www.egwald.com Egwald Web Services

Egwald Web Services
Domain Names
Web Site Design

Egwald Website Search Twitter - Follow Elmer Wiens Radio Podcasts - Geraldos Hour

 

Statistics Programs - Econometrics and Probability Economics - Microeconomics & Macroeconomics Operations Research - Linear Programming and Game Theory Egwald's Mathematics Egwald's Optimal Control
Egwald HomeEconomics Home PageOligopoly/Public Firm ModelRun Oligopoly ModelDerive Oligopoly ModelProduction FunctionsCost FunctionsDuality Production Cost FunctionsGraduate EssaysReferences & Links
 

Egwald Economics: Microeconomics

Duality and the Translog Production / Cost Functions
Non-Homothetic Generalized CES Technology

by

Elmer G. Wiens

Egwald's popular web pages are provided without cost to users.
Follow Elmer Wiens on Twitter: Twitter - Follow Elmer Wiens

Duality: Production / Cost Functions:   Cobb-Douglas Duality | CES Duality | Theory of Duality | Translog Duality - CES | Translog Duality - Generalized CES | References and Links

Cost Functions:   Cobb-Douglas Cost | Normalized Quadratic Cost | Translog Cost | Diewert Cost | Generalized CES-Translog Cost | Generalized CES-Diewert Cost | References and Links

Production Functions:   Cobb-Douglas | CES | Generalized CES | Translog | Diewert | Translog vs Diewert | Diewert vs Translog | Estimate Translog | Estimate Diewert | References and Links

S. Translog (Transcendental Logarithmic) Duality and the Generalized CES Technology

I. Profit (Wealth) Maximizing Firm.

Production and cost functions (and profit functions) can be used to model how a profit (wealth) maximizing firm hires or purchases inputs (factors), such as labour, capital (structures and machinery), and materials and supplies, and combines these inputs through its production process to produce the products (outputs) that the firm sells (supplies) to its customers. The theory of duality links the production function models to the cost function models by way of a minimization or maximization framework. The cost function is derived from the production function by choosing the combination of factor quantities that minimize the cost of producing levels of output at given factor prices. Conversely, the production function is derived from the cost function by calculating the maximum level of output that can be obtained from specified combinations of inputs.

II. The Production and Cost Functions.

The Translog production function:

ln(q) = ln(A) + aL*ln(L) + aK*ln(K) + aM*ln(M) + bLL*ln(L)*ln(L) + bKK*ln(K)*ln(K) + bMM*ln(M)*ln(M)
+ bLK*ln(L)*ln(K) + bLM*ln(L)*ln(M) + bKM*ln(K)*ln(M)   =   f(L, K, M).        

q = exp(f(L, K, M)) = F(L, K, M).

an equation in 10 parameters, A, aL, aK, aM, bLL, bKK, bMM, bLK, bLM, bKM, where L = labour, K = capital, M = materials and supplies, and q = product.

The Translog (total) cost function:

ln(C(q;wL,wK,wM)) = c + cq * ln(q) + cL * ln(wL) + cK * ln(wK) + cM * log(wM)              
                  + .5 * [dqq * ln(q)^2 + dLL * ln(wL)^2 + dKK * ln(wK)^2 + dMM * ln(wM)^2]
                  + .5 * [(dLK + dKL) * ln(wL)*ln(wK) + (dLM + dML) * ln(wL)*ln(wM) + (dKM + dMK) * ln(wK)*log(wM)]
                  + dLq * ln(wL)*ln(q) + dKq * ln(wK)*ln(q) + dMq * ln(wM)*ln(q)
       

an equation in 18 parameters, c, cq, cL, cK, cM, dqq, dLL, dKK, dMM, dLK, dKL, dLM, dML, dKM, dMK, dLq, dKq, and dMq, where wL, wK, and wM are the factor prices of L, K, and M respectively.

III. Duality between Production and Cost Functions.

Mathematically, the duality between a production function, q = F(L, K, M), and a cost function, C(q; wL, wK, wM), is expressed:

C*(q; wL, wK, wM) = minL,K,M{ wL * L + wK * K + wM * M   :   q - F(L,K,M) = 0,   q > 0, wL > 0; wK > 0, and wM > 0,   L > 0, K > 0, M > 0}     (*)

F*(L, K, M) = maxq {q   :   C(q; wL, wK, wM)   <=   wL * L + wK * K + wM * M,   L > 0, K > 0, M > 0,   for all wL >= 0 , wK >= 0, wM >= 0}     (**),

with the questions, F == F*, and C == C*?

The dual functions C*, and F* will be derived from the estimated functions F, and C.

If the functions F(L, K, M), and C(q; wL, wK, wM) obey sufficient conditions, the above minimization and maximization problems can be solved by nonlinear optimization techniques, such as Newton's Method. Furthermore, the implicit function theorem can be exploited to facilitate such calculations.

IV. Constrained Optimization.

The (*) minimization problems are solved for the Generalized CES production function, the Translog production function, and for the Diewert (Generalized Leontief) production function.

The maximization problem (**) is somewhat more difficult than the minimization problem (*), since as specified, the requirements   —   for all wL >= 0 , wK >= 0, wM >= 0   —   imply an infinite number of constraints.

This difficulty is overcome with cost functions that factor:

C(q; wL, wK, wM) = q^1/nu * c(wL, wK, wM)

by solving the reduced dimension problem:

F*(L, K, M)^1/nu = 1 / maxwL,wK,wM {c(wL, wK, wM)   :   wL * L + wK * K + wM * M = 1,   wL >= 0, wK >= 0, wM >= 0)

However, with cost functions that do not factor, such as the Translog cost function, the following method can be used to reduce the dimensionality of the constraints of (**), if the cost function is linear homogeneous in factor prices.

Cost Function Linear Homogeneous in Factor Prices.

When (L, K, M) is the least cost combination of inputs at the specific combination of factor prices (wL, wK, wM):

C(q; wL, wK, wM)   =   wL * L + wK * K + wM * M.

With the cost function, C(q; wL, wK, wM), linear homogeneous in factor prices:

C(q; wL / wM, wK / wM, 1)   =   wL / wM * L + wK / wM * K + M,

assuming that wM > 0. Writing vL = wL / wM, and vK = wK / wM, the maximization problem (**) becomes the minimization problem:

F*(L, K, M) = minq, vL, vK {q   :   C(q; vL, vK, 1)   >=   vL * L + vK * K + M,   L > 0, K > 0, M > 0,   for all vL >= 0 , vK >= 0}     (***),

an optimization problem with one constraint.

