Allocative Efficiency Across Multiple Goods

Why Efficiency Depends on the Mix of Goods

An economy can produce efficiently and still produce the wrong mix of goods. For example, it may be impossible to increase production of one good without reducing production of another, yet consumers may value the forgone good more than the additional output. Allocative efficiency asks whether the mix produced matches the relative value consumers place on the goods.

This comparison connects two margins. The marginal rate of substitution describes how much of one good a consumer would give up for a little more of another while keeping utility constant. The marginal rate of transformation describes how much of one good production must give up to make a little more of another. At an interior efficient allocation, those rates agree.

Definition: Let $X$ and $Y$ be two goods, and let a consumer have differentiable utility $U(X,Y)$ with positive marginal utilities $MU_X=\partial U/\partial X$ and $MU_Y=\partial U/\partial Y$. Define the marginal rate of substitution of $X$ for $Y$ as $$ MRS_{XY}=\frac{MU_X}{MU_Y}. $$ It is the amount of $Y$ the consumer is willing to give up for one more unit of $X$, locally, while holding utility constant. If the production possibility frontier is $Y=F(X)$, define the marginal rate of transformation of $X$ for $Y$ as $$ MRT_{XY}=-F'(X). $$ It is the amount of $Y$ production that must be forgone to produce one more unit of $X$ at the frontier.

Both rates use the same orientation: units of $Y$ per additional unit of $X$. With this convention, a positive $MRS_{XY}$ can be compared directly with a positive $MRT_{XY}$. The frontier is assumed to describe feasible, fully used resources; points below it may be feasible but leave some production capacity unused.

Deriving the Interior Efficiency Condition

Suppose a planner chooses a feasible production mix to maximize a consumer’s utility. Assume the allocation is interior, utility and the frontier are differentiable, marginal utilities are positive, and the choice is not restricted by a boundary. Along the production frontier, $Y=F(X)$, utility is $U(X,F(X))$. Its change when $X$ changes is

\frac{dU}{dX} = MU_X+MU_YF'(X) = MU_X-MU_Y MRT_{XY} = MU_Y(MRS_{XY}-MRT_{XY}).

At an interior maximum, a small increase or decrease in $X$ cannot improve utility, so the derivative along the frontier must be zero. Since $MU_Y>0$, the last expression is zero exactly when $MRS_{XY}=MRT_{XY}$. This is a local first-order condition. Concavity of preferences and an appropriately shaped feasible set help ensure that a point satisfying it is a maximum rather than a minimum or another stationary point.

Theorem: At an interior allocatively efficient allocation with differentiable preferences, positive marginal utilities, and a differentiable production frontier, consumers’ marginal rates of substitution between $X$ and $Y$ equal the frontier’s marginal rate of transformation: $$ MRS_{XY}=MRT_{XY}. $$ If several consumers consume both goods and goods can be reallocated between them, Pareto efficiency also requires their marginal rates of substitution to agree. Thus, at an interior Pareto-efficient allocation, each consumer’s $MRS_{XY}$ equals the common $MRT_{XY}$.

The multiple-consumer condition has a separate reason: if two consumers have different marginal rates of substitution, a small transfer of goods between them can make one better off without making the other worse off. For instance, a consumer with the higher $MRS_{XY}$ values an extra unit of $X$, relative to $Y$, more highly. If the transfer is small enough, the gain to that consumer can exceed the loss to the other. At an interior Pareto-efficient allocation, this transfer opportunity must be absent. The common rate must also match the production trade-off; otherwise a small change in production can create a net gain for consumers.

Finding the Efficient Mix

Worked Example: A Utility-Maximizing Production Mix

A planner chooses quantities of two divisible goods, $X$ and $Y$. The production frontier is $Y=120-2X$, with $0\leq X\leq60$. A representative consumer has utility $U(X,Y)=XY$. Find the utility-maximizing mix and verify the marginal-rate condition.

Method and assumptions. Treat the frontier as the feasible boundary and maximize utility on it. The goods are divisible, utility is differentiable, and the solution may be interior. The utility function is strictly concave along this frontier, as the substitution below will confirm.

Step 1: Find the marginal rates. The marginal utilities are $MU_X=Y$ and $MU_Y=X$. Therefore $MRS_{XY}=Y/X$. The frontier has slope $F'(X)=-2$, so $MRT_{XY}=2$.

