- MACRO ECONOMICS Chapter 3 Bernanke
- Assig6 Student
- Industrial Engineering
- Production Lecture
- YarlagaddaSandeep-Production-Ch4r.pdf
- Chicago UGM-8_ANSYS Conference - 08292011
- CEE218_Project3
- Algoritma Multi Objective
- Martin 2016 Working Paper
- mehran.pdf
- Dlg Are Final
- Ariba User Conference Training
- v
- aiq
- Economics of Production
- Term paper on Intelligent Ship Arrangements
- M1L3_LN
- Assign 2
- Sgm With Ces Tech (Jae)
- 1-s2.0-S0016003206000329-main
- Evolver7 En
- 4
- A Population Dynamics Approach for the Dispatch of Distributed Generators
- Ac 33155161
- Facility Layout
- Selsp Queueing System
- TEACHING AND LEARNING BASED OPTIMISATION
- Morphing 09
- Ansa for Lsopt
- Nr310206 Optimization Techniques
- Sapiens: A Brief History of Humankind
- The Unwinding: An Inner History of the New America
- Yes Please
- The Innovators: How a Group of Hackers, Geniuses, and Geeks Created the Digital Revolution
- Dispatches from Pluto: Lost and Found in the Mississippi Delta
- Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future
- John Adams
- Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America
- The Prize: The Epic Quest for Oil, Money & Power
- A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story
- Grand Pursuit: The Story of Economic Genius
- This Changes Everything: Capitalism vs. The Climate
- The Emperor of All Maladies: A Biography of Cancer
- Team of Rivals: The Political Genius of Abraham Lincoln
- The New Confessions of an Economic Hit Man
- Rise of ISIS: A Threat We Can't Ignore
- Smart People Should Build Things: How to Restore Our Culture of Achievement, Build a Path for Entrepreneurs, and Create New Jobs in America
- The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers
- The World Is Flat 3.0: A Brief History of the Twenty-first Century
- Bad Feminist: Essays
- How To Win Friends and Influence People
- Angela's Ashes: A Memoir
- Steve Jobs
- Leaving Berlin: A Novel
- The Silver Linings Playbook: A Novel
- The Sympathizer: A Novel (Pulitzer Prize for Fiction)
- Extremely Loud and Incredibly Close: A Novel
- The Light Between Oceans: A Novel
- The Incarnations: A Novel
- You Too Can Have a Body Like Mine: A Novel
- Life of Pi
- The Love Affairs of Nathaniel P.: A Novel
- We Are Not Ourselves: A Novel
- The First Bad Man: A Novel
- The Rosie Project: A Novel
- The Blazing World: A Novel
- Brooklyn: A Novel
- The Flamethrowers: A Novel
- A Man Called Ove: A Novel
- Bel Canto
- The Master
- Interpreter of Maladies
- Beautiful Ruins: A Novel
- The Kitchen House: A Novel
- The Art of Racing in the Rain: A Novel
- Wolf Hall: A Novel
- The Wallcreeper
- A Prayer for Owen Meany: A Novel
- The Cider House Rules
- The Perks of Being a Wallflower
- The Bonfire of the Vanities: A Novel
- Lovers at the Chameleon Club, Paris 1932: A Novel
- Little Bee: A Novel

ISSN 1657-7191 (Edición Electrónica)

MAYO DE 2006

**CEDE A CHARACTERIZATION OF HOMOGENEOUS PRODUCTION FUNCTIONS USING THE RATIO OF AVERAGE TO MARGINAL COSTS1
**

PIETRO BONALDI HERNÁN VALLEJO

Abstract This paper states a theorem that characterizes homogeneous production functions in terms of the ratio of average to marginal costs. The theorem claims that a production function is homogeneous of degree k if and only if the ratio of average costs to marginal costs is constant and equal to k. In order to prove the theorem two lemmas -with theoretical value of their own– are demonstrated before hand: the first one establishes that a production function is homogeneous of degree k if and only if its elasticity of scale is k; the second one determines the conditions on the production function under which any input vector can be an optimum, for some choice of the price vector and the level of production. Keywords: Elasticity of scale, homogeneous production functions, returns to scale, average costs, and marginal costs. JEL Classification: D24