V. Cost Function to Production Function: The Method of Langrange.

Using the Method of Lagrange, define the Langrangian function, H, of the minimization problem (***):

H(L, K, M, q, vL, vK, λ) = q + λ * (vL * L + vK * K + M - C(q; vL, vK, 1)),

where the new variable, λ, is called the Lagrange multiplier.

a.   First Order Necessary Conditions:

0.   Hλ(L, K, M, q, vL, vK, λ) = vL * L + vK * K + M - C(q; vL, vK, 1) = 0
1.   Hq(L, K, M, q, vL, vK, λ) = 1 - λ * ∂C(q; vL, vK, 1)/∂q = 0
2.   HvL(L, K, M, q, vL, vK, λ) = λ * (L - ∂C(q; vL, vK, 1)/∂vL) = 0
3.   HvK(L, K, M, q, vL, vK, λ) = λ * (K - ∂C(q; vL, vK, 1)/∂vK) = 0

b.   Solution Functions:

We want to solve, simultaneously, these four equations for the variables λ, q, vL, and vK as functions of the variables L, K, and M, and the parameters of the cost function.

λ = λ(L, K, M)
q = q(L, K, M)
vL = vL(L, K, M)
vK = vK(L, K, M)

c.   Jacobian Matrices:

The Jacobian matrix (bordered Hessian of H) of the four functions, Hλ, Hq, HvL, HvK, with respect to the choice variables, λ, q, vL, vK :

J3   =  

HλλHλqHλvLHλvK
HHqqHqvLHqvK
HvLλHvLqHvLvLHvLvK
HvKλHvKqHvKvLHvKvK
=
  0   -CqL - CvLK - CvK
  -Cq -λ*Cqq-λ*CqvL-λ*CqvK
L - CvL C-λ*CqvL-λ*CvLvL-λ*CvLvK
K - CvK -λ*CqvK-λ*CvKvL-λ*CvKvK

  =   Jλ, q, vL, vK

The bordered principal minor of the bordered Hessian of the Langrangian function, H:

J2   =  

  0   -CqL - CvL
  -Cq -λ*Cqq-λ*CqvL
L - CvL -λ*CqvL-λ*CvLvL

d.   Second Order Necessary Conditions:

The second order necessary conditions require that the Jacobian matrix (bordered Hessian of H) be positive definite at the solution vector W = (L, K, M, q, vL, vK, λ). The Jacobian matrix, J3, is a positive definite matrix if the determinant of J2 is negative, and the determinant of J3 is negative.

e.   Sufficient Conditions:

If the second order necessary conditions are satisfied, then the first order necessary conditions are sufficient for a minimum at W.

f. The Solution Functions' Comparative Statics.

The Jacobian matrix of the four functions, Hλ, Hq, HvL, HvK, with respect to the variables, L, K, M:

JL, K, M   =  

Hλ,L;Hλ,KHλ,M
Hq,LHq,KHq,M
HvL,LHvL,KHvL,M
HwK,LHwK,KHvK,M
=
vLvK1
000
λ00
0λ0

The Jacobian matrix of the four solution functions, Φ = {λ, q, vL, vK}, with respect to the variables, L, K, and M:

JΦ   =  

λLλKλM
qLqKqM
vLLvLKvLM
vKLvKKvKM

From the Implicit Function Theorem:

JL, K, M;   +   Jλ, q, vL, vM   * JΦ   =   0 (zero matrix)   →

JΦ   =   - (Jλ, q, vL, vK)-1   *   JL, K, M

for points (L, K, M) in a neighborhood of the point (L, K, M), and wL = wL(L, K, M), wK = wK(L, K, M), wM = wM(L, K, M), and λ = λ(L, K, M),

with:

F*(L, K, M) = q(L, K, M),
F*L(L, K, M) = qL(L, K, M),
F*K(L, K, M) = qK(L, K, M), and
F*M(L, K, M) = qM(L, K, M),

VI. Question: F == F*?

The obtained values for F*(L, K, M), F*L(L, K, M), F*K(L, K, M), and F*M(L, K, M) can be compared with the values of F(L, K, M), FL(L, K, M), FK(L, K, M), and FM(L, K, M), and with the corresponding values from the Generalized CES production function.

VII. Duality.  The Plan:

      1. Specify the parameters of a Generalized CES production function, and obtain the derived Generalized CES cost function using Newton's Method.

      2. Generate the Generalized CES data displayed in the table below, and estimate the parameters of a Translog production function, and the parameters of a Translog cost function.

      3. Using Newton's Method with the implicit function theorem, obtain the production function that is dual to the estimated Translog cost function. Check that the derived Translog production function corresponds with the estimated Translog production function, and the underlying Generalized CES production function.

      4. Using Newton's Method with the implicit function theorem, obtain the cost function that is dual to the estimated Translog production function. Check that the derived cost function corresponds with the estimated Translog cost function, and the underlying derived Generalized CES cost function.

VIII. Generate Generalized CES production / cost data.

The three factor Generalized CES production function is:

q = A * [alpha * L^-rhoL + beta * K^-rhoK + gamma *M^-rhoM]^(-nu/rho) = f(L,K,M).

where L = labour, K = capital, M = materials and supplies, and q = product.

The parameter nu permits one to adjust the returns to scale, while the parameter rho is the geometric mean of rhoL, rhoK, and rhoM:

rho = (rhoL * rhoK * rhoM)^1/3.

If rho = rhoL = rhoK = rhoM, we get the standard CES production function. If also, rho = 0, ie sigma = 1/(1+rho) = 1, we get the Cobb-Douglas production function.

The estimated coefficients of the Translog production and cost functions and will vary with the parameters nu, rho, rhoL, rhoK, rhoM, alpha, beta and gamma of the Generalized CES production function.

Set the parameters below to re-run with your own Generalized CES parameters.

The restrictions ensure that the least-cost problems can be solved to obtain the underlying Generalized CES cost function, using the parameters as specified.
Intermediate (and other) values of the parameters also work. The program might "time-out" for values of rho >> 0 and rho << 0, yielding "NAN" values.

Restrictions:
.8 < nu < 1.1;
-.6 < rhoL = rho < .6;
-.6 < rho < -.2 → rhoK = .95 * rho, rhoM = rho/.95;
-.2 <= rho < -.1 → rhoK = .9 * rho, rhoM = rho / .9;
-.1 <= rho < 0 → rhoK = .87 * rho, rhoM = rho / .87;
0 <= rho < .1 → rhoK = .7 * rho, rhoM = rho / .7;
.1 <= rho < .2 → rhoK = .67 * rho, rhoM = rho / .67;
.2 <= rho < .6 → rhoK = .6 * rho, rhoM = rho / .6;
rho = 0 → nu = alpha + beta + gamma (Cobb-Douglas)
4 <= wL* <= 11,   7<= wK* <= 16,   4 <= wM* <= 10

Generalized CES Production Function Parameters
nu:      
rho:      
Base Factor Prices
wL* wK* wM*
Distribution to Randomize Factor Prices
Use [-2, 2] Uniform distribution    
Use .25 * Normal (μ = 0, σ2 = 1)

The Generalized CES production function as specified:

q = 1 * [0.35 * L^- 0.17647 + 0.4 * K^- 0.11823 + 0.25 *M^- 0.26339]^(-1/0.17647) = f(L,K,M).

