Elsevier

Social Science & Medicine

Volume 75, Issue 11, December 2012, Pages 1981-1988
Social Science & Medicine

Productivity cost calculations in health economic evaluations: Correcting for compensation mechanisms and multiplier effects

https://doi.org/10.1016/j.socscimed.2012.07.012Get rights and content

Abstract

Productivity costs related to paid work are commonly calculated in economic evaluations of health technologies by multiplying the relevant number of work days lost with a wage rate estimate. It has been argued that actual productivity costs may either be lower or higher than current estimates due to compensation mechanisms and/or multiplier effects (related to team dependency and problems with finding good substitutes in cases of absenteeism). Empirical evidence on such mechanisms and their impact on productivity costs is scarce, however.

This study aims to increase knowledge on how diminished productivity is compensated within firms. Moreover, it aims to explore how compensation and multiplier effects potentially affect productivity cost estimates. Absenteeism and compensation mechanisms were measured in a randomized trial among Dutch citizens examining the cost-effectiveness of reimbursement for smoking cessation treatment. Multiplier effects were extracted from published literature. Productivity costs were calculated applying the Friction Cost Approach. Regular estimates were subsequently adjusted for (i) compensation during regular working hours, (ii) job dependent multipliers and (iii) both compensation and multiplier effects. A total of 187 respondents included in the trial were useful for inclusion in this study, based on being in paid employment, having experienced absenteeism in the preceding six months and completing the questionnaire on absenteeism and compensation mechanisms. Over half of these respondents stated that their absenteeism was compensated during normal working hours by themselves or colleagues. Only counting productivity costs not compensated in regular working hours reduced the traditional estimate by 57%. Correcting for multiplier effects increased regular estimates by a quarter. Combining both impacts decreased traditional estimates by 29%. To conclude, large amounts of lost production are compensated in normal hours. Productivity costs estimates are strongly influenced by adjustment for compensation mechanisms and multiplier effects. The validity of such adjustments needs further examination, however.

Highlights

► Most lost work due to absenteeism in the Netherlands is compensated in normal working hours. ► If assumed compensation in normal hours is costless common productivity cost estimates decrease by more than half. ► Oppositely, including effects on team output (multiplier effects) increases regular estimates by more than a quarter. ► Including both compensation mechanisms and multiplier effects decreases cost estimates with almost one-third. ► This is the first study which has simultaneously explored multiplier effects and compensation effects.

Introduction

Economic evaluations are increasingly used to inform decision-makers on cost-effectiveness of health interventions. Such economic evaluations can be performed from different perspectives. Most national health economic guidelines prescribe either a health care perspective or a societal perspective (Knies, Severens, Ament, & Evers, 2010). In economic evaluations performed from a health care perspective only costs falling on the health care budget are included. In analyses from the societal perspective all relevant costs and effects are included regardless of who bears the costs and who receives the benefits (Drummond, Sculpher, Torrance, O'Brien & Stoddart, 2005). Although no consensus seems to exist on which perspective to use for decisions regarding reimbursement of health interventions, there are strong arguments to perform economic evaluations from a societal perspective (Brouwer, Exel, Baltussen, & Rutten, 2006; Drummond et al., 2005; Gold, Siegel, Russell, & Weinstein, 1996; Johannesson, 1995; Jonsson, 2009). The most compelling argument is that costs falling outside the health care budget are equally real as costs that fall inside this budget and thus affect overall welfare of a society. Subsequently, ignoring these costs could lead to suboptimal decisions for the society as a whole, especially if these costs strongly affect cost-effectiveness outcomes, which is often the case for productivity costs (Koopmanschap & Rutten, 1994; Krol, Papenburg, Koopmanschap, & Brouwer, 2011). Productivity costs can be seen as “Costs associated with production loss and replacement costs due to illness, disability and death of productive persons, both paid and unpaid” (Brouwer, Koopmanschap, & Rutten, 1997). In this paper we focus on productivity costs in relation to paid work.

