📘 Lecture 40 — CONSTRAINED OPTIMIZATION ANALYSIS
📖 Overview: This lecture covers the comparative-static aspects of optimization, exploring how changes in parameters affect optimal solutions. It introduces the rationale for constrained optimization, methods for finding stationary values, and the interpretation of the Lagrange multiplier, concluding with second-order conditions using the bordered Hessian matrix.
🗂️ Topics Covered
This lecture covers comparative-static analysis through reduced form solutions, the rationale for constrained optimization, the substitution/elimination method and the Lagrange multiplier method for finding stationary values, interpretation of the Lagrange multiplier as shadow prices, and second-order conditions using the bordered Hessian determinant for verifying maxima and minima in constrained problems.
📝 Lecture Summary
TOPIC 193: COMPARATIVE-STATIC ASPECTS OF OPTIMIZATION
Example # 1 Reduced form Solutions
Comparative-static analysis examines how the optimal solution of an optimization problem changes when a parameter changes. This is done by partially differentiating the reduced form solutions (expressions for the optimal values of choice variables in terms of parameters) with respect to the parameters of interest.
🔑 Definition — Reduced form solution: An expression that gives the optimal value of a choice variable solely in terms of the exogenous parameters of the model.
📐 Formula: For optimal (x^* = x^(a, b)) and (y^ = y^(a, b)), comparative statics involve finding (\frac{\partial x^}{\partial a}, \frac{\partial x^}{\partial b}, \frac{\partial y^}{\partial a}, \frac{\partial y^*}{\partial b}).
The lecture presents a technical example showing the partial derivatives of a reduced form function with respect to parameters, signifying the technical relation in production. The signs of these derivatives (positive or negative) indicate the direction of change in the optimal solution due to a change in the parameter.
TOPIC 194: RATIONALE FOR CONSTRAINED OPTIMIZATION
Constraints can be considered as old as scarcity. A constraint is also known as a restraint, side relation, or subsidiary condition. It narrows the domain ((D)) and hence range ((R)) of the objective function (say (f(x))).
🔑 Definition — Constrained optimum: The maximum or minimum value of an objective function subject to a given constraint.
The free optimum (unconstrained maximum) is higher than the constrained optimum. Sometimes both can be the same, but the constrained optimum cannot be higher than the free optimum. This is because the constraint restricts the feasible set of solutions.
💡 Why this matters: This establishes the fundamental principle that adding a constraint can only lower (or at best maintain) the achievable optimum value, never increase it.
TOPIC 195: FINDING STATIONARY VALUES USING SUBSTITUTION/ELIMINATION METHOD
Assume an objective function: (z = f(x, y) = -4x^2 - 3y^2) A constraint function is: (x + y = 8)
Use the constraint to express one variable in terms of the other: (y = 8 - x)
Substitute this into the objective function to convert it into an unconstrained problem in one variable: (z = f(x, 8-x) = -4x^2 - 3(8-x)^2) (z = -4x^2 - 3(64 - 16x + x^2)) (z = -4x^2 - 192 + 48x - 3x^2) (z = -7x^2 + 48x - 192)
First order condition (F.o.C.): (\frac{dz}{dx} = -14x + 48 = 0) Solving gives: (x^* = \frac{48}{14} = \frac{24}{7})
Since (y = 8 - x): (y^* = 8 - \frac{24}{7} = \frac{56}{7} - \frac{24}{7} = \frac{32}{7})
Critical values: (x^* = \frac{24}{7}) and (y^* = \frac{32}{7})
Substitute (x^) and (y^) into the original objective function to find the constrained optimum value: (z^* = f\left(\frac{24}{7}, \frac{32}{7}\right) = -4\left(\frac{24}{7}\right)^2 - 3\left(\frac{32}{7}\right)^2) (z^* = -4\left(\frac{576}{49}\right) - 3\left(\frac{1024}{49}\right)) (z^* = -\frac{2304}{49} - \frac{3072}{49} = -\frac{5376}{49})
📌 Example: For (z = -4x^2 - 3y^2) subject to (x + y = 8), the substitution method gives (x^* = 24/7, y^* = 32/7), and (z^* = -5376/49).
TOPIC 196: FINDING STATIONARY VALUES USING METHOD OF LAGRANGE MULTIPLIER
Introduced by Joseph-Louis Lagrange, this method incorporates the constraint into the objective function using an auxiliary variable called the Lagrange multiplier.
Assume an objective function: (z = f(x, y) = -4x^2 - 3y^2) A constraint function is: (x + y = 8)
Form the Lagrange function (Z): (Z(x, y, \lambda) = f(x, y) + \lambda (8 - x - y)) (Z = -4x^2 - 3y^2 + \lambda(8 - x - y))
First order conditions (F.o.C.): (\frac{\partial Z}{\partial x} = -8x - \lambda = 0) (\frac{\partial Z}{\partial y} = -6y - \lambda = 0) (\frac{\partial Z}{\partial \lambda} = 8 - x - y = 0)
Rewriting the first two equations: (-8x = \lambda \implies \lambda = -8x) (-6y = \lambda \implies \lambda = -6y)
Equating the expressions for (\lambda): (-8x = -6y \implies x = \frac{3}{4}y)
Substitute into the 3rd F.o.C. ((x + y = 8)): (\frac{3}{4}y + y = 8 \implies \frac{7}{4}y = 8 \implies y^* = \frac{32}{7})
Then: (x^* = \frac{3}{4} \times \frac{32}{7} = \frac{24}{7})
Critical values: (x^* = \frac{24}{7}) and (y^* = \frac{32}{7})
Substitute into the objective function: (z^* = f\left(\frac{24}{7}, \frac{32}{7}\right) = -4\left(\frac{24}{7}\right)^2 - 3\left(\frac{32}{7}\right)^2 = -\frac{5376}{49})
This yields the same optimum as the substitution method but also provides the value of (\lambda^* = -8\left(\frac{24}{7}\right) = -\frac{192}{7}).
💡 Why this matters: The Lagrange multiplier method is more powerful than substitution because it treats all variables symmetrically and is easily extended to problems with multiple constraints and many variables.
TOPIC 197: INTERPRETATION OF THE LAGRANGE MULTIPLIER
The Lagrange multiplier answers the question: What is the impact of a per-unit change in the constraint constant on the objective function? The value of (\lambda^*) is also known as shadow prices.