1

We are grateful to the students of the Advanced Microeconomics course that Vallejo teaches at the Department of Economics of Universidad de los Andes for their questions –and their criticisms-, some of which motivated this paper. All remaining errors are ours. Contact addresses j-bonald@uniandes.edu.co or hvallejo@uniandes.edu.co.

algunas de las cuales motivaron este documento.edu. retornos a escala. se prueban de antemano dos lemas –con valor teórico propio-: el primero establece que una función de producción es homogénea de grado k sí y sólo si su elasticidad de escala es k.edu. costos medios y costos marginales.co o hvallejo@uniandes. Palabras clave: Elasticidad de escala. el segundo teorema establece las condiciones que debe cumplir una función de producción de tal forma que cualquier vector de insumos pueda ser un óptimo.co. Con el fin de demostrar el teorema. Clasificación JEL: D24 2 Agradecemos a los estudiantes del curso de Microeconomía Avanzada que Vallejo enseña en la Facultad de Economía de la Universidad de los Andes por sus preguntas –y sus críticas-.UNA CARACTERIZACIÓN DE LAS FUNCIONES DE PRODUCCIÓN HOMOGÉNEAS UTILIZANDO LA RAZÓN ENTRE LOS COSTOS MEDIOS Y LOS COSTOS MARGINALES2 Resumen Este documento enuncia un teorema que caracteriza a las funciones de producción homogéneas en términos de la razón entre los costos medios y los costos marginales. 2 . funciones de producción homogéneas. Todos los errores que quedan son nuestros. para algún vector de precios y un nivel dado de producción. Direcciones de contacto j-bonald@uniandes. El teorema establece que una función es homogénea de grado k si y sólo si la razón entre los costos medios y los costos marginales es constante e igual a k.

i.e. Another well known result on homogeneous functions is the Euler’s Theorem -also called the adding up theorem-. Some. kf ( x ) = ∑ i =1 n ∂f ( x ) ∂xi xi . exhibit decreasing. a homogenous function has decreasing. that the only production functions for which the elasticity of scale is 3 . Although the converse is also true. such as the elasticity of scale. less. constant or increasing returns to scale if its degree of homogeneity is. he points out that for any homogeneous function of degree k. it should not be surprising that a local measure of returns to scale takes the same value for every input vector in its domain. In particular. But it is not obvious that the converse also holds. It should be noticed that many functions representing technologies are not as wellbehaved as the homogeneous functions. the elasticity of scale is k at all input vectors. Since every homogeneous function has some kind of global returns to scale. for instance. As an initial result. That is why a local measure of returns to scale is needed. respectively.. equal or greater than 1. Hanoch (1975) makes a comparison between different concepts of returns to scale leading to the same measure of elasticity of scale. this is not considered in most of the textbooks. INTRODUCTION Almost every textbook in Microeconomic Theory includes a section on homogeneous functions. (2000) present a demonstration of both the theorem and its converse. which states that If f ( x ) is homogeneous of degree k then. As an exception to this peculiar omission Jehle et al. constant and increasing returns to scale for different choices of the input vector. and presents some well know relations between (global) returns to scale and the degree of homogeneity of the production function.I. at every optimum input vector where costs are minimized. for all x.

