efforts to improve the work environment (EE) on overall labour effectiveness (OLE) was assessed by:

where both OLE and EE are measured in percentage units, the intercept α₃ is interpreted as the level of OLE independent of EE, and β₃ is the marginal effect of EE on OLE (also known as the OLE elasticity of EE).

The impact of the employer's efforts to improve the work environment (EE) on informal work impairments (IWI); that is, reduced work capacity among workers which is not reflected in formal terms such as part-time working, was assessed by:

where both IWI and EE are measured in percentage units, the intercept α₄ is interpreted as the level of IWI independent of EE, and β₄ is the marginal effect of EE on IWI (and known as the IWI elasticity of EE).

The impact of the employer's efforts to improve the work environment (EE) on the working hours lost by ill workers (LWH) was assessed by:

where the intercept α₅ is interpreted as the amount of LWH independent on the level of EE, and β₅ is the marginal effect of EE on LWH (i.e. the responsiveness of LWH to changes in EE by one percentage unit).

Ordinary least squares (OLS) regression analysis was used to assess the important marginal values.

The marginal cost of labour ineffectiveness in terms of labour productivity loss was estimated to be 6647 SEK on average for each SMT worker during 2014–2015. This means that each additional percentage unit of ineffectiveness embodied in a worker (due to individual characteristics, insufficient skills, and non-optimized environmental factors at the workplace) cost SMT 6647 SEK on average during that period. The autonomous cost of 50,665 SEK is the estimated cost not caused by workforce ineffectiveness. As an overall evaluation of the regression model, 58% of the variation in labour productivity loss among SMT workers during the study period can be explained by fluctuations in the labour ineffectiveness.

The marginal benefit (economic advantage) of the employer's efforts in terms of decreasing illness-related labour productivity loss was estimated as 6241 SEK, meaning that SMT saved an average of 6241 SEK of productivity loss per worker during 2014–2015 for each additional percentage unit of its efforts to improve the work environment. The autonomous productivity loss of 637,895 SEK is the estimated productivity loss that cannot be saved by increasing the efforts. According to the regression model, 75% of the variation in labour productivity loss among the workers can be explained by fluctuations in the work environmental efforts.

The marginal effect of the employer's efforts on overall labour effectiveness was 1.09, meaning that each additional percentage unit of the employer's efforts increased the labour effectiveness by more than one percent. The employer's efforts in SMT are thus characterized by a (weak) increasing return to labour effectiveness, which confirms the economic advantage in expansion of the efforts. The autonomous labour effectiveness not explained by the efforts was not statistically significant. However, 72% of the variation in labour effectiveness among the workers of SMT during 2014–2015 can be explained by fluctuations in the work environmental efforts.

The marginal effect of the employer's work environmental efforts on workers' informal impairments in SMT during 2014–2015 was − 0.69, meaning that the informal impairments decreased by 0.69% for each additional percentage unit of the efforts. The autonomous impairments, which cannot be explained by these efforts, were 67.63% of the total. The regression analysis showed that 73% of the variation in labour informal work impairments at SMT during 2014–2015 can be explained by fluctuations in the work environmental efforts.

The marginal effect of the employer's efforts to improve the work environment on the workers' illness-related lost working hours was − 14.22, meaning that during 2014–2015 SMT could decrease workers' illness-related LWH by 14.22 man-hours for each additional percentage unit of their efforts to improve the work environment. The estimated number of 1460 h reflects external factors affecting the amount of illness-related LWH, since 35% of the variation in workers' LWH cannot be explained by fluctuations in EE.

The empirical results are summarized in Table 2, for an entire economic evaluation of the employer's efforts to improve the work environment.