Considering the example with the same objective function but a different constraint constant: (z = -4x^2 - 3y^2) subject to (x + y = 56)
Solving this new problem yields the optimum (z^{**} = -5376).
The change from a constraint constant of 8 (giving (z^* = -5376/49 \approx -109.7)) to a constant of 56 (giving (z^{**} = -5376)) represents a difference of (\Delta z = -5376 - (-109.7) = -5266.3). The change in the constant is (\Delta c = 56 - 8 = 48).
The Lagrange multiplier from the original problem, (\lambda^* = -192/7 \approx -27.43), predicts the rate of change. The predicted change would be (\lambda^* \times \Delta c = (-192/7) \times 48 = -9216/7 \approx -1316.6). The actual change is much larger, indicating that (\lambda) predicts the approximate change for a small (1-unit) change in the constraint.
🔑 Definition — Shadow price (in constrained optimization): The rate of change in the optimal value of the objective function per unit change in the constraint constant; equal to the Lagrange multiplier (\lambda^*).
For a 1-unit increase (decrease) in the constant of the constraint (from 56), the value of (\lambda^* = -192/7 \approx -27.43) predicts that (z) would increase (decrease) by approximately (-27.43) units.
The lecture verifies this by increasing the constraint constant from 56 to 57: Subject to (x + y = 57): (y^* = \frac{4}{7}(57) = \frac{228}{7}) and (x^* = \frac{3}{7}(57) = \frac{171}{7}) (z^{} = -4\left(\frac{171}{7}\right)^2 - 3\left(\frac{228}{7}\right)^2 = -\frac{116964}{49} - \frac{155952}{49} = -\frac{272916}{49} \approx -5569.7) (\Delta z = z^{} - z^{**} = -\frac{272916}{49} + \frac{263424}{49} = -\frac{9492}{49} \approx -193.7)
The actual change ((-193.7)) is not close to (\lambda^* (-27.43)) for this large previous change but is verified for a 1-unit change near the original point.
Economics Instances:
- In utility maximization subject to a budget constraint: (U(x, y)) subject to (P_x x + P_y y = I), (\lambda^*) estimates the marginal utility of an extra PKR of income.
- In output maximization subject to a cost constraint: (Q(K, L)) subject to (rK + wL = C), (\lambda^*) estimates the marginal product of an extra PKR of cost.
TOPIC 198: SECOND ORDER CONDITION: THE BORDERED HESSIAN
Attributed to German mathematician Ludwig Otto Hesse, the determinant of matrices can be used for the 2nd order condition in constrained optimization.
For (z = f(x, y)) subject to (g(x, y) = k), the 2nd order condition is tested with the bordered Hessian determinant (\bar{H}_2):
(\bar{H}2 = \begin{vmatrix} 0 & g_x & g_y \ g_x & f{xx} & f_{xy} \ g_y & f_{yx} & f_{yy} \end{vmatrix})
The order of a bordered principal minor equals the order of the principal minor being bordered (here, the first principal minor of the underlying Hessian).
(\bar{H}_2) represents a second bordered principal minor (since it borders a (1 \times 1) sub-matrix).
In algebraic form: (\bar{H}2 = \begin{vmatrix} 0 & g_x & g_y \ g_x & f{xx} & f_{xy} \ g_y & f_{yx} & f_{yy} \end{vmatrix} = (0)\begin{vmatrix} f_{xx} & f_{xy} \ f_{yx} & f_{yy} \end{vmatrix} - (g_x)\begin{vmatrix} g_x & g_y \ f_{yx} & f_{yy} \end{vmatrix} + (g_y)\begin{vmatrix} g_x & g_y \ f_{xx} & f_{xy} \end{vmatrix}) (\bar{H}2 = -g_x(g_x f{yy} - g_y f_{yx}) + g_y(g_x f_{xy} - g_y f_{xx})) (\bar{H}2 = -g_x^2 f{yy} + 2 g_x g_y f_{xy} - g_y^2 f_{xx})
For a function with (n) variables ((x_1, x_2, \dots, x_n)), subject to (g(x_1, x_2, \dots, x_n) = k): (\bar{H}2 = \begin{vmatrix} 0 & g_1 & g_2 & \dots & g_n \ g_1 & f{11} & f_{12} & \dots & f_{1n} \ g_2 & f_{21} & f_{22} & \dots & f_{2n} \ \vdots & \vdots & \vdots & \ddots & \vdots \ g_n & f_{n1} & f_{n2} & \dots & f_{nn} \end{vmatrix})
Where (\bar{H}_m) denotes the (m)th bordered principal minor (the determinant of the upper-left (m+1) rows and columns).
Conditions for maximum: (\bar{H}_2 > 0, \bar{H}_3 < 0, \bar{H}_4 > 0, \dots, (-1)^n \bar{H}_n > 0)
Conditions for minimum: (\bar{H}_2 < 0, \bar{H}_3 < 0, \bar{H}_4 < 0, \dots, \bar{H}_n < 0)
The condition ((-1)^n) is explained: For even (n), the final bordered principal minor should be positive: ((-1)^n \bar{H}_n > 0 \implies \bar{H}_n > 0). For odd (n), the final bordered principal minor should be negative: ((-1)^n \bar{H}_n > 0 \implies \bar{H}_n < 0). The factor ((-1)^n) helps to retain alternating signs.
The lecture provides an example checking the bordered Hessian for the earlier problem with constraint (x + y = 8): (g_x = 1, g_y = 1, f_{xx} = -8, f_{xy} = 0, f_{yx} = 0, f_{yy} = -6)
(\bar{H}2 = \begin{vmatrix} 0 & g_x & g_y \ g_x & f{xx} & f_{xy} \ g_y & f_{yx} & f_{yy} \end{vmatrix} = \begin{vmatrix} 0 & 1 & 1 \ 1 & -8 & 0 \ 1 & 0 & -6 \end{vmatrix})
(\bar{H}_2 = 0\begin{vmatrix} -8 & 0 \ 0 & -6 \end{vmatrix} - 1\begin{vmatrix} 1 & 1 \ 0 & -6 \end{vmatrix} + 1\begin{vmatrix} 1 & 1 \ -8 & 0 \end{vmatrix}) (\bar{H}_2 = -1(-6 - 0) + 1(0 - (-8)) = 6 + 8 = 14)
Since (\bar{H}_2 > 0), for a two-variable ((n = 2)) problem, the condition for a maximum (((-1)^2 \bar{H}_2 > 0 \implies \bar{H}_2 > 0)) is satisfied.
