Mathematical Economics I: Static Optimization

In the study of economics, the assumption of rationality is fundamental. Economic agents, whether they are consumers, firms, or governments, are assumed to make choices that lead to the best possible outcome given their limitations. Calculating these "best" outcomes is the domain of optimization. Specifically, Static Optimization deals with decision-making where time is not an explicit variable in the optimization process. Agents make decisions based on current parameters and constraints to maximize an objective functionsuch as utility or profitor to minimize a function, such as cost.

Unlike dynamic optimization, which considers paths of variables over time, static optimization focuses on a single point in time. It answers the question: "Given the current economic environment and constraints, what is the optimal level of activity?" This guide explores the core mathematical tools used in static optimization, ranging from unconstrained optimization to the complex method of inequality constraints.

The Nature of the Objective Function

Every optimization problem in economics centers on an objective function. This is a mathematical relationship that depends on choice variables. For a consumer, the objective function is a utility function, $U(x_1, x_2, \dots, x_n)$, which depends on the quantities of various goods consumed. For a firm, the objective function is often a profit function, $\pi(q, k, l)$, or a cost function, $C(w_1, w_2, y)$.

The goal is to find the values of the choice variables that either maximize or minimize this function. To determine whether we are looking for a maximum or a minimum, we rely on the First-Order Conditions (FOC) and Second-Order Conditions (SOC). The FOC helps locate potential stationary points (candidates for optima), while the SOC helps distinguish whether those points represent a peak, a trough, or a saddle point.

Unconstrained Optimization

The simplest form of optimization occurs when the economic agent faces no restrictions on their choice variables. While rare in real-world economicsresources are always finiteunconstrained optimization provides the foundational understanding necessary for tackling more complex models.

Univariate Optimization

Consider a firm that wants to maximize a profit function that depends on a single variable, quantity $q$. The profit function is given by $\pi(q)$. To find the maximum, we take the first derivative with respect to $q$ and set it equal to zero:

This condition implies that marginal profit is zero. In economic terms, the firm should produce up to the point where selling one more unit yields no additional net profit.

However, finding a point where the slope is zero is not enough to confirm a maximum. It could be a minimum. We must verify the Second-Order Condition. For a maximum, the second derivative must be negative:

This implies that the function is concave at that point; the rate of profit is decreasing. If the second derivative were positive, the function would be convex, indicating minimum profit (or maximum loss).

Multivariate Optimization

In reality, choices rarely depend on a single variable. A consumer chooses a bundle of goods, and a firm chooses combinations of labor and capital. Let the objective function be $f(x_1, x_2, \dots, x_n)$. The necessary condition for an optimum is that the partial derivative with respect to each variable is zero. The gradient vector must be the zero vector:

For a maximum, the Second-Order Condition involves the Hessian matrixa square matrix of second-order partial derivatives. For the stationary point to be a local maximum, the Hessian matrix must be negative definite. Intuitively, this means that regardless of the direction in which we move away from the stationary point, the function will curve downward. Conversely, for a local minimum, the Hessian must be positive definite.

Equality Constrained Optimization

Economic problems are almost always constrained. Consumers have limited budgets; firms have production capacities or input requirements. When the constraints are binding and must be satisfied exactly, we use the method of Lagrange Multipliers.

Consider a consumer who wants to maximize utility $U(x, y)$ subject to a budget constraint. The total expenditure on goods $x$ and $y$ at prices $P_x$ and $P_y$ must equal income $I$:

To solve this, we construct a Lagrangian function, which incorporates the objective function and the constraint multiplied by an unknown scalar, the Lagrange multiplier ($\lambda$):

We then treat this as an unconstrained optimization problem with three variables: $x$, $y$, and $\lambda$. We take the partial derivatives:

• $\partial \mathcal{L} / \partial x = 0$: implies that the marginal utility of $x$ equals its marginal cost (adjusted by $\lambda$). Specifically, $MU_x = \lambda P_x$.
• $\partial \mathcal{L} / \partial y = 0$: implies $MU_y = \lambda P_y$.
• $\partial \mathcal{L} / \partial \lambda = 0$: simply returns the constraint itself: $P_x x + P_y y = I$.