the converse does not follow directly. III. some usual conditions will be imposed on the production function gradually. established by Frisch (1965). since the result stated by Frisch does not tell us anything about the value of the elasticity of scale outside the optimum. marginal costs and average costs [for instance. in particular at the optimum. What follows is a detailed proof of these results. as they are needed to prove the lemmas required to demonstrate the 4 . which Frisch called the passus coefficient. As it will become clear later. the elasticity of scale will be k at all input vectors. nor the level of output. states that the elasticity of scale of a production function. Since technology can be described by production functions or by cost functions. Formally. A well known theorem. a geometrical approach can be found in Mas-Colell et al. Again. So. equals the ratio of average cost to marginal cost. and so the production function is homogeneous of degree k. where the amount of any input used is not allowed to be negative. some natural conditions can be imposed on the production function in order to guarantee that any input vector could be an optimum. evaluated at the optimum (at the input vector that minimizes costs). depending on the input prices and the level of production. In this paper. if the production function is homogeneous of degree k. an so it will be the ratio between average and marginal costs. if the ratio between average and marginal costs is a constant k. This implies that. many authors have studied the relationship between returns to scale. this result follows directly from the converse of Euler’s theorem. f : n + → + and y = f ( x ) . THEORETICAL FRAMEWORK A production function shows the maximum level of output y that can be achieved with an input vector x. 1995]. However.constant are the homogeneous functions. then the elasticity of scale equals k everywhere.

theorem stating the central result under consideration. but there could still be points over the axis.1. f ( 0 ) = 0 1. if x 0 x1 then f ( x0 ) > f ( x1 ) Elasticity of scale and homogeneity The elasticity of scale is a local measure of returns to scale. consequently. where f ( x ) = 0 . in percentage terms. A common way to define it is as in Varian (1992) ⎡⎛ df (tx) ⎞ ⎛ t ⎞ ⎤ e( x ) = ⎢⎜ ⎟⎥ ⎟⎜ ⎣⎝ dt ⎠ ⎝ f (tx) ⎠ ⎦ t =1 (1) where: e ( x ) is the elasticity of scale f is the production function t is the scale factor. restrict the definition of the elasticity of scale to the domain D. f is strictly increasing: If x 0 ≥ x1 then f ( x0 ) ≥ f ( x1 ) and. when all inputs are changed in one percent. normalized to 1 By assumptions 1. So we define D as the subset of n + where f ( x ) ≠ 0 and.2 the production function is never zero on n ++ .2. 5 . For now. Defined at a point x . different from 0. let us assume that the production function satisfies: 1.1 and 1. To illustrate this restriction. it shows the change in output.

the first part of the lemma has been verified. In such case.α β consider a two input Cobb-Douglas production function f ( x1 . Lemma 1: A production function -satisfying certain conditions. 1. 6 . A production function is homogeneous of degree k if. and different from 0 elsewhere. if one of the inputs is set to 0.2) is homogenous of degree k. its elasticity of scale is k. we can state and prove the following result.is homogeneous of degree k if and only if the elasticity of scale is k. f ( tx ) = t k f ( x ) (2) Given these definitions.1. for all t > 0. the elasticity of scale would not be defined only if any of the inputs is 0. where the total output is 0. Consider one direction of the lemma: If a production function (satisfying assumptions 1. x2 ) = Ax1 x2 . Proof: Note first that df ( tx ) dt = kt k −1 f ( x ) (3) Replacing (2) and (3) in equation (1) ⎡ ⎛ t ⎞⎤ e(x) = ⎢ kt k −1 f ( x ) ⎜ k ⎟⎥ ⎜ t f ( x ) ⎟⎥ ⎢ ⎝ ⎠ ⎦ t =1 ⎣ ( ) e( x) = k ∀x ∈ D Thus.