The lecture also tests a 2-variable minimum case: Suppose (f_{xx} = ?), (f_{yy} = ?), and (\bar{H}_2 = -10). Since (\bar{H}_2 < 0) and (n = 2) (even), the condition for a minimum (((-1)^2 \bar{H}_2 < 0 \implies \bar{H}_2 < 0)) is satisfied.
An economic instance given is the two-period model of utility maximization, where the bordered Hessian condition verifies the second-order condition for constrained optimization.
💡 Why this matters: The bordered Hessian provides a systematic algebraic method for verifying whether a stationary point is a maximum or minimum under constraints, avoiding reliance on graphical or intuitive reasoning.
⭐ Key Takeaways
The lecture establishes that constrained optimization problems can be solved using substitution or the Lagrange multiplier method, with the latter providing the shadow price (\lambda) that measures the sensitivity of the optimum to constraint changes. The second-order condition for constrained optimization is tested using the bordered Hessian determinant, where (\bar{H}_2 > 0) indicates a maximum and (\bar{H}_2 < 0) indicates a minimum for two-variable problems. The Lagrange multiplier's interpretation as a shadow price is crucial in economics for understanding the marginal value of relaxing a constraint. The comparative-static analysis extends these tools to predict how optimal solutions change with parameter shifts.
🧠 Quick Revision Questions
- What is the Lagrange multiplier method, and how does it differ from the substitution method for solving constrained optimization problems?
- In the example (z = -4x^2 - 3y^2) subject to (x + y = 8), what is the value of the Lagrange multiplier (\lambda^*) and what does it mean?
- What are the conditions for a maximum and a minimum in the bordered Hessian test for a 2-variable constrained optimization problem?
- How does the bordered Hessian determinant (\bar{H}_2) differ from the ordinary Hessian determinant (H_1)?
- In economic applications, what does the Lagrange multiplier represent in utility maximization subject to a budget constraint?
📘 Lecture 41 — UTILITY MAXIMIZATION ANALYSIS
📖 Overview: This lecture focuses on utility maximization for consumers, starting with a two-period model and then systematically developing the mathematical conditions for optimal consumer choice. It covers convexity/concavity of functions and derives both first-order and second-order conditions for utility maximization, culminating in the law of equi-marginal utility.
🗂️ Topics Covered
The lecture covers six topics: Two Period Model of Utility, Convexity and Concavity Using Second Order Derivative, Utility Maximization and Consumer Demand: First Order Condition, Utility Maximization and Consumer Demand: Second Order Condition, Numerical Example of Utility Maximization, and Law of Equi-Marginal Utility Using Lagrangian Multiplier.
📝 Lecture Summary
TOPIC 199: TWO PERIOD MODEL OF UTILITY
This topic introduces a consumer decision-making framework across two time periods, typically present and future consumption. The model establishes the foundation for intertemporal utility maximization under a budget constraint that spans both periods.
TOPIC 200: CONVEXITY AND CONCAVITY USING SECOND ORDER DERIVATIVE
Convexity and concavity are closely related to minimum and maximum of functions. A function (f) that is continuous and twice differentiable on interval (I) is convex on (I) if (f''(x) \geq 0) for all (x) in (I). It is concave on (I) if (f''(x) \leq 0) for all (x) in (I).
📐 Numerical: (f(x) = x^3 - 6x^2 + 5x + 12). Then (f'(x) = 3x^2 - 12x + 5) and (f''(x) = 6x - 12). Setting (f''(x) = 0 \Rightarrow x = 2). For (x < 2), (f''(x) < 0) (concave), and for (x > 2), (f''(x) > 0) (convex).
📐 Single input production function: (Q = f(L)) where (L) is labor.
📐 Case – I (Concave production function): (Q = 12L^{0.5}). Then (f'(L) = 6L^{-0.5}) and (f''(L) = -3L^{-1.5} < 0 \Rightarrow) Concave production function (diminishing returns).
📐 Case – II (Convex production function): (Q = L^2). Then (f'(L) = 2L) and (f''(L) = 2 > 0 \Rightarrow) Convex production function (increasing returns).
TOPIC 201: UTILITY MAXIMIZATION AND CONSUMER DEMAND: FIRST ORDER CONDITION
Consider a hypothetical consumer with two goods ((x_1, x_2)), with continuous, positive marginal utility ((MU_1 > 0, MU_2 > 0)). Market-determined prices ((p_1, p_2)) are exogenous. The budget constraint is (p_1 x_1 + p_2 x_2 = B). The utility (objective) function is (U = U(x_1, x_2)). The Lagrangian function is (L = U(x_1, x_2) + \lambda(B - p_1 x_1 - p_2 x_2)).
First order conditions (F.o.C):
- (\frac{\partial L}{\partial x_1} = U_1 - \lambda p_1 = 0 \Rightarrow U_1 = \lambda p_1)
- (\frac{\partial L}{\partial x_2} = U_2 - \lambda p_2 = 0 \Rightarrow U_2 = \lambda p_2)
- (\frac{\partial L}{\partial \lambda} = B - p_1 x_1 - p_2 x_2 = 0)
Extracting the value of (\lambda) from 1st and 2nd F.o.C and equating: [ \frac{U_1}{p_1} = \lambda = \frac{U_2}{p_2} \Rightarrow \frac{U_1}{p_1} = \frac{U_2}{p_2} ]
Interpretation: (\frac{U_1}{p_1} = \frac{U_2}{p_2}) is actually the law of equi-marginal utility. The marginal utility of money spent on each good is equal. (\lambda) can be termed as the marginal utility of (budget) money when utility is maximized. Alternatively, (U_1 = \lambda p_1) and (U_2 = \lambda p_2) gives: [ \frac{U_1}{U_2} = \frac{p_1}{p_2} = (\text{Slope of IC}) < 0 ] [ \frac{dx_2}{dx_1} = -\frac{U_1}{U_2} ]
Since (U_1 > 0) and (U_2 > 0), the indifference curve (IC) has negative slope. A demand function is the function that relates price and quantity, e.g., (x_1 = f(p_1, p_2, B)) or from (F.o.C): (x_1^* = f(p_1, p_2, B)).