Solving this system of equations yields the optimal consumption bundle. The Lagrange multiplier $\lambda$ has significant economic importance. It represents the shadow price of the constraint. In the consumer problem, $\lambda$ measures the marginal utility of income. It tells us how much total utility would increase if the budget constraint were relaxed by one unit (one dollar).

Inequality Constrained Optimization

In many economic scenarios, constraints are not always binding. For instance, a consumer might not be allowed to consume negative quantities of goods ($x \ge 0$), or they might be required to consume at least a minimum amount for survival ($x \geq x_{min}$). Furthermore, resource constraints often act as upper bounds (e.g., production cannot exceed capacity). To handle these, we use the Karush-Kuhn-Tucker (KKT) conditions.

The KKT conditions generalize the method of Lagrange multipliers to handle inequalities. Suppose we have a maximization problem subject to constraints $g(x) \le 0$. The KKT conditions introduce the concept of complementary slackness.

The Complementary Slackness condition states that for each inequality constraint, the product of the Lagrange multiplier and the constraint function must be zero:

This implies a logical split:
1. If the constraint is binding (active), meaning $g_i(x) = 0$, then the multiplier $\lambda_i$ can be positive. The restriction is effectively limiting the objective.
2. If the constraint is not binding (slack), meaning $g_i(x) < 0$, then the multiplier $\lambda_i$ must be zero. The restriction is irrelevant to the optimal solution.

This logic is essential for modeling corner solutions. For example, if the optimal price calculated by a firm is negative (which is impossible), the non-negativity constraint ($P \ge 0$) becomes binding. The KKT conditions ensure the mathematical solution respects these economic reality boundaries.

Economic Applications of Static Optimization

Consumer Theory

The cornerstone of consumer theory is the Utility Maximization Problem. As discussed earlier, a consumer maximizes satisfaction subject to a budget constraint. The result of this optimization is the set of demand functions. These functions express the optimal quantity of goods demanded as a function of prices and income. By applying the Envelope Theorem, economists can analyze how these optimal values change when parameters (like income or price) shift, leading to the derivation of concepts like the Roy's Identity or Slutsky equations.

Producer Theory

Firms face two primary optimization problems: Cost Minimization and Profit Maximization.
Cost Minimization: A firm chooses inputs (labor and capital) to minimize the total cost of producing a specific level of output ($y$). The constraint is the production function $f(k, l) \ge y$.
Profit Maximization: A firm chooses inputs and outputs to maximize total revenue minus total cost.
Solving the cost minimization problem yields the firm's cost functions, which are then used as inputs into the profit maximization problem. Duality theory shows that these two approaches yield consistent results regarding the firm's supply behavior.

Market Equilibrium

Static optimization is not limited to individual agents. Finding a market equilibrium often involves solving a system where the aggregate of individual optimizations leads to a clearing of markets. However, at the core, every agent's behavior within that equilibrium is derived from their specific static optimization problem. Walras' law and the proof of existence of competitive equilibrium rely heavily on the properties derived from convexity and concavitythe same properties analyzed in the Second-Order Conditions of static optimization.

Conclusion

Mathematical Economics I: Static Optimization provides the rigorous framework for understanding rational economic behavior. By translating qualitative concepts like "maximizing happiness" or "minimizing cost" into precise mathematical problems involving objective functions and constraints, economists can derive testable predictions and profound insights.

From the simplicity of finding the peak of a curve using derivatives to the complexity of handling inequality constraints via KKT conditions, these tools bridge the gap between abstract theory and real-world application. Mastering static optimization allows economists to quantify how consumers react to price changes, how firms respond to tax policies, and how resources are allocated efficiently across an economy. It is the bedrock upon which more advanced topics, including general equilibrium theory and dynamic optimization, are built.