Economic Principles of Constrained Optimization and Marginal Analysis

Conceptual Overview: Constrained optimization occurs when a decision-maker must choose the levels of two or more activities simultaneously while subject to at least one restriction.
The Decision-Maker: In an economic context, the decision-maker is typically identified as either an individual consumer or a firm.
Activity Selection:
- Consumers often go to stores (e.g., Publix) to buy a collection of items (a shopping basket) rather than a single item. This requires choosing quantities of various goods simultaneously.
- Marketing managers must determine a promotional mix, such as the number of TV ads, radio ads, newspaper ads, billboards, and internet ads, to maximize outreach.

Variety of Constraints: Resource limitations can take several forms:
- Income: For the consumer, this is the total money available to spend on goods.
- Budget: For a firm, this is the allocated spending for specific departments, such as marketing or production.
- Time: Total hours available for activities (e.g., studying vs. playing tennis).
- Productive Capacity: The physical limit of what a firm can produce.
- Legal Obligations: Regulations or laws that limit the scope of activity.
Monetization: This course focuses primarily on constraints that can be expressed as monetary values (Income and Budgetary constraints). This allows for a simplified mathematical analysis by assigning prices to activities.

Activity Variables: Let $x$ represent the level of activity x and $y$ represent the level of activity y.
Total Benefit Function: The total benefit ( $TB$ ) derived from both activities is expressed as a multivariable function: $TB(x, y)$ .
Marginal Benefit (MB): This is the incremental or additional benefit obtained from pursuing one more unit of an activity. Mathematically, these are partial derivatives of the total benefit function:
- Marginal Benefit of X ( $MB_x$ ): The partial derivative of total benefit with respect to activity x, holding y constant. $MB_x = \frac{\partial TB}{\partial x}$ .
- Marginal Benefit of Y ( $MB_y$ ): The partial derivative of total benefit with respect to activity y, holding x constant. $MB_y = \frac{\partial TB}{\partial y}$ .
Diminishing Marginal Returns: The more one pursues an activity, the lower the incremental increase in total benefit becomes. Consequently, both the $MB_x$ and $MB_y$ curves are downward-sloping when plotted against their respective activity levels on the horizontal axis.

Price Definition:
- $P_x$ denotes the price or cost per unit of activity x.
- $P_y$ denotes the price or cost per unit of activity y.
The Proposition: To compare different activities objectively, economists utilize the ratio of marginal benefit to price:
- Value per dollar spent on x: $\frac{MB_x}{P_x}$ .
- Value per dollar spent on y: $\frac{MB_y}{P_y}$ .
Interpretation: This metric measures the "bang for the buck," or how much extra benefit a decision-maker receives for the very last dollar spent on a specific activity.

To illustrate the ability to compare vastly different products using the value per dollar metric, consider the following hypothetical scenario:

Activity 1: Buying a New Car
- $P_{car} = $40,000$
- $MB_{car} = 60,000$ (representing the lifetime value of reliability and transportation)\n * Value per dollar: \frac{60,000}{40,000} = 1.5\n * Each dollar spent on the car yields 1.50 in benefit.\n\n* **Activity 2: Buying a Soda**\n * $P_{soda} = 2.50\n * $MB_{soda} = 5.00$ (representing the hiker's or consumer's satisfaction on a specific day)
- Value per dollar: $\frac{5}{2.50} = 2$
- Each dollar spent on a soda yields $2.00$ in benefit.
Economic Conclusion: Even though the car provides $12,000$ times more marginal benefit than the soda ( $60,000$ vs. $5$ ), the soda is a "better buy" in terms of value proposition per dollar ( $2$ vs. $1.5$ ). This helps explain why consumers purchase sodas frequently (multiple times per week) but rarely purchase cars (once every five or ten years).

Case 1: \frac{MB_x}{P_x} > \frac{MB_y}{P_y}
1. Incentive: The decision-maker gets more value from x. They have an incentive to increase $x$ and decrease $y$ .
2. Outcome of Action: As $x$ increases, $MB_x$ falls due to diminishing returns. As $y$ decreases, the decision-maker moves up the marginal benefit curve for y, causing $MB_y$ to rise.
3. Result: The ratio $\frac{MB_x}{P_x}$ falls and $\frac{MB_y}{P_y}$ rises until they become equal.
Case 2: \frac{MB_x}{P_x} < \frac{MB_y}{P_y}
1. Incentive: The decision-maker gets more value from y. They should decrease $x$ and increase $y$ .
2. Outcome of Action: Decreasing $x$ causes $MB_x$ to rise; increasing $y$ causes $MB_y$ to fall.
3. Result: $\frac{MB_x}{P_x}$ rises and $\frac{MB_y}{P_y}$ falls until equality is achieved.

The Rule: Optimal choice is achieved when the value per dollar spent is equalized across all options considered: $\frac{MB_x}{P_x} = \frac{MB_y}{P_y}$
Equilibrium state: At this point, the decision-maker has no incentive to change their allocation of $x$ or $y$ . Any modification would cause the marginal benefits to shift, creating an inequality that would prompt a return to the balanced state.

Objective: Maximize Total Benefit $TB(x, y)$ .
Constraint: $P_x x + P_y y \le K$ , where $K$ is the total budget or constraint constant.
The Lagrangian Function ( $L$ ): $L = TB(x, y) + \lambda (K - P_x x - P_y y)$
Finding the Critical Point:
1. Take the partial derivative with respect to x: $L_x = \frac{\partial TB}{\partial x} - \lambda P_x = 0$ .
2. Solve for $\lambda$ : $\lambda = \frac{MB_x}{P_x}$ .
3. Take the partial derivative with respect to y: $L_y = \frac{\partial TB}{\partial y} - \lambda P_y = 0$ .
4. Solve for $\lambda$ : $\lambda = \frac{MB_y}{P_y}$ .
Conclusion: Since $\lambda$ is equal for both expressions, it follows that $\frac{MB_x}{P_x}$ must equal $\frac{MB_y}{P_y}$ . The Lagrange multiplier $\lambda$ effectively represents the value per dollar spent at the optimal point.

Classroom Logistics: Discussion regarding the lack of erasers in certain campus buildings. The instructor noted that Cooper Hall (Social Science Building) often lacks stationery like erasers, unlike other colleges on campus.
Spring of 25 Anecdote: The instructor recounted a "classroom rejuvenation" process in Cooper Hall during the middle of a term. Construction crews removed all whiteboards during the semester, forcing an abrupt change in teaching conditions.
Mathematical Proof Responsibility: The instructor clarified that while the Lagrangian proof is useful for understanding, students are not required to set up a Lagrangian on the exam. They are responsible for understanding the relationship between incentives, relative values per dollar, and the resulting optimal outcome.