The factor prices are distributed about the base factor prices by adding a random number distributed uniformly in the [-2, 2] domain.

The Generalized CES cost function as derived from the Generalized CES production function:

Unlike the homothetic CES technology, the non-homothetic Generalized CES technology does not admit a closed form cost function. Consequently, the Generalized CES cost function is derived from the Generalized CES production function as demonstrated on the Generalized CES Production Function web page.

IX. Estimate the Translog Production and Cost Functions.

Estimating the Translog production function from the Generalized CES data using SVD least squares yields the coefficient estimates:

Generalized CES
Translog Production Function

SVD Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
lnA0.0066780.0019.061
aL0.358430815.302
aK0.269290.001532.989
aM0.3582280840.735
bLL-0.0191320-105.938
bKK-0.009810-50.775
bMM-0.0278560-183.247
bLK0.018733051.52
bLM0.014576085.228
bKM0.015605047.845
R2 = 1 R2b = 1 # obs = 28
Observation Matrix Rank: 10

aL + aK + aM = 0.986
-2*bLL = 0.038263 =~ 0.03331 = bLK + bLM
-2*bKK = 0.01962 =~ 0.034338 = bLK + bKLM
-2*bMM = 0.055712 =~ 0.030181 = bLM + bKM

The estimated Translog production function:

ln(q) = 0.006678 + 0.35843 * ln(L) + 0.26929 * ln(K) + 0.358228 * ln(M) + -0.019132 * ln(L)*ln(L) + -0.00981 * ln(K)*ln(K) + -0.027856 * ln(M)*ln(M)
+ 0.018733 * ln(L)*ln(K) + 0.014576 * ln(L)*ln(M) + 0.015605 * ln(K)*ln(M)   =   f(L,K,M).
       

The Translog cost factor share functions are:

sL(q;wL,wK,wM) = cL + dLq * ln(q) + dLL * ln(wL) + dLK * ln(wK) + dLM * ln(wM),

sK(q;wL,wK,wM) = cK + dKq * ln(q) + dKL * ln(wL) + dKK * ln(wK) + dKM * ln(wM),

sM(q;wL,wK,wM) = cM + dMq * ln(q) + dML * ln(wL) + dMK * ln(wK) + dMM * ln(wM),

three linear equations in their 15 parameters.

Estimating the Translog factor share functions simultaneously with the specified constraints, from the Generalized CES data using QR least squares, yields the coefficient estimates:

QR Restricted Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
cL0.361090.000221642.804313
dLq5.8E-56.5E-50.886905
dLL0.0343189.6E-5355.86694
dLK-0.013828.6E-5-161.377385
dLM-0.0204997.0E-5-293.928326
cK0.2722070.0002181245.888046
dKq0.0207546.7E-5311.975571
dKL-0.013828.6E-5-161.377385
dKK0.0307560.000113272.55637
dKM-0.0169367.8E-5-218.008585
cM0.3667030.0002171687.130471
dMq-0.0208116.7E-5-312.457944
dML-0.0204997.0E-5-293.928326
dMK-0.0169367.8E-5-218.008585
dMM0.0374358.7E-5431.310908
R2 = 1 R2b = 1 # obs = 84

dLK = dKL, dLM = dML, dKM = dMK
1 = cL + cK + cM
0 = dLL + dLK + dLM
0 = dKL + dKK + dKM
0 = dML + dMK + dMM
0 = dLq + dKq + dMq

The first three constraints are the usual across equation constraints on the factor share functions. The last five constraints are necessary conditions for the estimated Translog cost function to be linear homogeneous in factor prices. See the Generalized CES-Translog Cost Function for more details.

To obtain estimates of the remaining three parameters, c, cq, and dqq, of the Translog cost function write:

ln(C(q;wL,wK,wM)) - {cL * ln(wL) + cK * ln(wK) + cM * log(wM) + .5 * [dLL * ln(wL)^2 + dKK * ln(wK)^2 + dMM * ln(wM)^2]
+ .5 * [(dLK + dKL) * ln(wL)*ln(wK) + (dLM + dML) * ln(wL)*ln(wM) + (dKM + dMK) * ln(wK)*log(wM)]
+ dLq * ln(wL)*ln(q) + dKq * ln(wK)*ln(q) + dMq * ln(wM)*ln(q)}
= R(q;wL,wK,wM) = c + cq * ln(q) + .5 * dqq * ln(q)^2

Estimate the linear equation:

R(q;wL,wK,wM) = c + cq * ln(q) + .5 * dqq * ln(q)^2

to obtain c, cq, and dqq.

QR Restricted Least Squares
Parameter Estimates
Parameter Coefficient std error t-ratio
c1.1150620.0001427835.023451
cq101000
dqq0.0209962.4E-5859.622338
R2 = 1 R2b = 1 # obs = 28

1 = cq

The estimated Translog cost function:

ln(C(q;wL,wK,wM)) = 1.115062 + 1 * ln(q) + 0.36109 * ln(wL) + 0.272207 * ln(wK) + 0.366703 * log(wM)
+ .5 * [0.020996 * ln(q)^2 + 0.034318 * ln(wL)^2 + 0.030756 * ln(wK)^2 + 0.037435 * ln(wM)^2]
+ .5 * [-0.02764 * ln(wL)*ln(wK) + -0.040997 * ln(wL)*ln(wM) + -0.033873 * ln(wK)*log(wM)]
+ 5.8E-5 * ln(wL)*ln(q) + 0.020754 * ln(wK)*ln(q) + -0.020811 * ln(wM)*ln(q)          

X. Example: Cost Function to Production Function:

The dual Translog production function, F*, is obtained from the estimated Translog cost function, C, by:

F*(L, K, M) = minq, vL, vK {q   :   C(q; vL, vK, 1)   >=   vL * L + vK * K + M,   L > 0, K > 0, M > 0,   for all vL >= 0 , vK >= 0}     (***).

With L = 39.05, K = 21.4, and M = 36.07, (***) becomes:

F*(39.05, 21.4, 36.07) = minq, vL, vK {q   :   C(q; vL, vK, 1)   >=   vL * 39.05 + vK * 21.4 + 36.07,   for all vL >= 0 , vK >= 0}     (***).