The challenge in the management of manufacturing organizations, in many countries, is to improve OLE as a measure for the cumulative effect of availability, work performance, and work quality [30, 31]. The managers are usually focused on factors as machinery, attendance and quality control, and professional knowledge and task training. However, previous research has definitely shown that there are other opportunities to optimize OLE; these consist of efforts aimed at optimizing environmental factors that affect the mental, social, and physical well-being of the workforce: workers' attitudes and behaviours have a significant effect on their work performance and the production process [32, 33]; and lack of motivation affects the labour effectiveness not only directly, but also indirectly through the workers' knowledge and skills [30]. Production rate and labour effectiveness are also affected by health and safety of the labour force, ergonomic work conditions [9, 10], ‘extrinsic rewards' [29], and environmental factors such as noise, lighting, humidity, and temperature [7, 20]. The empirical results of this study have strengthen the early findings and assessed the effects of environmental factors at the workplace on OLE by using regression technique and economic theories; factors concerning the psychosocial, ergonomic, and physical dimensions of the work environment. Occupational health and job satisfaction are the most important sources of labour effectiveness, and can also strengthen the effectiveness of the traditional measures to increase availability, work performance, and product quality (i.e. efforts to improve knowledge, skills, and control systems for workers' attendance and work quality). Workers' health, well-being, and satisfaction can be improved by an investment in changing the psychosocial, ergonomic, and physical environments at the workplace; the cost of such an investment will be much less than its opportunity cost in terms of the costs of work-related disorders [1]. Thus, the employer's efforts to improve the work environment and occupational health cannot be less effective and profitable than the abovementioned measures applied in the manufacturing industries, when the huge socio-economic costs of the work environmental risk factors for illness and impairments consist substantially of labour productivity losses [1, 3–6].

Both the marginal economic cost of labour ineffectiveness (6647 SEK) and the marginal economic benefit of employer's efforts to improve the work environment (6241 SEK) were assessed in terms of the labour productivity loss (LPL). LPL was estimated on the basis of workers' sickness absenteeism and presenteeism. A comparison analysis of the estimated marginal values showed that the labour ineffectiveness was substantially caused by the environmental risk factors at the SMT workplace, and that the effects of individual characteristics and insufficient qualification/experience were negligible. A major part of the company's illness-related economic costs during the study period was due to the labour ineffectiveness caused by insufficiencies in the work environment. The association of OLE and EE in the regression model (3) explains why the difference between the marginal economic cost of labour ineffectiveness and the marginal benefit of the work environmental efforts was not considerable.

In development of the models introduced here it was assumed that the business relationships and technical performance remained unchanged and were not subject to any unexpected anomalies and defects affecting labour productivity and effectiveness. This means assuming that 1) machinery was well maintained and in good condition, 2) raw and intermediate materials were supplied and delivered in good time and of the desired quality, 3) the relevant markets were in equilibrium, meaning that the related supply and demand curves remained unchanged, and 4) the level of workers' qualification for and experience in their tasks remained constant during 2014–2015.

As usual in the econometrics literature, it must be emphasized that the estimated autonomous effects (effects not determined in the models) in these econometric analyses should not be used in decision-making regarding the work environment. In principle, they are either statistically non-significant or unexplainable and unrealistic. The estimated constant cost not explained by labour ineffectiveness (50,665 SEK) could be caused by sickness absenteeism, but the value has a weak significance level; the autonomous productivity loss not saved by work environmental efforts (637,895 SEK) is significant and caused by external factors, but is unrealistically high; the autonomous labour effectiveness with negative sign (− 10.69%) is non-significant, unexplainable, and unrealistic; the worker's autonomous informal impairments (67.63%) show more realistically the combined effects of individual characteristics, societal factors, and the environmental factors at the workplace omitted in this study; and finally, the autonomous lost working hours (1460 h) are unexplained since they are much higher than the average yearly working hours lost by a worker (465 h).

The employer's effort to improve the work environment was found to have an increasing but weak return to labour effectiveness, and a decreasing return to the workers' informal impairments. Although the differential effects need to be addressed in future research, the results of this study and other studies based on the same dataset [1, 2] may allow the formation of some hypotheses around this; hypotheses to be tested in future research. The first is about the effect of time. The work environmental efforts have short-term effects on labour effectiveness, while their effects on work impairments are seen in the medium and long run. Thus, a long-run study of the efforts may smooth out the differential effects. Further, the employer's efforts to improve the work environment have an impact not only on the quantity of work but also on the quality of work; this latter is reflected in the overall labour effectiveness but not in the work impairments. Moreover, a major part of the labour effectiveness can be explained by work environmental factors, compared with the illness-related informal work impairments.

The marginal effect of the employer's efforts to improve the work environment on working hours lost by ill workers in SMT during 2014–2015 (14.22 h) indicates a strong potential ability of the efforts to reduce illness-related labour productivity loss in the company during this period. Each SMT worker lost 465 working hours on average during the study period due to sickness absenteeism and informal work impairments [1, 2]. This potential ability is verified by the high marginal economic benefit of the efforts (near the marginal economic cost of labour ineffectiveness) and their increasing return to overall labour effectiveness.