It is important to note that the inclusion of productivity costs in economic evaluations is not without dispute. This relates to two broad issues: (i) which perspective to adopt in economic evaluations and (ii) the equity implications of including productivity costs. Regarding the first issue, it is clear that not all (national) guidelines for economic evaluations recommend adopting a societal perspective. Many prescribe a narrower perspective, i.e. a health care perspective. Then, only costs falling on the health care budget, and thus are the responsibility of the health care policy maker economic evaluations seek to advice, are deemed relevant (Brouwer et al., 2006; Claxton, Paulden, Gravelle, Brouwer, & Culyer, 2011). Second, inclusion of productivity costs (like many other cost categories, including medical costs) can have equity implications. Their inclusion may sometimes lead to favoring interventions targeted at the working age population (Olsen & Richardson, 1999; Williams, 1992), although it may also work in the opposite direction (Krol et al., 2011). The equity implications may be particularly pronounced when only costs of paid labor are included while costs of unpaid labor are ignored. To provide policy makers with information on the distributional consequences of including productivity costs, it has been recommended that final cost effectiveness ratios may be best presented both with and without productivity costs (Brouwer et al., 2006; Pritchard & Sculpher, 2000). This paper will not further address the issue of whether to include productivity costs. It rather focuses on the question how to do so. This is by no means a trivial question.

Productivity costs related to paid work can reflect over half of total costs (Krol et al., 2011; Tranmer, Guerriere, Ungar, & Coyte, 2005) and because of their substantial potential impact on cost-effectiveness outcomes estimates should be accurate. Most commonly, productivity costs related to paid work are estimated through measuring absenteeism and/or diminished productivity while at work. Productivity loss is estimated by calculating the number of days individuals are absent from work due to illness and treatment and/or the amount of diminished productivity at work. These estimates are then multiplied with a relevant value of production per time unit (e.g. hour), commonly some wage rate estimate. Such estimations implicitly assume a direct linear relation between diminished input (through absenteeism and diminished productivity while being at work) and diminished output (lost production). However, a linear translation from diminished productive input to diminished economic output may not be accurate.

It has been suggested that diminished productivity may partly be compensated during normal working hours by the ill worker itself or its colleagues. Koopmanschap, Rutten, van Ineveld, and van Roijen (1995) describe that sometimes absenteeism may not change production levels and lead to extra costs. “This situation may occur if work can be made up for by the sick employee on his return to work or if an internal labor reserve exists, allowing work to be taken over by colleagues without extra costs.” However, they note that “the existence of permanent internal labor reserves raises labor costs, which may have medium term macroeconomic implications”. The existence of such compensation mechanisms was explored further by Severens, Laheij, Jansen, Van der Lisdonk, and Verbeek (1998) and Jacob-Tacken, Koopmanschap, Meerding, and Severens (2005). It seemed that missed work was indeed partly compensated during normal working hours. Moreover, compensation mechanisms seemed related to occupational and job characteristics. Severens et al. (1998) and Jacob-Tacken et al. (2005) argued that, given these compensation mechanisms, traditional productivity cost calculations might overestimate true productivity losses.

Others (Pauly et al., 2002), on the contrary, suggested that common productivity estimates are, in fact, underestimations of ‘true’ productivity costs. They concluded that “when there is a team production and substantial team-specific human capital, the value of lost output to the firm from an absence will exceed the wage per day of the absent worker.” In other words, a person's absenteeism or diminished productivity at work can negatively affect the output of co-workers in cases of team dependencies. Moreover, substitution of ill workers can be difficult and imperfect. To incorporate effects of substitution difficulties and effects on team output it is suggested to multiply the individual's wage with a job-dependent multiplier (Nicholson et al., 2006; Pauly, Nicholson, Polsky, Berger, & Sharda, 2008). The multiplier is determined by considering the level of team dependency in realizing the teams' output, the time sensitivity of the output and the ease of substitution of the ill employee.

Traditional productivity cost estimates can be adjusted for such ‘multiplier effects’ and ‘compensation during normal working hours’, if this is deemed desirable. To date, however, only a limited amount of empirical research is available regarding the existence of multiplier effects or compensation mechanisms (Jacob-Tacken et al., 2005; Nicholson et al., 2006; Pauly et al., 2008; Severens et al., 1998), and there is no empirical study investigating both concepts integrally. Recently, a new productivity cost questionnaire was developed which is the first to simultaneously include questions on compensation mechanisms and the effects on co-workers (Zhang, Bansback, Boonen, Severens, & Anis, 2012). The validity of this questionnaire needs to be further investigated, however.

In this paper we present an empirical study on compensation mechanisms of reduced productivity among a large group of smokers in a smoking-cessation program. Moreover, based on this information and estimated effects of ill-health on co-workers, we calculated productivity costs for several scenarios. To our knowledge, this is the first study to simultaneously explore multiplier effects and compensation effects. The aim of this study is twofold. First, it aims to increase knowledge on how diminished productivity is actually compensated within firms and to further investigate the association between compensation mechanisms and occupational and job characteristics. Second, the study aims to explore how compensation and multiplier effects potentially affect productivity costs. In doing so, this study seeks to contribute to the discussion on how to increase accuracy of productivity cost estimates. This paper starts with presenting the methods of our empirical study on compensation mechanisms and describing the productivity costs scenarios. Subsequently, the results are presented and then discussed. Finally, some suggestions are made regarding future research and the estimation of productivity costs.