TOPIC 202: UTILITY MAXIMIZATION AND CONSUMER DEMAND: SECOND ORDER CONDITION
After developing and analyzing the 1st order condition, we develop the 2nd order condition. Given (L = U(x_1, x_2) + \lambda(B - p_1 x_1 - p_2 x_2)), the Bordered Hessian is: [ \bar{H} = \begin{vmatrix} L_{11} & L_{12} & -p_1 \ L_{21} & L_{22} & -p_2 \ -p_1 & -p_2 & 0 \end{vmatrix}
0 ] where (L_{11} = U_{11}), (L_{12} = U_{12}), (L_{21} = U_{21}), (L_{22} = U_{22}).
Diagrammatically, the shape of IC should be convex to origin. Algebraically: (\frac{d^2 x_2}{dx_1^2} > 0).
Using (\frac{dx_2}{dx_1} = -\frac{U_1}{U_2}), we differentiate: [ \frac{d^2 x_2}{dx_1^2} = -\left(\frac{U_2[U_{11} + U_{12}\frac{dx_2}{dx_1}] - U_1[U_{21} + U_{22}\frac{dx_2}{dx_1}]}{U_2^2}\right) ] [ = -\left(\frac{U_2[U_{11} + U_{12}(-\frac{U_1}{U_2})] - U_1[U_{21} + U_{22}(-\frac{U_1}{U_2})]}{U_2^2}\right) ] [ = -\left(\frac{U_2 U_{11} - U_{12}U_1 - U_1 U_{21} + \frac{U_1 U_{22}U_1}{U_2}}{U_2^2}\right) ] [ = -\frac{1}{U_2^3}\left(U_2^2 U_{11} - 2U_1 U_2 U_{12} + U_1^2 U_{22}\right) > 0 ]
🔑 Definition — Bordered Hessian ((\bar{H})): A determinant used to test for the second-order condition in constrained optimization, where the constraint is expressed as border elements.
Since both (\bar{H} > 0) and (\frac{d^2 x_2}{dx_1^2} > 0) are positive, this fulfills the condition for convexity (to origin). Both the S.o.C (Bordered Hessian (\bar{H} > 0)) and the convexity condition ((\frac{d^2 x_2}{dx_1^2} > 0)) are verified and inter-related. 💡 Why this matters: These conditions ensure that the solution from the first-order conditions yields a true maximum, not a minimum.
TOPIC 203: NUMERICAL EXAMPLE OF UTILITY MAXIMIZATION
This topic provides a worked numerical example demonstrating the application of the Lagrangian method and first-order conditions to find optimal consumption bundles.
TOPIC 204: LAW OF EQUI-MARGINAL UTILITY USING LAGRANGIAN MULTIPLIER
This topic demonstrates that the Lagrangian multiplier approach formally proves the law of equi-marginal utility, showing that at the optimum, the ratio of marginal utility to price is equal across all goods.
⭐ Key Takeaways
A student must remember that utility maximization involves setting up a Lagrangian to incorporate the budget constraint, deriving first-order conditions that equate the marginal utility per dollar spent across goods (the law of equi-marginal utility). The second-order condition requires the bordered Hessian to be positive, which is algebraically equivalent to the indifference curve being convex to the origin. Convexity and concavity of functions are determined by the sign of the second derivative, with positive indicating convexity and negative indicating concavity.
🧠 Quick Revision Questions
- What is the first-order condition for utility maximization with two goods, and what economic principle does it represent?
- How do you determine if a function is convex or concave using its second derivative?
- What is the bordered Hessian, and what sign must it have for a maximum in utility maximization?
- Derive the condition for convexity of an indifference curve algebraically.
- What does the Lagrangian multiplier (\lambda) represent in the utility maximization problem?
📘 Lecture 42 — Homogenous Production Function
📖 Overview: This lecture examines the economic applications of production function maximization, focusing on homogeneous functions and their properties. It explores how logarithmic transformations are used in production functions and introduces Euler's theorem, which is crucial for understanding factor payments and returns to scale in economic theory.
🗂️ Topics Covered
The lecture covers economic applications of production function maximization, logarithmic transformation of production functions, the definition and testing of homogeneous functions, homogeneous production functions in relation to average products and capital-labor ratio, homogeneous production functions in relation to marginal products and capital-labor ratio, and finally the Homogeneous Production Function and Euler's Theorem with its economic interpretation.
📝 Lecture Summary
TOPIC 205: ECONOMIC APPLICATION OF PRODUCTION FUNCTION MAXIMIZATION
[No specific content provided for this topic in the text, but it introduces the application of maximizing production functions in an economic context.]
TOPIC 206: ECONOMIC APPLICATION ON LOGARITHMICALLY TRANSFORMED PRODUCTION FUNCTION
[No specific content provided for this topic in the text, but it introduces the use of logarithmic transformations in production function analysis.]
TOPIC 207: HOMOGENEOUS FUNCTIONS
Etymology: The word comes from Greek homogenēs: homos meaning 'same' + genos meaning 'race, kind'. Here, genos refers to the degree of a term.
A function is considered homogenous if each term involved has the same degree. For a polynomial of degree r: each term has degree r.
📌 Example: Check if function ( f(x, y) = x^3 + x^2 y + 3xy^2 + 5y^3 ) is homogeneous.
To check, introduce a scalar ( \lambda ) on both sides:
- ( f(\lambda x, \lambda y) = (\lambda x)^3 + (\lambda x)^2(\lambda y) + 3(\lambda x)(\lambda y)^2 + 5(\lambda y)^3 )
- ( = \lambda^3 x^3 + \lambda^3 x^2 y + 3\lambda^3 x y^2 + 5\lambda^3 y^3 )
- ( = \lambda^3 (x^3 + x^2 y + 3xy^2 + 5y^3) )
- ( = \lambda^3 [f(x, y)] ) Thus, the function is homogenous of degree 3 in variables ( x ) and ( y ).
