Human Capital as a Driver of Growth

Why Human Capital Matters for Growth

A worker’s education, training, and experience can increase the amount or quality of output produced in a given period. These capabilities are called human capital. They matter for growth because they raise productive capacity: an economy can produce more with the same number of workers and the same stock of machines when workers have more effective skills.

Human capital is not the same as physical capital. Physical capital consists of produced inputs such as equipment, buildings, and infrastructure. Human capital is embodied in people: it cannot be separated from a worker and assigned to another worker in the same way that a machine can be reassigned. Both can contribute to production, but they accumulate through different processes. Investment in physical capital includes acquiring or maintaining equipment; investment in human capital includes education, training, and the time spent developing skills.

Definition: Human capital is the productive knowledge, skills, and capabilities embodied in people. Physical capital is the stock of produced assets used as inputs in production. Human-capital investment uses resources now, such as tuition, training time, or forgone work, in expectation of higher productive capacity later.

Human capital also differs from general knowledge. Knowledge may be available for many people to use, while a specific worker’s skills are embodied in that worker. The earlier material on nonrival knowledge and knowledge spillovers helps explain why the distinction matters: a person’s education can raise that person’s productivity and may also help others learn or work more effectively. The treatment here is aligned with the SOA VEE Economics (Validation by Educational Experience) syllabus.

Human Capital in a Production Function

A simple production function can represent workers with different skill levels. Let $Y$ be output, $A$ be total factor productivity, $K$ be physical capital, $L$ be the number of workers, and $h$ be human capital per worker. In this illustration, each worker supplies $h$ efficiency units of labor, so total effective labor is $hL$. Consider the Cobb-Douglas form below, where $0<\alpha<1$.

Y=A K^\alpha (hL)^{1-\alpha}

This is a model of production, not a claim that every economy has exactly this technology. It treats workers as having a common human-capital level $h$, holds other determinants of productivity in $A$, and assumes the displayed inputs can be combined according to this functional form. It is useful for isolating the contribution of skills. If physical capital, the number of workers, and $A$ are fixed, an increase in $h$ raises output. It does not imply that education is the only determinant of a worker’s productivity or that every additional unit of education has the same effect.

The exponent on effective labor also has a direct interpretation. Taking logarithms gives $\ln Y=\ln A+\alpha\ln K+(1-\alpha)(\ln h+\ln L)$. With $A$, $K$, and $L$ held constant, differentiating with respect to $\ln h$ gives $d\ln Y/d\ln h=1-\alpha$. Thus $1-\alpha$ is the elasticity of output with respect to human capital per worker in this model: a 1% increase in $h$ produces approximately a $(1-\alpha)$% increase in output for a small change. This is a model-specific result, not a universal estimate of education’s effect.

Worked Example: Skills and Output in a Cobb-Douglas Economy

Method: Hold physical capital and the number of workers fixed, then calculate output before and after the change in human capital. Assume $A=1$, $\alpha=0.5$, $K=64$, and $L=16$. Initially, $h=4$; after training, $h=5$. Treat these inputs as certain and constant within each calculation. There is no interest rate or mortality model in this static production comparison.

Initially, effective labor is $hL=4(16)=64$. Therefore,

Y_0=1(64)^{0.5}(64)^{0.5}=8(8)=64

After training, effective labor is $hL=5(16)=80$. Output becomes

Y_1=1(64)^{0.5}(80)^{0.5}=8\sqrt{80}\approx 71.5542

The increase is $71.5542-64=7.5542$ output units. As a percentage of initial output, it is $7.5542/64 \times 100\approx 11.8034\%$. Human capital per worker rose by $25\%$, but output rose by less because the labor elasticity is $1-\alpha=0.5$. The calculation checks against the ratio $Y_1/Y_0=(80/64)^{0.5}=\sqrt{1.25}\approx1.1180$, or about an 11.80% increase.

