11. Cost minimization Econ 494 Spring 2013.

Slides:



Advertisements
Similar presentations
Nonlinear Programming McCarl and Spreen Chapter 12.
Advertisements

1 Chapter 9 PROFIT MAXIMIZATION Copyright ©2005 by South-Western, a division of Thomson Learning. All rights reserved.
Optimization using Calculus
THE MATHEMATICS OF OPTIMIZATION
Maximization without Calculus Not all economic maximization problems can be solved using calculus –If a manager does not know the profit function, but.
Constrained Maximization
6 © 2004 Prentice Hall Business PublishingPrinciples of Economics, 7/eKarl Case, Ray Fair The Production Process: The Behavior of Profit-Maximizing Firms.
Economics D10-1: Lecture 12 Cost minimization and the cost function.
Expenditure Minimization
THE MATHEMATICS OF OPTIMIZATION
Definition and Properties of the Cost Function
Definition and Properties of the Production Function Lecture II.
1 13. Expenditure minimization Econ 494 Spring 2013.
The Production Process: The Behavior of Profit-Maximizing Firms
Applied Economics for Business Management
The Firm: Optimisation
Applied Economics for Business Management
6. Optimization with 2 variables: Price discrimination Econ 494 Spring 2013 See Handout 5 for general case with two variables.
THE MATHEMATICS OF OPTIMIZATION
Lecture #7. Lecture Outline Review Go over Exam #1 Continue production economic theory.
Optimization Techniques Methods for maximizing or minimizing an objective function Examples –Consumers maximize utility by purchasing an optimal combination.
Managerial Economics Managerial Economics = economic theory + mathematical eco + statistical analysis.
Microeconomia Corso E John Hey. Compito a casa/Homework CES technology with parameters c 1 =0.4, c 2 =0.5, ρ=0.9 and s=1.0. The production function: y.
10.1 Chapter 10 –Theory of Production and Cost in the Long Run(LR)  The theory of production in the LR provides the theoretical basis for firm decision-making.
1 Cost Minimization and Cost Curves Beattie, Taylor, and Watts Sections: 3.1a, 3.2a-b, 4.1.
8. Production functions Econ 494 Spring 2013
Short-run Production Function
Ch 4 THE THEORY OF PRODUCTION
Chapter 6 Production. ©2005 Pearson Education, Inc. Chapter 62 Topics to be Discussed The Technology of Production Production with One Variable Input.
1 Chapter 9 PROFIT MAXIMIZATION Copyright ©2005 by South-Western, a division of Thomson Learning. All rights reserved.
6.1 Chapter 7 – The Theory of Consumer Behavior  The Theory of Consumer behavior provides the theoretical basis for buyer decision- making and the foundation.
INCOME AND SUBSTITUTION EFFECTS
Profit Maximization Single variable unconstrained optimization Econ 494 Spring 2013 Hands Ch
The Supply Curve and the Behavior of Firms
WEEK 8 SYSTEMS OF EQUATIONS DETERMINANTS AND CRAMER’S RULE.
Managerial Economics Prof. M. El-Sakka CBA. Kuwait University Managerial Economics in a Global Economy Chapter 2 Optimization Techniques and New Management.
Microeconomics 2 John Hey. Health Warning There is a LOT of detail in the Maple html file. Do not be swamped by this detail. Stand back and try and understand.
ECONOMICS OF OPTIMAL INPUT USE AAE 575 Paul D. Mitchell.
Comparative Statics and Duality of the Cost Function Lecture VII.
7. Concavity and convexity
Applied Economics for Business Management Lecture #8.
1 Profit Maximization Molly W. Dahl Georgetown University Econ 101 – Spring 2009.
12. Consumer Theory Econ 494 Spring 2013.
Managerial Economics Managerial Economics = economic theory + mathematical eco + statistical analysis.
Faculty of Economics Optimization Lecture 3 Marco Haan February 28, 2005.
Comparative statics Single variable unconstrained optimization Multiple parameters Econ 494 Spring 2013.
Optimization unconstrained and constrained Calculus part II.
9. Perfect competition Econ 494 Spring 2013 Most of these notes are taken directly from Silb §4.4.
10. Envelope Theorem Econ 494 Spring Agenda Indirect profit function Envelope theorem for general unconstrained optimization problems Theorem Proof.
Microeconomics 2 John Hey. Plan for today 1.We look at the Homework I set at the end of Lecture We will quickly revise the relationships between.
© 2003 McGraw-Hill Ryerson Limited. Production and Cost Analysis I Chapter 9.
Calculus-Based Optimization AGEC 317 Economic Analysis for Agribusiness and Management.
© 2007 Prentice Hall Business Publishing Principles of Economics 8e by Case and Fair Prepared by: Fernando & Yvonn Quijano 7 Chapter The Production Process:
C opyright  2007 by Oxford University Press, Inc. PowerPoint Slides Prepared by Robert F. Brooker, Ph.D.Slide 1 1.
Review Class Seven Producer theory  Key sentence: A representative, or say, typical firm will maximize his profit under the restriction of technology.
Linear Programming Chap 2. The Geometry of LP  In the text, polyhedron is defined as P = { x  R n : Ax  b }. So some of our earlier results should.
Econ 102 SY Lecture 7 Supply and Cost Minimization Sept 2, 2008.
1 Chapter 11 PROFIT MAXIMIZATION. 2 Profit Maximization A profit-maximizing firm chooses both its inputs and its outputs with the sole goal of achieving.
CASE FAIR OSTER ECONOMICS P R I N C I P L E S O F
Production.
Chapter 6 Production.
Short-run Production Function
Principles and Worldwide Applications, 7th Edition
Labor demand in the long run
Choice.
Molly W. Dahl Georgetown University Econ 101 – Spring 2008
7 The Production Process: The Behavior of Profit-Maximizing Firms
Choice.
7 The Production Process: The Behavior of Profit-Maximizing Firms
Presentation transcript:

11. Cost minimization Econ 494 Spring 2013

Agenda Cost minimization Readings Primal approach Envelope theorem for constrained optimization Link to profit maximization Readings Silberberg Chapter 8 Also review constrained optimization

Quick optimization review For the constrained optimization problem: Set up the Lagrangian function: To get correct sign of l, For maximization problems, write the constraints to be non-negative For minimization problems, write the constraints to be non-positive

FONC Note that l is a choice variable, hence there are 3 FONC. If the SOSC are satisfied, you can use the IFT to solve the FONC simultaneously for x1*, x2* and l*.

See Silb table 6-1, p. 138; Chiang table 12.1, p. 362 SOSC With one constraint, the SOSC require that all border-preserving principal minors be: Sign (–1)r, r = 2,…n for a maximum starts with positive second order principal minor All minors of order r will have same sign (no need to check all) Sign is negative for a minimum A border-preserving principal minor of order r is the determinant of the matrix obtained by deleting n-r rows and corresponding columns – but you cannot delete the row/column with the border

Border-preserving principal minors (one constraint) n =2, r = 1  n – r =1  delete one row and column. Sign unknown n =2, r = 2  n – r =0  delete no rows or columns. Max: >0 Min: <0

Question For a profit maximization problem, what must we assume about the shape of the production function? How do you know this? Profit max requires that the production function be strictly concave. This is implied by the SOSC: |H|=f11·f22-(f12)^2>0

Cost minimization We want to determine the properties of a function that specifies the total cost of producing any given level of output. Note that this cost function differs from the “cost function” we used for P-max. How?

Cost minimization To assert the existence of a well-defined cost function, we need a theory of the firm. The cost function depends upon what the firm intends to do, and the constraints they face. The production function is a constraint We cannot directly derive the cost function from the postulate of P-max. (see Silb p. 176) Need y to enter as a parameter, not a variable

Cost minimization Cost functions must be derived from a model with output y as a parameter Clearly, in order to maximize profits, the firm must minimize the cost producing the optimal level of output. This cost min assertion is consistent with P-max

Cost minimization We assert that the firm will minimize the costs of producing a given level of output Costs are simply the sum of all input levels times their respective (fixed) prices Constraint: have to produce some level of output, y, which is related to inputs by the production function