Section snippets

Study population

In order to meet our first aim to increase knowledge on compensation of diminished productivity within firms, absenteeism and compensation mechanisms were measured in 2002 in a large randomized trial conducted among smokers living in the North of the Netherlands examining the cost-effectiveness of including smoking cessation treatments in the basic benefit package. Participants in the smoking cessation study had to be daily smokers, aged over 18 years. A detailed description of the randomized

Results

In total, 1266 persons were included in the smoking cessation trial (see Kaper et al., 2005). The response on the absenteeism and compensation questionnaire was 66% (n = 836) of the initial response. In total, 501 respondents were in paid employment. Of these, 187 respondents had experienced absenteeism in the past 6 months and, as a result, completed the questions on compensation mechanisms. Baseline characteristics of these 187 respondents are shown in Table 1.

Discussion

The first aim of our study was to increase knowledge on how diminished productivity is compensated within firms. Over half of the respondents in the smoking cessation trial indicated that their absenteeism was compensated in normal working hours by either themselves or their colleagues. These findings are in line with the results from previous research by Severens et al. (1998) and Jacob-Tacken et al. (2005). The way lost work was compensated was associated with work type characteristics, such

Acknowledgments

The authors have no conflicts of interest to declare.

References (29)

  • M.F. Drummond et al.

    Methods for the economic evaluation of health care programmes (3rd ed.)

    Oxford/New York: Oxford Medical Publications/Oxford University Press

    (2005)
  • M. Gold et al.

    Cost-effectiveness in health and medicine

    (1996)
  • L. Hakkaart-van Rooijen et al.

    Dutch costing manual

    (2010)
  • K.H. Jacob-Tacken et al.

    Correcting for compensating mechanisms related to productivity costs in economic evaluations of health care programmes

    Health Economics

    (2005)
  • Cited by (42)

    • Advances in the methodological approach to friction period estimation: A European perspective

      2020, Social Science and Medicine
      Citation Excerpt :

      In addition, some authors have argued that the true productivity costs of work absence is greater than the wage rate due to the impact on team-based time sensitive production where a perfect substitute worker is not readily available (Nicholson et al., 2006; Strömberg et al., 2017), however European-wide estimates for these multipliers are lacking. Other authors have noted the existence of compensation effects whereby diminished productivity due to absenteeism can be made up for either by the sick employee upon return to work or by colleagues using an internal labour reserve (Krol et al., 2012). These compensation mechanisms appear to be related to occupational and job characteristics (Krol et al., 2012) but similar to the team-based multipliers, extensive country-specific European-wide estimates of these are lacking.

    • A history that goes hand in hand: Reflections on the development of health economics and the role played by Social Science & Medicine, 1967–2017

      2018, Social Science and Medicine
      Citation Excerpt :

      Work published in the journal has, for example, gone beyond the accepted approach to economic evaluation, embracing work on alternatives to the standard (for health economists) health-focused normative extra-welfarist approach to evaluation, including methods rooted both in welfarism (Birch and Donaldson, 2003; Lee et al., 2013; Mooney and Lange, 1993; Olsen and Donaldson, 1998) and capability (Anand and Dolan, 2005; Coast et al., 2008a; Greco et al., 2015; Grewal et al., 2006; Lorgelly et al., 2015; Mitchell et al., 2015; Simon et al., 2013). Research has focused on going beyond the patient in terms of outcomes (Al-Janabi et al., 2008; van den Berg et al., 2005) and beyond the health care system in terms of costs (Koopmanschap and van Ineveld, 1992; Krol et al., 2012). In terms of understanding the determinants of health, research has gone beyond the health care system (Amin et al., 2015; French et al., 2012) and there has been an extensive focus on equity issues (Ngalesoni et al., 2016; Wagstaff et al., 1991), going beyond the usual economic focus on efficiency.

    • The Estimation and Inclusion of Presenteeism Costs in Applied Economic Evaluation: A Systematic Review

      2017, Value in Health
      Citation Excerpt :

      Multiplier effects are additional costs that could result from the negative impact of sick coworkers on productivity particularly when team work is involved [59]. Compensation mechanisms are adjustments for productivity loss through internal employee substitution mechanisms or as a result of ill employees compensating for lost time [60]. These have been reported to have a significant impact on overall productivity costs [61].

    View all citing articles on Scopus
    View full text