XI. Constrained Optimization (Minimum): The Method of Lagrange:

H(39.05, 21.4, 36.07, q, vL, vK, λ) = q + λ * (vL * 39.05 + vK * 21.4 + 36.07 - C(q; vL, vK, 1)),

where λ is the Lagrange multiplier.

a. First Order Necessary Conditions:

0.   Hλ(39.05, 21.4, 36.07, q, vL, vK, λ) = vL * 39.05 + vK * 21.4 + 36.07 - C(q; vL, vK, 1) = 0
1.   Hq(39.05, 21.4, 36.07, q, vL, vK, λ) = 1 - λ * ∂C(q; vL, vK, 1)/∂q = 0
2.   HvL(39.05, 21.4, 36.07, q, vL, vK, λ) = λ * (39.05 - ∂C(q; vL, vK, 1)/∂vL) = 0
3.   HvK(39.05, 21.4, 36.07, q, vL, vK, λ) = λ * (21.4 - ∂C(q; vL, vK, 1)/∂vK) = 0

b.   Solution Functions:

Solve these four equations simultaneously (using Newton's Method) for λ, q, vL, and vK as functions of the variables L, K, and M, and the parameters of the cost function, so that:

λ = λ(39.05, 21.4, 36.07)
q = q(39.05, 21.4, 36.07)
vL = vL(39.05, 21.4, 36.07)
vK = vK(39.05, 21.4, 36.07)

We know that the ranges of factor prices are: 4 <= wL <= 11, 7 <= wK <= 16, and 4 <= wM <= 10. Estimating wL = 7.5, wK = 11.5, and wM = 7 yields vL = wL / wM = 7.5 / 7, and vK = wK / wM = 11.5/7. With a range of 20 to 45 for output, estimate q = 33. From the first order condition 1., estimate λ = 1 / ∂C(33; 7.5/7, 11.5/7, 1)/∂q = 0.218.

Newton's Method:

Using these estimates, Newton's Method provides:

Nonlinear Optimization: Newton's Method
Parameter Estimates
Iter #λqvLvK
  0   0.218 331.0714 1.6429
10.2122 26.791.0601 1.8585
20.1956 26.991.1515 2.1159
30.1945 271.1657 2.1647
40.1944 271.1661 2.166
50.1944 271.1661 2.166

c. Solution Vector:

W = (L, K, M, q, vL, vK, λ) = (39.05, 21.4, 36.07, 27, 1.16611, 2.16603, 0.19444)

With λ = 0.19444, q = 27, vL = 1.16611, and vK = 2.16603, the first order conditions are:

0.   Hλ(39.05, 21.4, 36.07, q, vL, vK, λ) = 1.16611 * 39.05 + 2.16603 * 21.4 + 36.07 - C(q; 1.16611, 2.16603, 1) = 0
1.   Hq(39.05, 21.4, 36.07, q, vL, vK, λ) = 1 - 0.19444 * ∂C(q; 1.16611, 2.16603, 1)/∂q = 0
2.   HvL(39.05, 21.4, 36.07, q, vL, vK, λ) = 0.19444 * (39.05 - ∂C(q; 1.16611, 2.16603, 1)/∂vL) = 0
3.   HvK(39.05, 21.4, 36.07, q, vL, vK, λ) = 0.19444 * (21.4 - ∂C(q; 1.16611, 2.16603, 1)/∂vK) = -0

d. Second Order Necessary Conditions:

The second order necessary conditions require that the Jacobian matrix (bordered Hessian of H) be positive definite at the solution vector W = (L, K, M, q, vL, vK, λ). The Jacobian matrix, J3, is a positive definite matrix if the determinant of J2 is negative, and the determinant of J3 is negative.

Jλ, q, vL, vK =

J3   =  

  0   -CqL - CvLK - CvK
  -Cq -λ*Cqq-λ*CqvL-λ*CqvK
L - CvL C-λ*CqvL-λ*CvLvL-λ*CvLvK
K - CvK -λ*CqvK-λ*CvKvL-λ*CvKvK
=
0-5.1430-0
-5.143-0.004-0.305-0.176
0-0.3053.567-1.134
-0-0.176-1.1341.062

  Determinant(J3) = -66.19419

 

J2   =  

0-5.1430
-5.143-0.004-0.305
0-0.3053.567

  Determinant(J2) = -94.33794

e. Maximum Output:

With L = 39.05, K = 21.4, M = 36.07,

          Dual Translog production function:

F*(39.05, 21.4, 36.07) = q(39.05, 21.4, 36.07) = 27,

          Estimated Translog production function:

F(39.05, 21.4, 36.07) = 27.

          Specified Generalized CES production function:

f(39.05, 21.4, 36.07) = 27.

f. The Solution Functions' Comparative Statics.

JL, K, M   =  

vLvK1
000
λ00
0λ0
  =  
1.166112.166031
000
0.194400
00.19440

From the Implicit Function Theorem:

JΦ   =   - (Jλ, wL, wK, wM, q)-1   *   JL, K, M

JΦ   =  

λLλKλM
qLqKqM
vLLvLKvLM
vKLvKKvKM
  =  
0.0019050.003554-0.005153
0.226740.421160.19444
-0.035066.0E-50.0407
0.00016-0.113180.07568

g. Comparing Partial Derivatives:

          Partial Derivates of the dual Translog production function F*(L,K,M):

F*L(L,K,M) = F*(39.05, 21.4, 36.07) = qL(39.05, 21.4, 36.07) = 0.227,
F*K(L,K,M) = F*(39.05, 21.4, 36.07) = qK(39.05, 21.4, 36.07) = 0.421,
F*M(L,K,M) = F*(39.05, 21.4, 36.07) = qM(39.05, 21.4, 36.07) = 0.194.

          Partial Derivatives of the estimated Translog production function F(L,K,M):

FL(39.05, 21.4, 36.07) = 0.227,
FK(39.05, 21.4, 36.07) = 0.421,
FM(39.05, 21.4, 36.07) = 0.194.

          Partial Derivatives of the specified Generalized CES production function f(L,K,M):

fL(39.05, 21.4, 36.07) = 0.227,
fK(39.05, 21.4, 36.07) = 0.421,
fM(39.05, 21.4, 36.07) = 0.194.

 

XII. Table of Results.

Check that the derived dual production function corresponds with the estimated Translog production function by comparing the values for output, q, and the partial derivatives of the production function.