Graphical illustration Choose some combination of x1 and x2 to produce y0 output  Isoquant

Graphical illustration Choose (x1, x2) such that the firm produces y0 at least cost. Remember w1 and w2 are fixed, as is y0 Isocost line – “same cost” There are an infinite number of isocost lines given factor prices w1 and w2, but only one that will produce y0

Graphical illustration Cost is minimized where isocost line is tangent to isoquant

Graphical illustration What if isocost and isoquant were not tangent? Consider: C2 > C0 > C1

Mathematics of cost minimization Set up Lagrangian: Adding zero

Optimality conditions

Evaluate SOSC

Evaluate SOSC (cont.) SHAPE OF PRODUCTION FUNCTION: Note that this result is the same as the slope of the isoquant dx2/dx1. We derived this earlier in the semester when we talked about production functions. SHAPE OF PRODUCTION FUNCTION: Cost min requires quasi-concavity Profit max requires strict concavity (stronger requirement) Isoquants are convex

Conditional factor demands There are 3 parameters in the FONC (y,w1,w2) By the IFT, assuming the SOSC hold, we can solve the FONC simultaneously for the explicit choice functions: x1c(y,w1,w2), x2c (y,w1,w2), lc (y,w1,w2) xic are known as conditional factor demands The factor demands from a cost min problem are conditioned on the level of output y. To distinguish between conditional factor demands and profit-max factor demands, look at the arguments: Conditional factor demands: xic (y,w1,w2) Profit max factor demands: xi* (p,w1,w2)

Interpreting lc Later, using the Envelope Thm, we will see that lc(y,w1,w2) is the marginal cost function From the first two FONC:

lc(y,w1,w2) is marginal cost function At cost min input mix, if the firm were to increase xi “a little bit” Dxi Total cost would rise by wi×Dxi Output would also rise by MPi×Dxi = fi×Dxi Hence l represents the incremental cost of increasing output through the use of xi.

Cost min comparative statics The approach is the same as always. Substitute the explicit choice functions x1c(y,w1,w2), x2c (y,w1,w2), lc (y,w1,w2) back into the FONC to get the identities:

Differentiate wrt w1 If we differentiate the identities with respect to w1, and then express this system of equations using matrix notation, we get:

Use Cramer’s rule to solve

Interpretations ¶ x1c / ¶w1 < 0 Law of demand – but different from P-max Cost min move along isoquant (y fixed) pure substitution effect P-max also move to new isoquant (y varies) substitution and output effects

Interpretations ¶ x2c / ¶w1 > 0 The firm will use more x2 when w1 increases. We already know that x1 will decrease when w1 increases If output is held constant, and x1 decreases, then x2 must increase This result will not necessarily hold if there are more than two inputs.

Graphical illustration (w1 increases)

Interpretations ¶ lc / ¶w1 indeterminate Marginal costs may increase or decrease when a factor price increases May seem counter-intuitive Later, we will show that this depends upon whether x1 is a normal or inferior input See Silb fig 8-11

Comparative statics for y Differentiate identities wrt y and express in matrix notation, apply Cramer’s rule

Interpretation ¶ x1c / ¶y indeterminate If x1c is normal (¶ x1c / ¶y > 0), then MC increase as the price of factor 1 rises (¶ lc / ¶w1 > 0) If x1c is inferior (¶ x1c / ¶y < 0), then MC decreases as the price of factor 1 rises (¶ lc / ¶w1 < 0) Does this mean that an increase in w1 is beneficial for the firm? At least one factor must be normal motivation: see Silb p. 200 “digging ditches” and Figures 8-10 and 8-11 Fig 8-10  when the price of an inferior factor increases, MC decreases, but AC increases. The firm does NOT view a factor price increase as beneficial

Interpretation Look familiar? ¶ lc / ¶y is slope of MC MR $ y p Cost min allows for either, P-max requires ¶lc / ¶y > 0 ¶ lc / ¶y is slope of MC Cost min allows for possibility of U-shaped MC f11f22 – f122 > 0 for P-max MC can only slope up for P-max Cost min is “weaker” than P-max Only require quasi-concave production fctn P-max has stronger restrictions and is not implied by cost min Requires strictly concave production function Some students have asked whether this graph violates the quasi-concavity condition in the SOSC. It is true that this graph is NOT quasi-concave. Does this violate the SOSC? No. The SOSC state that the production function must be quasi-concave, the graph above is the marginal cost function, not the production function.