we know that the elasticity of scale is a constant k for all x ∈ D (where it is defined). but we want to conclude that the production function is homogeneous of degree k in its entire domain n +. There 7 . By hypothesis. consider the following questions that can be found in some textbooks. 21] b. Constant Elasticity of Substitution production function.. Note however.. p. that this assumption is not sufficient. p.. Perfect substitutes production function.. f ( x ) = A∏ xiα i [Varian (1992). The production function is C1 (it has continuous first order derivatives in its entire domain).As an illustration of the usefulness of this result. the same: e(x) = k . since we have an additional problem to deal with. f ( x ) = ∑ ai xi i =1 n d. f ( x ) = min [ a1 x1 . But the first part of the lemma implies that the answer to all of them is. 17] i =1 n c. so. trivially. Perfect complements production function.3. What is the elasticity of scale of the: ⎛ n ⎞ρ a. the degree of homogeneity of the production function. we assume further that: 1. to ensure that the partial derivatives exist and are continuous. Cobb Douglas production function. To prove the second part of lemma 1 we will take the partial derivatives of the production function with respect to all its inputs. f ( x ) = ⎜ ∑ xiρ ⎟ [Varian ⎝ i =1 ⎠ ε (1992). an xn ] All of the above questions are useful as ways of applying the definition of elasticity of scale with specific production functions that have the same returns to scale for all possible values of x.

4 implies assumption 1.5.2. The domain of f is n ++ (instead of n + ).3. df ( tx ) dt =∑ i =1 n ∂f ( tx ) ∂xi xi Evaluating at t = 1. the production function is homogeneous of degree k (assuming conditions 1.4. Furthermore.1 – 1. Any of these. 8 . f is strongly increasing: if x0 ≥ x1 and x0 ≠ x1 then f ( x0 ) > f ( x1 ) 1.are various ways to avoid this difficulty. ⎛ df ( tx ) ⎞ ⎛ t ⎞ e (x) = ⎜ ⎟ =k ⎟⎜ ⎜ ⎟ ⎝ dt ⎠ ⎝ f ( tx ) ⎠ t =1 For fixed x ∈ D . Proof: Let the elasticity of scale be a constant k.6 is the one that holds. if 1. f ( x1 )} 1. D = else D = n + \ {0} .1) : if x = tx0 + (1 − t ) x1 then f ( x ) > min { f ( x0 ) . n ++ . consider the following tentative assumptions on the production function f: 1. In particular. For example.3 and 1.4. Then.63). f is strictly quasiconcave: ∀t ∈ ( 0. 1. We can prove now the other direction of lemma 1: If the elasticity of scale of a production function is a constant k. for all x ∈ D . can be used to prove lemma 1.5 separately imply that f ( x ) = 0 if and only if x = 0 . given assumption 1.4 or 1.5 or 1. taken alone. 1.6. we get df ( x ) dt =∑ i =1 n ∂f ( x ) ∂xi xi 3 Note that assumption 1.

for all x [Jehle et al.To obtain k we must multiply by t evaluated at t = 1. Optimality and production functions The price of the n inputs used by a typical firm with production function f: n + → + can be represented by a price vector w ∈ n ++ . if the production function has a constant elasticity of scale k for all x ∈ D . However. instead of 1. the total costs faced by the firm are x ⋅ w . which gives f ( tx ) ⎛ n ∂f ( x ) ⎞ ⎛ 1 ⎞ k = ⎜∑ xi ⎟ ⎜ ⎟ ⎜ ⎟ ⎝ i =1 ∂xi ⎠ ⎝ f (x) ⎠ kf ( x ) = ∑ i =1 n ∂f ( x ) ∂xi xi Finally. p. and the result is already proved. a profit maximizing firm will solve the problem: 9 .4 or 1. In this case f ( x ) = 0 and so. and thus. This completes the proof of lemma 1..e. then there is a case we have not considered yet. (4) Note that if assumption 1. if we assume 1. D = n ++ as f is strictly increasing.5. 471]. the homogeneity of the function f follows directly from Euler’s Theorem which states that f is homogeneous of degree k if and only if ∂f ( x ) ∂xi kf ( x ) = ∑ i =1 n xi . Given a price vector w and a level of production y. x = 0 . But then both sides of equation (4) equal cero and the equality holds trivially. i.6.6. Thus. holds. x ∉ D . it is homogeneous of degree k in all of its domain. (2000). as stated previously.