Common mistake avoided: Treating a 25% increase in $h$ as a 25% increase in output ignores the production function’s elasticity. Here the output change follows the exponent on effective labor.

Private Returns to Education and Training

A person’s private return to human-capital investment compares the costs that person bears with the benefits the person receives. A common benefit is a higher stream of after-tax earnings. Costs can include tuition and materials, the value of time spent studying, and income forgone while studying or training. Because costs may occur before benefits, the comparison should account for timing.

For a deterministic example with a cost paid at time 0 and an annual benefit received at the end of each year for $n$ years, the net present value (NPV) at effective annual discount rate $i$ is

\operatorname{NPV}=-C+\sum_{t=1}^{n}\frac{B_t}{(1+i)^t}

Here, $C$ is the time-0 cost and $B_t$ is the benefit at the end of year $t$. The formula follows because a dollar received at time $t$ is worth $1/(1+i)^t$ dollars at time 0 under the stated discount rate. When the annual benefit is constant at $B$, the sum is $B a_{\overline{n}|}$, where $a_{\overline{n}|}=\sum_{t=1}^{n}(1+i)^{-t}$. A positive NPV means discounted benefits exceed the included costs under the assumptions; it does not guarantee that every student will receive those benefits.

Worked Example: Evaluating a Training Program Privately

Method: Discount the earnings gains to the present and subtract the participant’s full opportunity cost. Assume a worker pays $9,000 for a training program at time 0 and forgoes $5,000 of earnings during training, also valued at time 0. After training, the worker receives an additional $5,000 in after-tax earnings at each year-end for four years. The gains are certain, employment continues throughout, and there is no mortality or tax change during the period. Use an effective annual discount rate $i=10\%$.

The total time-0 cost is $C=\$9,000+\$5,000=\$14,000$. The discount factors for years 1 through 4 are $1/1.1=0.909091$, $1/1.1^2=0.826446$, $1/1.1^3=0.751315$, and $1/1.1^4=0.683013$. The present values of the four earnings gains are therefore:

Year 1: $\$5,000(0.909091)=\$4,545.45$.
Year 2: $\$5,000(0.826446)=\$4,132.23$.
Year 3: $\$5,000(0.751315)=\$3,756.57$.
Year 4: $\$5,000(0.683013)=\$3,415.07$.

Adding the present values gives $\$4,545.45+\$4,132.23+\$3,756.57+\$3,415.07=\$15,849.32$. Subtracting the cost gives an NPV of $\$15,849.32-\$14,000=\$1,849.32$. The result is positive: under these assumptions, the discounted earnings gains exceed the included private costs by $1,849.32.

Common mistake avoided: Subtracting tuition but omitting the $5,000 of forgone earnings understates the investment’s cost. The calculation includes both direct cost and the value of time spent training.

Social Returns and Spillovers

A social return includes effects on others as well as effects received by the person who invests. If a more-skilled worker helps coworkers solve problems, shares useful methods, or improves the productivity of a team, some of the gains may not appear in that worker’s own wage. These additional effects are spillovers. The gap between a private and a social return depends on whether such external effects exist and how large they are.

A social calculation also needs care with costs. Tuition is a payment by the student and revenue to a provider; the payment itself is not necessarily a net resource cost to society. The resources used to provide instruction, along with the value of output forgone while someone studies, are social costs. For a simplified comparison, it is acceptable to use stated resource costs directly, but the analyst should not count a transfer payment as if it were automatically a resource consumed.

A private wage gain is not always equal to the full increase in the learner’s productive contribution. Wages may reflect bargaining, institutions, or the value employers place on credentials. A social-return calculation should therefore identify what each benefit represents and avoid counting the same productivity gain twice. The example below stipulates that the learner’s wage gain represents the learner’s own productivity benefit and that the separate spillover is additional output.

