Golden Rule of Capital Accumulation

In the Solow growth model, the economy converges to a steady state in which output, capital, and labour all grow at the same constant rate. But there is a wide range of possible steady states, each with a different capital-to-labour ratio and consumption per worker. The Golden Rule identifies the saving rate that achieves the steady state with the highest consumption per worker—the most efficient capital stock.

Why different steady states exist

In Solow’s framework, the capital-to-labour ratio (capital intensity) adjusts until investment equals depreciation plus population growth. But there is no single steady state—there are infinitely many, corresponding to different saving rates.

If households save 10% of income, capital accumulates until returns fall to 10%. If they save 30%, capital accumulates until returns fall to 30%. Both are valid steady states. The difference is that the high-saving economy has much more capital per worker, higher output per worker, but lower consumption per worker, because more income is devoted to replacing and expanding the capital stock.

The puzzle: is there a saving rate that is best? The answer is yes. The Golden Rule identifies it.

The consumption maximisation problem

Consumption per worker in steady state equals output per worker minus investment per worker. Output per worker depends on capital intensity, via the production function. Investment per worker must cover depreciation of existing capital plus additions needed for population growth.

Formally, if capital intensity is k and the depreciation rate is δ, then investment per worker is (n + δ)k, where n is the population growth rate. Consumption per worker is f(k) − (n + δ)k, where f(k) is output per worker.

To find the optimal k, take the derivative with respect to k and set it to zero:

df/dk − (n + δ) = 0

Or equivalently, df/dk = n + δ.

This condition says: the marginal product of capital (the output gained by adding one more unit of capital) should equal the rate at which capital is depleted through depreciation and population growth. This is the Golden Rule.

Intuition: balancing investment and consumption

Why does this make sense? If the marginal product of capital exceeds n + δ, adding more capital yields more output than is needed to maintain the capital stock. The gains could be consumed—so the current saving rate is too low.

Conversely, if the marginal product falls short of n + δ, extra capital adds less output than is needed to maintain it. Saving at the current rate is wasteful; you could cut saving, boost current consumption, and still maintain the capital stock. So the saving rate is too high.

At the Golden Rule, the marginal product exactly equals the break-even rate. No reallocation can raise consumption per worker. It is the socially optimal steady state.

How does the saving rate adjust the economy there?

The model does not tell us that economies automatically reach the Golden Rule. In fact, most do not. The Solow framework is silent on what determines the saving rate—it is typically treated as an exogenous preference.

An economy can undershoot the Golden Rule: if households are impatient and save too little, capital accumulation is sluggish, the capital intensity is low, and output per worker is depressed. Higher saving would lift future consumption, but current savers do not benefit—a classic coordination failure.

An economy can also overshoot: if households are forced to save heavily (through government mandates, corporate retention, or deeply ingrained thrift), capital accumulates excessively, returns collapse, and most income goes to maintaining capital rather than funding consumption. This was the narrative around the Soviet Union in the 1970s—high investment, but low consumption and declining returns on marginal units of capital.

In reality, the saving rate is determined by a mix of time preferences, income expectations, interest rates, and policy. If a country’s saving rate puts it below the Golden Rule, growth in output per capita and living standards is slower than it could be. If it overshoots, the population is essentially accumulating capital that yields little and could be better used for current wellbeing.

The Golden Rule and the interest rate

Another way to state the condition: in the Golden Rule steady state, the real interest rate equals the population growth rate plus the depreciation rate.

This is a clean result. It says that the return on capital should exactly equal the rate at which the economy is expanding (through population growth) and the physical loss of capital (depreciation). Any higher, and capital is scarce relative to labour—more saving should occur. Any lower, and capital is abundant—saving should fall.

This also connects to intergenerational fairness. If generations are equally weighted in a social welfare function, no generation should consume less than the previous one. The Golden Rule achieves constant consumption per capita across time, making it a natural benchmark for equity.

Does the real world obey the Golden Rule?

Empirically, the answer is mixed. Rich countries with low population growth (below 2% per year) should have a Golden Rule return on capital around 2–4%, depending on depreciation. Measured returns on capital are typically 5–8%, suggesting these economies are below the Golden Rule—they could save more and reach a higher steady-state consumption level.

Conversely, some very poor countries with high population growth and low capital stocks show high marginal returns to capital (15%+). This suggests they are also below the Golden Rule, but in a different direction: their capital scarcity is extreme.

The puzzle is that most economists do not advocate for a massive increase in saving to reach the Golden Rule, especially in rich countries. Why? Because the Golden Rule is a steady-state benchmark. Getting there requires a transition period of high saving and low consumption. Depending on discount rates and time horizons, such a transition may not be welfare-improving for current generations, even if it lifts steady-state consumption.

This is why endogenous growth theory and newer models relax Solow’s assumptions: they ask whether dynamic paths toward a higher growth rate (not just a higher steady state) might be preferable, even if they involve near-term sacrifice.

See also