🔑 D.I.Y.: Check if functions are homogenous or not. If yes, then of what degree. (Function provided: ( x^3 + x^2 y + 3xy^2 + 5y^3 ))
TOPIC 208: HOMOGENEOUS PRODUCTION FUNCTION, AVERAGE PRODUCTS AND CAPITAL-LABOR RATIO
Consider the Cobb-Douglas production function: ( Q = A K^\alpha L^\beta )
Testing homogeneity:
- ( Q(\lambda K, \lambda L) = A (\lambda K)^\alpha (\lambda L)^\beta )
- ( = A \lambda^\alpha K^\alpha \lambda^\beta L^\beta )
- ( = \lambda^{\alpha+\beta} A K^\alpha L^\beta )
- ( = \lambda^{\alpha+\beta} Q(K, L) )
The degree of homogeneity is ( \alpha + \beta ). For a Cobb-Douglas production function, when ( \alpha + \beta = 1 ), it is linearly homogeneous or shows linear homogeneity.
💡 Why this matters: Linear homogeneity implies that if all independent variables (inputs) are increased by the same proportion ( \lambda ), the dependent variable (output) will increase by the same proportion. This implies constant returns to scale. (If ( \lambda > 1 \Rightarrow Q > 1 \Rightarrow ) output increases by the same proportion).
Average product of labor (( AP_L )):
- Formula: ( AP_L = \frac{Q}{L} = \frac{A K^\alpha L^\beta}{L} = A K^\alpha L^{\beta-1} )
- Since ( \alpha + \beta = 1 \Rightarrow \beta = 1 - \alpha )
- ( AP_L = A K^\alpha L^{-\alpha} = A \left(\frac{K}{L}\right)^\alpha )
- This shows a relationship between average product of labor and capital-labor ratio ( \left(\frac{K}{L}\right) ).
Degree of homogeneity of ( AP_L ):
- ( AP_L = A \left(\frac{K}{L}\right)^\alpha )
- ( AP_L(\lambda K, \lambda L) = A \left(\frac{\lambda K}{\lambda L}\right)^\alpha = A \left(\frac{K}{L}\right)^\alpha )
- ( = \lambda^0 \cdot A \left(\frac{K}{L}\right)^\alpha = \lambda^0 [AP_L(K, L)] )
- Answer: Homogeneous of degree zero.
Average Product of Capital (( AP_K )):
- Formula: ( AP_K = \frac{Q}{K} = \frac{A K^\alpha L^{1-\alpha}}{K} = A K^{\alpha-1} L^{1-\alpha} )
- ( AP_K = A \left(\frac{L}{K}\right)^{1-\alpha} = A \left(\frac{K}{L}\right)^{\alpha-1} )
- This shows a relationship between average product of capital and capital-labor ratio.
Degree of homogeneity of ( AP_K ):
- ( AP_K = A \left(\frac{K}{L}\right)^{\alpha-1} )
- ( AP_K(\lambda K, \lambda L) = A \left(\frac{\lambda K}{\lambda L}\right)^{\alpha-1} = A \left(\frac{K}{L}\right)^{\alpha-1} )
- ( = \lambda^0 \cdot AP_K(K, L) )
- Answer: Homogeneous of degree zero.
🔑 Interpretation: The degree of homogeneity for ( AP_L ) and ( AP_K ) is zero. This means both are homogeneous of degree zero in the variables ( K ) and ( L ). Since equal proportionate changes in ( K ) and ( L ) (maintaining a constant ( \frac{K}{L} )) will not change the magnitudes of the average products.
TOPIC 209: HOMOGENEOUS PRODUCTION FUNCTION, MARGINAL PRODUCTS AND CAPITAL-LABOR RATIO
Marginal product of capital (( MP_K )):
- Formula: ( MP_K = \frac{\partial Q}{\partial K} = \frac{\partial}{\partial K} [A K^\alpha L^{1-\alpha}] )
- ( MP_K = \alpha A K^{\alpha-1} L^{1-\alpha} = \alpha A \left(\frac{L}{K}\right)^{1-\alpha} )
- ( MP_K = \alpha A \left(\frac{K}{L}\right)^{\alpha-1} )
- This shows a relationship between marginal product of capital and capital-labor ratio.
Degree of homogeneity of ( MP_K ):
- ( MP_K = \alpha A \left(\frac{K}{L}\right)^{\alpha-1} )
- ( MP_K(\lambda K, \lambda L) = \alpha A \left(\frac{\lambda K}{\lambda L}\right)^{\alpha-1} = \alpha A \left(\frac{K}{L}\right)^{\alpha-1} )
- ( = \lambda^0 \cdot MP_K(K, L) )
- Answer: Homogeneous of degree zero.
Marginal product of labor (( MP_L )):
- Formula: ( MP_L = \frac{\partial Q}{\partial L} = \frac{\partial}{\partial L} [A K^\alpha L^\beta] )
- Since ( \alpha + \beta = 1 \Rightarrow \beta = 1 - \alpha )
- ( MP_L = \frac{\partial}{\partial L} [A K^\alpha L^{1-\alpha}] )
- ( MP_L = (1-\alpha) A K^\alpha L^{-\alpha} = (1-\alpha) A \left(\frac{K}{L}\right)^\alpha )
- This shows marginal product of capital and capital-labor ratio are related.
Degree of homogeneity of ( MP_L ):
- ( MP_L = (1-\alpha) A \left(\frac{K}{L}\right)^\alpha )
- ( MP_L(\lambda K, \lambda L) = (1-\alpha) A \left(\frac{\lambda K}{\lambda L}\right)^\alpha = (1-\alpha) A \left(\frac{K}{L}\right)^\alpha )
- ( = \lambda^0 \cdot MP_L(K, L) )
- Answer: Homogeneous of degree zero.
🔑 Interpretation: The degree of homogeneity for ( MP_K ) and ( MP_L ) is zero. Both are homogeneous of degree zero in the variables ( K ) and ( L ). Equal proportionate changes in ( K ) and ( L ) (maintaining a constant ( \frac{K}{L} )) will not change the magnitudes of the marginal products.