Envelope Theorem for general constrained problems Consider the two-variable, one constraint problem: By the IFT, assuming the SOSC hold, we can solve the FONC simultaneously for the explicit choice functions x1*(a), x2*(a), l*(a) Form the Lagrangian: FONC

Envelope Theorem for one constraint Define the indirect objective function Substitute the explicit choice functions back into the objective function to get the indirect objective function Note that the indirect objective function, g(x*(a)), by definition incorporates the constraint. x*(a) is the optimal x that solves the constrained optimization problem. So when writing the indirect objective function do NOT make the mistake of saying the indirect obj fctn is “g(x*(a)) subject to h(x*(a)).”

Proof Note that we are differentiating the indirect obj function with respect to a; we are NOT differentiating L evaluated at the optimal solution.

Proof (cont)

Apply ET to cost minimization Define the indirect cost function We already solved the FONC for the explicit choice functions. Substitute these back into the objective function to get the indirect cost function

Envelope Theorem Results (y) By definition, marginal cost is the change in cost when output changes: ¶C* / ¶y  lc = ¶C* / ¶y must be the marginal cost function

Envelope Theorem Results (w1) Derivative of indirect cost function wrt an input price yields conditional factor demand Shephard’s Lemma

Symmetry and reciprocity

Conditional factor demands are HOD(0) in factor prices Multiply factor prices in FONC by t: Clearly, the t will cancel out. Since these FONC, with prices (tw1, tw2) are identical to the FONC with prices (w1, w2), the solutions must also be the same. Thus: xic(tw1, tw2) = xic(w1, w2)

Indirect cost function is HOD(1) in factor prices For MC is HOD(1), use FONC: lambda* = w1 / f(x1*,x2*) and xi* are HOD(0). Stranlund 712 notes, p. 1.23 or Silb p. 213. On your own…show that Avg Cost and Marginal Cost are HOD(1). Silb p. 213

Relationship between Cost min and profit max Recall the profit max problem Instead consider the following: Clearly, a solution to P-max, requires a solution to cost min Lagrangian for cost min problem

Relationship between factor demands from cost min and profit max The solutions to the profit max and cost min FONC are the factor demand functions The profit max factor demand functions: xi*(p, w1, w2) The conditional factor demands from cost min: xic(y, w1, w2) These functions are different. As w1 increases: P-max xi* holds w2 and p constant  movement to new isoquant Cost-min xic holds w2 and y constant  movement along isoquant How can we tie these two demand functions together?

Identity linking cost-min and profit max factors demands Same idea as Le Châtlier Says that the value of the profit maximizing and cost minimizing factor demand functions are the same (when the conditional factor demand is evaluated at the P-max output level). Does not say that the functions are the same

Interpretation Economic interpretation: Alternatively: The value of the profit maximizing demand function is identically equal to the value of the cost minimizing demand function, if the parametric level of output in the cost min problem is taken as the profit maximizing level of supply, given identical input and output prices for each model Alternatively: If, in the cost min model, output is allowed to adjust to price changes so as to maintain a profit max (i.e., adjust via profit maximizing supply response), then the value of profit max and cost min factor demands will be identical

Differentiate wrt w1 Using Young’s theorem and the Envelope theorem for the P-max model, we showed: Now differentiate identity wrt p:

Combine <0 >0

Interpret x1* is more negatively sloped in own price than x1c Multiply last inequality by w1/x1  elasticities P-max factor demands are more price elastic

Graph The cost min and profit max factor demand functions intersect w1 Be careful… Note that this graph has w1 on the vertical axis, so x1* more price elastic means x1* is flatter than x1c. w1 x1 x1*(p, w1, w2) x1c(y, w1 , w2)

Decomposing the effect OUTPUT EFFECT Output responds to changes in w1.  movement to new isoquant SUBSTITUTION EFFECT Holds output constant.  movement along isoquant