Worked Example: Comparing Private and Social NPV

Method: Calculate the learner’s private NPV, then add the separately identified spillover to obtain the social benefit. Assume a training investment uses $14,000 of resources at time 0: $7,500 for instruction and materials and $6,500 of output forgone while studying. The learner receives an additional $4,000 in wages at each year-end for four years. The training also generates an additional $1,200 of output for others at each year-end for those same four years. These amounts are certain, there is no mortality or employment interruption, and the effective annual discount rate is $i=10\%$.

The four-year annuity factor is

a_{\overline{4}|}=\frac{1-(1.10)^{-4}}{0.10}=3.169865

The present value of the learner’s wage gains is $\$4,000(3.169865)=\$12,679.46$. The learner’s private NPV is $\$12,679.46-\$14,000=-\$1,320.54$. On these assumptions, the learner alone would not recover the stated resource cost through the wage gains.

The spillover’s present value is $\$1,200(3.169865)=\$3,803.84$. Total social benefits are the learner’s wage benefit plus this additional output benefit: $\$12,679.46+\$3,803.84=\$16,483.30$. Social NPV is $\$16,483.30-\$14,000=\$2,483.30$. The social NPV exceeds the private NPV by exactly the spillover’s present value, $\$3,803.84$. The arithmetic checks: $-\$1,320.54+\$3,803.84=\$2,483.30$.

Common mistake avoided: The spillover is added only once because it is defined as output beyond the learner’s own productivity benefit. Counting it both within the wage gain and again as an external benefit would overstate social NPV.

Interpreting Returns Carefully

The calculations above isolate benefits and costs under simplified assumptions. In practice, returns can vary with the learner’s ability, field of study, program quality, labor demand, and the length of time skills remain useful. Skills can become less valuable when technology or job requirements change. A wage difference between more- and less-educated workers is not, by itself, proof that education caused the entire difference: people with different education levels may also differ in other ways.

For an actuarial or economic analysis, keep three distinctions clear:

Private versus social: private returns accrue to the learner; social returns include effects on other people and use social resource costs.
Costs versus transfers: a payment is a cost to the payer, but may be income to another party rather than a net use of societal resources.
Association versus causation: higher earnings among educated workers do not alone establish that education caused the full earnings difference.

Key takeaway: Human capital is a productive capability embodied in people. It can raise output by making labor more effective, while its private return depends on the learner’s discounted benefits and costs. When skills also benefit others, the social return can exceed the private return.

Check Your Understanding

For each problem, state the calculation or distinction that supports your answer.

In $Y=K^{0.5}(hL)^{0.5}$, let $K=64$, $L=16$, and $h=4$. If $h$ rises by 25% while the other inputs remain fixed, calculate output before and after the change and the percentage increase. (Answer: Initial output $=64$; final output $\approx71.55$; increase $\approx11.80\%$.)
A course costs $8,000 at time 0 and requires $12,000 of forgone earnings at time 0. It produces an additional $6,000 in after-tax earnings at each year-end for four years. Assuming certain gains, continuous employment, no mortality, and $i=5\%$ effective annually, calculate private NPV. (Answer: $\$1,275.70$.)
A skills program has a time-0 social resource cost of $18,000. It generates an additional $4,500 of annual wages for the learner and $900 of additional annual output for others, at each year-end for five years. Assuming certain benefits, no mortality, and $i=8\%$ effective annually, calculate private and social NPV. (Answer: Private NPV $\approx-\$32.80$; social NPV $\approx\$3,560.63$.)
An employer pays a worker for training, while the training also improves coworkers’ productivity. Identify one potential private benefit, one potential social spillover, and explain why the employer’s payment should not automatically be counted as a net social resource cost. (Answer: The worker may receive higher wages; coworkers may produce more output; the payment is a transfer to the training provider, while the resources used to provide training are the relevant social cost.)
Two groups have different average wages, and the group with more education earns more. Explain why this comparison alone does not establish the causal private return to education. (Answer: The groups may differ in other productivity-related characteristics, so the observed wage gap does not isolate the effect caused by education.)

What You'll Learn