TOPIC 210: HOMOGENEOUS PRODUCTION FUNCTION AND EULER'S THEOREM
Euler's theorem (by Leonhard Euler, a Swiss mathematician) is a property of a function ( f(x, y) ) with degree ( r ) of homogeneity: [ x \cdot f_x + y \cdot f_y = r \cdot f(x, y) ] In production context: [ K \cdot \frac{\partial Q}{\partial K} + L \cdot \frac{\partial Q}{\partial L} = r \cdot Q(K, L) ] In another notation: [ K \cdot MP_K + L \cdot MP_L = r \cdot Q(K, L) ] Or: [ \left(\frac{K}{Q}\right) MP_K + \left(\frac{L}{Q}\right) MP_L = r ]
For a linearly homogeneous production function (( r = 1 )): [ K \cdot MP_K + L \cdot MP_L = Q ]
Verification for Cobb-Douglas: From Topic 209: ( MP_K = \alpha A \left(\frac{K}{L}\right)^{\alpha-1} ) and ( MP_L = (1-\alpha) A \left(\frac{K}{L}\right)^\alpha ) [ K \cdot MP_K + L \cdot MP_L = K \cdot \alpha A \left(\frac{K}{L}\right)^{\alpha-1} + L \cdot (1-\alpha) A \left(\frac{K}{L}\right)^\alpha ] [ = A K^{\alpha} L^{1-\alpha} [\alpha + (1-\alpha)] ] Restoring ( Q = A K^\alpha L^{1-\alpha} ): [ = Q ]
🔑 Economic Interpretation: Under constant returns to scale (CRS) , if an input factor is paid the amount of its marginal product (MP), the total product (TP) will be exhausted by the distributive shares for all the input factors. This means pure economic profit will be zero.
⚠️ Caveat: Euler's theorem holds only if perfect competition holds in the factors market.
⭐ Key Takeaways
This lecture establishes that homogeneous functions are central to understanding production theory, particularly the Cobb-Douglas function. Under linear homogeneity (constant returns to scale), both average and marginal products are homogeneous of degree zero, meaning they depend only on the capital-labor ratio and are unaffected by proportionate input changes. Euler's theorem provides the critical economic insight that under perfect competition, paying factors their marginal products exactly exhausts total output, resulting in zero pure economic profit. Students must be able to test for homogeneity, derive relationships between product functions and the capital-labor ratio, and apply Euler's theorem to verify factor exhaustion.
🧠 Quick Revision Questions
- What condition must a function satisfy to be classified as homogeneous of degree r?
- For a linearly homogeneous Cobb-Douglas production function, what is the relationship between the average product of labor and the capital-labor ratio?
- What is the economic significance of average and marginal products being homogeneous of degree zero?
- State Euler's theorem for a linearly homogeneous production function. What does it imply in a perfectly competitive factor market?
- What condition must hold for Euler's theorem to be valid in an economic context?
📘 Lecture 43 — COBB DOUGLAS PRODUCTION FUNCTION
📖 Overview: This lecture explores the Cobb-Douglas production function and its properties, focusing on returns to scale, homogeneity, and least-cost combinations. It extends the analysis to three-input production functions, expansion paths, and introduces the concept of homothetic functions, which are crucial for understanding firm behavior and production optimization.
🗂️ Topics Covered
The lecture covers the Cobb-Douglas production function and returns to scale, including homogeneity for both two and three input functions. It then examines least-cost combinations using the CES production function and Lagrangian optimization, the expansion path derived from first-order conditions, the concept of homothetic functions, and finally the homotheticity of the Cobb-Douglas production function itself.
📝 Lecture Summary
TOPIC 211: COBB-DOUGLAS PRODUCTION FUNCTION AND RETURNS TO SCALE
The Cobb-Douglas production function is a fundamental tool in economics for modeling the relationship between inputs (like capital and labor) and output. The returns to scale of this function are determined by the sum of its exponents.
TOPIC 212: HOMOGENEITY AND RETURNS TO SCALE OF THREE INPUT PRODUCTION FUNCTION
This topic extends the analysis to a production function with three inputs: Capital (K), Labor (L), and Material (M). The function is given as Q = f(K, L, M).
-
a + b + c is the degree of homogeneity.
-
Introducing k on both sides, Euler’s Theorem requires: K(∂Q/∂K) + L(∂Q/∂L) + M(∂Q/∂M) = k(a+b+c) * f(K, L, M).
k(a+b+c) * f(K, L, M) = f(kK, kL, kM).
k(a+b+c) * Q = k is completely factored out, showing it’s a homogenous production function. (a+b+c) is the exponent showing degree of homogeneity.
-
If (a+b+c) = 1, then CRS (Constant Returns to Scale) prevails.
-
If (a+b+c) > 1, then IRS (Increasing Returns to Scale).
-
If (a+b+c) < 1, then DRS (Decreasing Returns to Scale).
TOPIC 213: LEAST-COST COMBINATION IN COBB-DOUGLAS PRODUCTION FUNCTION
This topic uses the CES production function to find the least-cost input combination. The Lagrangian function is formed: L = bKαLβ + λ( C - rK - wL ) where b is a constant.
The First-Order Conditions (F.O.Cs) are:
- ∂L/∂K = bαKα-1Lβ - λr = 0
- ∂L/∂L = bβKαLβ-1 - λw = 0
- ∂L/∂λ = C - rK - wL = 0
Solving for λ in the 1st and 2nd F.O.C.s and equating them yields: bαKα-1Lβ / r = bβKαLβ-1 / w
The marginal products are:
- MPK = bαKα-1Lβ
- MPL = bβKαLβ-1
Reverting to the previous equation, we get the least cost input ratio: αL / r = βK / w, which simplifies to: K/L = (αw) / (βr)
This ratio is the optimal proportion of capital to labor at the least cost. 🔑 Definition — Least Cost Input Ratio (K/L): The ratio of capital to labor that minimizes the cost of producing a given level of output, found by equating the marginal product per dollar spent on each input. 📐 Formula: K/L = (α/β) * (w/r) → The optimal capital-labor ratio equals the ratio of the output elasticities of capital and labor multiplied by the inverse of the input price ratio.
TOPIC 214: EXPANSION PATH USING FIRST ORDER CONDITION
This topic explores the comparative-static aspect of producer equilibrium. With a fixed ratio of K and L (i.e., K/L = constant), postulate successive increases of Q (ascent to higher isoquants) to trace the effect on the least-cost combination (K*, L*).
Each shift of the isoquant gives a new point of tangency with a higher isocost line. The locus of such points is the expansion path of the firm, describing the least-cost combinations for varying levels of Q.
From the 1st order condition: MPL/MPK = w/r. For the Cobb-Douglas function Q = AKαLβ: MPL = AβKαLβ-1 MPK = AαKα-1Lβ
Substituting into the condition: (AβKαLβ-1) / (AαKα-1Lβ) = w/r (βK) / (αL) = w/r K/L = (α/β) * (w/r)
At equilibrium: K/L = (α/β) * (w/r). The answer is numerical as α, β, w, r are all constants. Therefore, all points on the expansion path will show a fixed input ratio, meaning the expansion path will be a straight line starting from the origin. A homogeneous production function gives rise to a straight line expansion path.
