Human Capital Approach

Human Capital Approach: Concept Architecture

What the approach values

The human capital approach values productivity lost because illness, disability, or premature death prevents people from undertaking activities that would otherwise produce goods or services. In health economics, it commonly uses earnings and time away from paid work as a practical proxy for forgone production. It is used in cost-of-illness studies and, when the chosen perspective permits productivity costs, alongside health and healthcare costs in economic evaluations.

The literature also calls this the human capital method. Those labels do not denote established rival techniques. The approach is the overall valuation idea; a particular application still has to specify which activities, people, time periods, and prices it counts. It is distinct from the broader theory of human capital as education, skills, and experience embodied in people, and from valuing a human life by its earnings. Lost earnings measure one possible component of economic production; they are not the value of a person's life or health.

Define the decision and the boundary of costs

Begin with the perspective required by the decision maker. A healthcare-payer perspective may exclude productivity costs, whereas a societal perspective can include them, subject to the jurisdiction's guidance. Identify the comparator, target population, observation period, price year, currency, and whether the question is the burden of an illness or the difference in productivity costs between interventions. For an intervention comparison, include changes caused by the alternatives on a consistent basis; a high total burden is not itself evidence that a treatment is cost-effective.

Lost production may arise from absence from paid work (absenteeism), reduced output while at work (presenteeism), reduced hours or early retirement, and premature mortality. The analyst can separately consider unpaid household production, care provided by family or friends, and informal work. These require defensible measures and valuations rather than assuming a recorded wage of zero means no production. State whose time is included—patients, caregivers, or both—and avoid counting the same hour under multiple categories.

Connect time lost to an economic value

For a simple paid-work absence, the human capital calculation multiplies attributable working time lost by a value per unit of productive time. If periods differ in wages, employment, survival, or time worked, calculate period-specific amounts and, where appropriate, discount future costs to a common date. A general representation is:

$$C_{\mathrm{HC}}=\sum_{t=0}^{T}\frac{\sum_{g} L_{gt}v_{gt}}{(1+r)^t}.$$

Here $L_{gt}$ is attributable productive time lost by group or activity $g$ in period $t$, $v_{gt}$ is the corresponding value per unit of time, $r$ is the discount rate appropriate to the analysis, and $T$ is its horizon. Units must agree: hours times currency per hour, for example. This is a framework, not a claim that all lost time has one wage or that an individual's actual pay equals the social value of output. Future mortality-related losses should reflect the assumed counterfactual survival, employment and hours, and stop at the stated horizon; do not automatically assign full-time earnings to every person until an arbitrary retirement age.

For paid work, studies use gross wages, sometimes adding employer-paid benefits or non-wage labour costs, as proxies for marginal production. Wages may depart from the value of forgone output owing to imperfect labour markets, substitution, team production, unpaid overtime, or changes in employment. Presenteeism needs an estimate of reduced effective output, not simply hours present while ill. Unpaid work can be valued using a replacement-service price or the opportunity cost of time, but those reflect different questions and should not be mixed silently.

A small intervention example

Suppose a trial finds that, on average, the comparator causes five attributable paid-work days of absence per person, while treatment causes two. If each missed day is valued at £160, the human capital estimates are £800 for the comparator and £320 for treatment, giving £480 less lost paid-work production per treated person. This is an illustrative one-period calculation: $(5-2)\times £160=£480$. It assumes equal working days and a suitable £160 daily valuation; it does not establish that the treatment saves £480 in employer payroll or public spending.

If treatment itself costs £600 more per person, adding this productivity component alone leaves £120 additional cost before considering any other costs, outcomes, uncertainty, or perspective restrictions. In a full economic evaluation, report healthcare and productivity components separately, then combine those permitted by the decision perspective. The health benefit is assessed in its own outcome measure, such as QALYs, rather than being replaced by the £480 figure.

Understand the counterfactual

The method asks what productive activity would have occurred without the illness or under the comparator. Collecting days absent is not enough when some absence would have occurred for unrelated reasons. Patient reports, employer records, administrative data, or validated questionnaires can support measurement, but each may miss informal work, short episodes, productivity while present, or work changes after death. Estimate employment probabilities and expected hours when projecting into future years, and make the assumptions explicit.

For long absences and premature mortality, a gross human capital estimate can count potential production over many years. It need not equal net production lost to society if another worker replaces the person or colleagues compensate. The friction cost approach usually focuses on losses during the period required to restore production, with context-specific replacement and training costs. Neither method is universally correct: the counterfactual labour market, perspective, and decision question matter. Show a friction-cost scenario where it could change the conclusion rather than presenting one estimate as a physical fact.

Interpret fairly and avoid double counting

Using actual or average wages mechanically can assign a lower monetary productivity value to people who are unemployed, retired, in unpaid care roles, or paid less because of structural inequality. That is a limitation of the wage-based valuation, not evidence that those people's health matters less. Report distributional consequences and unpaid work transparently; where relevant, test common wage rates or alternative values, while keeping the chosen perspective and costing rules clear.

Do not label transfers such as sickness benefits or disability payments as societal resource losses merely because money changes hands. Separate genuine administrative or other resource costs from transfers. Examine whether lost income or work impairment is already reflected in the preference-based health outcome used in a cost-utility analysis; include productivity costs under applicable guidance and discuss any overlap rather than assuming either that double counting is inevitable or that it is impossible. Caregiver time, forgone leisure, and household tasks likewise need an explicit inclusion rule.

Report a credible estimate

Provide the illness or intervention comparison, costing perspective, whose activities are counted, absenteeism and presenteeism instruments, counterfactual work and survival, treatment of unpaid labour and caregivers, wage source, employer-cost adjustments, horizon, discounting, price year, and uncertainty. Break out mortality and morbidity costs and paid and unpaid components. Test the wage proxy, employment assumptions, presenteeism measure, inclusion of future years, and the alternative friction-cost approach where relevant.

The resulting number is a modeled monetary value of specified forgone production under stated assumptions. It is useful for showing how illness affects economic activity and for informing a decision with a suitable perspective; it should never be read as a complete measure of welfare or human worth.

Sources and further reading

• Pike J, Grosse SD. Friction cost estimates of productivity costs in cost-of-illness studies in comparison with human capital estimates: A review. Applied Health Economics and Health Policy (2018). Explains gross human capital estimates, replacement, and differences from friction-cost estimates.
• Setiawan E, et al. A systematic review of methods for estimating productivity costs in health economic evaluations. PharmacoEconomics (2024). Documents measurement and valuation choices and challenges for informal and unpaid work.
• Jiang S, et al. Incorporating productivity loss in health economic evaluations. Discusses perspective, absenteeism, presenteeism, and the human capital and friction-cost approaches.
• Rissanen I, et al. Adjusted productivity costs of stroke by human capital and friction cost methods. European Journal of Health Economics (2021). Illustrates the importance of empirical adjustments and method choice.