Step 2: Apply the interior condition. Set $Y/X=2$, which gives $Y=2X$. Combine this with $Y=120-2X$:

2X=120-2X,\qquad 4X=120,\qquad X=30,\qquad Y=60.

Step 3: Check that it is a maximum. Substituting the frontier into utility gives $U(X,120-2X)=120X-2X^2$. Its first derivative is $120-4X$, which is zero at $X=30$, and its second derivative is $-4<0$. The endpoints $X=0$ and $X=60$ give utility zero, while the interior solution gives $U=30(60)=1800$.

Verification and mistake avoided. At $(30,60)$, $MRS_{XY}=60/30=2=MRT_{XY}$. A common mistake is to use the frontier’s slope $-2$ directly as the MRT. The negative slope records that $Y$ falls as $X$ rises; the MRT is the positive opportunity cost, $-F'(X)=2$.

Using a Mismatch to Find the Direction of Change

The equality condition is not only a way to solve for an optimum. It also diagnoses whether moving along the frontier raises or lowers utility. From the derivative,

\frac{dU}{dX}=MU_Y(MRS_{XY}-MRT_{XY}).

Because $MU_Y>0$, if $MRS_{XY}>MRT_{XY}$, a small increase in $X$ raises utility: consumers’ willingness to trade $Y$ for $X$ exceeds the production cost of that change. If $MRS_{XY}<MRT_{XY}$, a small decrease in $X$ and increase in $Y$ raises utility. This conclusion concerns a local change at the stated point; it does not by itself identify the final optimum.

Worked Example: Diagnosing a Welfare-Improving Change

A consumer has utility $U(X,Y)=\sqrt{XY}$. The frontier is $Y=90-X$, for $0\leq X\leq90$. At the current frontier allocation $(20,70)$, determine whether a small increase in $X$ improves utility, then find the interior efficient mix.

Method and assumptions. Use the sign of utility’s derivative along the differentiable frontier to determine the direction of improvement, then solve the interior equality. The quantities are positive at the current allocation and at the solution.

Step 1: Calculate the rates at the current allocation. The marginal utilities are $MU_X=\frac12\sqrt{Y/X}$ and $MU_Y=\frac12\sqrt{X/Y}$, so $MRS_{XY}=Y/X$. At $(20,70)$, this is $70/20=3.5$. Since $F'(X)=-1$, the MRT is $1$.

Step 2: Determine the direction. Here $MRS_{XY}-MRT_{XY}=3.5-1=2.5>0$. Therefore $dU/dX>0$, and increasing $X$ slightly while remaining on the frontier raises utility. For a numerical check, current utility is $\sqrt{20\cdot70}=\sqrt{1400}\approx37.4166$; moving to $(21,69)$ gives $\sqrt{1449}\approx38.0657$, an increase.

Step 3: Find the efficient mix. Set $Y/X=1$, so $Y=X$. Combining with $Y=90-X$ gives $2X=90$, hence $X=45$ and $Y=45$. Utility there is $\sqrt{45\cdot45}=45$.

Check and mistake avoided. At $(45,45)$, $MRS_{XY}=45/45=1=MRT_{XY}$. The initial point is not efficient: its rates differ, and the calculated move improves utility. Do not infer that $X$ should increase merely because its quantity is smaller than $Y$; the direction follows from comparing marginal rates, not quantities alone.

Applying the Condition with Multiple Consumers

With multiple consumers, checking only the aggregate production mix is not enough. If goods are assigned so that consumers’ marginal rates of substitution differ, there may be an efficiency-improving transfer even when the aggregate mix is right. The following planner problem shows both requirements in one calculation.

Worked Example: Two Consumers and a Production Frontier

Two consumers share goods $X$ and $Y$. The production frontier for total quantities is $Y=120-2X$, where $X=X_1+X_2$ and $Y=Y_1+Y_2$. Consumer $i$ has utility $U_i(X_i,Y_i)=\ln X_i+\ln Y_i$, and a planner maximizes $U_1+U_2$. Find the efficient quantities for each consumer and verify both marginal-rate conditions.

Method and assumptions. Quantities are positive and divisible, and the planner’s objective is explicitly the sum of the two utilities. The logarithmic utility functions are strictly concave, so equal division maximizes the sum for any fixed positive total of either good.

Step 1: Divide each total efficiently. For a fixed total $X$, maximizing $\ln X_1+\ln X_2$ subject to $X_1+X_2=X$ gives $1/X_1=1/X_2$, hence $X_1=X_2=X/2$. Similarly, $Y_1=Y_2=Y/2$.

