The Marginal Rate of Technical Substitution Explained

Every producer juggles two questions constantly: how much labour to hire, and how much capital to invest in machines. The marginal rate of technical substitution, or MRTS, is the concept that tells a firm exactly how these two choices trade off against each other without disturbing the output level. Once you understand MRTS, isoquants, cost minimisation, and a good chunk of production theory start making a lot more sense.

What is the marginal rate of technical substitution? 🔗

MRTS measures how much of one input a firm can give up in exchange for one more unit of another input, while keeping total output exactly the same. It is usually discussed in the context of labour (L) and capital (K), the two most common inputs in production theory. If a firm is willing to give up 3 units of capital to hire 1 more unit of labour, and output stays unchanged, the MRTS of labour for capital is 3.

This is not a random trade. It reflects how productive each input actually is at the current combination. MRTS is closely related to the idea of an isoquant, which is a curve showing every combination of labour and capital that produces the same quantity of output. In fact, MRTS is simply the slope of the isoquant at any given point, taken as a positive value since the curve itself slopes downward.

The formula: connecting MRTS to marginal products 🔗

The formal definition ties MRTS directly to the marginal products of the two inputs involved:

MRTS_LK = −ΔK/ΔL = MP_L / MP_K

Here, MP_L is the marginal product of labour (the extra output from one more unit of labour) and MP_K is the marginal product of capital. The logic behind this formula is fairly intuitive: if a firm reduces labour slightly, output falls by roughly MP_L multiplied by the size of that reduction. To keep output constant, the firm must add capital until output rises back by the same amount, which requires adding output divided by MP_K units of capital. Dividing one by the other gives the MRTS.

A quick numerical illustration 🔗

Suppose a small garment manufacturing unit can produce 100 shirts a day using different combinations of tailors (labour) and stitching machines (capital):

Notice how each additional tailor allows the unit to give up fewer and fewer machines while still producing 100 shirts a day. That falling ratio is exactly what the law of diminishing MRTS describes.

Why MRTS diminishes 🔗

As a firm keeps substituting labour for capital along an isoquant, MRTS does not stay constant, it keeps falling. This happens because of diminishing marginal productivity. As the firm hires more tailors and uses fewer machines, each additional tailor contributes less to output because there are fewer machines to work with, so MP_L declines. At the same time, each remaining machine becomes more heavily used per tailor, so MP_K rises. Since MRTS equals MP_L divided by MP_K, a falling numerator and a rising denominator together push MRTS down.

This decline reflects a simple production reality: labour and capital are rarely perfect substitutes. Beyond a certain point, adding more of one input without a matching supply of the other yields shrinking benefits.

MRTS and the convex shape of isoquants 🔗

The diminishing nature of MRTS is precisely why isoquants curve inward toward the origin instead of being straight lines. Since MRTS is the slope of the isoquant, a falling MRTS means the curve gets flatter as you move rightward along it, which produces the familiar bowed-in, convex shape economists refer to constantly in production theory. A well-behaved isoquant is always convex to the origin, downward sloping, and never touches either axis, because a firm needs at least some of both inputs to produce anything.

Two special cases worth knowing 🔗

Not every isoquant is smoothly convex. Two extreme cases show up often in textbooks:

• Perfect substitutes: The isoquant is a straight line and MRTS stays constant throughout, because one input can fully replace the other at a fixed rate.
• Perfect complements: The isoquant is L-shaped, and inputs must be used in a fixed ratio, so there is effectively no substitution possible.

Most real production processes, from garment units to IT services, fall somewhere between these extremes, which is why the standard convex isoquant with diminishing MRTS is the more realistic model.

How MRTS guides a firm’s input decisions 🔗

Knowing the rate at which inputs can be substituted is only half the story. A firm also cares about cost. This is where the isocost line comes in, representing every combination of labour and capital that costs the same total amount, given wage rate and rental price of capital. A firm minimises its cost for a given output level at the point where the isoquant is tangent to the isocost line. At that tangency, the following condition holds:

MRTS_LK = w/r

where w is the wage rate and r is the rental price (or cost) of capital. In plain terms, the rate at which a firm can technically substitute labour for capital should equal the rate at which the market lets it substitute the two, based on their relative prices. If MRTS is higher than the wage-rental ratio, the firm is using too much capital relative to labour and can lower costs by hiring more workers and buying fewer machines, and vice versa.

MRTS in everyday business decisions 🔗

This is not just a diagram in a textbook. Indian firms make these calls constantly. A textile unit in a region with low wages and expensive credit will likely operate at a high MRTS, relying heavily on labour before automating. A software company, by contrast, faces a very different input mix, where skilled programmers and computing infrastructure substitute for each other quite differently than tailors and machines. A firm employing very little of one input can often replace a unit of the scarce input for a large amount of the abundant one, and this saving shrinks as the input mix becomes more balanced.

The same logic explains why, as wages rise relative to the cost of capital, firms tend to shift toward more machine-intensive production. Understanding MRTS helps explain real shifts in how Indian manufacturing and services increasingly automate certain tasks while keeping others labour-intensive.

Common mistakes students make with MRTS 🔗

A few points often trip up students working through this topic for the first time:

• Confusing MRTS with MRS: Marginal rate of substitution (MRS) applies to consumer indifference curves between two goods; MRTS applies to producer isoquants between two inputs. They look similar mathematically but describe entirely different decisions.
• Assuming MRTS is always constant: This is only true for perfect substitutes. In the standard case, MRTS falls as one input is used more heavily.
• Forgetting the negative sign convention: MRTS is technically the negative of the isoquant’s slope, but is reported as a positive number by convention, since the isoquant itself always slopes downward.

Bringing it all together 🔗

MRTS is essentially the exchange rate between two inputs, capital and labour, that keeps output unchanged. It equals the ratio of their marginal products, it diminishes as a firm substitutes more of one input for the other, and this diminishing pattern is exactly why isoquants curve the way they do. Combined with the isocost line, MRTS becomes the tool that tells a firm the cheapest way to produce a given quantity of output. From garment factories choosing between tailors and machines to tech firms weighing headcount against infrastructure, this single ratio quietly shapes a large share of production decisions across industries.

What do you think? If wages in a particular industry rose sharply while the cost of machinery stayed flat, how would you expect the MRTS at a typical firm’s cost-minimising point to shift? Can you think of an Indian industry where inputs behave more like perfect complements than smoothly substitutable inputs?

References

1. https://en.wikipedia.org/wiki/Marginal_rate_of_technical_substitution
2. https://www.econgraphs.org/textbooks/intermediate_micro/scarcity_and_choice/production/mrts
3. https://fiveable.me/intermediate-microeconomic-theory/unit-2/isoquants-isocost-lines/study-guide/Xj3rKc42W1sqAJ0f
4. https://en.wikipedia.org/wiki/Isoquant
5. https://corporatefinanceinstitute.com/resources/economics/marginal-rate-of-technical-substitution-mrts/
6. https://www.economicshelp.org/blog/glossary/isoquant-and-isocosts/