Translog Production / Cost Function Duality
Cost Function to Production Function
Generalized CES: Returns to Scale = 1, rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
  —       Generalized CES Data   —    —    Estimated Translog Cost   —   — Derived Dual Production — — Estimated Translog Production —
obs #qwLwKwM LK Mcost LKMcostqF*LF*KF*MqFLFKFM
1177.06 14.326.72 24.9412.34 21.86499.6324.9512.3421.86499.8116.9950.2230.4530.212170.2230.4520.212
2185.96 13.025.86 27.3612.73 23.23464.9727.3712.7323.23465.0817.9960.2140.4680.21180.2140.4670.21
3197.06 12.265.26 25.314.4 26.91496.6725.3114.3926.92496.7518.9970.250.4340.186190.250.4340.186
4208.62 12.566.2 25.116.7 27.46596.3825.116.6927.48596.4419.9980.2680.3910.193200.2680.3910.193
5216.98 12.66.94 30.516.92 25.47602.8730.5116.9225.46602.88210.2260.4080.224210.2260.4070.224
6228.72 12.464.28 25.4317.23 37.62597.4925.4117.2337.64597.4222.0020.2950.4220.145220.2950.4220.145
7235.76 14.76.54 38.7815.99 28.69646.0238.7715.9928.68645.9823.0010.1890.4840.215230.1890.4830.215
8246.62 14.767.38 38.7118.01 29.05736.4138.718.0129.04736.3424.0020.20.4450.222240.20.4450.222
9255.98 11.67.46 39.7120.95 27.22683.639.7320.9627.19683.4925.0040.2030.3940.253250.2030.3940.253
10268.86 13.487.8 35.3823.05 32.21875.4335.3823.0432.21875.3226.0030.2440.3710.215260.2440.3710.214
11275.5 14.546.48 47.0919 33.38751.5647.0719.0133.37751.4627.0030.1820.4810.214270.1820.4810.214
12285.5 14.984.6 45.8517.99 42.71718.1745.8417.9942.72718.0628.0040.1960.5340.164280.1960.5340.164
13296.9 12.326.52 42.5924.29 36.2829.1142.5824.2836.2828.9929.0040.2230.3980.21290.2230.3980.21
14307 136 43.7924.14 40.13861.1743.7824.1440.13861.0530.0040.2240.4170.192300.2240.4170.192
15315.68 11.264.12 44.7123.28 46.69708.4244.723.2746.7708.3231.0040.2270.4510.165310.2270.4510.165
16327.16 13.825.64 46.7124.96 45.56936.2846.6924.9645.56936.1632.0040.2240.4330.177320.2240.4330.177
17335.04 14.365 57.8322.06 46.31839.7357.8122.0546.33839.6433.0030.1810.5150.18330.1810.5150.18
18346.34 14.965.36 54.1724.37 49.45973.1254.1624.3749.47973.0334.0030.2020.4770.171340.2020.4770.171
19358.28 13.946.98 49.9730.28 45.981156.8849.9630.2845.991156.8135.0020.230.3870.194350.230.3880.194
20366.32 14.364.68 55.3525.79 56.04982.4855.3625.7956.04982.4236.0020.2110.4790.156360.2110.480.156
21378.84 14.54.12 46.4828.72 68.71110.3346.4928.7468.611110.3536.9990.2670.4390.124370.2680.440.124
22386.84 12.567.16 58.9833.36 45.211146.1758.9833.3745.211146.237.9990.2080.3830.218380.2090.3830.219
23398.64 14.824.4 50.8930.36 69.691196.2150.9130.3769.621196.3138.9970.2560.4390.13390.2560.440.13
24405.22 12.75.68 69.0630.62 50.771037.7369.0530.6250.81037.8339.9970.1840.4470.2400.1840.4480.201
25417.98 12.26.94 57.9238.56 51.471289.7757.9238.5651.491289.9340.9960.2330.3560.202410.2330.3560.203
26426.48 135.22 63.9833.58 601164.336433.5860.021164.541.9940.2130.4270.172420.2130.4280.172
27436.36 13.724.24 64.6731.82 70.391146.4364.7231.8270.371146.6442.9930.2160.4660.144430.2170.4670.144
28447.36 14.485.66 66.2435.44 64.281364.4366.2635.4464.31364.7343.9910.2160.4250.166440.2160.4260.166

 

XIII. Production Function to Cost Function: The Method of Langrange

The dual cost function, C*(q; wL, wK, wM), is obtained from a production function, q = F(L, K, M), by the constrained optimization:

C*(q; wL, wK, wM) = minL,K,M{ wL * L + wK * K + wM * M   :   q - F(L,K,M) = 0,   q > 0, wL > 0; wK > 0, and wM > 0,   L > 0, K > 0, M > 0 }     (*)

Define the Langrangian function, G, of the least-cost problem (*):

G(q; wL, wK, wM, L, K, M, μ) = wL * L + wK * K + wM * M + μ * (q - F(L,K,M))

where the new variable, μ, is called the Lagrange multiplier.

a.   First Order Necessary Conditions:

0.   Gµ(q; wL, wK, wM, L, K, M, μ) = q - F(L, K, M) = 0
1.   GL(q; wL, wK, wM, L, K, M, μ) = wL - µ * FL(L, K, M) = 0
2.   GK(q; wL, wK, wM, L, K, M, μ) = wK - µ * FK(L, K, M) = 0
3.   GM(q; wL, wK, wM, L, K, M, μ) = wM - µ * FM(L, K, M) = 0

b.   Solution Functions:

We want to solve, simultaneously, these four equations for the variables µ, L, K, and M as functions of the variables q, wL, wK, and wM, and the parameters of the production function.

µ = µ(q; wL, wK, wM)
L = L(q; wL, wK, wM)
K = K(q; wL, wK, wM)
M = M(q; wL, wK, wM)

c.   Jacobian Matrices:

The Jacobian matrix (bordered Hessian of G) of the four functions, Gµ, GL, GK, GM, with respect to the choice variables, μ, L, K, and M:

J3   =  

GµµGµLGµKGµM
GGLLGLKGLM
GGKLGKKGKM
GGMLGMKGMM
=
0-FL-FK-FM
-FL-µ * FLL-µ * FLK-µ * FLM
-FK-µ * FKL-µ * FKK-µ * FKM
-FM-µ * FML-µ * FMK-µ * FMM

  =   Jµ, L, K, M

The bordered principal minor of the bordered Hessian of the Langrangian function, G:

J2   =  

0-FL-FK
-fL-µ * FLL-µ * FLK
-fK-µ * FKL-µ * FKK

d.   Second Order Necessary Conditions:

The second order necessary conditions require that the Jacobian matrix (bordered Hessian of G) be positive definite at the solution vector Z = (q, wL, wK, wM, L, K, M, µ). The Jacobian matrix, J3, is a positive definite matrix if the determinants of J2 and J3 are both negative.

e.   Sufficient Conditions:

If the second order necessary conditions are satisfied, then the first order necessary conditions are sufficient for a minimum at Z.

f. The Solution Functions' Comparative Statics.

The Jacobian matrix of the four functions, Gµ, GL, GK, GM, with respect to the variables, q, wL, wK, and wM:

Jq, wL, wK, wM;   =  

GµqGµwLGµwKGµwM
GLqGLwLGLwKGLwM
GKqGKwLGKwKGKwM
GMqGMwLGMwKGMwM
=
1000
0100
0010
0001

The Jacobian matrix of the four solution functions, Φ = {µ, L, K, M}, with respect to the variables, q, wL, wK, and wM:

JΦ   =  

µqµwLµwKµwM
LqLwLLwKLwM
KqKwLKwKKwM
MqMwLMwKMwM

From the Implicit Function Theorem:

Jq, wL, wK, wM;   +   Jµ, L, K, M   * JΦ   =   0 (zero matrix)   →

JΦ   =   - (Jµ, L, K, M)-1

for points (q; wL, wK, wM) in a neighborhood of the point (q; wL, wK, wM), and L = L(q; wL, wK, wM), K = K(q; wL, wK, wM), M = M(q; wL, wK, wM), and µ = µ(q; wL, wK, wM).