Step 2: Find the efficient aggregate mix. For each consumer, $MU_{Xi}=1/X_i$ and $MU_{Yi}=1/Y_i$, so $MRS_{XY,i}=Y_i/X_i$. Under equal division, this is $Y/X$ for each consumer. The frontier MRT is $2$. Set $Y/X=2$ and combine with $Y=120-2X$:

2X=120-2X,\qquad X=30,\qquad Y=60.

Step 3: State the individual allocations. Equal division gives $X_1=X_2=15$ and $Y_1=Y_2=30$. Each consumer’s MRS is $30/15=2$, equal to the other consumer’s MRS and to the production MRT of $2$.

Check and mistake avoided. The aggregate quantities satisfy the frontier, $60=120-2(30)$, and all three marginal rates agree. Solving only $Y/X=MRT$ would not establish that a proposed division between consumers is efficient; their individual marginal rates must also agree. Here the specified planner objective and the equal-division calculation establish that allocation.

When Equality Is Not the Right Test

The equality $MRS_{XY}=MRT_{XY}$ is an interior condition, not a rule to apply mechanically to every feasible allocation. At a boundary, the planner may be unable to make a small change in both directions. The appropriate test then compares feasible one-sided changes. For example, if $X$ cannot be reduced below zero, an optimum at $X=0$ need not satisfy the interior equality.

The standard condition also relies on the stated two-good setting and on the relevant marginal rates reflecting social costs and benefits. If production or consumption affects others outside the model, private marginal values or costs may not capture all consequences. In that case the simple equality using private rates may not characterize social efficiency. Do not add such effects implicitly: first specify whose utility and which production opportunities are included.

For an interior competitive allocation, prices can provide a supporting interpretation: a consumer choosing both goods equates the consumer’s MRS to the relative price $p_X/p_Y$, while firms’ production choices under the relevant competitive assumptions align the MRT with that relative price. The shared ratio links consumer choice to production choice. The essential efficiency test remains the comparison of marginal rates, not the mere observation that markets clear.

Key takeaway: Use a consistent orientation for both rates. With $MRS_{XY}=MU_X/MU_Y$ and $MRT_{XY}=-dY/dX$, an interior efficient mix requires consumers’ MRS values to agree with one another and with the production MRT. If the rates differ, their ordering identifies the locally improving direction along the frontier.

Check Your Understanding

For each problem, use the stated $X$-for-$Y$ convention and check whether an interior solution is appropriate.

A planner has $U(X,Y)=X^2Y$ and frontier $Y=120-4X$, with nonnegative quantities. Find the interior efficient mix. (Answer: $X=20,\ Y=40$.)
A consumer has $U(X,Y)=\sqrt{XY}$, and the frontier is $Y=96-2X$. Find the interior efficient mix and the common marginal rate at that point. (Answer: $X=24,\ Y=48;\ MRS_{XY}=MRT_{XY}=2$.)
For $U(X,Y)=X^{1/3}Y^{2/3}$ and frontier $Y=75-X$, test the allocation $(25,50)$ using the marginal-rate condition. (Answer: It satisfies the interior condition; $MRS_{XY}=50/(2\cdot25)=1=MRT_{XY}$.)
Two consumers have bundles $(X_1,Y_1)=(2,8)$ and $(X_2,Y_2)=(4,2)$, with utilities $U_1=X_1Y_1$ and $U_2=\sqrt{X_2Y_2}$. The production MRT is $2$. Are the stated interior bundles Pareto efficient? Give each consumer’s MRS. (Answer: No; $MRS_{XY,1}=8/2=4$, while $MRS_{XY,2}=2/4=0.5$, and neither equals the MRT of $2$.)
A planner has $U(X,Y)=XY$ and frontier $Y=144-X^2$, for $0\leq X\leq12$. Find the interior efficient mix and verify the common rate. (Answer: $X=4\sqrt{3},\ Y=96;\ MRS_{XY}=96/(4\sqrt{3})=8\sqrt{3}=MRT_{XY}$.)
A consumer has $U(X,Y)=X+2Y$, and the frontier is $Y=10-X$, with $0\leq X\leq10$. Find the maximizing allocation and explain why the interior equality condition does not apply. (Answer: $X=0,\ Y=10$; utility falls as $X$ increases along the frontier, so the maximum is a corner, not an interior point requiring $MRS_{XY}=MRT_{XY}$.)

What You'll Learn