10 .1]: if x = tx0 + (1 − t ) x1 then f ( x ) ≥ min { f ( x0 ) . From now on.6 then. the cost function.3.5. for any n x ∈ R+ \ {0} there exist n w ∈ R+ + and y>0 such that x ∈ Argmin x ⋅ w s...1.4. both of these conditions are usually imposed on the production function to derive most of its desirable properties.6 The marginal productivities are positive: ∀x ∈ n + : ∂f ( x ) ∂xi > 0.n Lemma 2: If the production function satisfies 1.min x ⋅ w s.5(b) and 1.t. However.4 or 1. 1.5(b) The production function is quasiconcave: ∀t ∈ [ 0. 1. f ( x ) ≥ y x∈ n + Correspondingly. f ( x ) ≥ y .4 and we will suppose further that: 1. which expresses the minimum expenditure needed to achieve a level of production y at input prices w . y ) = min x ⋅ w : x ∈ { n + ∧ f (x) ≥ y } Previously we showed that lemma 1 can be demonstrated using either assumption 1. 1.. we are going to make weaker assumptinons: we will keep assumption 1. f ( x1 )} 1. i = 1.t.. is usually defined as: c ( w.

w ∗ ∈ R+ + . Now. for any x and x’ such that c(x') < c(x ) ...Let x ∗ n ∈ R+ \ {0} .n λ ( f (x ) − y ∗ ) = 0 ..3] if: (1) There are no equality constraints (2) The inequality constraint is given by a quasiconcave function. Note first that the objective function is c(x ) = x ⋅ w ∗ . with equality if x i > 0 . since for n all i. x∗ meets the necessary conditions for an optimum.e.n . where λ ≥ 0 1 It can be easily verified. whenever c(x') < c(x ) 11 . (1995) [Theorem M.. f x ( ) ∗ ⎛ f x∗ ⎜ 1. n = y and w = ⎜ f1 x∗ ⎝ ∗ ∗ ( ) ⎞ . ⎟ ∂x ( )⎟ ⎠ ∗ ∗ i i Given that f is strongly increasing and f ( 0 ) = 0 .. then y ∗ = f x ∗ > 0 and. and ∇c ( x ) = w ∗ .. where f (x ) = ∂f (x ) . replacing in (a) and (b).K. as proved in Mas-Collel et al. To complete this demonstration we must prove that x∗ also satisfies the sufficient conditions for a global minimizer of the problem considered. for all i = 1. i = 1. ( ) ( ) Now consider the problem min x ⋅ w ∗ s. and (3) The objective function satisfies ∇c(x ) ⋅ (x'− x ) < 0.. So..t. i.. The Kuhn Tucker conditions x∈ n + for an optimum (a local minimizer) of this problem are: (a) (b) wi∗ ≥ λ fi ( x ) .. and then. for all x . ∇c(x ) ⋅ (x'−x ) < 0 . f ( x ) ≥ y∗ . that x∗ and λ ∗ = f1 x∗ ( ) > 0 satisfy these conditions. c ( x ) = ∇c ( x ) ⋅ x ... f i x ∗ > 0 ..

3 1. 1.5(b) and 1. x∗ is a global minimizer. x∗ is a solution to the minimization problem given. y ) (5) Using this theorem and lemmas 1 and 2.4. that satisfies 1. Homogeneity and costs Starting from the cost function.Then. and this completes the proof of lemma 2. the following statement can be proved. if x∗ satisfies the Kunh-Tucker conditions. at the optimum vector x∗ that minimizes costs: e ( x∗ ) = AC (w. y ) ∂y As mentioned before. y ) y ∂c(w . y ) = c (w. is homogeneous of degree k if and only if the ratio of average costs to marginal costs is constant and equal to k. the objective function and x∗ meet the conditions required. as it was previously defined. y ) = Marginal cost: MC (w .6. Frisch (1965) proved a theorem relating average and marginal costs with the elasticity of scale which states that. Theorem: A production function. 1. y ) MC (w. Average cost: AC (w. 12 .1. two concepts that are commonly used to describe the behavior of the firm are the average and marginal costs. As the constraint.

