Optimal Consumption: Lagrangian Method

1. The Geometric Intuition Behind the Lagrangian

Before diving into the algebra, let's build a strong intuition for why the Lagrangian method works. The method is a clever algebraic representation of a simple geometric condition.

As we discussed, the optimum occurs where the budget line is tangent to the highest possible indifference curve. From vector calculus, you know that the gradient of a function, $\nabla f$, at a point is perpendicular to the level curve (or surface) of $f$ at that point.

• The gradient of the utility function, $\nabla U(x_1,x_2)$, is perpendicular to the indifference curve. It points in the direction of the fastest increase in utility.
• The gradient of the budget constraint function, $g(x_1,x_2)=p_1{x}_1+p_2{x}_2-I=0$, is perpendicular to the budget line.

At the point of tangency, the indifference curve and the budget line have the same slope. This means their perpendicular vectors—their gradients—must be parallel.

Two vectors are parallel if one is a scalar multiple of the other. So, at the optimal bundle $(x_1^{\ast },x_2^{\ast })$, there must exist some scalar, which we'll call $\lambda$, such that:

$\nabla U(x_1^{\ast },x_2^{\ast })=\lambda \nabla g(x_1^{\ast },x_2^{\ast })$

Let's write this out component by component:

• $\frac{\partial U}{\partial x_1}=\lambda \frac{\partial g}{\partial x_1}⟹\frac{\partial U}{\partial x_1}=\lambda p_1$
• $\frac{\partial U}{\partial x_2}=\lambda \frac{\partial g}{\partial x_2}⟹\frac{\partial U}{\partial x_2}=\lambda p_2$

Rearranging these gives precisely the first two first-order conditions from our Lagrangian setup:

1. $\frac{\partial U}{\partial x_1}-\lambda p_1=0$
2. $\frac{\partial U}{\partial x_2}-\lambda p_2=0$

The third condition, $\frac{\partial \mathcal{L}}{\partial \lambda }=0$, simply recovers the budget constraint $p_1{x}_1+p_2{x}_2=I$, ensuring our solution is affordable.

The following video provides a clear explanation of this connection between gradients and the Lagrangian.

2. Solving for the Optimal Bundle: A Step-by-Step Example

Now that we have the intuition, let's walk through the mechanics of solving the system of first-order conditions. We'll use the canonical Cobb-Douglas utility function, which has the form $U(x,y)=x^a{y}^b$. This functional form is widely used in economics due to its convenient mathematical properties.

The general procedure is as follows:

1. Set up the Lagrangian: $\mathcal{L}=U(x_1,x_2)+\lambda (I-p_1{x}_1-p_2{x}_2)$.
2. Find the First-Order Conditions (FOCs): Take the partial derivative of $\mathcal{L}$ with respect to $x_1$, $x_2$, and $\lambda$, and set each equal to zero.
3. Solve the System of Equations: Use the three FOCs to solve for the three unknowns: $x_1$, $x_2$, and $\lambda$.

The following video provides a complete, step-by-step solution for a specific Cobb-Douglas utility function.

Let's recap the algebraic strategy you saw in the video, which is a standard approach:

1. Isolate $\lambda$ using the first two FOCs (the ones for $x$ and $y$). This gives you two expressions for $\lambda$.
   • From $\frac{\partial \mathcal{L}}{\partial x}=M{U}_x-\lambda p_x=0$, we get $\lambda =\frac{M{U}_x}{p_x}$.
   • From $\frac{\partial \mathcal{L}}{\partial y}=M{U}_y-\lambda p_y=0$, we get $\lambda =\frac{M{U}_y}{p_y}$.
2. Equate the expressions for $\lambda$. This yields the familiar "bang for the buck" condition: $\frac{M{U}_x}{p_x}=\frac{M{U}_y}{p_y}$. Rearranging gives the tangency condition: $\frac{M{U}_x}{M{U}_y}=\frac{p_x}{p_y}$.
3. Solve for one variable in terms of the other. The equation from step 2 gives you a relationship between $x$ and $y$. You can rearrange it to express, for example, $y$ as a function of $x$, prices, and the utility parameters.
4. Substitute into the budget constraint. Take the expression for $y$ from step 3 and substitute it into the third FOC ($I-p_{x}x-p_{y}y=0$). This leaves you with one equation with one unknown, $x$.
5. Solve for the optimal quantities. Solve the equation from step 4 to find the optimal quantity $x^{\ast }$. Then, plug $x^{\ast }$ back into the expression from step 3 to find the optimal quantity $y^{\ast }$.

The final results, $x^{\ast }(p_x,p_y,I)$ and $y^{\ast }(p_y,p_x,I)$, represent the optimal consumption bundle. These are general solutions that tell you the optimal amount to consume for any given set of prices and income.

3. The Economic Interpretation of the Lagrange Multiplier ($\lambda$)

We've solved for $x^{\ast }$ and $y^{\ast }$, but what about the third variable, $\lambda$? It's not just an algebraic tool; it has a crucial economic interpretation.

The Lagrange multiplier, $\lambda$, represents the marginal utility of income.

It tells you how much your maximum utility would increase if your income $I$ were to increase by one unit (e.g., one dollar). Mathematically, at the optimal solution, it can be shown that:
${\lambda }^{\ast }=\frac{d{U}^{\ast }(x_1^{\ast }(I),x_2^{\ast }(I))}{dI}$
where $U^{\ast }$ is the utility evaluated at the optimal bundle. This is a direct consequence of a powerful result in optimization theory called the Envelope Theorem.

To better understand this, please read the following section from the MIT OpenCourseWare notes you saw in the last lesson.

As the text explains, we can also see this interpretation directly from the FOCs. We established that $\lambda =\frac{M{U}_1}{p_1}=\frac{M{U}_2}{p_2}$. This shows that the value of $\lambda$ is the marginal utility gained from the last dollar spent on any good. It makes intuitive sense that this value should be equal to the marginal utility you would get from having an extra dollar of income to spend. This is why $\lambda$ is sometimes called the "shadow price" of income—it's the utility value of relaxing the budget constraint by one dollar.

Conclusion

In this lesson, you have learned how to move from formulating the consumer's problem to solving it. By applying the Lagrangian method, you can now derive the specific quantities of goods that a consumer will choose to purchase to maximize their utility.

Key Takeaways:

• The Lagrangian method is a systematic procedure for solving constrained optimization problems. Its logic is rooted in the geometric condition that the gradients of the objective function and the constraint must be parallel at the optimum.
• The solution process involves setting up the Lagrangian, deriving the three first-order conditions, and solving this system of equations for $x_1^{\ast }$, $x_2^{\ast }$, and ${\lambda }^{\ast }$.
• The optimal quantities $(x_1^{\ast },x_2^{\ast })$ depend on the consumer's income and the prices of the goods.
• The Lagrange multiplier, $\lambda$, is not just a mathematical artifact; it represents the marginal utility of income, a concept with significant economic meaning.

Preview for the Next Lesson:
The functions we derived today, $x_1^{\ast }(p_1,p_2,I)$ and $x_2^{\ast }(p_1,p_2,I)$, are of central importance. They are the consumer's demand functions. In our next lesson, we will formally define these Marshallian demand functions and begin to analyze their properties, exploring how consumer choices change when prices or income fluctuate.