XIV. Question C == C*?

The obtained values of the solution functions at q, wL, wK, and wM:

µ = µ(q; wL, wK, wM)
L = L(q; wL, wK, wM)
K = K(q; wL, wK, wM)
M = M(q; wL, wK, wM),

and the calculated value of the dual Translog cost function:

C*(q; wL, wK, wM) = wL * L + wK * K + wM * M

can be compared with the corresponding values of the estimated Translog cost function, and the underlying CES cost function. Moreover, these functions' corresponding comparative static values can also be compared.

XV. Example: Production Function to Cost Function:

The dual Translog cost function, C*, is obtained from the estimated Translog production function, F, by:

C*(q; wL, wK, wM) = minL,K,M{ wL * L + wK * K + wM * M   :   q - F(L,K,M) = 0,   q > 0, wL > 0; wK > 0, and wM > 0,   L > 0, K > 0, M > 0 }     (*)

With q = 30, wL = 7, wK = 13, and wM = 6, (*) becomes:

C*(30; 7, 13, 6) = minL,K,M{ 7 * L + 13 * K + 6 * M   :   30 - F(L,K,M) = 0,     L > 0, K > 0, M > 0 }     (*)

The Langrangian function, G, of the least-cost problem (*):

G(30; 7, 13, 6, L, K, M, μ) = 7 * L + 13 * K + 6 * M + μ * (30 - F(L,K,M))

where the new variable, μ, is called the Lagrange multiplier.

a.   First Order Necessary Conditions:

0.   Gµ(30; 7, 13, 6, L, K, M, μ) = 30 - F(L, K, M) = 0
1.   GL(30; 7, 13, 6, L, K, M, μ) = 7 - µ * FL(L, K, M) = 0
2.   GK(30; 7, 13, 6, L, K, M, μ) = 13 - µ * FK(L, K, M) = 0
3.   GM(30; 7, 13, 6, L, K, M, μ) = 13 - µ * FM(L, K, M) = 0

b.   Solution Functions:

We want to solve, simultaneously, these four equations for the variables µ, L, K, and M as functions of the variables q, wL, wK, and wM, and the parameters of the production function, so that.

µ = µ(30; 7, 13, 6)
L = L(30; 7, 13, 6)
K = K(30; 7, 13, 6)
M = M(30; 7, 13, 6)

Suppose we think that the factor inputs for q = 30 are about 20 units. Estimate L = 20, K = 20, and M = 20, and guess µ = 10 to start Newton's Method.

Newton's Method:

Using these estimates, Newton's Method provides:

Nonlinear Optimization: Newton's Method
Parameter Estimates
Iter #µLKM
  0   10 2020 20
131.748 48.788.3107 45.3267
236.9687 48.2316.9224 43.2498
332.0725 45.1922.0266 41.3943
431.2736 43.8423.9718 40.1977
531.2088 43.7824.1426 40.1468
631.2085 43.7824.1436 40.1464
731.2085 43.7824.1436 40.1464

c. Solution Vector:

Z = (q, wL, wK, wM, L, K, M, µ) = (30, 7, 13, 6, 43.78, 24.14, 40.15, 31.21)

With µ = 31.21, L = 43.78, K = 24.14, and M = 40.15, the first order conditions are:

0.   Gµ(30; 7, 13, 6, L, K, M, μ) = 30 - F(43.78, 24.14, 40.15) = -0
1.   GL(30; 7, 13, 6, L, K, M, μ) = 7 - µ * FL(43.78, 24.14, 40.15) = -0
2.   GK(30; 7, 13, 6, L, K, M, μ) = 13 - µ * FK(43.78, 24.14, 40.15) = 0
3.   GM(30; 7, 13, 6, L, K, M, μ) = 13 - µ * FM(43.78, 24.14, 40.15) = -0

d. Second Order Necessary Conditions:

The second order necessary conditions require that the Jacobian matrix (bordered Hessian of G) be positive definite at the solution vector Z = (q, wL, wK, wM, L, K, M, µ). The Jacobian matrix, J3, is a positive definite matrix if the determinants of J2 and J3 are both negative.

Jµ, L, K, M   =  

J3   =  

0-FL-FK-FM
-FL-µ * FLL-µ * FLK-µ * FLM
-FK-µ * FKL-µ * FKK-µ * FKM
-FM-µ * FML-µ * FMK-µ * FMM
=
0-0.224-0.417-0.192
-0.2240.126-0.114-0.053
-0.417-0.1140.389-0.098
-0.192-0.053-0.0980.143

  Determinant(J3) = -0.01602

 

J2   =  

0-0.224-0.417
-0.2240.126-0.114
-0.417-0.1140.389

  Determinant(J2) = -0.06276

e. The Solution Functions' Comparative Statics:

JΦ   =  

C*q,qC*q,wLC*q,wKC*q,wM
C*wL,qC*wL,wLC*wL,wKC*wL,wM
C*wK,qC*wK,wLC*wK,wKC*wK,wM
C*wM,qC*wM,wLC*wM,wKC*wM,wM
=
µqµwLµwKµwM
LqLwLLwKLwM
KqKwLKwKKwM
MqMwLMwKMwM
=
0.1071.5850.9211.357
1.585-3.4351.1021.621
0.9211.102-1.0250.935
1.3571.6210.935-3.918

The estimated Translog cost function's comparative statics.

Cq,qCq,wLCq,wKCq,wM
CwL,qCwL,wLCwL,wKCwL,wM
CwK,qCwK,wLCwK,wKCwK,wM
CwM,qCwM,wLCwM,wKCwM,wM
=
0.1111.5870.9211.355
1.587-3.4251.0961.62
0.9211.096-1.0230.938
1.3551.620.938-3.923

The underlying specified Generalized CES cost function's comparative statics:

Cq,qCq,wLCq,wKCq,wM
CwL,qCwL,wLCwL,wKCwL,wM
CwK,qCwK,wLCwK,wKCwK,wM
CwM,qCwM,wLCwM,wKCwM,wM
=
0.1071.5870.9211.355
1.587-3.4241.0981.616
0.9211.098-1.0240.937
1.3551.6160.937-3.915

f. Comparing Least-Cost Combinations of Inputs:

With q = 30, wL = 7, wK = 13, and wM = 6,

          Derived Dual Translog cost function factor demands: (L, K, M) = (43.78, 24.14, 40.15)

          Estimated Translog cost function factor demands: (L, K, M) = (43.78, 24.14, 40.13)

          Specified Generalized CES cost function factor demands: (L, K, M) = (43.79, 24.14, 40.13)

g. Comparing total costs and marginal costs:

          Derived Dual Translog cost function: total cost = 861.18; marginal cost = µ = 31.21.

          Estimated Dual Translog cost function: total cost = 861.05; marginal cost = 31.21.

          Specified Generalized CES cost function: total cost = 861.17; marginal cost = 31.21.

XVI. Table of Results.

Check that the derived dual cost function corresponds with the estimated Translog cost function by comparing the values of the inputs, L, K, and M.