TOPIC 215: HOMOTHETIC FUNCTIONS
Homothetic functions generate radial expansions, preserving both angles and ratios of distances.
Consider a production function T: Q = f(K, L). Its monotonic transformation generates a new production (composite) function G: G(K, L) = F[f(K, L)], where F is a monotonic function.
While f(K, L) may be homogeneous, F[f(K, L)] is not necessarily homogeneous. For example: f(K, L) has a homogeneous degree of r, however, G(K, L) = F[f(K, L)] may not be homogeneous.
The expansion path of G(K, L) is linear like that of f(K, L). As observed, the slope of isoquants (MRTS) remains the same at a given K:L ratio: The slope for the original function is found from the ratio of marginal products. For the monotonic transformation, the slope is given by derivative of the composite function, which will also be constant at a given input ratio. 🔑 Definition — Monotonic Transformation: A strictly increasing function applied to an existing function, which preserves the ordering of values.
TOPIC 216: HOMOTHETICITY OF COBB-DOUGLAS PRODUCTION FUNCTION
Consider the Cobb-Douglas production function T: f(K, L) = AKαLβ. It is homogeneous. A monotonic transformation (e.g., squaring) generates a new function G: G(K, L) = [f(K, L)]² = (AKαLβ)² = A²K2αL2β.
The slope of the isoquant (MRTS) for G is: MRTS = MPL/MPK = [∂(A²K2αL2β)/∂L] / [∂(A²K2αL2β)/∂K] = (2βA²K2αL2β-1) / (2αA²K2α-1L2β) = (β/α) * (K/L)
With a given (K, L), the slope of G(K, L) will be constant.
- f(K, L) is homogeneous of degree (α + β).
- F[f(K, L)] is also homogeneous, but of degree p(α + β) (where p is the power of the transformation).
However, a homothetic function is not necessarily homogeneous. All homogeneous functions are homothetic, but not all homothetic functions are homogeneous.
⭐ Key Takeaways
The degree of homogeneity in a Cobb-Douglas function determines its returns to scale: sum of exponents equals 1 for CRS, more than 1 for IRS, and less than 1 for DRS. For a least-cost combination, the optimal capital-labor ratio is derived by equating the marginal rate of technical substitution to the input price ratio. The expansion path for a homogeneous production function is a straight line from the origin, showing a fixed input ratio at all output levels. Finally, a homothetic function, which is a monotonic transformation of a homogeneous function, retains a linear expansion path but is not necessarily homogeneous itself.
🧠 Quick Revision Questions
- How do you determine if a Cobb-Douglas production function with three inputs (K, L, M) exhibits increasing, constant, or decreasing returns to scale?
- What is the formula for the least-cost input ratio (K/L) in a two-input Cobb-Douglas production function?
- What shape is the expansion path for a homogeneous production function, and why?
- What is a homothetic function, and how is it related to a homogeneous function?
- If a Cobb-Douglas function f(K, L) = K⁰·⁵L⁰·⁵ is squared, is the new function homogeneous? If so, what is its degree of homogeneity?
📘 Lecture 44 — CES Production Function
📖 Overview: This lecture introduces the Constant Elasticity of Substitution (CES) production function as a more general form of production functions, of which the Cobb-Douglas is a specific case. It examines the function's homogeneity properties, its returns to scale characteristics, and derives the marginal products of capital and labor.
🗂️ Topics Covered
The lecture begins by introducing the CES production function attributed to Solow, Minhas, Arrow and Chenery (SMAC), along with its parameters. It then examines the homogeneity of the CES function and how the degree of homogeneity determines returns to scale. Finally, the lecture derives the marginal products of capital and labor from the CES production function.
📝 Lecture Summary
TOPIC 217: INTRODUCING CES PRODUCTION FUNCTION
The CES production function is a relatively new and more general form of production functions, attributed to Solow, Minhas, Arrow, and Chenery, also known as the SMAC production function. The Cobb-Douglas production function can be a specific case of it. The standard form is:
📐 Formula: Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ)
Where:
A= Efficiency parameterδ= Distribution parameter (0 < δ < 1)ρ= Substitution parameter (ρ ≥ -1)μ= Degree of homogeneity
The CES function generates a constant (not variable) value of elasticity of substitution (σ) between capital and labor, hence its name. The elasticity of substitution formula is:
📐 Formula: σ = 1/(1+ρ)
Where:
δis the distribution parameter1-δis the distribution parameter for laborμis the degree of homogeneityμ = 1implies Constant Returns to Scale (CRS)
📌 Example: As σ changes, σ = 0.71 → 1 implies less elastic – lower substitution between capital and labor.
TOPIC 218: HOMOGENEITY OF CES PRODUCTION FUNCTION
Starting from the CES production function Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ), for simplicity let Q = [δK^(-ρ) + (1-δ)L^(-ρ)]^(-1/ρ). This can be written as Q = (Q_0)^(-1/ρ) where Q_0 = δK^(-ρ) + (1-δ)L^(-ρ).
Introducing J on both sides yields Q(JK, JL) = A[δ(JK)^(-ρ) + (1-δ)(JL)^(-ρ)]^(-μ/ρ).
This simplifies to Q(JK, JL) = J^μ * A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ) = J^μ * Q(K, L).
As μ is the degree of homogeneity, the CES production function is homogenous of degree μ. The value of μ can be any positive number:
- If
μ < 1: implies decreasing returns to scale —Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ)withμ < 1. - If
μ = 1: implies constant returns to scale —Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-1/ρ). - If
μ > 1: implies increasing returns to scale —Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ)withμ > 1.
💡 Why this matters: The homogeneity property allows us to quickly determine returns to scale from the parameter μ.
TOPIC 219: MARGINAL PRODUCTS OF CES PRODUCTION FUNCTION
For the CES production function Q = A[δK^(-ρ) + (1-δ)L^(-ρ)]^(-μ/ρ), where δ is the distribution parameter:
🔑 Definition — Marginal product of capital (MP_K): The additional output produced by using one more unit of capital, holding labor constant.
📐 Formula: MP_K = ∂Q/∂K = μ A^(-ρ/μ) δ Q^((ρ/μ)+1) K^(-(ρ+1))
🔑 Definition — Marginal product of labor (MP_L): The additional output produced by using one more unit of labor, holding capital constant.