and then. by (5). Andrew. 3. e(x) = k . 1995. 13 . Whinston and Jerry R. y ) = k . pp. Michael D.t. V. y ) . there exist w ∈ R+ + and y > 0 such that x ∈ Argmin x ⋅ w s. New York : Addison Wesley. Vol. 471. y ≤ f ( x ) . MC (w. 21. at the optimum x∗ that minimizes costs with prices w and output y. and so AC (w. Holland: D. Jehle. 3d Edition. y ) Now suppose that the ratio of average costs to marginal costs is constant and n n equal to k. 2nd ed. so x is an optimum that minimizes costs at (w. Hanoch. in particular. REFERENCES Frisch. New York: Oxford University Press. Green. This completes the demonstration of the theorem. Geoffrey A. p. 2000. n As x ∈ R+ \ {0} was arbitrarily chosen. Norton. (1975) “The Elasticity of Scale and the Shape of Average Costs” The American Economic Review. the last result holds for all x. Reidel Publishing Co. Boston . using lemma 1. 65. and Philip J.Proof: If the production function is homogeneous of degree k. p. Mas-Collel.. Hal (1992) Microeconomic Analysis. it follows that the production function is homogeneous of degree k. No. y ) and. then by lemma 1 e(x) = k everywhere. then by lemma 2. Microeconomic Theory. 492-497. by (5). Reny (2000) Advanced Microeconomic Theory. Let x ∈ R+ \ {0} . 17 and p. Ragnar (1965) Theory of Production. G. for all (w. Varian. 1965.

- MACRO ECONOMICS Chapter 3 BernankeUploaded bysirfanalizaidi
- Assig6 StudentUploaded byStaticoce
- Industrial EngineeringUploaded byJoel Dsouza
- Production LectureUploaded bykiranvja
- YarlagaddaSandeep-Production-Ch4r.pdfUploaded byDiego Fernando González
- Chicago UGM-8_ANSYS Conference - 08292011Uploaded byKi Chun
- CEE218_Project3Uploaded bykaren dejo
- Algoritma Multi ObjectiveUploaded byekatsukimath
- Martin 2016 Working PaperUploaded byBains M Collinz
- mehran.pdfUploaded byMehran
- Dlg Are FinalUploaded byAnonymous kqK4X9ZgGm
- Ariba User Conference TrainingUploaded byyankees1
- vUploaded byPeter Mikhail
- aiqUploaded byAbrar Malik
- Economics of ProductionUploaded byprasadparulekar
- Term paper on Intelligent Ship ArrangementsUploaded byNikhil Ambattuparambil Gopi
- M1L3_LNUploaded byyotaladro
- Assign 2Uploaded byAshok Yogi
- Sgm With Ces Tech (Jae)Uploaded bySlava Yurkov
- 1-s2.0-S0016003206000329-mainUploaded byHuynh Quoc Viet
- Evolver7 EnUploaded byLihet Pasere Andrea Anca
- 4Uploaded byapi-3833671
- A Population Dynamics Approach for the Dispatch of Distributed GeneratorsUploaded byGhaith A. Abdul-Rahim
- Ac 33155161Uploaded byAnonymous 7VPPkWS8O
- Facility LayoutUploaded byamitnd55
- Selsp Queueing SystemUploaded byujangketul62
- TEACHING AND LEARNING BASED OPTIMISATIONUploaded byuday wankar
- Morphing 09Uploaded byPatrick Perez
- Ansa for LsoptUploaded byjoshimohil
- Nr310206 Optimization TechniquesUploaded bySRINIVASA RAO GANTA