Translog Production / Cost Function Duality
Production Function to Cost Function
Generalized CES: Returns to Scale = 1, rho = 0.17647, rhoL = 0.17647, rhoK = 0.11823, rhoM = 0.26339
—   Generalized CES Data   — — Est. Translog Production — — Derived Dual Translog Cost —   — Estimated Translog Cost —  
obs #qwLwKwMLKMcost q LK ML KMcostLKMcost
117 7.06 14.326.72 24.9412.34 21.86499.63 1724.9412.3421.86 24.9612.34 21.83499.63 24.9512.3421.86499.81
218 5.96 13.025.86 27.3612.73 23.23464.97 1827.3612.7323.23 27.3812.73 23.2464.97 27.3712.7323.23465.08
319 7.06 12.265.26 25.314.4 26.91496.67 1925.314.426.91 25.314.4 26.9496.69 25.3114.3926.92496.75
420 8.62 12.566.2 25.116.7 27.46596.38 2025.116.727.46 25.0916.71 27.46596.39 25.116.6927.48596.44
521 6.98 12.66.94 30.516.92 25.47602.87 2130.516.9225.47 30.5116.93 25.45602.87 30.5116.9225.46602.88
622 8.72 12.464.28 25.4317.23 37.62597.49 2225.4317.2337.62 25.4117.25 37.62597.47 25.4117.2337.64597.42
723 5.76 14.76.54 38.7815.99 28.69646.02 2338.7815.9928.69 38.815.99 28.67646.02 38.7715.9928.68645.98
824 6.62 14.767.38 38.7118.01 29.05736.41 2438.7118.0129.05 38.7218.01 29.03736.41 38.718.0129.04736.34
925 5.98 11.67.46 39.7120.95 27.22683.6 2539.7120.9527.22 39.7420.96 27.19683.58 39.7320.9627.19683.49
1026 8.86 13.487.8 35.3823.05 32.21875.43 2635.3823.0532.21 35.3723.06 32.21875.43 35.3823.0432.21875.32
1127 5.5 14.546.48 47.0919 33.38751.56 2747.091933.38 47.119 33.37751.57 47.0719.0133.37751.46
1228 5.5 14.984.6 45.8517.99 42.71718.17 2845.8517.9942.71 45.8617.99 42.71718.18 45.8417.9942.72718.06
1329 6.9 12.326.52 42.5924.29 36.2829.11 2942.5924.2936.2 42.5824.29 36.21829.12 42.5824.2836.2828.99
1430 7 136 43.7924.14 40.13861.17 3043.7924.1440.13 43.7824.14 40.15861.18 43.7824.1440.13861.05
1531 5.68 11.264.12 44.7123.28 46.69708.42 3144.7123.2846.69 44.723.28 46.7708.41 44.723.2746.7708.32
1632 7.16 13.825.64 46.7124.96 45.56936.28 3246.7124.9645.56 46.6924.96 45.57936.28 46.6924.9645.56936.16
1733 5.04 14.365 57.8322.06 46.31839.73 3357.8322.0646.31 57.8322.05 46.32839.72 57.8122.0546.33839.64
1834 6.34 14.965.36 54.1724.37 49.45973.12 3454.1724.3749.45 54.1624.37 49.46973.12 54.1624.3749.47973.03
1935 8.28 13.946.98 49.9730.28 45.981156.88 3549.9730.2845.98 49.9430.29 46.011156.87 49.9630.2845.991156.81
2036 6.32 14.364.68 55.3525.79 56.04982.48 3655.3525.7956.04 55.3525.8 56.03982.47 55.3625.7956.04982.42
2137 8.84 14.54.12 46.4828.72 68.71110.33 3746.4828.7268.7 46.4728.77 68.551110.35 46.4928.7468.611110.35
2238 6.84 12.567.16 58.9833.36 45.211146.17 3858.9833.3645.21 58.9633.36 45.241146.21 58.9833.3745.211146.2
2339 8.64 14.824.4 50.8930.36 69.691196.21 3950.8930.3669.69 50.8830.4 69.561196.23 50.9130.3769.621196.31
2440 5.22 12.75.68 69.0630.62 50.771037.73 4069.0630.6250.77 69.0430.61 50.811037.76 69.0530.6250.81037.83
2541 7.98 12.26.94 57.9238.56 51.471289.77 4157.9238.5651.47 57.8738.56 51.521289.76 57.9238.5651.491289.93
2642 6.48 135.22 63.9833.58 601164.33 4263.9833.5860 63.9633.58 60.021164.3 6433.5860.021164.5
2743 6.36 13.724.24 64.6731.82 70.391146.43 4364.6731.8270.39 64.6931.84 70.321146.43 64.7231.8270.371146.64
2844 7.36 14.485.66 66.2435.44 64.281364.43 4466.2435.4464.28 66.2135.44 64.291364.39 66.2635.4464.31364.73




Mathematical Notes

1. The Translog (Transcendental Logarithmic) Production Function:

ln(q) = ln(A) + aL * ln(L) + aK * ln(K) + aM * ln(M)
              + bLL * ln(L) * ln(L) + bKK * ln(K) * ln(K) + bMM * ln(M) * ln(M)
                  + bLK * ln(L) * ln(K) + bLM * ln(L) * ln(M) + bKM * ln(K) * ln(M)   =   f(L,K,M).

an equation in 10 parameters, A, aL, aK, aM, bLL, bKK, bMM, bLK, bLM, bKM, where L = labour, K = capital, M = materials and supplies, and q = product.

2. The partial derivatives of f(L,K,M):

fL(L,K,M) = (1/L) * [aL + 2 * bLL * ln(L) + bLK * ln(K) + bLM * ln(M)] = (1/L) * vL ,
fK(L,K,M) = (1/K) * [aK + 2 * bKK * ln(K) + bLK * ln(L) + bKM * ln(M)] = (1/K) * vK,
fM(L,K,M) = (1/M) * [aM + 2 * bMM * ln(M) + bLM * ln(L) + bKM * ln(K)] = (1/M)* vM,

fLL = (1/L^2) * [2 * bLL - vL],   fLK = bLK / (L*K),   fLM = bLM / (L*M),  
fKK = (1/K^2) * [2 * bKK - vK],   fKL = bLK / (L*K),   fKM = bKM / (K*M),
fMM = (1/M^2) * [2 * bMM - vM],   fML = bLM / (L*M),   fMK = bKM / (K*M).