📐 Formula: MP_L = ∂Q/∂L = μ A^(-ρ/μ) (1-δ) Q^((ρ/μ)+1) L^(-(ρ+1))
⭐ Key Takeaways
The CES production function is a general form that includes the Cobb-Douglas as a specific case and generates constant elasticity of substitution between capital and labor. The function is homogenous of degree μ, where μ determines the returns to scale: μ<1 indicates decreasing, μ=1 constant, and μ>1 increasing returns to scale. The marginal products of capital and labor involve the parameters A, δ, ρ, μ, and the output level Q. The elasticity of substitution parameter σ is calculated as 1/(1+ρ).
🧠 Quick Revision Questions
- What does CES stand for and what is the formula for the elasticity of substitution (σ) in terms of ρ?
- How does the degree of homogeneity μ determine returns to scale in the CES production function?
- What are the formulas for the marginal product of capital (MPK) and marginal product of labor (MPL)?
- Why is the Cobb-Douglas production function considered a specific case of the CES function?
- What does σ = 0.71 imply about the elasticity of substitution between capital and labor?
📘 Lecture 45 — CES Production Function (Continued)
📖 Overview: This lecture continues the analysis of the CES production function, focusing on the shares of labor and capital in output, the application of Euler’s theorem to verify constant returns to scale, and a numerical example demonstrating how to calculate output and verify Euler's theorem empirically.
🗂️ Topics Covered
This lecture is divided into three topics. First, it derives the share of labor and capital in the CES production function. Second, it connects the CES production function to Euler's theorem for verifying constant returns to scale. Third, it provides a numerical calculation to demonstrate the function’s properties in practice.
📝 Lecture Summary
TOPIC 220: SHARE OF LABOR AND CAPITAL IN CES PRODUCTION FUNCTION
The CES production function is given by:
Q = A [δ K^{-ρ} + (1-δ) L^{-ρ}]^{-ν/ρ}
Where, δ is the distribution parameter, ρ is the substitution parameter, ν is the degree of homogeneity, and A is efficiency.
Let capital share of output = (∂Q/∂K) * (K/Q).
Let labor share of output = (∂Q/∂L) * (L/Q).
For the CES function, these shares depend on the substitution parameter ρ and the distribution parameter δ.
TOPIC 221: CES PRODUCTION FUNCTION AND EULER'S THEOREM
The CES production function is: Q = A [δ K^{-ρ} + (1-δ) L^{-ρ}]^{-ν/ρ}
Where ν is the degree of homogeneity (ν=1 for constant returns to scale).
Euler's theorem requires: (∂Q/∂K) K + (∂Q/∂L) L = νQ
The lecture derives the marginal products:
MP_K = ∂Q/∂K = (ν δ K^{-(ρ+1)} Q) / (δ K^{-ρ} + (1-δ) L^{-ρ})
MP_L = ∂Q/∂L = (ν (1-δ) L^{-(ρ+1)} Q) / (δ K^{-ρ} + (1-δ) L^{-ρ})
🔑 Definition — Euler's Theorem for homogeneous functions: A function f(x,y) is homogeneous of degree r if f(tx,ty) = t^r f(x,y), and Euler's theorem states that x·(∂f/∂x) + y·(∂f/∂y) = r·f(x,y).
📐 Formula: (∂Q/∂K) K + (∂Q/∂L) L = νQ
→ This means the sum of capital's and labor's marginal products weighted by their quantities equals the degree of homogeneity times total output.
📌 Example: When ν=1 (constant returns to scale), Euler's theorem implies (∂Q/∂K)K + (∂Q/∂L)L = Q, meaning total output is exactly exhausted by factor payments.
TOPIC 222: NUMERICAL CES PRODUCTION FUNCTION CALCULATION
This section provides a worked numerical example. Given: A=1, ν=1, δ=0.6, ρ=2, K=10, L=20.
Step 1: Calculate K^{-ρ} = 10^{-2} = 0.01
Step 2: Calculate L^{-ρ} = 20^{-2} = 0.0025
Step 3: δ K^{-ρ} = 0.6(0.01) = 0.006
Step 4: (1-δ) L^{-ρ} = 0.4(0.0025) = 0.001
Step 5: Sum = 0.006 + 0.001 = 0.007
Step 6: -ν/ρ = -1/2 = -0.5
Step 7: Q = 1 * (0.007)^{-0.5} = 1 / √0.007 ≈ 11.95 units
Next, compute MP_K and MP_L:
MP_K = (ν δ K^{-(ρ+1)} Q) / (δ K^{-ρ} + (1-δ) L^{-ρ}) = (1)(0.6)(10^{-3})(11.95)/0.007
MP_K = (0.6 * 0.001 * 11.95) / 0.007 = 0.00717 / 0.007 ≈ 1.0243
MP_L = (ν (1-δ) L^{-(ρ+1)} Q) / (δ K^{-ρ} + (1-δ) L^{-ρ}) = (1)(0.4)(20^{-3})(11.95)/0.007
MP_L = (0.4 * 0.000125 * 11.95) / 0.007 = 0.0005975 / 0.007 ≈ 0.08536
Verify Euler's theorem: MP_K * K + MP_L * L = 1.0243(10) + 0.08536(20) = 10.243 + 1.7072 ≈ 11.95 = Q ✓
⭐ Key Takeaways
The CES production function allows the share of output going to labor and capital to vary with the substitution parameter ρ. Euler's theorem confirms that when ν=1, total output equals the sum of factor payments at marginal productivity rates. The numerical example demonstrates that for constant returns to scale CES function, the marginal products correctly exhaust total product, making it consistent with neoclassical distribution theory. The capital share depends on δ and ρ, with higher ρ reducing substitutability between factors.
🧠 Quick Revision Questions
- What is the formula for capital share of output in the CES production function?
- State Euler's theorem for a production function homogeneous of degree ν.
- In the CES function Q = A[δK^{-ρ} + (1-δ)L^{-ρ}]^{-ν/ρ}, what does the parameter ρ represent?
- In the numerical example with K=10, L=20, δ=0.6, ρ=2, and ν=1, verify that capital's share (MP_KK/Q) adds with labor's share (MP_LL/Q) to exactly 1.
- What happens to factor shares as ρ approaches 0 in the CES function?