3. The partial derivatives of F(L,K,M) = exp(f(L,K,M)):

FL = fL * exp(f(L,K,M)),   FK = fK * exp(f(L,K,M)),   FM = fM * exp(f(L,K,M)),
FLL = [fLL + fL * fL] * exp(f(L,K,M)),   FLK = [fLK + fL * fK] * exp(f(L,K,M)),   FLM = [fLM + fL * fM] * exp(f(L,K,M)),  
FKL = FLK,   FKK = [fKK + fK * fK] * exp(f(L,K,M)),   FKM = [fKM + fK * fM] * exp(f(L,K,M)),  
FML = FLM,   FMK = FKM,   FMM = [fMM + fM * fM] * exp(f(L,K,M)),  

4. The Translog Cost Function:

ln(C(q;wL,wK,wM)) = c + cq * ln(q) + cL * ln(wL) + cK * ln(wK) + cM * log(wM)              
                  + .5 * [dqq * ln(q)^2 + dLL * ln(wL)^2 + dKK * ln(wK)^2 + dMM * ln(wM)^2]
                  + .5 * [(dLK + dKL) * ln(wL)*ln(wK) + (dLM + dML) * ln(wL)*ln(wM) + (dKM + dMK) * ln(wK)*log(wM)]
                  + dLq * ln(wL)*ln(q) + dKq * ln(wK)*ln(q) + dMq * ln(wM)*ln(q)   =   C(q;wL,wK,wM)        

an equation in 18 parameters, c, cq, cL, cK, cM, dqq, dLL, dKK, dMM, dLK, dKL, dLM, dML, dKM, dMK, dLq, dKq, and dMq, where wL, wK, and wM are the factor prices of L, K, and M respectively.

Note: C(q; wL, wK, wM) = exp(C(q; wL, wK, wM)), a change in the notation above where C = CES cost function.

5. The Factor Share Functions:

∂ln(C)/∂ln(wL) = (∂ln(C)/∂wL )/ (∂ln(wL)/∂wL) = (1 / C(q;wL,wK,wM)) * ∂ln(C)/∂wL * wL = wL * L(q; wL, wK, wM) / C(q;wL,wK,wM) = sL(q;wL,wK,wM),

∂ln(C)/∂ln(wK) = (∂ln(C)/∂wK )/ (∂ln(wK)/∂wK) = (1 / C(q;wL,wK,wM)) * ∂ln(C)/∂wK * wK = wK * K(q; wL, wK, wM) / C(q;wL,wK,wM) = sK(q;wL,wK,wM),

∂ln(C)/∂ln(wM) = (∂ln(C)/∂wM )/ (∂ln(wM)/∂wM) = (1 / C(q;wL,wK,wM)) * ∂ln(C)/∂wM * wM = wM * M(q; wL, wK, wM) / C(q;wL,wK,wM) = sM(q;wL,wK,wM).

 

sL(q;wL,wK,wM) = cL + dLq * ln(q) + dLL * ln(wL) + dLK * ln(wK) + dLM * ln(wM),

sK(q;wL,wK,wM) = cK + dKq * ln(q) + dKL * ln(wL) + dKK * ln(wK) + dKM * ln(wM),

sM(q;wL,wK,wM) = cM + dMq * ln(q) + dML * ln(wL) + dMK * ln(wK) + dMM * ln(wM),

6. The Factor Demand Functions:

L(q;wL,wK,wM) = sL(q;wL,wK,wM) * C(q;wL,wK,wM) / wL,

K(q;wL,wK,wM) = sK(q;wL,wK,wM) * C(q;wL,wK,wM) / wK,

M(q;wL,wK,wM) = sM(q;wL,wK,wM) * C(q;wL,wK,wM) / wM.

7. The partial derivatives of C(q; wL, wK, wM) = exp(C(q; wL, wK, wM)):

CwL(q;wL,wK,wM) = L(q;wL,wK,wM) = sL(q;wL,wK,wM) * C(q;wL,wK,wM) / wL,
CwK(q;wL,wK,wM) = K(q;wL,wK,wM) = sK(q;wL,wK,wM) * C(q;wL,wK,wM) / wK,
CwM(q;wL,wK,wM) = M(q;wL,wK,wM) = sM(q;wL,wK,wM) * C(q;wL,wK,wM) / wM,

Writing:

Cq(q;wL,wK,wM) = (ecq + edqq * log(q) + edLq * log(wL) + edKq * log(wK) + edMq * log(wM) ) / q,
Cq,q(q;wL,wK,wM) = (edqq - (ecq + edqq * log(q) + edLq * log(wL) + edKq * log(wK) + edMq * log(wM) )) / (q * q),

Cq(q;wL,wK,wM) = Cq(q;wL,wK,wM) * C(q;wL,wK,wM),
Cq,q(q;wL,wK,wM) = ( Cq,q(q;wL,wK,wM) + Cq(q;wL,wK,wM) * Cq(q;wL,wK,wM) ) * C(q;wL,wK,wM);

then:

CwL,q(q;wL,wK,wM) = ( edLq * C(q; wL, wK, wM) / q + sL(q,wL,wK,wM) * Cq(q;wL,wK,wM) ) / wL,
CwL,wL(q;wL,wK,wM) = ( L(q;wL,wK,wM) / wL ) * (-1 + sL(q,wL,wK,wM) + edLL / sL(q,wL,wK,wM)),
CwL,wK(q;wL,wK,wM) = ( L(q,wL,wK,wM) / wK ) * (sK(q,wL,wK,wM) + edLK / sL(q,wL,wK,wM)),
CwL,wM(q;wL,wK,wM) = ( L(q,wL,wK,wM) / wM ) * (sM(q,wL,wK,wM) + edLM / sL(q,wL,wK,wM)),

CwK,q(q;wL,wK,wM) = ( edKq * C(q; wL, wK, wM) / q + sK(q,wL,wK,wM) * Cq(q;wL,wK,wM) ) / wK,
CwK,wL(q;wL,wK,wM) = ( K(q;wL,wK,wM) / wL ) * (sL(q,wL,wK,wM) + edKL / sK(q,wL,wK,wM)),
CwK,wK(q;wL,wK,wM) = ( K(q,wL,wK,wM) / wK ) * (-1 + sK(q,wL,wK,wM) + edKK / sK(q,wL,wK,wM)),
CwK,wM(q;wL,wK,wM) = ( K(q,wL,wK,wM) / wM ) * (sM(q,wL,wK,wM) + edKM / sK(q,wL,wK,wM)),

CwM,q(q;wL,wK,wM) = ( edMq * C(q; wL, wK, wM) / q + sM(q,wL,wK,wM) * Cq(q;wL,wK,wM) ) / wM,
CwM,wL(q;wL,wK,wM) = ( M(q;wL,wK,wM) / wL ) * (sL(q,wL,wK,wM) + edML / sM(q,wL,wK,wM)),
CwM,wK(q;wL,wK,wM) = ( M(q,wL,wK,wM) / wK ) * (sK(q,wL,wK,wM) + edMK / sM(q,wL,wK,wM)),
CwM,wM(q;wL,wK,wM) = ( M(q,wL,wK,wM) / wM ) * (-1 + sM(q,wL,wK,wM) + edMM / sM(q,wL,wK,wM)),

 

 
   

      Copyright © Elmer G. Wiens:   Egwald Web Services       All Rights Reserved.    Inquiries